tag:tagteam.harvard.edu,2005:/hubs/1186/user/amarashar/atom 2021-07-29T14:15:51-04:00 TagTeam social RSS aggregrator tag:tagteam.harvard.edu,2005:FeedItem/2724986 2020-07-06T09:26:56-04:00 2021-07-29T14:15:51-04:00

<p> </p> <p><span>Juan Ortiz Freuler </span></p> <p><span>berkman affiliate</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2848709 2020-12-31T16:02:32-05:00 2020-12-31T16:02:32-05:00

<p>Google’s appalling treatment of Gebru exposes a dual crisis in AI research. The field is dominated by an elite, primarily white male workforce, and it is controlled and funded primarily by large industry players—Microsoft, Facebook, Amazon, IBM, and yes, Google. With Gebru’s firing, the civility politics that yoked the young effort to construct the necessary guardrails around AI have been torn apart, bringing questions about the racial homogeneity of the AI workforce and the inefficacy of corporate diversity programs to the center of the discourse. But this situation has also made clear that—however sincere a company like Google’s promises may seem—corporate-funded research can never be divorced from the realities of power, and the flows of revenue and capital.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2838643 2020-12-15T17:02:53-05:00 2020-12-15T17:02:53-05:00

<p>We look forward to seeing statements of solidarity from Stanford’s AI leaders and institutions, followed by actions that defend the integrity of research and the dignity of underrepresented researchers.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2838516 2020-12-15T16:42:05-05:00 2020-12-15T16:42:05-05:00

<p>In their paper, ‘Why Fairness Cannot Be Automated: Bridging the gap between EU non-discrimination law and AI’ Professor Sandra Wachter, Associate Professor, and Dr Brent Mittelstadt of the Oxford Internet Institute, University of Oxford and Dr Chris Russell, propose a new test for ensuring fairness in algorithmic modelling and data driven decisions, called ‘Conditional Demographic Disparity’ (CDD). The test is significant as it aligns with the approach used by courts across Europe in applying non-discrimination law.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2837545 2020-12-14T08:21:24-05:00 2020-12-14T08:21:24-05:00

<p>Why do biased predictions arise about human capital? What interventions can prevent them? We evaluate 8.2 million algorithmic predictions of math skill from ~400 AI engineers, each of whom developed an algorithm under a randomly assigned experimental condition. Our treatment arms modified programmers' incentives, training data, awareness, and/or technical knowledge of AI ethics. We then assess out-of-sample predictions from their algorithms using randomized audit manipulations of algorithm inputs and ground-truth math performance for 20K subjects. We find that biased predictions are mostly caused by biased training data. However, one-third of the benefit of better training data comes through a novel economic mechanism: Engineers exert greater effort and are more responsive to incentives when given better training data. We also assess how performance varies with programmers' demographic characteristics, and their performance on a psychological test of implicit bias (IAT) concerning gender and careers. We find no evidence that female, minority and low-IAT engineers exhibit lower bias or discrimination in their code. However we do find that prediction errors are correlated within demographic groups, which creates performance improvements through cross-demographic averaging. Finally, we quantify the benefits and tradeoffs of practical managerial or policy interventions such as technical advice, simple reminders and improved incentives for decreasing algorithmic bias.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2835470 2020-12-11T09:10:26-05:00 2020-12-11T09:10:26-05:00

<p>However, many AI systems are designed with embedded conflicts of interests, acting in ways that subtly benefit their creators (or funders) at the expense of users. Unlike the relationship between an individual and a doctor, lawyer, or financial advisor, there is no requirement that AI systems act in ways that are consistent with the users’ best interests. To address this problem, in this paper we introduce the concept of AI loyalty. AI systems are loyal to the degree that they are designed to minimize, and make transparent, conflicts of interest, and to act in ways that prioritize the interests of users. Properly designed, such systems could have considerable functional and competitive – not to mention ethical – advantages relative to those that do not. Loyal AI products hold an obvious appeal for the end-user and could serve to promote the alignment of the long-term interests of AI developers and customers. To this end, we suggest criteria for assessing whether an AI system is sufficiently transparent about conflicts of interest, and acting in a manner that is loyal to the user, and argue that AI loyalty should be deliberately considered during the technological design process alongside other important values in AI ethics such as fairness, accountability privacy, and equity. We discuss a range of mechanisms, from pure market forces to strong regulatory frameworks, that could support incorporation of AI loyalty into a variety of future AI systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2830189 2020-12-04T13:17:03-05:00 2020-12-04T13:17:03-05:00

<p>The paper recommends solutions like working with impacted communities, value sensitive design, improved data documentation, and adoption of frameworks such as Bender’s data statements for NLP or the datasheets for datasets approach Gebru coauthored while at Microsoft Research.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2830076 2020-12-04T10:25:13-05:00 2020-12-04T10:25:13-05:00

<p>A core theme of this report is that we now have the opportunity to adopt a more rigorous and proactive approach to identifying and mitigating bias in key areas of life, such as policing, social services, finance and recruitment. Good use of data can enable organisations to shine a light on existing practices and identify what is driving bias. There is an ethical obligation to act wherever there is a risk that bias is causing harm and instead make fairer, better choices.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2827796 2020-12-01T11:15:13-05:00 2020-12-01T11:15:13-05:00

<p>Regulations will have widespread effects on supply chains and consumer access to global platforms, as well as on data and privacy rules. Models for technology governance may be driven by security or social stability concerns, human rights (as in the EU) or economic factors such as antitrust. Whatever the case, decisions taken in Beijing, Brussels or Washington will ripple across the world.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2827401 2020-11-30T20:38:56-05:00 2020-11-30T20:38:56-05:00

<p>The goal of the three procurement roundtables held with a group of experts from academia, business and policy is to: identify concrete gaps in existing procurement and development processes map them to the risks that will arise from widespread use of these technologies create a new framework for public procurement guidelines and standards for public-use data science applications</p>

tag:tagteam.harvard.edu,2005:FeedItem/2827053 2020-11-30T11:19:26-05:00 2020-11-30T11:19:26-05:00

<p>To ground our discussion, we chose a hypothetical scenario: an electricity retailer uses an AI-powered tool to decide how to offer its products to customers, and on what terms. The general principles and solutions for mitigating the problem, however, will be relevant far beyond this specific situation. Because algorithmic bias can result in unlawful activity, there is a legal imperative to address this risk. However, good businesses go further than the bare minimum legal requirements, to ensure they always act ethically and do not jeopardise their good name. Rigorous design, testing and monitoring can avoid algorithmic bias. This technical paper offers some guidance for companies to ensure that when they use AI, their decisions are fair, accurate and comply with human rights. On behalf of the Australian Human Rights Commission, I pay tribute to our partner organisations in this project for the deep expertise they provided throughout this work: Gradient Institute, Consumer Policy Research Centre, CHOICE and CSIRO’s Data61.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2827048 2020-11-30T10:58:49-05:00 2020-11-30T10:58:49-05:00

<p>In 2020, starting from the “7 Key Requirements for a Trustworthy AI” published by the European Commission in April 2019, AI4People’s Committees members have assessed the impact of these requirements on 7 business sectors (Automotive, Banking and Finance, Energy, Healthcare, Insurance, Legal Services Industry and Media and Technology) whilst having suggested some practical steps that businesses operating within these industries must undertake to be compliant with the Requirements.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2826961 2020-11-30T09:46:21-05:00 2020-11-30T09:46:21-05:00

<p>To Shah, the problem of algorithmic fairness is most concerning when it leads to unfair treatment in the clinic. A recent paper by Ziad Oberyemer received a lot of attention for exactly this reason, Shah says. There, a healthcare provider had used a cost predictive algorithm to decide which patients should be referred to a special high-risk care management program. The algorithm was one that used historic healthcare costs to predict future healthcare costs (and did so in an unbiased way). But when the healthcare provider used future healthcare cost projections as a proxy for healthcare need, the impact of that usage led to unfair treatment: Black patients had to be a lot sicker than white patients before they received the extra care.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2826960 2020-11-30T09:44:44-05:00 2020-11-30T09:44:44-05:00

<p>In a new paper by Ho and Alice Xiang, the head of Fairness, Transparency, and Accountability Research at the Partnership on AI, a research group that focuses on responsible AI, the authors warn that many of the strategies for increasing fairness clash directly with the high court’s push for “anti-classification.” That’s the principle of remaining “blind” toward categories like race, gender, or religion.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2826941 2020-11-30T09:28:54-05:00 2020-11-30T09:28:54-05:00

<p>These efforts reflect a recognition that existing research norms have failed to address the impacts of AI research [7] [8] [9] [10], and take place against the backdrop of a larger reckoning with the role of AI in perpetuating injustice [11] [12] [13] [14] [15]. The changes have been met with both praise and criticism [16] some within and outside the community see them as a crucial first step towards integrating ethical reflection and review into the research process, fostering necessary changes to protect populations at risk of harm. Others worry that AI researchers are not well placed to recognize and reason about the potential impacts of their work, as effective ethical deliberation may require different expertise and the involvement of other stakeholders.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2822935 2020-11-24T15:57:09-05:00 2020-11-24T15:57:09-05:00

<p>1. authors @KLdivergence @alexdamour 2. evaluating a policy vs evaluating a prediction 3. fairness using data alone vs using additional context 4. examples 5. input from many, please keep pointing out what we missed</p>

tag:tagteam.harvard.edu,2005:FeedItem/2821812 2020-11-23T11:26:12-05:00 2020-11-23T11:26:12-05:00

<p>For years, technology ethicists have considered how to square the interests of major companies with the interests of society as a whole, and recent approaches to ethics in the technology industry provide a cautionary tale for antitrust policy. Our data ethics research shows that Big Tech companies tend to approach challenges from the perspective of compliance: that as long as a company ticks the boxes on a checklist, it is in the clear. While checklists can highlight concerns, they do not necessarily lead to ethical actions. Current approaches to antitrust threaten to extend a compliance mindset, which will likely result in efforts that ostensibly tackle market power but fail to protect consumers.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2821788 2020-11-23T10:49:54-05:00 2020-11-23T10:49:54-05:00

<p>Accused of enabling genocide, the social media giant has temporarily changed its rules in Myanmar to prevent abuse and misinformation.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2821776 2020-11-23T10:26:53-05:00 2020-11-23T10:26:53-05:00

<p>Image recognition systems offer the promise to learn from images at scale without requiring expert knowledge. However, past research suggests that machine learning systems often produce biased output. In this article, we evaluate potential gender biases of commercial image recognition platforms using photographs of U.S. members of Congress and a large number of Twitter images posted by these politicians. Our crowdsourced validation shows that commercial image recognition systems can produce labels that are correct and biased at the same time as they selectively report a subset of many possible true labels. We find that images of women received three times more annotations related to physical appearance. Moreover, women in images are recognized at substantially lower rates in comparison with men. We discuss how encoded biases such as these affect the visibility of women, reinforce harmful gender stereotypes, and limit the validity of the insights that can be gathered from such data.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2819889 2020-11-20T10:20:15-05:00 2020-11-20T10:20:15-05:00

<p>This article examines how modes of governance are reconfigured as a result of using algorithms in the governance process. We argue that deploying algorithmic systems creates a shift toward a special form of design‐based governance, with power exercised ex ante via choice architectures defined through protocols, requiring lower levels of commitment from governing actors. We use governance of three policy problems – speeding, disinformation, and social sharing – to illustrate what happens when algorithms are deployed to enable coordination in modes of hierarchical governance, self‐governance, and co‐governance. Our analysis shows that algorithms increase efficiency while decreasing the space for governing actors' discretion. Furthermore, we compare the effects of algorithms in each of these cases and explore sources of convergence and divergence between the governance modes. We suggest design‐based governance modes that rely on algorithmic systems might be re‐conceptualized as algorithmic governance to account for the prevalence of algorithms and the significance of their effects.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2819386 2020-11-19T19:59:21-05:00 2020-11-19T19:59:21-05:00

<p>There has been some great work on framing AI ethics issues as ultimately about power. I want to elaborate on *why* power imbalances are a problem. 1/</p>

tag:tagteam.harvard.edu,2005:FeedItem/2817027 2020-11-17T11:20:13-05:00 2020-11-17T11:20:13-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2811725 2020-11-10T10:54:31-05:00 2020-11-10T10:54:31-05:00

<p>As university budgets are squeezed and student debt loads rise, an era of close-knit relationships between companies and universities is getting under way</p>

tag:tagteam.harvard.edu,2005:FeedItem/2811162 2020-11-09T17:29:46-05:00 2020-11-09T17:29:46-05:00

<p>Our challenge to American philanthropy is twofold. First, it must use its power, influence, and money to dismantle“techno-solutionism”—the idea that technological solutions are the key to strengthening American (or any) democracy. Technology is not and never will be the solution to social and political problems, but the fetishism of technology in the public sector has been extremely lucrative for many private companies. All too often we have seen how technology, adopted in the name of progress, actually exacerbates and entrenches society’s problems. Modern capitalism and the technologies arising from it are strangling the very ideals of democracy in the United States. Combined with a rise in nationalism and nativism, this version of techno-solutionism is reinforcing structural inequities, systemic racism, and economic injustice.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2806331 2020-11-02T20:17:23-05:00 2020-11-02T20:17:23-05:00

<p>The process is known as algorithmic decision making, or ADS. It is also used to guide judgements that can change lives forever—whether a person is likely to commit a crime, for example, or whether their children should be taken from them. “As tools of algorithmic decision making appear in every field of our lives, it’s going to be critical that communities understand what these tools are,” said Bryan Mercer, Executive Director of Movement Alliance Project (MAP). The Philadelphia nonprofit unites community organizations around initiatives where technology, race, and inequality intersect.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2799002 2020-10-23T14:53:17-04:00 2020-10-23T14:53:17-04:00

<p>Where Responsible AI meets Reality: PractitionerPerspectives on Enablers for shi�ing OrganizationalPractices</p>

tag:tagteam.harvard.edu,2005:FeedItem/2799001 2020-10-23T14:52:49-04:00 2020-10-23T14:52:49-04:00

<p>A unified framework of five principles for ethical AI</p>

tag:tagteam.harvard.edu,2005:FeedItem/2798927 2020-10-23T13:10:02-04:00 2020-10-23T13:10:02-04:00

<p>“AI has enormous positive potential for society,” the professor said in her opening remarks, noting the value of AI-driven solutions related to education, training, remote work, government services, medicine, and other areas. Here are several key takeaways from her testimony (or read her full report to Congress). Digitization Means Cheaper AI It is becoming cheaper to provide AI-based solutions in multiple domains, from supporting caregivers to helping rural residents find employment opportunities beyond their local communities. “Digitization of everything from consumer interactions to supply chains has enabled creation of data that optimizes how AI is implemented,” Athey said. “Cloud computing now allows companies to rent computing as they need it. Software-as-a-service lets companies subscribe to the best products for their needs on a case-by-case basis.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2798237 2020-10-22T17:52:17-04:00 2020-10-22T17:52:17-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2797456 2020-10-21T20:37:04-04:00 2020-10-21T20:37:04-04:00

<p>“If you compare today to five years ago, there is a much different awareness among policymakers and the public,” said Vera Franz, deputy director of the Open Society Foundations, an organization backed by Mr. Soros that has spent $24 million this year on groups focused on privacy, online discrimination and other tech topics. “The key question is how to translate that awareness to real change and real accountability.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2797282 2020-10-21T17:08:47-04:00 2020-10-21T17:08:47-04:00

<p>DukeMTMC dataset of videos was taken down in June 2019 due to ethics issues (no consent of ppl in video &amp; being used for surveillance tech), yet was still used in 135 research papers AFTER it was taken down, at confs including CVPR &amp; AAAI</p>

tag:tagteam.harvard.edu,2005:FeedItem/2797270 2020-10-21T16:46:18-04:00 2020-10-21T16:46:18-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2797236 2020-10-21T15:54:08-04:00 2020-10-21T15:54:08-04:00

<p>Yesterday @Europarl_EN adopted 3 reports on #AI. Our @djleufer , @edri ’s @sarahchander &amp; @ellajakubowska1 dive into why ethics &amp; soft-law instruments won’t protect us from AI-driven mass surveillance, predictive policing &amp; other harmful “innovations”:</p>

tag:tagteam.harvard.edu,2005:FeedItem/2796967 2020-10-21T10:50:13-04:00 2020-10-21T10:50:13-04:00

<p>The large-scale implementation of AI in governments requires corresponding efforts from the AI Ethics community to provide a method to audit and oversee AI at scale, across complex enterprises. In the past, the AI Ethics community has focused on high-level, and recently on bias mitigation. It is easy to assume that an ML model trained on a representative dataset tested for statistical bias and embedded fairness metrics will continue to perform ethically without oversight. In reality, the environment in which the model is operating is constantly changing, and auditors need to periodically reassess model performance and outcomes with thorough, multidisciplinary auditing teams to avoid unethical outcomes seeping into the models over time since ethics is not a ‘bolt-on’, but a continuous process. Ensuring that ML algorithms behave ethically requires regulation, measurement, and consistent auditing.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2792242 2020-10-14T13:30:14-04:00 2020-10-14T13:30:14-04:00

<p>Access Now resigns from PAI, stating, "there is an increasingly smaller role for civil society to play within PAI" and "we did not find that PAI influenced or changed the attitude of member companies." Good on @accessnow for calling this out + recognizing the power in resigning.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2789252 2020-10-09T14:59:31-04:00 2020-10-09T14:59:31-04:00

<p>We conclude the essay by discussing how institutions can be designed to bolster stability, the relationship between social mobility and institutions, and the interplay between culture and institutions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2789251 2020-10-09T14:57:57-04:00 2020-10-09T14:57:57-04:00

<p>As the use of artificial intelligence (AI) systems became significantly more prevalent in recent years, the concerns on how these systems collect, use and process big data also increased. To address these concerns and advocate for ethical and responsible development and implementation of AI, non-governmental organizations (NGOs), research centers, private companies, and governmental agencies published more than 100 AI ethics principles and guidelines. This first wave was followed by a series of suggested frameworks, tools, and checklists that attempt a technical fix to issues brought up in the high-level principles. Principles are important to create a common understanding for priorities and are the groundwork for future governance and opportunities for innovation. However, a review of these documents based on their country of origin and funding entities shows that private companies from US-West axis dominate the conversation. Several cases surfaced in the meantime which demonstrate biased algorithms and their impact on individuals and society. The field of AI ethics is urgently calling for tangible action to move from high-level abstractions and conceptual arguments towards applying ethics in practice and creating accountability mechanisms. However, lessons must be learned from the shortcomings of AI ethics principles to ensure the future investments, collaborations, standards, codes or legislation reflect the diversity of voices and incorporate the experiences of those who are already impacted by the biased algorithms.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2789092 2020-10-09T12:30:38-04:00 2020-10-09T12:30:38-04:00

<p>This analysis shows that algorithmic censorship is distinctive for two reasons: (1) in potentially bringing all communications carried out on social platforms within reach and (2) in potentially allowing those platforms to take a more active, interventionist approach to moderating those communications. Consequently, algorithmic censorship could allow social platforms to exercise an unprecedented degree of control over both public and private communications. Moreover, commercial priorities would be inserted further into the everyday communications of billions of people. Due to the dominance of the web by a few social platforms, this may be difficult or impractical to escape for many people, although opportunities for resistance do exist.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2782621 2020-10-01T13:11:18-04:00 2020-10-01T13:11:18-04:00

<p>This article sets out our perspective on how to begin the journey of decolonising computational fields, such as data and cognitive sciences. We see this struggle as requiring two basic steps: a) realisation that the present-day system has inherited, and still enacts, hostile, conservative, and oppressive behaviours and principles towards women of colour (WoC); and b) rejection of the idea that centering individual people is a solution to system-level problems. The longer we ignore these two steps, the more "our" academic system maintains its toxic structure, excludes, and harms Black women and other minoritised groups. This also keeps the door open to discredited pseudoscience, like eugenics and physiognomy. We propose that grappling with our fields' histories and heritage holds the key to avoiding mistakes of the past. For example, initiatives such as "diversity boards" can still be harmful because they superficially appear reformatory but nonetheless center whiteness and maintain the status quo. Building on the shoulders of many WoC's work, who have been paving the way, we hope to advance the dialogue required to build both a grass-roots and a top-down re-imagining of computational sciences -- including but not limited to psychology, neuroscience, cognitive science, computer science, data science, statistics, machine learning, and artificial intelligence. We aspire for these fields to progress away from their stagnant, sexist, and racist shared past into carving and maintaining an ecosystem where both a diverse demographics of researchers and scientific ideas that critically challenge the status quo are welcomed.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2780386 2020-09-28T17:34:20-04:00 2020-09-28T17:34:20-04:00

<p>Artificial intelligence (AI) is expected to play a major role in shaping global competitiveness and productivity over the next couple of decades, granting early adopters significant societal, economic, and strategic advantages. As the pace of AI innovation and development picks up—underpinned by advancements in big data and high-performance computing—the United States and China are both in the driver’s seat. Europe, meanwhile, despite having certain advantages such as a strong industrial base and leading AI research and talent, is punching far below its weight. This state of affairs is especially due to the fragmentation of the EU’s digital market, difficulties in attracting human capital and external investment, and the lack of commercial competitiveness.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2780385 2020-09-28T17:34:10-04:00 2020-09-28T17:34:10-04:00

<p>I believe advances in AI are likely to be among the most impactful global developments in the coming decades, and that AI governance will become among the most important global issue areas. AI governance is a new field and is relatively neglected. I’ll explain here how I think about this as a cause area and my perspective on how best to pursue positive impact in this space. The value of investing in this field can be appreciated whether one is primarily concerned with contemporary policy challenges or long-term risks and opportunities (“longtermism”); this piece is primarily aimed at a longtermist perspective. Differing from some other longtermist work on AI, I emphasize the importance of also preparing for more conventional scenarios of AI development.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2778534 2020-09-25T11:13:51-04:00 2020-09-25T11:13:51-04:00

<p>nline social networks provide a platform for sharing information and free expression. However, thesenetworks are also used for malicious purposes, such as distributing misinformation and hate speech, sellingillegal drugs, and coordinating sex trafficking or child exploitation. This paper surveys the state of the artin keeping online platforms and their users safe from such harm, also known as the problem of preservingintegrity.This survey comes from the perspective of having to combat a broad spectrum of integrity violations atFacebook. We highlight the techniques that have been proven useful in practice and that deserve additionalattention from the academic community. Instead of discussing the many individual violation types, we identifykey aspects of the social-media eco-system, each of which is common to a wide variety violation types.Furthermore, each of these components represents an area for research and development, and the innovationsthat are found can be applied widely</p>

tag:tagteam.harvard.edu,2005:FeedItem/2776781 2020-09-23T12:17:56-04:00 2020-09-23T12:17:56-04:00

<p>When I read these articles, I think about social construction, bounded rationality, bureaucratic discretion in an era of #AI #ML in public sector decisionmaking.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2776774 2020-09-23T12:09:18-04:00 2020-09-23T12:09:18-04:00

<p>I am very excited to finally announce something I've worked on during the past few months. The European AI Fund is a new philanthropic initiative to shape the direction of AI in Europe by strengthening the voice of civil society. Please share widely!! And do apply.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2769842 2020-09-14T10:50:09-04:00 2020-09-14T10:50:09-04:00

<p>The German anti-discrimination agency published<a href="https://urldefense.proofpoint.com/v2/url?u=http-3A__x3ysn.mjt.lu_lnk_AMEAAH-5Fm4V8AAAAE27oAAATCtrYAAAAAKxoAAB1gABB0KQBfWyEiMaZZ9KB2TK-2DGvIRO3Bh4hQAQJ0I_15_uLb2N1XIwFct8ZUos4zwhQ_aHR0cHM6Ly93d3cuYW50aWRpc2tyaW1pbmllcnVuZ3NzdGVsbGUuZGUvU2hhcmVkRG9jcy9Eb3dubG9hZHMvRU4vcHVibGlrYXRpb25lbi9TdHVkaWVfZW5fRGlza3JpbWluaWVydW5nc3Jpc2lrZW5fZHVyY2hfVmVyd2VuZHVuZ192b25fQWxnb3JpdGhtZW4uaHRtbA&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=9UuICCKr8iot_lFkr20FKAXGYOtiPqUTwAf0nK6g_MQ&amp;s=msX_x84bQ_lOSF4VWfrU1kTHnpIJNrmub_rtJTIKcno&amp;e=" title="https://urldefense.proofpoint.com/v2/url?u=http-3A__x3ysn.mjt.lu_lnk_AMEAAH-5Fm4V8AAAAE27oAAATCtrYAAAAAKxoAAB1gABB0KQBfWyEiMaZZ9KB2TK-2DGvIRO3Bh4hQAQJ0I_15_uLb2N1XIwFct8ZUos4zwhQ_aHR0cHM6Ly93d3cuYW50aWRpc2tyaW1pbmllcnVuZ3NzdGVsbGUuZGUvU2hhcmVkRG9jcy9Eb3dubG9hZHMvRU4vcHVibGlrYXRpb25lbi9TdHVkaWVfZW5fRGlza3JpbWluaWVydW5nc3Jpc2lrZW5fZHVyY2hfVmVyd2VuZHVuZ192b25fQWxnb3JpdGhtZW4uaHRtbA&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=9UuICCKr8iot_lFkr20FKAXGYOtiPqUTwAf0nK6g_MQ&amp;s=msX_x84bQ_lOSF4VWfrU1kTHnpIJNrmub_rtJTIKcno&amp;e="><span> </span>a translated version of its excellent report</a><span> </span>on<span> </span><strong>algorithmic discrimination</strong>, originally released in German last October. They reviewed 47 cases studies (only 6 of which, unfortunately, focus on Europe) and plenty of academic literature on the topic. They offer a precise yet comprehensive overview of why algorithmic discrimination matters, what it looks like and how it can be observed.</p> <p>They conclude by advocating strongly for more research and investigation powers for anti-discrimination authorities. In particular, they suggest that operators of online services should document how they ensure that they respect anti-discrimination legislation. They also argue that the current legislation, which puts the burden of identifying discrimination on affected individuals, cannot work for online platforms, where users have no way to compare their experience with that of others.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2769772 2020-09-14T09:21:35-04:00 2020-09-14T09:21:35-04:00

<p>Too many councils and advisory boards still consist mostly of people based in Europe or the United States.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2763770 2020-09-04T10:27:12-04:00 2020-09-04T10:27:12-04:00

<p>A useful start may be shifting the emphasis from public space—often framed exclusively as an achievement of design—to publicness—as a sociopolitical practice. How do people “do” public space online, and under what conditions? Who dictates these conditions, and to what ends? What kinds of digital spaces feel “public” to their occupants? What are the thresholds of participation? Who is in, who is out, and who is somewhere in between? Katherine McKittrick and Clyde Woods powerfully remind us of the inextricable link between place, segregation, and oppressive inequality, particularly when it comes to race.6 Ruha Benjamin notes that technology’s (re-)production of racism is just one side of the coin, and that race itself is a technology in the sense that it, too, is productive.7 Our reading of the “practice of publicness in digital space,” therefore, is incomplete—if not itself complicit with injustice—without centering the productive forces of inequality and oppression, and racial oppression specifically. These forces work on a macro-level, but the pandemic has shown how localized they are.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2763174 2020-09-03T13:21:15-04:00 2020-09-03T13:21:15-04:00

<p>Catalogue existing key initiatives undertaken by various stakeholders to promote the responsible research and development of beneficial AI systems and applications, including: projects and frameworks to operationalize AI ethical principles and the application of AI for social good; mechanisms and processes to identify and mitigate bias, discrimination, and inequities in AI systems; tools, certifications, assessments, and audit mechanisms to evaluate AI systems for responsibility and trustworthiness, based on metrics such as safety, robustness, accountability, transparency, fairness, respect for human rights, and the promotion of equity.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2762564 2020-09-02T11:34:38-04:00 2020-09-02T11:34:38-04:00

<p>It’s not due to a lack of imagination that these scholarly communities have continuously circled the drain on questions such as the presence of racial bias in particular systems—this is a political arrangement. It’s a structural condition of how the grants that fund their work are allocated, and the relationships they have to industry and to government institutions. For decades, research questions have been staged to these scholarly communities in very particular ways by carceral institutions. There is a given-ness to the problems that these researchers are failing to interrogate. For instance, it's no accident that for years everyone was like, “We need explainable AI,” and then DARPA started handing out millions of dollars worth of grants to develop explainable AI. Historically, certain academic disciplines have had moments when they decided to reexamine their relationship with the military and police industrial complex. Consider anthropologists refusing to participate in the US military’s human terrain systems in Iraq and Afghanistan, for instance. But the ethics-in-technology communities haven’t had that kind of reckoning yet, where they start to deeply interrogate why they're asking the questions that they're asking. Because these technologies are moving so quickly, I think people in these research communities haven't had a chance to reflect on why they keep asking the questions that they're asking. Where do the questions come from? And why is it that they’re asking the exact same questions that DARPA is asking? And why isn't that entanglement ethically complicated for them?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2761927 2020-09-01T16:03:02-04:00 2020-09-01T16:03:02-04:00

<p>We need a fundamentally interdisciplinary approach, with the knowledge and insights of historically disadvantaged communities and social-science expertise embedded into technology development. Critics might argue that this approach is too time-consuming, especially during a crisis. But we need solutions that actually work.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2761757 2020-09-01T12:11:13-04:00 2020-09-01T12:11:13-04:00

<p>Autonomy has been central to moral and political philosophy for millennia, and has been positioned as a critical aspect of both justice and wellbeing. Research in psychology supports this position, providing empirical evidence that autonomy is critical to motivation, personal growth and psychological wellness. Responsible AI will require an understanding of, and ability to effectively design for, human autonomy (rather than just machine autonomy) if it is to genuinely benefit humanity. Yet the effects on human autonomy of digital experiences are neither straightforward nor consistent, and are complicated by commercial interests and tensions around compulsive overuse. This multi-layered reality requires an analysis that is itself multidimensional and that takes into account human experience at various levels of resolution. We borrow from HCI and psychological research to apply a model (“METUX”) that identifies six distinct spheres of technology experience. We demonstrate the value of the model for understanding human autonomy in a technology ethics context at multiple levels by applying it to the real-world case study of an AI-enhanced video recommender system. In the process we argue for the following three claims: (1) There are autonomy-related consequences to algorithms representing the interests of third parties, and they are not impartial and rational extensions of the self, as is often perceived; (2) Designing for autonomy is an ethical imperative critical to the future design of responsible AI; and (3) Autonomy-support must be analysed from at least six spheres of experience in order to appropriately capture contradictory and downstream effects.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2761159 2020-08-31T17:06:51-04:00 2020-08-31T17:06:51-04:00

<p>That’s why today we are announcing a new research partnership to better understand the impact of Facebook and Instagram on key political attitudes and behaviors during the US 2020 elections, building on the initiative we launched in 2018. It will examine the impact of how people interact with our products, including content shared in News Feed and across Instagram, and the role of features like content ranking systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2761158 2020-08-31T17:03:54-04:00 2020-08-31T17:03:54-04:00

<p>The second problem was that it is hard to conduct a rigorous scientific study on social media’s impact after the fact. With the US 2020 election upon us, it is clear we need a new approach. For this election, we are integrating Facebook researchers’ intimate knowledge and access to internal Facebook data with the expertise of independent scholars to conduct scientifically rigorous research that allows these academics to ensure that the research asks the right questions and that the questions are answered with the right methods.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2761105 2020-08-31T14:06:49-04:00 2020-08-31T14:06:49-04:00

<p>Our work shows how human face matching is affected by prior identity decision labels and we discuss how this may limit the total accuracy of human-algorithm teams performing face matching tasks.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2759725 2020-08-28T14:23:24-04:00 2020-08-28T14:23:24-04:00

<p>Google’s new offerings will test whether a lucrative but increasingly distrusted industry can boost its business by offering ethical pointers. The company is a distant third in the cloud computing market behind Amazon and Microsoft, and positions its AI expertise as a competitive advantage. If successful, the new initiative could spawn a new buzzword: EaaS, for ethics as a service, modeled after cloud industry coinages such as SaaS, for software as a service.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2759148 2020-08-27T15:37:54-04:00 2020-08-27T15:37:54-04:00

<p>We live in a world where technological corporations hold unprecedented power and influence. Technological solutions to social, political, and economic challenges are rampant. In the Global South, technology that is developed with Western perspectives, values, and interests is imported with little regulation or critical scrutiny. This work examines how Western tech monopolies, with their desire to dominate, control and influence social, political, and cultural discourse, share common characteristics with traditional colonialism. However, while traditional colonialism is driven by political and government forces, algorithmic colonialism is driven by corporate agendas. While the former used brute force domination, colonialism in the age of AI takes the form of ‘state-of-the-art algorithms’ and ‘AI driven solutions’ to social problems. Not only is Western-developed AI unfit for African problems, the West’s algorithmic invasion simultaneously impoverishes development of local products while also leaving the continent dependent on Western software and infrastructure. By drawing examples from various parts of the continent, this paper illustrates how the AI invasion of Africa echoes colonial era exploitation. This paper then concludes by outlining a vision of AI rooted in local community needs and interests.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2759142 2020-08-27T14:55:23-04:00 2020-11-12T10:05:08-05:00

<p>In other words, there’s too much toxic content for Facebook to ever really see it all (hello, scale) so Facebook has a formula for determining the amount based on random sampling. Abstractly, perhaps everyone understands this, but to think about it another way: There’s so much sewage on the platform that the company must continually guess how much of it people are actually seeing.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2757820 2020-08-25T17:15:04-04:00 2020-08-25T17:15:04-04:00

<p>These matters have come to light because Epstein was a criminal, but Harvard is not alone in accepting tainted money. Universities need to develop policies to ensure that research funding is based on merit, not cronyism, and researchers who are seeking public trust must be able to show that their own ethical compasses are not deflected by the magnetism of money.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2757814 2020-08-25T16:18:23-04:00 2020-08-25T16:18:23-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2757497 2020-08-25T11:10:05-04:00 2020-08-25T11:10:05-04:00

<p>This is a much-needed intervention in the field of machine learning, which can be excessively hierarchical and homogenous. But it is no silver bullet: in fact, “participation-washing” could become the field's next dangerous fad. That’s what I, along with my coauthors Emanuel Moss, Olaitan Awomolo, and Laura Forlano, argue in our recent paper “Participation is not a design fix for machine learning.” Ignoring patterns of systemic oppression and privilege leads to unaccountable machine-learning systems that are deeply opaque and unfair. These patterns have permeated the field for the last 30 years. Meanwhile, the world has watched the exponential growth of wealth inequality and fossil-fuel-driven climate change. These problems are rooted in a key dynamic of capitalism: extraction. Participation, too, is often based on the same extractive logic, especially when it comes to machine learning.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2756921 2020-08-24T10:11:59-04:00 2020-08-24T10:11:59-04:00

<p>The new Program on Platform Regulation, led by Daphne Keller, offers lawmakers, academics, and civil society groups ground-breaking analysis and research to support wise governance of Internet platforms. As a world-renowned expert on both platform regulation and the operational realities of tech companies, Keller is uniquely positioned to assess legislative proposals for their impact on Internet users’ rights under the U.S. Constitution and international human rights laws, as well as laws’ likely consequences for companies and competition. This work is critical now, as lawmakers in Washington, D.C., Brussels, and around the world rush to enact laws – some well-considered, some less so – that will shape the digital future.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2755554 2020-08-21T08:31:58-04:00 2020-08-21T08:31:58-04:00

<p>Despite all of these changes, the importance of a strong relationship between the United States and the European Union has been a constant. The transatlantic disagreements that have characterized the past few years—and have hampered a united front on emerging technologies like 5G and AI5—are not the first time US-EU relations have suffered, but they should not further divide allies that share common values.6 Deepened US-EU cooperation across the entire AI ecosystem7 is necessary to advance a more secure, safe, and prosperous world, but to do this the current level of AI-related coordination and partnership needs to be increased. This report’s purpose is twofold: first, to inform policymakers and researchers about the current state of transatlantic AI efforts; and second, to recommend specific areas where transatlantic AI collaboration should be strengthened. Based on a comprehensive study of over 260 documents and reports covering the period from December 1997 to June 2020, we proposes more than 16 recommendations to increase US-EU AI collaboration across the entire AI ecosystem, as well as 9 recommendations for AI cooperation in the healthcare, environmental sciences, and defense sectors. Greater transatlantic efforts are needed to prevent the advancement of an AI vision that is adversarial and harmful to the wellbeing of the United States, the European Union, and allies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2754974 2020-08-20T10:16:47-04:00 2020-08-20T10:16:47-04:00

<p>To be trained in virtue is to be given the opportunity to develop the right habits and frames of mind. This often comes though having good routines and being faced with challenges that exercise one’s moral faculties. Some ideas for how workplaces can cultivate this include: Regular workplace exercises that have individuals and teams solve complex moral problems from day to day life would allow for critical inquiry of values and enable character development. Michael Sandel’s Harvard Justice course provides a lot of insight on how to organise such problems and engage people with sound ethical reasoning. Encouraging time for reflection and philosophical reading would allow workers to develop their moral faculties. Scheduled unstructured time is critical for children’s moral development, but it is often lacking in schools and homes. There is no reason to believe that providing it for adults would fail to bring about the same results in allowing for improved communication, empathy, and social concern. Interdisciplinary work environments, which not only have diverse teams, but have individuals go out of their comfort zone to work on tasks on which they are less skilled, would aid in character building. When the Apollo missions had infighting about what the ideal approach to landing on the moon would be, each competing team was asked to complete a research document arguing for the option they disagreed with. This approach enabled greater understanding of others and a concern for full knowledge of the circumstance. Adapting these lessons to more workplaces can help take individuals outside of themselves.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2754508 2020-08-19T14:22:13-04:00 2020-08-19T14:22:13-04:00

<p>Schaake’s speech was a hairpin turn in rhetorical space. Where Schmidt emphasized the future, Schaake stressed the present; where Schmidt dwelled on the private and the technological, Schaake foregrounded the public and the political. “I’m going to talk about governance,” she began, before beginning a forceful critique of Big Tech’s extreme aversion to regulation, its unchecked collection of personal data, and its steady erosion of democracy. She brought up the ethics statements that committees and conferences like the one she was addressing tend to generate. “It’s a very popular topic,” she said, referring to A.I. ethics. “It’s also hard to be against ethics. . . . How do we make sure it’s meaningful and enforceable, and not just window dressing?”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2754296 2020-08-19T09:39:41-04:00 2020-08-19T09:39:41-04:00

<p>It’s with this ‘act + collaborate’ bias in mind that we are embarking on a request for comments on v0.9 of the Mozilla Trustworthy AI Whitepaper. The paper talks about how industry, regulators and citizens of the internet can work together to build more agency and accountability into our digital world. It also talks briefly about some of the areas where Mozilla will focus, knowing that Mozilla is only one small actor in the bigger picture of shifting the AI tide.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2753071 2020-08-17T14:37:12-04:00 2020-08-17T14:37:12-04:00

<p>From each of the examples I will be looking at - including the UK’s Warnock Committee and Human Embryology Act 1990, the Internet Corporation for assigned Names and Numbers (ICANN), the UN Space Treaty and the European ban on genetically modified crops - it is possible to glean excitingly prescient information to take with us in the quest to ensure AI remains safe, beneficial and under meaningful human control.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2753067 2020-08-17T14:21:14-04:00 2020-08-17T14:21:14-04:00

<p>Imagine you are a patient who has been diagnosed with prostate cancer. The two main approaches to treating it in the United States are active surveillance versus the surgical option of radical prostatectomy. Your physician recommends the surgical option, and spends considerable time explaining the steps in the surgery, the benefits of (among other things) eliminating the tumor and the risks of (among other things) erectile dysfunction and urinary incontinence after the surgery. What your physician does not tell you is that she has arrived at her recommendation of prostatectomy over active surveillance based on the analysis of an Artificial Intelligence (AI)/Machine Learning (ML) system, which recommended this treatment plan based on analysis of your age, tumor size, and other personal characteristics found in your electronic health record. Has the doctor secured informed consent from a legal perspective? From an ethical perspective? If the doctor actually chose to “overrule” the AI system, and the doctor fails to tell you that, has she violated your legal or ethical right to informed consent? If you were to find out that the AI/ML system was used to make recommendations on your care and no one told you, how would you feel? Well, come to think of it, do you know whether an AI/ML system was used the last time you saw a physician?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2753069 2020-08-17T14:21:43-04:00 2020-08-17T14:21:43-04:00

<p>AI Watch Artificial Intelligencein public servicesOverview of the use and impact of AI in public servicesin the EU</p>

tag:tagteam.harvard.edu,2005:FeedItem/2748992 2020-08-10T15:03:17-04:00 2020-08-10T15:03:17-04:00

<p>Yet the IoB also raises new challenges for data governance that concern not only individual privacy and autonomy but also new risks of discrimination and bias in employment, education, finance, access to health insurance and other important areas for the distribution of social resources.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2748840 2020-08-10T09:54:33-04:00 2020-08-10T09:54:33-04:00

<p>rain AI algorithms that computers use to solve visual recognition problems and recognize people, places, and things. Smartphone photo apps, for instance, use similar algorithms to automatically identify photos of skylines or beaches in your collection. The dataset was just one of many used in AI development. While the pre-print’s shocking findings about Tiny Images</p>

tag:tagteam.harvard.edu,2005:FeedItem/2748842 2020-08-10T09:54:50-04:00 2020-08-10T09:54:50-04:00

<p>The overwhelming 'Whiteness' of artificial intelligence—from stock images and cinematic robots to the dialects of virtual assistants—removes people of colour from the way humanity thinks about its technology-enhanced future.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2746078 2020-08-05T11:24:54-04:00 2020-08-05T11:24:54-04:00

<p>These Principles supplement the Principles of Professional Ethics for the IC and do not modify or supersede applicable laws, executive orders, or policies. Instead, they articulate the general norms that IC elements should follow in applying those authorities and requirements.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2746074 2020-08-05T10:47:47-04:00 2020-08-05T10:47:47-04:00

<p>To advance this conversation, the R Street Institute and the Center for Democracy and Technology launched the Governance Research in Artificial Intelligence Leadership (GRAIL) Network to connects top-tier academic researchers working on artificial intelligence issues with the policy conversation happening in Washington D.C.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2746007 2020-08-05T10:09:22-04:00 2020-08-05T10:09:22-04:00

<p>Algorithms are automatic decision-making processes used by computer programmes – to identify patterns in data. As in most countries, algorithms play an increasingly essential role in supporting the services the New Zealand government provides. Algorithms can help deliver new, innovative, and well-targeted policies that make New Zealand a better place to live.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2745453 2020-08-04T14:12:10-04:00 2020-08-04T14:12:10-04:00

<p>Newman embarks on the lonely and brave journey of investigating how to put AI governmental principles into action. To do this, 3 case studies are considered, ranging from ethics committees, publication norms and intergovernmental agreement. While all 3 of those aspects have their benefits, none of them are perfect, and Newman eloquently explains why. The challenges presented are numerous, but the way forward is visible, and that way is called practicality.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2744011 2020-08-02T14:39:12-04:00 2020-08-02T14:39:12-04:00

<p>Xi’s pronouncements on AI have a sinister edge. Artificial intelligence has applications in nearly every human domain, from the instant translation of spoken language to early viral-outbreak detection. But Xi also wants to use AI’s awesome analytical powers to push China to the cutting edge of surveillance. He wants to build an all-seeing digital system of social control, patrolled by precog algorithms that identify potential dissenters in real time.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2741932 2020-07-30T11:37:31-04:00 2020-07-30T11:37:31-04:00

<p>The Media Lab professes to be working in an anti-disciplinary way; it aspires to mix the autonomous experimentation of the arts with a paradigmatic mandate to launch not just new technologies, but new fields of technological research. Judging from the informal surveys that I take when I speak there, the students who actually get in have very little or no training in the arts or humanities. I’ve learned to recognize the mostly straightforward and narrow claim for doing “art” that’s common at the Lab and among tech folks generally: They mean to indicate the non-useful, beautiful, or elegant results from their technical work, and they want to encourage some wacky or whimsical side projects on the path to technical feasibility.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2741428 2020-07-29T17:21:55-04:00 2020-07-29T17:21:55-04:00

<p>We need new paradigms, not more new tech. We need fair and equitable implementations of public policy that bolster our collective good. We need to center the most vulnerable among us — the working poor and the disabled, those who live under racial and religious tyranny, the discriminated against and the oppressed. We need to house people and provide health, employment, creative arts and educational resources. We need to close the intersectional racial wealth gap.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2741171 2020-07-29T11:43:34-04:00 2020-07-29T11:43:34-04:00

<p>A parameter representing the cognitive cost of exploration controls the trade-off between quality and popularity. Below and above a critical exploration cost, popularity bias is more likely to hinder quality. But we find a narrow intermediate regime of user attention where an optimal balance exists: choosing what is popular can help promote high-quality items to the top. These findings clarify the effects of algorithmic popularity bias on quality outcomes, and may inform the design of more principled mechanisms for techno-social cultural markets.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2741052 2020-07-29T09:59:57-04:00 2020-07-29T09:59:57-04:00

<p>What are the human actions and predispositions that initially spark our curiosity? What is the kind of self-knowledge that we would cherish as a foundation for enriching our sociality, our civil and public institutions, and our democratic process? Readily resorting to platform data analytics for such knowledge risks taking platform environments as our entire world. Instead, when dealing with platform data we should aspire to “put the platforms in perspective,” foregrounding rather than obscuring their interventions in how we behave.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2740612 2020-07-28T17:46:55-04:00 2020-07-28T17:46:55-04:00

<p>Bridging AI’s trust gapsAligning policymakers and companies</p>

tag:tagteam.harvard.edu,2005:FeedItem/2740215 2020-07-28T09:28:42-04:00 2020-07-28T09:28:42-04:00

<p>The Oxford Commission on AI &amp; Good Governance mission Investigate the Artificial Intelligence implementation challenges faced by governments around the world Identify best practices for evaluating and managing risks and benefits Recommend strategies for taking full advantage of technical capacities while mitigating potential harms of AI-enabled public policy</p>

tag:tagteam.harvard.edu,2005:FeedItem/2739563 2020-07-27T13:33:27-04:00 2020-07-27T13:33:27-04:00

<p>The Arts and Humanities Research Council and the Institute for Government have been working in partnership for six years on the Engaging with Government programme – a three-day course for researchers in the arts and humanities. This programme helps academics develop the knowledge and skills they need to engage effectively with government and parliamentary bodies at all levels, along with the other organisations involved in the policy-making process. We, in turn, have learned a huge amount from our participants, who now form an active alumni network brimming with expertise about how to engage with policy in practice. This guide brings together some of that learning.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2737050 2020-07-23T13:18:33-04:00 2020-07-23T13:18:33-04:00

<p>The promises and limitations of government oversight over algorithmic systems</p>

tag:tagteam.harvard.edu,2005:FeedItem/2736954 2020-07-23T11:55:01-04:00 2020-07-23T11:55:01-04:00

<p>The industry AI paradigm assumes that it's possible for companies to build models using personal data scooped up from the web—without asking who shared it and why—and somehow make it ethical using technical tricks. If we reject this premise the trillion dollar castle falls apart.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2736328 2020-07-22T15:56:40-04:00 2020-07-22T15:56:40-04:00

<p>The school, named the US Digital Service Academy, looks to rival Stanford and MIT, two established tech talent pools.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2736283 2020-07-22T13:57:45-04:00 2020-07-22T13:57:45-04:00

<p>Ethics has powerful teeth, but these are barely being used in the ethics of AI today – it is no wonder the ethics of AI is then blamed for having no teeth. This article argues that ‘ethics’ in the current AI ethics field is largely ineffective, trapped in an ‘ethical principles’ approach and as such particularly prone to manipulation, especially by industry actors. Using ethics as a substitute for law risks its abuse and misuse. This significantly limits what ethics can achieve and is a great loss to the AI field and its impacts on individuals and society. This article discusses these risks and then highlights the teeth of ethics and the essential value they can – and should – bring to AI ethics now.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2736192 2020-07-22T12:24:20-04:00 2020-07-22T12:24:20-04:00

<p>Drawing on in-depth studies of major uses of AI, the authors detail how the technology actually works. They outline a policy and governance blueprint for gaining the benefits of artificial intelligence while minimizing its potential downsides.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2735229 2020-07-21T09:41:28-04:00 2020-07-21T09:44:15-04:00

<p>Achieving the global benefits of artificial intelligence (AI) will require international cooperation on many areas of governance and ethical standards, while allowing for diverse cultural perspectives and priorities. There are many barriers to achieving this at present, including mistrust between cultures, and more practical challenges of coordinating across different locations. This paper focuses particularly on barriers to cooperation between Europe and North America on the one hand and East Asia on the other, as regions which currently have an outsized impact on the development of AI ethics and governance. We suggest that there is reason to be optimistic about achieving greater cross-cultural cooperation on AI ethics and governance. We argue that misunderstandings between cultures and regions play a more important role in undermining cross-cultural trust, relative to fundamental disagreements, than is often supposed. Even where fundamental differences exist, these may not necessarily prevent productive cross-cultural cooperation, for two reasons: (1) cooperation does not require achieving agreement on principles and standards for all areas of AI; and (2) it is sometimes possible to reach agreement on practical issues despite disagreement on more abstract values or principles. We believe that academia has a key role to play in promoting cross-cultural cooperation on AI ethics and governance, by building greater mutual understanding, and clarifying where different forms of agreement will be both necessary and possible. We make a number of recommendations for practical steps and initiatives, including translation and multilingual publication of key documents, researcher exchange programmes, and development of research agendas on cross-cultural topics.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2735230 2020-07-21T09:42:25-04:00 2020-07-21T09:42:25-04:00

<p>The world has changed a lot since January, and in many ways our Principles have become even more important to the work of our researchers and product teams. As we develop AI we are committed to testing safety, measuring social benefits, and building strong privacy protections into products. Our Principles give us a clear framework for the kinds of AI applications we will not design or deploy, like those that violate human rights or enable surveillance that violates international norms. For example, we were the first major company to have decided, several years ago, not to make general-purpose facial recognition commercially available.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2734606 2020-07-20T14:08:19-04:00 2020-07-20T14:08:19-04:00

<p>THE ASSESSMENT LIST FOR TRUSTWORTHY ARTIFICIAL INTELLIGENCE (ALTAI)</p>

tag:tagteam.harvard.edu,2005:FeedItem/2734593 2020-07-20T13:53:12-04:00 2020-07-20T13:53:12-04:00

<p>Data is needed not just to develop new technology but also to enable the evaluation of it. The UK will be unable to embrace the opportunities presented by AI unless data held by the government and wider public sector is shared. The sharing of personal data must be conducted in a way that is trustworthy, aligned with society’s values and people’s expectations. Public consent is crucial to the long term sustainability of data sharing activity.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2734561 2020-07-20T12:35:05-04:00 2020-07-20T12:35:05-04:00

<p>The idea of artificial intelligence for social good (henceforth AI4SG) is gaining traction within information societies in general and the AI community in particular. It has the potential to tackle social problems through the development of AI-based solutions. Yet, to date, there is only limited understanding of what makes AI socially good in theory, what counts as AI4SG in practice, and how to reproduce its initial successes in terms of policies. This article addresses this gap by identifying seven ethical factors that are essential for future AI4SG initiatives. The analysis is supported by 27 case examples of AI4SG projects. Some of these factors are almost entirely novel to AI, while the significance of other factors is heightened by the use of AI. From each of these factors, corresponding best practices are formulated which, subject to context and balance, may serve as preliminary guidelines to ensure that well-designed AI is more likely to serve the social good.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2732911 2020-07-17T13:57:24-04:00 2020-07-17T13:57:24-04:00

<p>Combating Anti-Blackness in the AI Community</p>

tag:tagteam.harvard.edu,2005:FeedItem/2732902 2020-07-17T13:38:32-04:00 2020-07-17T13:38:32-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2732901 2020-07-17T13:32:23-04:00 2020-07-17T13:32:23-04:00

<p>Ethical and societal implications of algorithms, data, and artificial intelligence: a roadmap for research</p>

tag:tagteam.harvard.edu,2005:FeedItem/2732900 2020-07-17T13:31:55-04:00 2020-07-17T13:31:55-04:00

<p>Turning shared interests to new ways of Governing in the age of Emergence</p>

tag:tagteam.harvard.edu,2005:FeedItem/2732348 2020-07-16T15:46:06-04:00 2020-07-16T15:46:06-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2730972 2020-07-14T14:05:29-04:00 2020-07-14T14:05:29-04:00

<p>An innovation and accountability-led government must involve and empower the public in co-creating policies, above and beyond the existing rules that engage individuals (consent forms etc.). In the paper I propose a Public Interest Data Bill that addresses this need: the rules of the digital highway used as a negotiation between the public and regulators, between private data consumers and data generators. Specifically: clear transparency, public participation and realistic sanctions when things go wrong.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2730783 2020-07-14T10:44:46-04:00 2020-07-14T10:44:46-04:00

<p>As part of our commitment to democratize #AI, @GoogleOrg is making a $2.5M grant &amp; joining @RockefellerFdn , @IDRC_CRDI &amp; @fair_forward to co-found the Lacuna Fund, a groundbreaking consortium to deliver equity-first open datasets:</p>

tag:tagteam.harvard.edu,2005:FeedItem/2730773 2020-07-14T10:07:28-04:00 2020-07-14T10:07:28-04:00

<p>ion Retrieval (IR): These are complex systems, such as the ones that power web search engines. It is possible that IR amplifies existing bias against PWD. For example, search results can return stereotypical content for PWD, while targeted-advertising may eventually exclude PWD from products or even employment opportunities.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2730770 2020-07-14T10:07:09-04:00 2020-07-14T10:07:09-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2730769 2020-07-14T10:06:53-04:00 2020-07-14T10:06:53-04:00

<p>This paper seeks to investigate how we might foster an inclusive, foresight-informed responsible AI governance framework. The paper discusses the gaps and opportunities in current AI initiatives across various stakeholders and acknowledges the importance of anticipation and agility. This paper also posits that it is important for legal, policy, industry and academia to understand the specificities of each other’s domains better to build an inclusive governance framework.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2730396 2020-07-13T17:20:13-04:00 2020-07-13T17:20:13-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2730186 2020-07-13T13:01:14-04:00 2020-07-13T13:01:14-04:00

<p>On the google policy team we did this as well, we approached academics who we knew what their findings would be because they were 'like-minded', we had a whole team set up to 'cultivate relationships with academics.'</p>

tag:tagteam.harvard.edu,2005:FeedItem/2730163 2020-07-13T12:17:12-04:00 2020-07-13T12:17:12-04:00

<p>Many scholars have an interest in maintaining a relationship with companies because it is difficult to answer key policy questions without access to their data. This can lead researchers to adopt more favorable assumptions when there is legitimate debate about how to handle a methodological question.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2730111 2020-07-13T09:44:48-04:00 2020-07-13T09:44:48-04:00

<p>The DeepMind paper interrogates how colonial features are found in algorithmic decision-making systems and what the authors call “sites of coloniality,” or practices that can perpetuate colonial AI. These include beta testing on disadvantaged communities — like Cambridge Analytica conducting tests in Kenya and Nigeria or Palantir using predictive policing to target Black residents of New Orleans. There’s also “ghost work,” the practice of relying on low-wage workers for data labeling and AI system development. Some argue ghost work can lead to the creation of a new global underclass.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2728566 2020-07-10T11:14:26-04:00 2020-07-10T13:14:58-04:00

<p>From Rationality to Relationality: Ubuntu as an Ethical and Human Rights Framework for Artificial Intelligence Governance</p>

tag:tagteam.harvard.edu,2005:FeedItem/2728361 2020-07-10T09:33:03-04:00 2020-07-10T09:33:03-04:00

<p>Policy evaluation is a critical element of good governance, as it promotes public accountability and contributes to citizens’ trust in government. Evaluation helps ensure that decisions are rooted in trustworthy evidence and deliver desired outcomes. Drawing on the first significant cross-country survey of policy evaluation practices covering 42 countries, this report offers a systemic analysis of the institutionalisation, quality and use of evaluation across countries and looks at how these three dimensions interrelate. The report also covers cross-cutting aspects related to regulatory assessment and performance budgeting. The analysis illustrates the role and functions of key institutions within the executive, such as centres of government and ministries of finance. It also underlines the role of supreme audit institutions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2727960 2020-07-09T17:49:43-04:00 2020-07-09T17:49:43-04:00

<p>What does that mean during the pandemic? AI-reliant moderation being “turned on” globally means that the further you get away from Silicon Valley, there’s a higher likelihood the quality of the content moderation would have gotten worse. While data is scarce on these points, it’s hard to escape the question of how much of the world is currently effectively unmoderated on Facebook, Twitter, and other platforms. That means the potential of unchecked hate speech, overmoderation of certain communities and dissenting voices, and increased vulnerability to disinformation campaigns. Data has not been forthcoming on this from the social media companies, but we know that the Secretary General of the U.N. has cited an unchecked “tsunami of hate” globally, in particular online. The silencing of free expression by those targeted should alarm us all. As with Issue 1, it’s not necessarily that we simply want a return to the pre-pandemic way of doing things—those were bad, too. But some transparency, access to relevant data for researchers and the public, and a good honest conversation could do a world of good.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2727942 2020-07-09T15:49:20-04:00 2020-07-09T15:49:20-04:00

<p>AI is more than a technology. It reflects the values in its system, suggesting that any ethical lapses simply mirror our own deficiencies. And yet, there’s hope: AI can also inspire us, augment us, and make us go deeper. AI’s goals need to be society’s goals. As opposed to the market-driven, profit-making ones that dominate its use today, applying AI responsibly is to use it to support systems that have human goals. We need a new rule-making system to guide its responsible development. Self-regulation simply isn’t enough. Cross-sector oversight must start with transparency and access to meaningful information, as well as an ability to expose harm.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2727775 2020-07-09T14:31:35-04:00 2020-07-09T14:31:35-04:00

<p>Excited to share a new paper on Decolonisation and AI. With the amazing @png_marie @wsisaac Star-struck we explore why Decolonial theory matters for our field, and tactics to decolonise and reshape our field into a Decolonial AI. Feedback please 🙏🏾 Thread 👇🏾 https://arxiv.org/abs/2007.04068</p>

tag:tagteam.harvard.edu,2005:FeedItem/2727754 2020-07-09T14:11:07-04:00 2020-07-09T14:11:07-04:00

<p>Those who could be exploited by AI should be shaping its projects.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2727200 2020-07-08T17:15:50-04:00 2020-07-08T17:15:50-04:00

<p>This paper critically examines existing modes of participation in design practice and machine learning. Cautioning against 'participation-washing', it suggests that the ML community must become attuned to possibly exploitative and extractive forms of community involvement and shift away from the prerogatives of context-independent scalability.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2725034 2020-07-06T11:25:41-04:00 2020-07-06T11:25:41-04:00

<p>AI4SDGs Think Tank is an online open service for everyone, a global repository and an analytic engine of AI projects and proposals that impacts UN Sustainable Development Goals, both positively and negatively. The goal is to promote the positive use of AI for Sustainable Development, and to investigate on the negative impact of AI on Sustainable Development. Detailed evaluation on each project is provided based on our rating scheme. You are welcome to share your project to the world and get evaluated by submitting your project or proposal information here.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2725033 2020-07-06T11:08:20-04:00 2020-07-06T11:08:20-04:00

<p>This study on ‘Opportunities of Artificial Intelligence’ aims to assess the state of play of Artificial Intelligence (AI) adoption in European industry from a technological, impact and regulatory perspective, before presenting a methodology to scrutinise the EU policy and regulatory framework in the context of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2725032 2020-07-06T11:07:50-04:00 2020-07-06T11:07:50-04:00

<p>The Government’s Online Harms programme presents a significant first step towards this goal. It needs to happen; it needs to happen fast; and the necessary draft legislation must be laid before Parliament for scrutiny without delay. The Government must not flinch in the face of the inevitable and powerful lobbying of Big Tech and others that benefit from the current situation.</p> <p>Well drafted Online Harms legislation can do much to protect our democracy. Issues such as misinformation and disinformation must be included in the Bill. The Government must make sure that online platforms bear ultimate responsibility for the content that their algorithms promote. Where harmful content spreads virally on their service or where it is posted by users with a large audience, they should face sanctions over their output as other broadcasters do.</p> <p>Individual users need greater protection. They must have redress against large platforms through an ombudsman tasked with safeguarding the rights of citizens.</p> <p>Transparency of online platforms is essential if democracy is to flourish. Platforms like Facebook and Google seek to hide behind ‘black box’ algorithms which choose what content users are shown. They take the position that their decisions are not responsible for harms that may result from online activity. This is plain wrong. The decisions platforms make in designing and training these algorithmic systems shape the conversations that happen online. For this reason, we recommend that platforms be mandated to conduct audits to show how in creating these algorithms they have ensured, for example, that they are not discriminating against certain groups. Regulators must have the powers to oversee these decisions, with the right to acquire the information from platforms they need to exercise those powers.</p> <p>Platforms’ decisions about what content they remove or stop promoting through their algorithms set the de facto limits of free expression online. As it currently stands the rules behind these decisions are poorly defined. Their practical operation should reflect what the public needs. In order to protect free and open debate online, platforms should be obliged to publish their content decisions making clear what the actual rules of online debate are.</p> <p>Alongside establishing rules in the online world, we must also empower citizens, young and old, to take part as critical users of information. We need to create a programme of lifelong education that will equip people with the skills they need to be active citizens. People need to be taught from a very young age about the ways in which platforms shape their online experience.</p> <p>The public needs to have access to high quality public interest journalism to help inform them about current events. This requires fair funding to support such journalism.</p> <p>Platforms must also be forced to ensure that their services empower users to exercise their rights online. The public need to understand how their data is being used. We propose that this obligation of fairness by design should be a core element in ensuring platforms meet their duty of care to their users.</p> <p>Parliament and government at all levels need to invest in technology to engage better with the public.</p> <p>Electoral law must be completely updated for an online age. There have been no major changes to electoral law since the invention of social media and the rise of online political advertising. As the Law Commission recently pointed out, a wholesale revision of the relevant law is now needed. This should include rules that set standards for online imprints on political advertisements so that people can see who they come from and advert libraries that enable researchers and the public to see what campaigns are saying. The Electoral Commission needs the powers to obtain the information necessary to understand when individuals are breaking the rules and to be able to set fines that act as a real deterrent against flagrant breaches. We also need to ensure that there is greater clarity around the use of personal data in political campaigns; the Information Commissioner’s guidance should be put on statutory footing.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2725030 2020-07-06T11:00:00-04:00 2020-07-06T11:00:00-04:00

<p>The DoD inspector general suggested the Joint Artificial Intelligence Center, established to facilitate the adoption of artificial intelligence tools across the department, take several steps to improve project management, including determining a standard definition of artificial intelligence, improving data sharing and developing a process to accurately track artificial intelligence programs. The JAIC missed a March 2020 deadline to release a governance framework. It still plans to do so, according to the report, but that date is redacted in the report.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2725029 2020-07-06T10:39:06-04:00 2020-07-06T10:39:06-04:00

<p><strong><span>Evaluating the present and future of AI in pretrial risk assessment.<span> </span></span></strong><span>Alexandra Chouldechova and Kristian Lum argue that today’s actuarial risk assessment tools are just a warning of what is to come. They say that the more complex, AI-based tools of the future will raise even more profound legal and ethical concerns because they will rely on more expansive data sets. (</span><a href="https://urldefense.proofpoint.com/v2/url?u=http-3A__www.safetyandjusticechallenge.org_wp-2Dcontent_uploads_2020_06_AI-2Din-2DPre-2DTrial-2DRisk-2DAssessment-2DBrief-2DJune-2D2020-2DR2.pdf&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=6B6xixkugH1KOzzLHF9ip1p7dNPwmGksfNCVmoaBs-U&amp;s=aGVnD_wf23Cw1aOHZsvsazgDDo8mhoXAQQYzyBwx-YQ&amp;e=" title="https://urldefense.proofpoint.com/v2/url?u=http-3A__www.safetyandjusticechallenge.org_wp-2Dcontent_uploads_2020_06_AI-2Din-2DPre-2DTrial-2DRisk-2DAssessment-2DBrief-2DJune-2D2020-2DR2.pdf&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=6B6xixkugH1KOzzLHF9ip1p7dNPwmGksfNCVmoaBs-U&amp;s=aGVnD_wf23Cw1aOHZsvsazgDDo8mhoXAQQYzyBwx-YQ&amp;e=">Safety and Justice Challenge</a><span>) — Logan</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2721670 2020-07-01T13:32:52-04:00 2020-07-01T13:32:52-04:00

<p>The problematic training library in question is 80 Million Tiny Images, which was created in 2008 to help produce advanced object detection techniques. It is, essentially, a huge collection of photos with labels describing what's in the pics, all of which can be fed into neural networks to teach them to associate patterns in photos with the descriptive labels. So when a trained neural network is shown a bike, it can accurately predict a bike is present in the snap. It's called Tiny Images because the pictures in library are small enough for computer-vision algorithms in the late-2000s and early-2010s to digest.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2720717 2020-06-30T13:15:07-04:00 2020-06-30T13:15:07-04:00

<p>So there has been a brain drain of computer scientists from universities to the big tech companies, lured by access to their cloud data centers as well as lucrative pay packages. The worry is that academic research — the seed corn of future breakthroughs — is being shortchanged. Academic work can be crucial particularly in areas where profits are not on the immediate horizon. That was the story with deep learning, which dates to the 1980s. A small band of academics nurtured the field for years. Only since 2012, with enough computing power and data, did deep learning really take off. There have been smaller efforts for university research to tap into the big tech clouds. But the current concept of an ambitious public-private partnership for a National Research Cloud came in March from John Etchemendy and Fei-Fei Li, co-directors of the Stanford Institute for Human-Centered Artificial Intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2720714 2020-06-30T13:05:23-04:00 2020-06-30T13:05:23-04:00

<p>Machine learning programs are not neutral; research agendas and the data sets they work with often inherit dominant cultural beliefs about the world. These research agendas reflect the incentives and perspectives of those in the privileged position of developing machine learning models, and the data on which they rely. The uncritical acceptance of default assumptions inevitably leads to discriminatory design in algorithmic systems, reproducing ideas which normalize social hierarchies and legitimize violence against marginalized groups.[9]</p>

tag:tagteam.harvard.edu,2005:FeedItem/2719698 2020-06-29T12:06:45-04:00 2020-06-29T12:06:45-04:00

<p>Through all of this, we should not lose sight of what is humanity, our own intelligence and what it is we truly want. It is often easier to blame technology, focusing on our fears and ‘what if’ scenarios rather than discussing our core values and charting a beneficial path forward for humankind. If we do not know what we want for our future, how c</p>

tag:tagteam.harvard.edu,2005:FeedItem/2714392 2020-06-22T12:07:35-04:00 2020-07-06T14:14:08-04:00

<p>Such a large-scale reliance on these technologies has reduced human bodies into evidence. As a result, and in an effort to regain some control over their bodies and their data, there have been a number of disturbing reports about refugees burning off their own fingerprints out of fear of being tracked and returned to countries of origin, or to entry-point countries in the EU due to the Dublin Regulation. …there have been a number of disturbing reports about refugees burning off their own fingerprints Like facial recognition technology, fingerprint readers were built on racialised assumptions of users, and only have high accuracy with middle aged men with lighter skin tones. There has been a pattern of centralised collection and storage of this biometric and personal data across sectors and agencies, seen with public-private partnerships like Palantir-WFP, Microsoft-International Committee for the Red Cross, and Facebook Libra-Mercy Corps.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2711597 2020-06-18T09:55:16-04:00 2020-06-18T09:55:16-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2711592 2020-06-18T09:38:16-04:00 2020-06-18T09:38:16-04:00

<p>GPAI is an international and multistakeholder initiative to guide the responsible development and use of AI, grounded in human rights, inclusion, diversity, innovation, and economic growth. In order to achieve this goal, the initiative will look to bridge the gap between theory and practice on AI by supporting cutting-edge research and applied activities on AI-related priorities.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2711146 2020-06-17T17:29:59-04:00 2020-06-17T17:29:59-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2709824 2020-06-16T10:41:21-04:00 2020-06-16T10:41:21-04:00

<p>JUST RELEASED: Free Expression, Harmful Speech, and Censorship in a Digital World: As the COVID-19 crisis deepens, Americans realize the threat of misinformation to people’s health, and the country’s democracy, however are undecided on how best to remedy the problem while preserving their deeply-held value of free expression. Download the executive summary and report. View the data and questionnaire.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2709814 2020-06-16T10:38:15-04:00 2020-06-16T10:38:15-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2709797 2020-06-16T10:26:23-04:00 2020-06-16T10:26:23-04:00

<p>Algorithm Watch, a research and advocacy organisation focused on algorithmic decision-making, collected data from 26 volunteer Instagram users to see which pictures from a select group of 37 creators were placed highest in their main feeds. This was done using a browser add-on, which automatically opened the Instagram homepage at regular intervals and analyzed the photos it saw there. Overall 2,400 photos from February to May 2020 were analyzed.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2709251 2020-06-15T16:27:48-04:00 2020-06-15T16:27:48-04:00

<p>Check out our new study: in one of the largest international experiments conducted in higher education we show the importance of context when scaling behavioral science interventions in online education and the limits of AI for targeting @bjfr @emyeom</p>

tag:tagteam.harvard.edu,2005:FeedItem/2708937 2020-06-15T09:40:14-04:00 2020-06-15T09:40:14-04:00

<p>This report lays out a roadmap in which all stakeholders play a role in advancing a safer, more equitable digital world, one which will lead to a brighter and more prosperous future for all.</p> <p> </p>

tag:tagteam.harvard.edu,2005:FeedItem/2706468 2020-06-11T13:11:41-04:00 2020-06-11T13:11:41-04:00

<p>To supplement experiments with public media platforms (Recommendation 5.2), establish a public-interest mandate for for-profit social media platforms. Analogous to zoning requirements, this mandate would require such for-profit digital platform companies to support the development of designated public-friendly digital spaces on their own platforms.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2705451 2020-06-10T12:56:44-04:00 2020-06-10T12:56:44-04:00

<p>The tech workers who long for political righteousness are mired in a fundamental trap. They conflate their identity with their workplace to an extreme, partly because they’ve been enculturated to believe that technological innovation brings about social benefit rather than hindering it</p>

tag:tagteam.harvard.edu,2005:FeedItem/2705447 2020-06-10T12:39:06-04:00 2020-06-10T12:39:06-04:00

<p>The authors utilize a modernity-coloniality framework to highlight practices of whiteness and white dominance within the academy and the field of communication studies in particular. The article grounds its framework in personal experience and theories in the flesh to build comradery with feminist of color scholarly forebears and to share with those who struggle to create spaces of change within the academy today and in the future. The authors advocate for an active consciousness of “included exclusion” as the first step toward decolonization. Decolonizing discursive and embodied knowledges creates avenues for social justice and change within the academy and beyond.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2704536 2020-06-09T09:43:54-04:00 2020-06-09T09:43:54-04:00

<p>I then outline a taxonomy of transparency for accountable algorithms, building on the work of earlier authors and my own research on the European Union’s General Data Protection Regulation (GDPR).</p>

tag:tagteam.harvard.edu,2005:FeedItem/2703664 2020-06-08T10:13:35-04:00 2020-06-08T10:13:35-04:00

<p>We suggest that academia has a particularly important role to play in supporting cross-cultural cooperation on AI ethics and governance. Academic communities are particularly well-suited to building greater mutual understanding between regions and cultures, due to their tradition of free-flowing, international, and intercultural exchange of ideas. We make a number of practical steps that academic institutions and individual researchers can take: including translation of key papers and reports; alternating the location of key conferences; and establishing exchange programmes. We also identify several promising research topics in AI ethics and governance that would be well-suited to international collaboration, such as conducting comparative foresight exercises exploring how different cultures perceive the future impact of AI on society.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2701829 2020-06-05T10:14:16-04:00 2020-06-05T10:14:16-04:00

<p>This paper investigates how the use of AI in relation to data governance regimes and regulatory practices can intensify power asymmetries. • The paper examines legal and policy instruments asssociated with case studies on the use of AI for the public sector, in Canada, Finland and Poland. • The paper advocates the need for a common framework to evaluate the potential impact of the use of AI in the public sector. • The findings are set to understand how to exploit the potential of technology, while negative effects are possibly avoided. • It discusses the effects of automated decision support systems and the crucial role of governments in the digital society. • In light of the current COVID-19 crisis, the paper debates how AI could strengthen – or harm - trust in governance systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2701826 2020-06-05T10:04:59-04:00 2020-06-05T10:04:59-04:00

ment failed his sibling and called Centrelink's claim that he had been overpaid by $300 a week 'absurd'. The 36-year-old mechanic, who lives in Sydney, said he didn't see much of his brother in the last year of his life because of his anxiety about travel.

tag:tagteam.harvard.edu,2005:FeedItem/2700001 2020-06-03T11:09:09-04:00 2020-06-03T11:09:09-04:00

<p>Ultimately, in the coming weeks, many of the same companies that tweeted “BLM” or claimed “support for the Black community” in the last few days will happily have their tech used in the service of tracking down and punishing Black protestors. It’s not technically impossible for a tech company to commit to Black liberation, but it’s going to take a lot more than a social media post.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2699999 2020-06-03T11:07:36-04:00 2020-06-03T11:07:36-04:00

<p>Our full report contains in-depth findings and analyses around these research questions, as well as recommendations and insights for policymakers and industry practitioners working in privacy-related spaces.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2699119 2020-06-02T11:09:08-04:00 2020-06-02T11:09:08-04:00

<p>RT definition.22This divergence further demonstrates howdifferent AI definiti</p>

tag:tagteam.harvard.edu,2005:FeedItem/2695405 2020-05-29T13:50:19-04:00 2020-05-29T13:50:19-04:00

<p>The coronavirus will likely transform other pillars of democratic governance—such as electoral processes, civilian control of militaries, and civic mobilization—and potentially reset the terms of the global debate on the merits of authoritarianism versus democracy. The pandemic will almost certainly usher in broader effects on governance by overburdening countries’ basic governance functions, taxing their sociopolitical cohesion, exacerbating corruption, unsettling relations between national and local governments, and transforming the role of nonstate actors. This article surveys this wide spectrum of effects. Of course, much remains uncertain as long as the ultimate scope and severity of the crisis are unknown. As the pandemic penetrates lower-income and fragile states, it will likely have even more profound and unpredictable effects than those visible thus far. This article focuses on the first-order political effects of the pandemic and governmental responses to it. Powerful second-order effects resulting from the unfolding global economic slowdown will pack a further governance punch.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2695240 2020-05-29T10:45:09-04:00 2020-05-29T10:45:09-04:00

<p>The group, called the Global Partnership on AI, will study and provide recommendations to encourage the creation of AI technologies that respect privacy and civil liberties, said Michael Kratsios, the chief technology officer of the U.S. He called the partnership “a first of its kind global organization that sees the future of AI as something that can uplift Americans and people around the world.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2693625 2020-05-27T15:12:40-04:00 2020-05-27T15:12:40-04:00

<p>The second in a two-part series aimed at U.S. policymakers and anybody concerned with the question of how internet platforms should be regulated, this report reinforces the need to adopt a human rights framework for platform accountability. We propose concrete areas where Congress needs to act to mitigate the harms of misinformation and other dangerous speech without compromising free expression and privacy: transparency and accountability for online advertising, starting with political ads; federal privacy law; and corporate governance reform.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2693621 2020-05-27T15:09:08-04:00 2020-05-27T15:09:08-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2693408 2020-05-27T10:45:17-04:00 2020-05-27T10:45:17-04:00

<p>did not answer the detailed questions we sent them.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2688184 2020-05-20T12:21:31-04:00 2020-05-20T12:21:31-04:00

<p>The crisis has made clear the asymmetry between private profits and the public interest</p>

tag:tagteam.harvard.edu,2005:FeedItem/2688114 2020-05-20T10:51:04-04:00 2020-05-20T10:51:04-04:00

<p>The Stanford Institute for Human-Centered Artificial Intelligence's second virtual conference on COVID will address these questions. Scholars from across disciplines will discuss their research using AI, data science and/or informatics to help us understand how we emerge from this crisis. Sessions will examine, among other topics: Preparing for the 2020 election, protecting privacy during contact tracing, and assessing COVID infections.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2687548 2020-05-19T15:54:49-04:00 2020-05-19T15:54:49-04:00

<p>Their findings are presented in “Co-Designing Checklists to Understand Organizational Challenges and Opportunities around Fairness in AI” and “Interpreting Interpretability: Understanding Data Scientists’ Use of Interpretability Tools for Machine Learning.” The publications received ACM CHI Conference on Human Factors in Computing Systems (CHI 2020) best paper recognition and honorable mention, respectively.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2687529 2020-05-19T12:59:33-04:00 2020-05-19T12:59:33-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2686675 2020-05-18T11:07:13-04:00 2020-05-18T11:07:13-04:00

<p>Innovations in data science and artificial intelligence (AI) have a central role to play in supporting global efforts to combat Covid-19. The versatility of AI technologies is enabling scientists and technologists to address an impressively broad range of biomedical, epidemiological, and socio-economic challenges. This wide-reaching scientific capacity is, however, also raising a diverse array of ethical challenges. The need for researchers to act quickly and globally in tackling Coronavirus demands unprecedented practices of open research and responsible data sharing at a time when innovation ecosystems are hobbled by proprietary protectionism and a lack of public trust. Moreover, societally impactful interventions like digital contact tracing are raising fears of “surveillance creep” and are challenging widely-held commitments to privacy, autonomy, and civil liberties. Pre-pandemic concerns that data-driven innovations may function to reinforce entrenched dynamics of social inequality are likewise now only more intensified given the life-and-death consequences of biased and discriminatory public health outcomes. The following offers five steps toward responsible research and innovation that need to be taken to address these concerns. It presents a practice-based path to responsible AI design and discovery centered on open, accountable, equitable, and democratically governed processes and products. When taken from the start, these steps will not only enhance the capacity of innovators to tackle Covid-19 responsibly, they will help to set the data science and AI community down a path that is both better prepared to cope with future pandemics and better equipped to support a more humane, rational, and just society of tomorrow.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2683689 2020-05-14T10:20:51-04:00 2020-05-14T10:20:51-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2683633 2020-05-14T09:33:33-04:00 2020-05-14T09:33:33-04:00

<p><span>We have collected </span><strong>global examples of how the UNDP is providing digital solutions to support governments in response COVID-19. </strong></p>

tag:tagteam.harvard.edu,2005:FeedItem/2682936 2020-05-13T09:56:09-04:00 2020-05-13T09:56:09-04:00

<p><span>Preliminary research suggests that<span> </span><a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__georgetown.us20.list-2Dmanage.com_track_click-3Fu-3D248119f7a6386759c3a4f7155-26id-3D9287be95de-26e-3Df8cfc530f1&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=GRvzhAZYzLEN23pFgnCuR5wBSLtg2U6CMAx-ocWiVLA&amp;s=mlGWvZkknGQoMV6Ed6PODFClNb9uHw9GQCynVq2ifAw&amp;e=" title="https://urldefense.proofpoint.com/v2/url?u=https-3A__georgetown.us20.list-2Dmanage.com_track_click-3Fu-3D248119f7a6386759c3a4f7155-26id-3D9287be95de-26e-3Df8cfc530f1&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=GRvzhAZYzLEN23pFgnCuR5wBSLtg2U6CMAx-ocWiVLA&amp;s=mlGWvZkknGQoMV6Ed6PODFClNb9uHw9GQCynVq2ifAw&amp;e=">most models using AI for COVID-19 diagnosis and prognosis are at high risk of bias and likely misleading</a>. The researchers reviewed 31 models that use AI for online diagnosis, diagnosis based on CT scans and case prognosis prediction. According to their assessment, the reported performance of all 31 models was “probably optimistic.” Given the lack of external validation of most models, the authors recommend against relying on any current AI-based COVID-19 diagnostic tools.</span><span><span> </span></span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2681441 2020-05-11T13:19:30-04:00 2020-05-11T13:19:30-04:00

<p>At CDT, we’re committed to advancing these priorities in the national and global pandemic response. Our Coronavirus: Data for Life and Liberty Task Force is bringing together individuals from companies and civil society organizations, lawyers, academics, health professionals, and technologists to explore whether, and if so how, data can be used to responsibly fight the pandemic while preserving civil liberties and promoting civil rights.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2681288 2020-05-11T10:25:42-04:00 2020-05-11T10:25:42-04:00

<p>This in-depth analysis discusses the opportunities and challenges brought by the recent and the foreseeable developments of Artificial Intelligence into online platforms and marketplaces. The paper advocates the importance to support trustworthy, explainable AI (in order to fight discrimination and manipulation, and empower citizens), and societal-aware AI (in order to fight polarization, monopolistic concentration and excessive inequality, and pursue diversity and openness). This document was prepared by the Policy Department A at the request of the Committee on the Internal Market and Consumer Protection.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2681287 2020-05-11T10:22:37-04:00 2020-05-11T10:22:37-04:00

<p> </p> <p>While researching the very challenging and open ended <a href="https://www.kaggle.com/allen-institute-for-ai/CORD-19-research-challenge/tasks?taskId=563">questions posed by this Kaggle task</a>, we take an approach inspired by Qualitative Social Science. Therefore we conducted a few unstructured interviews with policy-makers, ethics counselors, and front-line medical practitioners in order to better understand what are their biggest needs and how our work could be most helpful for them. The details that experts shared with us informed the data exploration and modeling presented in this notebook.</p> <p>The goal of this notebook is to link public health measures and social and ethical concerns for individual articles in the corpus. In order to extract such information, we will utilize a language grammar. In our case we've defined a grammar that aims to capture how the literature discusses <strong>factors such as the enablers, barriers, and implications of different public health measures</strong>. Our current grammar is in English but our language grammar approach could be evolved into covering other languages as well.</p> <p> </p>

tag:tagteam.harvard.edu,2005:FeedItem/2681284 2020-05-11T10:15:51-04:00 2020-05-11T10:15:51-04:00

<p>Gordon says, “Covid has proved we need institutions. And yet, our bureaucracies are proving increasingly ineffective, in part because the speed of decision-making that is required. We need to reimagine and rebuild. Who are the leaders who are going to come in and do that? My catastrophic event is the failure of bureaucracy to provide the governance our society needs, keeping true to our values.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2679568 2020-05-08T13:09:03-04:00 2020-05-08T13:09:03-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2678817 2020-05-07T16:31:46-04:00 2020-05-07T16:31:46-04:00

<p>We’re releasing an analysis showing that since 2012 the amount of compute needed to train a neural net to the same performance on ImageNet1 classification has been decreasing by a factor of 2 every 16 months. Compared to 2012, it now takes 44 times less compute to train a neural network to the level of AlexNet2 (by contrast, Moore’s Law3 would yield an 11x cost improvement over this period). Our results suggest that for AI tasks with high levels of recent investment, algorithmic progress has yielded more gains than classical hardware efficiency.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2677897 2020-05-06T16:31:13-04:00 2020-05-06T16:31:13-04:00

<p>Module 2: Building AI Projects Here’s the series of steps that all machine learning projects go through: First, you gather data. Then you train the algorithm using the data you collected. Finally, you deploy the algorithm in the real world to learn more so you can improve it again. Data science is slightly different: you gather data, extract important insights, and then create a report that can help a business leader improve their managerial decision-making. Specific examples of each are provided in the course.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2677752 2020-05-06T14:12:01-04:00 2020-05-06T14:12:01-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2676727 2020-05-05T14:31:20-04:00 2020-05-05T14:31:20-04:00

<p>Workers in the U.S. moved to unemployment by the COVID-19 pandemic may find themselves displaced by automation during the eventual recovery, according to new research by Georgia State University economists.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2676733 2020-05-05T14:31:47-04:00 2020-05-05T14:31:47-04:00

<p>Addressing this question is particularly difficult partly because the AI industry community currently has no standardized process for documenting AI systems. As opposed to the software industry, where companies have long-established processes created to create transparency and accountability with their services (including design documents, performance tests and cybersecurity protocols), AI practitioners are still trying to figure out what these processes should be. This lack of standardized protocols is a reminder and an expression of the youth of this field.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2675935 2020-05-04T17:13:18-04:00 2020-05-04T17:13:18-04:00

<p>It is a societal failure when we undervalue institutions and the critical services they provide. What we need is citizens and customers to demand co-ordination and partnership across sectors.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2675792 2020-05-04T15:51:39-04:00 2020-05-04T15:51:39-04:00

<p>A Year in Review: Observations from 50 Global Experts</p>

tag:tagteam.harvard.edu,2005:FeedItem/2673825 2020-05-01T13:26:29-04:00 2020-05-01T13:26:29-04:00

<p>Across the country, policymakers are putting aside politics and leaning on the best quality data and evidence to bring innovative solutions into practice for cities and states wrestling with this unfolding pandemic.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2673676 2020-05-01T10:42:20-04:00 2020-05-01T10:42:20-04:00

<p>nd the use of algorithmic systems, in particular those that underscore the large tech platforms. In the UK, the Centre for Data Ethics &amp; Innovation has even recognised a growing market opportunity for the UK to be at the forefront of ‘an AI audit market’, capitalising on increasing interest in audit as a mechanism for assessment, accountability and public trust and confidence.6However, through literature review and conversations with experts, we find the term ‘audit’ is used by different actors in different ways.We surmise that confusion about algorithm audits comes from the two relevant meanings of ‘audit’:1. Audit from the perspective of the computer science community, which proposes adopting the social science practice of an audit study and applying</p>

tag:tagteam.harvard.edu,2005:FeedItem/2672087 2020-04-29T14:08:02-04:00 2020-04-29T14:08:02-04:00

<p>From May to July 2020, UNESCO is organizing extensive multi-stakeholder online consultations at the national, regional and international levels. The purpose of these consultations is to ensure that all stakeholders, including scientists, civil society and the public, take part in developing the first global standard-setting instrument on the ethical dimensions of AI. The expert group will then deliver a first draft of an inclusive normative instrument on the ethics of AI, which will be considered by UNESCO Member States when they draft a final text for approval as a global Recommendation in November 2021.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2671925 2020-04-29T11:24:58-04:00 2020-04-29T11:24:58-04:00

<p><span>ee a compendium of 80 cases of technology-led responses to the COVID-19 pandemic from our partner companies<span> </span></span><a href="https://urldefense.proofpoint.com/v2/url?u=http-3A__click.e.weforum.org_-3Fqs-3Dc568059101d279f59301e3500b01490f218b4933e7fc97d1f423de0a0d289c585e46e0563c5376ff7932d5feeead10534d5817b2f1c4bdfe&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=FlFqp4Twe-5hB61IgwqK0JqA4RZ4yinBq6RDw1QUmgo&amp;s=F3Povp1Rrt1BmVZKNmyc_Sz--TvNALylxkNzhxP_iXA&amp;e=" title="here">here</a><span>. A related webinar on the role of technology in ending the COVID crisis featured Kevin Sneader, McKinsey; Steven Mnuchin, Government of United States; Ivan Duque, Government of Colombia; Rob Shuter, MTN and Shobana Kamineni, Apollo Hospitals. </span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2672089 2020-04-29T14:20:36-04:00 2020-07-06T10:02:12-04:00

<p>UNESCO has embarked on a two-year process to elaborate the first global standard-setting instrument on ethics of artificial intelligence, following the decision of UNESCO’s General Conference at its 40th session in November 2019. In 2020, the focus will be on preparing the draft text for the Recommendation with the assistance of an Ad Hoc Expert Group. This phase will include inclusive and multidisciplinary consultations with a wide range of stakeholders. These broad consultations are extremely important to ensure that the draft text is as inclusive as possible. Towards the end of 2020 and in 2021, the focus will be on an intergovernmental process and on negotiation on the draft text to produce a final version of the Recommendation for possible adoption by UNESCO’s General Conference at its 41st session at the end of 2021. UNESCO’s work on the Recommendation is building on the preliminary study on ethics of artificial intelligence of UNESCO’s World Commission on the Ethics of Scientific Knowledge and Technology (COMEST). This study emphasizes that currently no global instrument covers all the fields that guide the development and application of AI in a human-centred approach.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2669901 2020-04-27T11:19:52-04:00 2020-04-27T11:19:52-04:00

<p>Optimizing systems can also incorporate the perspectives of non-users. A ride sharing platform may cause traffic jams for non-riders, and environmental effects are potentially global. There is a whole field of “multi-stakeholder optimization” that studies algorithms designed to account for these types of issues. A product recommendation system is typically designed to present users with products they are most likely to buy, but could also incorporate climate change concerns through estimates of product carbon footprint. These could be used to increase the visibility of alternative low-carbon products, with the goal of minimizing the metric of carbon intensity per unit revenue.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2669869 2020-04-27T10:49:28-04:00 2020-04-27T10:49:28-04:00

<p>A stereotype is an over-generalized belief about a particular group of people, e.g., Asians are good at math or Asians are bad drivers. Such beliefs (biases) are known to hurt target groups. Since pretrained language models are trained on large real world data, they are known to capture stereotypical biases. In order to assess the adverse effects of these models, it is important to quantify the bias captured in them. Existing literature on quantifying bias evaluates pretrained language models on a small set of artificially constructed bias-assessing sentences. We present StereoSet, a large-scale natural dataset in English to measure stereotypical biases in four domains: gender, profession, race, and religion. We evaluate popular models like BERT, GPT-2, RoBERTa, and XLNet on our dataset and show that these models exhibit strong stereotypical biases. We also present a leaderboard with a hidden test set to track the bias of future language models at this https URL</p>

tag:tagteam.harvard.edu,2005:FeedItem/2668072 2020-04-24T14:58:19-04:00 2020-04-24T14:58:19-04:00

<p>Responding to growing demands for more accountability in the development and deployment of artificial intelligence, public policymakers have signed on to the fashionable push for "ethical AI." Yet by adopting what amounts to a euphemism for inaction, they are playing directly into the industry's hands.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2667878 2020-04-24T11:23:04-04:00 2020-04-24T11:23:04-04:00

<p>Theresa Züger: Another goal for the first years of the Lab is to develop and establish a new perspective on AI and society by focussing on the concept of Public Interest AI. We want to address the question of how AI can serve the public interest on different levels, reaching from its social consequences to AI’s technical design and user-experience design. This approach highlights a political perspective on AI – meaning a perspective that sees AI as a concern to society as a whole – and thereby broadens the often discussed ethical considerations on AI which mainly addresses developers and their organizations as the main decisions makers. Thinking about public interest always implies considering the well-being of society and all its parts and concerns. The discussion about AI in the line of public interest should be open to all who are affected, especially since AI will increasingly become a part of our daily lives and society’s infrastructure.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2667872 2020-04-24T11:21:51-04:00 2020-04-24T11:21:51-04:00

<p>If the app is not carefully planned and designed, the digital divide may become a biological divide. It may be objected that a small percentage of people voluntarily adopting the app is better than nothing. I doubt it, given that such a small percentage tends to live together in the same areas. It will be great, but only for them (benefits and positive incentives included), and useless for all the others. Indeed, the unfairness of a voluntary app adopted by a small percentage of the population may be used to support the case for a mandatory solution, that is, negative incentives. Where the primary benefit is collective, not individual, it may be contextually ethical to require that those who can afford more do more, that is, those who can adopt a digital solution are asked to adopt it. This "negative" incentive (the mandatory app) could "positively" discriminate the better-off (the digitally-educated owners of a Bluetooth-enabled mobile phone) to help, now as a very large group, the worst off. Metaphorically, a compulsory app would work like a tax. Those who can contribute more digitally may be required to do so. Consider that in some cultures this may be a de facto social outcome, not a legal requirement, whenever peer pressure and expectations make a behaviour so strongly expected to be almost compulsory.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2667835 2020-04-24T10:48:50-04:00 2020-04-24T10:49:00-04:00

<p>To this end, the TUM Institute for Ethics in Artificial Intelligence has led the way in the creation of the Global AI Ethics Consortium (GAIEC), joining forces with academic institutions, research centers and distinguished members of academia worldwide in order to foster trust in data and technology, maximize the potential of AI while limiting its harms, help all the involved parties navigate current uncertainty and create ethical frameworks.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2667724 2020-04-24T08:35:55-04:00 2020-04-24T08:35:55-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2664378 2020-04-20T11:16:27-04:00 2020-04-20T11:16:27-04:00

<p>With the recent wave of progress in artificial intelligence (AI) has come a growing awareness of the large-scale impacts of AI systems, and recognition that existing regulations and norms in industry and academia are insufficient to ensure responsible AI development. In order for AI developers to earn trust from system users, customers, civil society, governments, and other stakeholders that they are building AI responsibly, they will need to make verifiable claims to which they can be held accountable. Those outside of a given organization also need effective means of scrutinizing such claims. This report suggests various steps that different stakeholders can take to improve the verifiability of claims made about AI systems and their associated development processes, with a focus on providing evidence about the safety, security, fairness, and privacy protection of AI systems. We analyze ten mechanisms for this purpose--spanning institutions, software, and hardware--and make recommendations aimed at implementing, exploring, or improving those mechanisms.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2664429 2020-04-20T12:46:17-04:00 2020-04-21T11:09:48-04:00

<p>Ben Green/Salome</p> <p>As tech firms team up with governments to fight the coronavirus pandemic, we’re being asked to accept a trade-off between our digital privacy and our health. It’s a false choice: we can achieve the public health benefits of data without accepting abusive and illicit surveillance.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2662156 2020-04-16T15:30:50-04:00 2020-04-16T15:30:50-04:00

<p>In order for AI developers to earn trust from users, civil society, governments, and other stakeholders, there is a need to move beyond principles to a focus on mechanisms for demonstrating responsible behavior. Making and assessing verifiable claims, to which developers can be held accountable, is one step in this direction.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2661983 2020-04-16T12:20:46-04:00 2020-04-16T12:20:46-04:00

<p>Inclusion. One of the great promises of AI is its potential for improving quality of life. But without the right planning and oversight, we risk exacerbating problems of inequality or marginalizing groups of people. As an example, India’s AI strategy is focused on leveraging the technology not only for economic growth, but also for social inclusion. The approach is called #AIforAll and outlines a strategy that aims to empower Indians with the skills to find the quality jobs, invest in research, and scale Indian-made AI solutions to the rest of the developing world. India is not the first country to incorporate AI and inclusion. Canada and France, for example, recently announced a task force to develop an international study group on inclusive and ethical AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2661311 2020-04-15T14:09:19-04:00 2020-04-15T14:09:19-04:00

<p>The report looks at existing strategies adopted by governments and regional organisations since 2017. It assesses the extent to which human rights considerations have been incorporated and makes a series of recommendations to policymakers looking to develop or revise AI strategies in the future.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2661172 2020-04-15T12:14:50-04:00 2020-04-15T12:14:50-04:00

<p>Is data privacy the price we must pay to survive a pandemic?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2658117 2020-04-10T11:39:00-04:00 2020-04-10T11:39:00-04:00

<p>r</p>

tag:tagteam.harvard.edu,2005:FeedItem/2656731 2020-04-08T13:46:27-04:00 2020-04-08T13:46:27-04:00

<p>The scientists say they want to become a “trusted intermediary” between private companies and public officials—sidestepping worries about giving the government direct access to information on Americans’ movements. Officials could ask for help in making critical policy decisions, but they wouldn’t be able to muck around in the data directly or reuse it for another purpose, like policing or immigration enforcement.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2656144 2020-04-07T16:47:13-04:00 2020-04-07T16:47:13-04:00

<p>This paper discusses the establishment of a governance framework to secure the development and deployment of “good AI”, and describes the quest for a morally objective compass to steer it. Asserting that human rights can provide such compass, this paper first examines what a human rights-based approach to AI governance entails, and sets out the promise it propagates. Subsequently, it examines the pitfalls associated with human rights, particularly focusing on the criticism that these rights may be too Western, too individualistic, too narrow in scope and too abstract to form the basis of sound AI governance. After rebutting these reproaches, a plea is made to move beyond the calls for a human rights-based approach, and start taking the necessary steps to attain its realisation. It is argued that, without elucidating the applicability and enforceability of human rights in the context of AI; adopting legal rules that concretise those rights where appropriate; enhancing existing enforcement mechanisms; and securing an underlying societal infrastructure that enables human rights in the first place, any human rights-based governance framework for AI risks falling short of its purpose.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2653151 2020-04-03T10:38:53-04:00 2020-04-03T10:38:53-04:00

<p>The Community Data Science Collective, in collaboration with Pushshift and others, is launching a new collaborative project to create a digital observatory for socially produced COVID-19 information. The observatory has already begun the process of collecting, and aggregating public data from multiple online communities and platforms. We are publishing reworked versions of these data in forms that are well-documented and more easily analyzable by researchers with a range of skills and computation resources. We hope that these data will facilitate analysis and interventions to improve the quality of socially produced information and public health.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2652276 2020-04-02T10:23:33-04:00 2020-04-02T10:23:33-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2651471 2020-04-01T11:04:42-04:00 2020-04-01T11:04:42-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2650886 2020-03-31T17:03:04-04:00 2020-03-31T17:03:04-04:00

<p>This work considered features whose unfairness rests on a broad ethical consensus. In general, however, there is disagreement on what is fair and unfair. This disagreement is rooted in political, and ultimately ethical disagreement among human beings. Real progress on algorithmic fairness entails not only formalizing human ethical consensus on what is fair and unfair, but explicitly considering, within the marketplace of ideas, differences in our ethical commitments as well.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2650594 2020-03-31T12:33:50-04:00 2020-03-31T12:33:50-04:00

<p>There are many potential benefits to the application of Artificial Intelligence (AI) technologies, including the reduction of economic inefficiencies and increase in high-skilled jobs. There are also significant risks that must be managed — through both technical design and policy-making instruments— to maximize these benefits for any given society while protecting its important ethical values.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2649858 2020-03-30T13:25:24-04:00 2020-03-30T13:25:24-04:00

<p>According to Irina Orssich, team leader AI at the unit Technologies and Systems for Digitizing Industry at the European Commission’s DG CONNECT, the COVID-19 crisis has highlighted the importance of digital and how critical it is to society. The first priority should be to acknowledge and further explore how digital and AI in particular can contribute to the fight against the virus, such as with 3D printing of respiratory masks, or by saving scientists months of experimentation to find cures or treatments. Regarding the white paper on AI, the Commission is conducting a public consultation until May 31, and does not currently foresee an extension to this deadline. This could change depending on how the crisis evolves. The Commission will hold stakeholders meetings, public hearings, and targeted consultations with experts, including online, to assess the impact of legal provisions on various stakeholders. A large public consultation event will be organized mid-September. The Commission would still present proposals for a regulatory framework in December. In parallel, there will be an update of the coordinated action plan with member states, focusing on national AI strategies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2646380 2020-03-25T09:49:00-04:00 2020-03-25T09:49:00-04:00

<p>The very idea of AI might create a diversion that makes it easier for a small group of technologists and investors to claim all rewards from a widely distributed effort. Computation is an essential technology, but the AI way of thinking about it can be murky and dysfunctional.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2645717 2020-03-24T13:12:45-04:00 2020-03-24T13:12:45-04:00

<p>Here is my claim. Steering technology markets back towards the interests of democracy with changes to law and regulation could help turn around the negative trends we face across the board. If the collective impact of digital media on politics is to weaken democratic outcomes, then it follows that if we remove this force (or alter its trajectory) we’ll do better. This should hold true for all of the problems we face. Although we may achieve only incremental progress, it will be systemic.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2644799 2020-03-23T10:20:28-04:00 2020-03-23T10:20:28-04:00

<p>It is possible that the AI prowess of China and the United States in fact stood in their way. Both have a technocratic, top-down vision of the future of AI, in which a small digital elite, concentrated in a few tech hubs and largely separated from the concerns of the rest of the population, produces tools meant to be used by the rest of the population. While the locus of this elite is the Communist Party in one case and West Coast tech hubs in the other, the logic is similar.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2641736 2020-03-18T17:10:20-04:00 2020-03-18T17:10:20-04:00

<p>Since 2016, more than 80 AI ethics documents – including codes, principles, frameworks, and policy strategies – have been produced by corporations, governments, and NGOs. In this paper, we examine three topics of importance related to our ongoing empirical study of ethics and policy issues in these emerging documents. First, we review possible challenges associated with the relative homogeneity of the documents’ creators. Second, we provide a novel typology of motivations to characterize both obvious and less obvious goals of the documents. Third, we discuss the varied impacts these documents may have on the AI governance landscape, including what factors are relevant to assessing whether a given document is likely to be successful in achieving its goals.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2641483 2020-03-18T11:17:55-04:00 2020-03-30T10:21:46-04:00

<p>… academia focuses too much on outputs. Only a fraction of the academic knowledge that reaches society can be attributed to a single output, such as a patent or a report. Most forms of impact are not attributable to a single output; they are based on diffuse processes. For example, scientists advise politics and business; they initiate public discussions. How do you measure the deliberative impact of science? Not by counting patents but rather by enabling young researchers to communicate their results effectively. We avoid complexity by pretending that impact manifests in outputs.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2640567 2020-03-17T11:51:46-04:00 2020-03-31T16:49:37-04:00

<p>This report, the first in a two-part series, articulates the connection between surveillance-based business models and the health of democracy. Drawing from Ranking Digital Rights’s extensive research on corporate policies and digital rights, we examine two overarching types of algorithms, give examples of how these technologies are used both to propagate and prohibit different forms of online speech (including targeted ads), and show how they can cause or catalyze social harm, particularly in the context of the 2020 U.S. election. We also highlight what we don’t know about these systems, and call on companies to be much more transparent about how they work.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2640767 2020-03-17T15:03:44-04:00 2020-03-17T15:03:44-04:00

<p>The African continent will be forced to face unique challenges with negative AI. Weaker governance structures, transitioning democratic norms, and highly salient ethnic and religious divisions will likely struggle to minimize the damage associated with negative AI. This requires strategies that are inclusive of African researchers and experts who can recognize both threats and opportunities for positive use. Intergovernmental organizations should seek to partner with AI entrepreneurs and corporate actors in Africa while simultaneously providing training to government officials in AI and digital forensics. If officials can recognize the nuanced nature of AI use, they may be better able to recognize and support that technology with positive use in mind. Additionally, the effective use of digital forensics may help governmental institutions more quickly identify fabricated media so as to remove inflammatory content and deescalate violence after exposing the fraudulent content.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2637684 2020-03-13T11:37:01-04:00 2020-03-13T11:37:01-04:00

<p>Through this directory, you will be able to browse algorithms with a proven social impact by typing free text or using the preexistent filters defined as implemented location, kind of discrimination produced and/or impacted sector. It pretends to be a useful informational tool about algorithms that can help to raise awareness around current algorithms implementation practices, so feel free to contact us for any improvement you may have.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2634289 2020-03-09T16:43:53-04:00 2020-03-09T16:43:53-04:00

<p>Heavily centralized, hierarchical, hyper-capitalistic, market based solutions were much more effective in extracting and redistributing value from the shared, distributed resources, than the communities that produced them under the naive dream of building a "commons". If the rise of fascism is the failure of the left, then the rise of the platform economy, and the “sharing economy” was the failure of the commons-based peer production movement, because it refused to think about value.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2627151 2020-02-28T15:44:55-05:00 2020-02-28T15:44:55-05:00

<p>UN High-Level Report – discussing Architectures for global digital Cooperation</p>

tag:tagteam.harvard.edu,2005:FeedItem/2624288 2020-02-24T14:03:59-05:00 2020-02-24T14:03:59-05:00

<p>The White Paper is the European Commission’s first concrete attempt at discussing AI policy beyond the high-level statements of previous Communications. In this sense, the Commission takes up a rule setting role (rather than a referee role). In our opinion, this is a good first step. If we were to draw the analogy with a game, independently of who is playing the game, without rules no one wins. Moreover, the potential impact of AI, both positive and negative, is too large to be left outside of democratic oversight. While the ideas of the Commission need further elaboration and depth, the true the leap forward would be not only to focus on “Trustworthy AI made in Europe” as an alternative to AI made by the existing tech giants, but to promote trustworthy AI as a competitive advantage and incentivize and invest in the institutions, research and frameworks that can set this new AI playing field.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2622650 2020-02-21T11:08:50-05:00 2020-02-21T11:08:50-05:00

<p>Ethical considerations and governance approaches of AI are at a crossroads. Either one tries to convey the impression that one can bring back a status quo ante of our given "onlife"-era [1,2], or one accepts to get responsibly involved in a digital world in which informational self-determination can no longer be safeguarded and fostered through the old fashioned data protection principles of informed consent, purpose limitation and data economy [3,4,6]. The main focus of the talk is on how under the given conditions of AI and machine learning, data sovereignty (interpreted as controllability [not control (!)] of the data subject over the use of her data throughout the entire data processing cycle [5]) can be strengthened without hindering innovation dynamics of digital economy and social cohesion of fully digitized societies. In order to put this approach into practice the talk combines a presentation of the concept of data sovereignty put forward by the German Ethics Council [3] with recent research trends in effectively applying the AI ethics principles of explainability and enforceability [4-9].</p>

tag:tagteam.harvard.edu,2005:FeedItem/2619399 2020-02-18T12:04:20-05:00 2020-02-18T12:04:20-05:00

<p>The Convention's Committee underlines that the protection of human rights, including the right to protection of personal data, should be an essential pre-requisite when developing or adopting AI applications, in particular when they are used in decision-making processes, and be based on the principles of the updated data protection convention, Convention 108 +, opened for signature in October 2018. In addition, any innovation in the field of AI should pay close attention to avoiding and mitigating the potential risks of processing of personal data, and should allow meaningful control by data subjects over the data processing and its effects. These guidelines about AI refer to important issues previously identified in the Guidelines on the Protection of Individuals with regard to the Processing of Personal Data in a World of Big Data and to the necessity “to secure the protection of personal autonomy based on a person’s right to control his or her personal data and the processing of such data, the nature of this right to control should be carefully addressed” in this context.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2619303 2020-02-18T09:44:23-05:00 2020-02-18T09:44:23-05:00

<p>The Seven Principles of Public Life (also known as the Nolan Principles) apply to anyone who works as a public office-holder. This includes all those who are elected or appointed to public office, nationally and locally, and all people appointed to work in the Civil Service, local government, the police, courts and probation services, non-departmental public bodies (NDPBs), and in the health, education, social and care services. All public office-holders are both servants of the public and stewards of public resources. The principles also apply to all those in other sectors delivering public services.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2616643 2020-02-14T10:25:35-05:00 2020-02-14T10:25:35-05:00

<p>With national innovation strategies focused primarily on achieving dominance in artificial intelligence, the problem of actually regulating AI applications has received less attention. Fortunately, cities and other local jurisdictions are picking up the baton and conducting policy experiments that will yield lessons for everyone.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2616642 2020-02-14T10:25:20-05:00 2020-02-14T10:25:20-05:00

<p>Deep learning techniques are growing in popularity within the field of artificial intelligence (AI). These approaches identify patterns in large scale datasets, and make classifications and predictions, which have been celebrated as more accurate than those of humans. But for a number of reasons, including nonlinear path from inputs to outputs, there is a dearth of theory that can explain why deep learning techniques work so well at pattern detection and prediction. Claims about “superhuman” accuracy and insight, paired with the inability to fully explain how these results are produced, form a discourse about AI that we call enchanted determinism. To analyze enchanted determinism, we situate it within a broader epistemological diagnosis of modernity: Max Weber’s theory of disenchantment. Deep learning occupies an ambiguous position in this framework. On one hand, it represents a complex form of technological calculation and prediction, phenomena Weber associated with disenchantment. On the other hand, both deep learning experts and observers deploy enchanted, magical discourses to describe these systems’ uninterpretable mechanisms and counter-intuitive behavior. The combination of predictive accuracy and mysterious or unexplainable properties results in myth-making about deep learning’s transcendent, superhuman capacities, especially when it is applied in social settings. We analyze how discourses of magical deep learning produce techno-optimism, drawing on case studies from game-playing, adversarial examples, and attempts to infer sexual orientation from facial images. Enchantment shields the creators of these systems from accountability while its deterministic, calculative power intensifies social processes of classification and control.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2616640 2020-02-14T10:25:02-05:00 2020-02-14T10:25:02-05:00

<p>The moral authority of ethics codes stems from an assumption that they serve a unified society, yet this ignores the political aspects of any shared resource. The sociologist Howard S. Becker challenged researchers to clarify their power and responsibility in the classic essay: Whose Side Are We On. Building on Becker's hierarchy of credibility, we report on a critical discourse analysis of data ethics codes and emerging conceptualizations of beneficence, or the "social good", of data technology. The analysis revealed that ethics codes from corporations and professional associations conflated consumers with society and were largely silent on agency. Interviews with community organizers about social change in the digital era supplement the analysis, surfacing the limits of technical solutions to concerns of marginalized communities. Given evidence that highlights the gulf between the documents and lived experiences, we argue that ethics codes that elevate consumers may simultaneously subordinate the needs of vulnerable populations. Understanding contested digital resources is central to the emerging field of public interest technology. We introduce the concept of digital differential vulnerability to explain disproportionate exposures to harm within data technology and suggest recommendations for future ethics codes.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2614852 2020-02-12T13:54:23-05:00 2020-02-12T13:54:23-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2613432 2020-02-10T14:41:59-05:00 2020-02-10T14:41:59-05:00

<p>But there is a less well-recognized but equally important application for AI in helping to detect misinformation and limit its spread. This dual role will be particularly important in geopolitics, which is closely tied to how governments shape and react to public opinion both within and beyond their borders. And it is important for another reason as well: While nation-state interest in information is certainly not new, the incorporation of AI into the information ecosystem is set to accelerate as machine learning and related technologies experience continued advances.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2613343 2020-02-10T12:41:56-05:00 2020-02-10T12:41:56-05:00

<p>The Office of the Privacy Commissioner of Canada (OPC) is currently engaged in legislative reform policy analysis of both federal privacy laws. We are examining artificial intelligence (AI) as a subset of this work as it relates specifically to the Personal Information Protection and Electronic Documents Act (PIPEDA). We are of the view that PIPEDA falls short in its application to AI systems and we have identified several areas where PIPEDA could be enhanced. We are seeking to consult with experts in the field to validate our understanding of how privacy principles should apply and whether our proposals would be consistent with the responsible development and deployment of these systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2611858 2020-02-07T15:01:50-05:00 2020-02-07T15:02:03-05:00

<p>In 2018, for example, the German Chancellor’s Office, working in collaboration with Harvard University's Berkman Center for Internet and Society and the New York University Governance Lab, set up an “annotation platform” and invited dozens of experts—leading computer science, law and ethics scholars—to comment online on the draft of its artificial intelligence policy, providing much-needed perspectives that informed the final draft. Imagine if, instead of oversight hearings with a handful of experts, congressional committees or federal agencies invited a hundred experts to review a draft rule online?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2610306 2020-02-05T12:59:08-05:00 2020-02-05T12:59:08-05:00

<p>Fifty-two floors below the top of Salesforce Tower, I meet Paula Goldman in a glass-paneled conference room where the words EQUALITY OFFICE are spelled out on a patchwork bunting banner, the kind of decoration you might buy for a child's birthday party. Goldman has a master's degree from Princeton and a Ph.D. from Harvard, where she studied how controversial ideas become mainstream. She arrived at Salesforce just over a year ago to become its first-ever Chief Ethical and Humane Use Officer, taking on an unprecedented and decidedly ambiguous title that was created specifically for her unprecedented, ambiguous, yet highly specific job: see to it that Salesforce makes the world better, not worse.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2609677 2020-02-04T13:54:51-05:00 2020-02-04T13:54:51-05:00

<p>Postdoctoral Research Fellows in AI and Sustainable Development</p>

tag:tagteam.harvard.edu,2005:FeedItem/2609672 2020-02-04T13:15:15-05:00 2020-02-04T13:15:15-05:00

<p>AI education for end users must provide leaders with a firm grasp of the basic underlying principles of AI and a grounded sense of its development trajectory. Akin to the combustion engine or electricity, AI is an enabling technology that will augment currently existing systems and enable new functions, rather than a singular piece of technology with a specific set of uses. Recognizing the general-purpose character of AI will help policymakers avoid the logical fallacy that AI is a sort of widget to bolt on to an existing capability to give it autonomous power. Coupled with recognizing the limits of algorithms, framing AI in this manner will allow less-technical security leaders to start to better envision where AI can fit into existing institutions and systems, and where it would be most useful to integrate or introduce. Finally, it is essential for U.S. leaders to be aware of the status of AI development in other countries, as well as AI-centric policies and regulations, funding mechanisms, organizations, educational initiatives, human capital systems, and best practices that other countries have employed to cultivate their own nascent AI programs.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2602466 2020-01-23T13:37:03-05:00 2020-01-23T13:37:03-05:00

<p>The emergence of artificial intelligence (AI) and its progressively wider impact on many sectors requires an assessment of its effect on the achievement of the Sustainable Development Goals. Using a consensus-based expert elicitation process, we find that AI can enable the accomplishment of 134 targets across all the goals, but it may also inhibit 59 targets. However, current research foci overlook important aspects. The fast development of AI needs to be supported by the necessary regulatory insight and oversight for AI-based technologies to enable sustainable development. Failure to do so could result in gaps in transparency, safety, and ethical standards.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2602369 2020-01-23T12:36:52-05:00 2020-01-23T12:36:52-05:00

<p>Artificial intelligence (AI) is a transformative technology that offers both great benefits and poses a number of challenges. Effective governance is important to ensure that the ethical concerns are met. James Butcher and Irakli Beridze provide an overview of current AI governance activities, covering notable examples across the private sector, public sector, research and multi-stakeholder organisations, and the UN. They take a deeper look at regulation, then reflect on the future of AI governance, touching on some frameworks that may offer guidance for the beneficial development of AI technologies while mitigating the risks.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2602360 2020-01-23T12:22:18-05:00 2020-01-23T12:22:18-05:00

<p>What’s next for the News Provenance Project? The Times says its next phase “will shift from exploration to execution to show how an end-to-end solution can help users share trusted news with confidence. The team will explore other potential solutions in emerging technology and work across newsrooms and industry to imagine what a better internet might look like.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2601481 2020-01-22T10:42:56-05:00 2020-01-22T10:42:56-05:00

<p>Principles are admirable and can help communicate a company’s commitments to citizens and consumers. But it’s past time to move from principles to policy. Requiring disclosure — as appropriate based on use-case and end-user — should be the default expectation for many companies creating, distributing, or commercializing AI systems. In an earlier Policy Lab essay, we articulated a disclosure requirement for law enforcement use-cases of facial recognition technology. Something similar should be required of AI more generally in order to provide the public with appropriate assurances that they are being treated fairly and equitably by AI-based determinations in sensitive use-cases.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2601469 2020-01-22T09:45:14-05:00 2020-01-22T09:45:14-05:00

<p>The first index to benchmark nations on their level of investment, innovation and implementation of artificial intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2598108 2020-01-16T16:10:25-05:00 2020-01-16T16:10:25-05:00

<p>The 'robot rights' debate, and its related question of 'robot responsibility', invokes some of the most polarized positions in AI ethics. While some advocate for granting robots rights on a par with human beings, others, in a stark opposition argue that robots are not deserving of rights but are objects that should be our slaves. Grounded in post-Cartesian philosophical foundations, we argue not just to deny robots 'rights', but to deny that robots, as artifacts emerging out of and mediating human being, are the kinds of things that could be granted rights in the first place. Once we see robots as mediators of human being, we can understand how the `robots rights' debate is focused on first world problems, at the expense of urgent ethical concerns, such as machine bias, machine elicited human labour exploitation, and erosion of privacy all impacting society's least privileged individuals. We conclude that, if human being is our starting point and human welfare is the primary concern, the negative impacts emerging from machinic systems, as well as the lack of taking responsibility by people designing, selling and deploying such machines, remains the most pressing ethical discussion in AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2592713 2020-01-09T10:22:38-05:00 2020-01-09T10:22:38-05:00

<p>Even if code is modified with the aim of securing procedural fairness, however, we are left with the deeper philosophical and political issue of whether neutrality constitutes fairness in background conditions of pervasive inequality and structural injustice. Purportedly neutral solutions in the context of widespread injustice risk further entrenching existing injustices. As many critics have pointed out, even if algorithms themselves achieve some sort of neutrality in themselves, the data that these algorithms learn from is still riddled with prejudice. In short, the data we have—and thus the data that gets fed into the algorithm—is neither the data we need nor the data we deserve. Thus, the cure for algorithmic bias may not be more, or better, algorithms. There may be some machine learning systems that should not be deployed in the first place, no matter how much we can optimize them.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2591124 2020-01-07T12:34:11-05:00 2020-01-07T12:34:11-05:00

<p>The principles also promote a light-touch regulatory approach. The White House is directing federal agencies to avoid preemptive, burdensome or duplicative rules that would needlessly hamper AI innovation and growth. Agencies will be required to conduct risk assessments and cost-benefit analyses prior to regulatory action to evaluate the potential tradeoffs of regulating a given AI technology. Given the pace at which AI will continue to evolve, agencies will need to establish flexible frameworks that allow for rapid change and updates across sectors, rather than one-size-fits-all regulations. Automated vehicles, drones, and AI-powered medical devices all call for vastly different regulatory considerations.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2582404 2019-12-20T17:03:24-05:00 2019-12-20T17:03:24-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2580478 2019-12-18T13:46:41-05:00 2019-12-18T13:46:41-05:00

<p>rtificial intelligence: Human rights, social justice and development</p>

tag:tagteam.harvard.edu,2005:FeedItem/2579478 2019-12-17T13:15:32-05:00 2019-12-17T13:15:32-05:00

<p>Taking up Schumacher's call, this article considers the possibility of “small” digital platforms in the context of China's contemporary ride‐hailing industry. Through an ethnographic study of V Taxi, a grassroots community of taxi drivers, it explores how a human‐centric platform for transportation services emerged, developed, and survived in a rapidly shifting economic and political landscape. The beauty and resilience of this platform lie in its ability to harness local social networks, enabling individual actors to share economic opportunities, skills, and knowledge. Through these practices of sharing, V Taxi members produce relational value, which strengthens community bonds. Drawing upon analytic frameworks from economic anthropology in conjunction with insights from science and technology studies, this article argues that “small” digital platforms can empower economic actors by enabling the production of relational value, thus protecting human agency and expanding human capabilities.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2579357 2019-12-17T11:37:40-05:00 2019-12-17T11:37:40-05:00

<p>One reason for the persistent bias is that Facebook’s modified algorithm appears to rely on proxy characteristics that correlate with age or gender, said Alan Mislove, a Northeastern University professor of computer science and one of the study’s co-authors. “Our research shows that simply removing a few protected features from an algorithm is unlikely to provide any meaningful protection against discrimination,” Mislove said. If the advertiser provides a sample audience of software engineers, for example, that might be considered a proxy for male profiles. Facebook’s ad delivery algorithm further skews the audience based on the content of the ad itself, the researchers previously found. As a result, even when advertisers try to reach a diverse audience, they aren’t always able to. Dolese’s ad, for example, could have reached a predominantly male audience because it featured a man, or because an interest in trucking acts as a proxy for maleness, or both. (A Dolese spokeswoman said the ad targeted categories “that would appeal to someone in this line of work.”) The settlement did not resolve the potential bias from proxies and ad content, but said Facebook would study the issue.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2575739 2019-12-12T14:52:52-05:00 2019-12-12T14:52:52-05:00

<p>A recent normative turn in computer science has brought concerns about fairness, bias, and accountability to the core of the field. Yet recent scholarship has warned that much of this technical work treats problematic features of the status quo as fixed, and fails to address deeper patterns of injustice and inequality. While acknowledging these critiques, we posit that computational research has valuable roles to play in addressing social problems --- roles whose value can be recognized even from a perspective that aspires toward fundamental social change. In this paper, we articulate four such roles, through an analysis that considers the opportunities as well as the significant risks inherent in such work. Computing research can serve as a diagnostic, helping us to understand and measure social problems with precision and clarity. As a formalizer, computing shapes how social problems are explicitly defined --- changing how those problems, and possible responses to them, are understood. Computing serves as rebuttal when it illuminates the boundaries of what is possible through technical means. And computing acts as synecdoche when it makes long-standing social problems newly salient in the public eye. We offer these paths forward as modalities that leverage the particular strengths of computational work in the service of social change, without overclaiming computing's capacity to solve social problems on its own.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2575738 2019-12-12T14:52:23-05:00 2019-12-12T14:52:23-05:00

<p>An expert on artificial intelligence has called for all algorithms that make life-changing decisions – in areas from job applications to immigration into the UK – to be halted immediately. Prof Noel Sharkey, who is also a leading figure in a global campaign against “killer robots”, said algorithms were so “infected with biases” that their decision-making processes could not be fair or trusted. A moratorium must be imposed on all “life-changing decision-making algorithms” in Britain, he said. 'We are hurtling towards a surveillance state’: the rise of facial recognition technology Read more Sharkey has suggested testing AI decision-making machines in the same way as new pharmaceutical drugs are vigorously checked before they are allowed on to the market.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2575695 2019-12-12T14:04:50-05:00 2019-12-12T14:04:50-05:00

<p>The AI Index Report tracks, collates, distills, and visualizes data relating to artificial intelligence. Its mission is to provide unbiased, rigorous, and comprehensive data for policymakers, researchers, journalists, executives, and the general public to develop a deeper understanding of the complex field of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2575613 2019-12-12T12:43:32-05:00 2019-12-12T12:43:32-05:00

<p>TikTok, a social network video app with more than 1 billion downloads globally, admitted Tuesday to a set of policies that had suppressed the reach of content created by users assumed to be “vulnerable to cyberbullying.” As examples of users “susceptible to bullying or harassment,” the policy listed people with facial disfigurement, autism, Down syndrome, and “Disabled people or people with some facial problems such as birthmark, slight squint and etc.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2575611 2019-12-12T12:23:47-05:00 2019-12-12T12:23:47-05:00

<p>To date, all stakeholders are working intensively on policy design for artificial intelligence. All initiatives center around the requirement that AI algorithms should be fair. But what exactly does it mean? And how algorithmic fairness can be translated to legal and policy terms? These are the main questions that this paper aims to explore. Each discipline approaches those questions differently. While computer scientists may favor one notion of fairness over others across the board, this paper argues in favor of a case-by-case analysis and application of the relevant fairness notion. The paper discusses the legal limitations of the computer science (CS) notions of fairness and suggests a typology of matching each CS notions to its corresponding legal mechanism. The paper concludes that fairness is contextual. The fact that each notion, or a group of notions correspond with a different legal mechanism, make them suitable for a certain policy domain more than others. Thus, throughout the paper, examples for possible applicability of the CS notions to some policy domains will be introduced. In addition, the paper will highlight for both developers and policy makers the practical steps that need to be taken in order to better address algorithmic fairness.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2574723 2019-12-11T12:19:34-05:00 2019-12-11T12:19:34-05:00

<p>Many legal scholars have explored how courts can apply legal doctrines, such as procedural due process and equal protection, directly to government actors when those actors deploy artificial intelligence (AI) systems. But very little attention has been given to how courts should hold private vendors of these technologies accountable when the government uses their AI tools in ways that violate the law. This is a concerning gap, given that governments are turning to third-party vendors with increasing frequency to provide the algorithmic architectures for public services, including welfare benefits and criminal risk assessments. As such, when challenged, many state governments have disclaimed any knowledge or ability to understand, explain, or remedy problems created by AI systems that they have procured from third parties. The general position has been “we cannot be responsible for something we don’t understand.” This means that algorithmic systems are contributing to the process of government decisionmaking without any mechanisms of accountability or liability. They fall within an accountability gap. In response, we argue that courts should adopt a version of the state action doctrine to apply to vendors who supply AI systems for government decisionmaking. Analyzing the state action doctrine’s public function, compulsion, and joint participation tests, we argue that—much like other private actors who perform traditional core government functions at the behest of the state—developers of AI systems that directly influence government decisions should be found to be state actors for purposes of constitutional liability. This is a necessary step, we suggest, to bridge the current AI accountability gap.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2570526 2019-12-05T14:24:05-05:00 2019-12-05T14:24:05-05:00

<p>SEX LEX MACHINA: INTIMACY AND ARTIFICIAL INTELLIGENCE Jeannie Suk Gersen* Share on Twitter Share on Facebook SEX LEX MACHINA: INTIMACY AND ARTIFICIAL INTELLIGENCE Sex robots are here. Created specifically to allow individuals to simulate erotic and romantic experiences with a seemingly alive and present human being, sex robots will soon force lawmakers to address the rise of digisexuality and the human–robot relationship. The extent to which intimacy between a human and robot can be regulated depends on how we characterize sex with robots—as a masturbatory act, an in­timate relationship, or nonconsensual sexual contact—and whether sexual activity with robots makes us see robots as more human or less human. A robot sex panic may be driven primarily by the idea that ro­bots are servile by nature. Critics argue that an inherently nonrecipro­cal dynamic between humans and robots will translate into exploitative relationships that may fuel abuse of human partners, or that sex robots may further social isolation and retreat from human intimacy. Con­versely, sex robots may function as safe—and otherwise unavailable—sexual and emotional outlets for those who may otherwise harm others. They may even train individuals to be more respectful in human rela­tionships. At this point, we do not know how our relationships with ro­bots will inform our relationships with humans, for better or for worse. This Essay explores the consequences of sex robots on society and argues that questions of how sex robots will improve or worsen humans’ treat­ment of one another is the key to regulation to come. What is clear is that sex robots will require us to grapple with our vulnerabilities in re­lation­ships, reconsider fundamental rights, and question what it means to be intimate and to be human.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2568959 2019-12-03T16:59:01-05:00 2019-12-03T16:59:01-05:00

<p>Advances in artificial intelligence and its increasing application in more spheres of life demand attention to the impact of algorithms on human well-being. It is little surprise, then, that calls for sector-specific regulation or the creation of a federal agency or commission to guide and constrain AI development are increasing. Such a turn to administrative law is understandable because AI’s regulatory challenges seem facially similar to those in other technocratic domains, such as the pharmaceutical industry or environmental law. But when it comes to AI, an “FDA for algorithms” or a federal robotics commission is of limited utility in constructing enduring governance solutions. AI is unique, even if it is not different in kind. AI’s distinctiveness comes from technical attributes (speed, complexity, and unpredictability) that strain traditional administrative law tactics, in conjunction with the institutional settings and incentives, or strategic context, that affect its development path. This Article assesses algorithmic governance strategies in light of the attributes and institutional factors that make AI unique. In addition to technical attributes and the contemporary imbalance of public and private resources and expertise, AI governance must contend with a fundamental conceptual challenge: algorithmic applications permit seemingly technical decisions to de facto regulate human behavior, with a greater potential for physical and social impact than ever before. Applying this point, this Article warns that the current trajectory of AI development, which is dominated by large private firms, augurs an era of private governance. Without rethinking regulatory strategies to ensure that public values inform AI research, development, and deployment, we risk losing the democratic accountability that is at the heart of public law.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2568951 2019-12-03T16:56:07-05:00 2019-12-03T16:56:07-05:00

<p>The UK’s Data Protection Authority just issued much-anticipated guidance that clarifies the complicated issue of the GDPR’s ‘right to explanation’. Here is some background on the issue and what the new information means.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2568857 2019-12-03T15:36:12-05:00 2019-12-03T15:36:12-05:00

<p>Most risk managers in financial services say they aren’t capable of assessing the risks associated with artificial intelligence, according to a survey from Accenture</p>

tag:tagteam.harvard.edu,2005:FeedItem/2568819 2019-12-03T14:48:37-05:00 2019-12-03T14:48:37-05:00

<p>Despite widespread enthusiasm among computer scientists to contribute to “socialgood,” the field’s efforts to promote good lack a rigorous foundation in politicsor social change. There is limited discourse regarding what “good” actuallyentails, and instead a reliance on vague notions of what aspects of society aregood or bad. Moreover, the field rarely considers the types of social changethat result from algorithmic interventions, instead following a “greedy algorithm”approach of pursuing technology-centric incremental reform at all points. In orderto reason well about doing good, computer scientists must reflexively evaluatetheir normative commitments, consider the long-term impacts of technologicalinterventions, evaluate algorithmic interventions against alternative reforms, and nolonger prioritize technical considerations as superior to other forms of knowledge</p>

tag:tagteam.harvard.edu,2005:FeedItem/2567631 2019-12-02T10:17:07-05:00 2019-12-02T10:17:07-05:00

<p>Significant differences exist between medicine and AI development that suggest a principled approach for the latter may not enjoy success comparable to the former. Compared to medicine, AI development lacks (1) common aims and fiduciary duties, (2) professional history and norms, (3) proven methods to translate principles into practice, and (4) robust legal and professional accountability mechanisms. These differences suggest we should not yet celebrate consensus around high-level principles that hide deep political and normative disagreement.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2567619 2019-12-02T09:30:08-05:00 2019-12-02T09:30:08-05:00

<p>This edition of Global Information Society Watch (GISWatch) provides a perspective from the global South on the application of AI to our everyday lives. It includes 40 country reports from countries as diverse as Benin, Argentina, India, Russia and Ukraine, as well as three regional reports. These are framed by eight thematic reports dealing with topics such as data governance, food sovereignty, AI in the workplace, and so-called “killer robots”.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2564042 2019-11-26T15:29:48-05:00 2019-11-26T15:32:50-05:00

<p>Rather than providing a thoughtful critique of how specific systems succeed and fail, the document gives a passing reference to an array of concerns, ranging from bias, to funding, to regulatory burden. And thus died this first valiant effort at municipal algorithmic accountability.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2564009 2019-11-26T12:35:56-05:00 2019-11-26T12:35:56-05:00

<p>Mark Zuckerberg completed his 2019 “personal challenge,” in which he set out to “host a series of public discussions about the future of technology in society — the opportunities, the challenges, the hopes, and the anxieties.” Here’s who he spoke to: 8 men, 1 woman. All white.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2563329 2019-11-25T15:51:03-05:00 2019-11-25T15:51:03-05:00

<p>Whether it’s knowing the nutritional content in our food, the conditions of our roads, or a medication’s interaction warnings, we rely on information to make responsible decisions. But what about AI? Despite its potential to transform so much of the way we work and live, machine learning models are often distributed without a clear understanding of how they function. For example, under what conditions does the model perform best and most consistently? Does it have blind spots? If so, where? Traditionally, such questions have been surprisingly difficult to answer.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2563328 2019-11-25T15:47:30-05:00 2019-11-25T15:47:30-05:00

<p>Which systems really deserve to be built? Which problems most need to be tackled? Who is best placed to build them? And who decides? We need genuine accountability mechanisms, external to companies and accessible to populations. Any A.I. system that is integrated into people’s lives must be capable of contest, account, and redress to citizens and representatives of the public interest. While the first wave of algorithmic accountability focuses on improving existing systems, a second wave of research has asked whether they should be used at all—and, if so, who gets to govern them.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2563327 2019-11-25T15:38:27-05:00 2019-11-25T15:38:27-05:00

<p>Over the past six months, we’ve partnered with Nesta to explore some of these questions. In the “Foundation Horizon Scan,” unveiled at an event today with sector leaders, we take the long view to explore the future of philanthropic giving. In compiling the report, we reviewed relevant literature and spoke to over 30 foundation leaders and critics internationally to understand what the challenges to foundations’ legitimacy and impact mean in practice and how foundations are responding to them today.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2563266 2019-11-25T13:51:47-05:00 2019-11-25T13:51:47-05:00

<p>Based on the ideas put forward in the book "Towards a Global Framework for CyberPeace and Digital Cooperation: An Agenda for the 2020s" (download full draft) stakeholders from across the globe and from civil society, private sector and public sector are deliberating and bootstrapping on an equal footing how to create #NextGenIG. The IGF 2019 in Berlin is an ideal starting point for discussions on #NextGenIG with a view towards the 75th Anniversary of the UN (UNGA in 2020) and WSIS+20 possibly becoming a major Internet Governance event in 2025.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2561498 2019-11-22T10:43:02-05:00 2019-11-25T16:01:24-05:00

<p><span>As of October 2019, some 30 indus-</span><span>trialized and middle-income countries </span><span>have developed plans, incentives, and </span><span>policies to stimulate national, rather </span><span>than international, AI research and </span><span>adoption. In general these countries </span><span>have firms developing AI systems as </span><span>well as rules governing the use of some </span><span>types of data such as personal data.</span><span>(See Table 1) But most countries are </span><span>in the early stages of learning how to </span><span>govern data, let alone data-driven tech</span><span>-</span><span>nologies.</span><span> Nonetheless, the people of </span><span>the world won’t be able to reap the ben-</span><span>efits of AI unless policymakers work at </span><span>the national and international levels to </span><span>create an effective enabling and regula-</span><span>tory environment for AI.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2562050 2019-11-22T22:07:05-05:00 2019-11-25T12:12:16-05:00 Gennie Gebhart

<p>Twitter, Google, and Facebook are trying approaches across the spectrum: from Twitter’s combined ban and limited microtargeting, to Google’s limited microtargeting and fact-checking, to Facebook’s explicit exception for political ads from its content policies. It’s not worth comparing which one is better or worse, because each of these different policies has the potential to worsen the problem it is trying to solve.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2561101 2019-11-21T15:58:30-05:00 2019-11-21T15:58:30-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2560241 2019-11-20T14:22:34-05:00 2019-11-20T14:22:34-05:00

<p>Big idea here. 👇🏼 Machine learning outperforms judges at predicting recidivism, but only because judges have multiple objectives -- they don't want to incarcerate young people.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2560240 2019-11-20T14:18:25-05:00 2019-11-20T14:18:25-05:00

<p>This research consists of a desk study of how countries are approaching algorithmic regulation, starting with the Library of Congress’ Regulation of Artificial Intelligence in Selected Jurisdictions.7A general search on algorithmic regulation, regulation of AI, and personal data protection laws that may contain regulations for the use of algorithms in making decisions was also performed. Sector-specific regulations that cover technologies like autonomous vehicles or weapons systems were not considered within the scope of this research</p>

tag:tagteam.harvard.edu,2005:FeedItem/2560215 2019-11-20T13:00:08-05:00 2019-11-20T13:00:08-05:00

<p>As the effort began to pick up steam at Facebook, however, it bumped into a series of problems. Privacy experts working for and with Facebook, along with other privacy experts working with Social Science One, Facebook attorneys and the firm’s policy and technical experts advised the company that it did not have systems that could easily give researchers around the world access to specific data sets without exposing other data about or risking identification of users: Even without names or user IDs, intricate data can still reveal the identities of the people within them.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2560214 2019-11-20T12:50:14-05:00 2019-11-20T12:50:14-05:00

<p>Still, he stresses that strong internet governance—and new rules—needn’t divide people along predictable liberal-conservative lines. “What’s interesting about public health as an analogy is, the fact that it became a problem at a public scale didn’t mean that the solution was exclusively public. We still have a private healthcare system. It just meant that at that moment, there were questions that we had to ask as a society, management questions to which companies and hospitals and insurers and government, doctors, patients, et cetera, were all party.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2559473 2019-11-19T12:07:41-05:00 2019-11-19T12:07:48-05:00

<p>T H E S T A T E O F H A T E F U LA C T I V I T I E S O N T H E P L A T F O R M</p>

tag:tagteam.harvard.edu,2005:FeedItem/2558946 2019-11-18T16:31:21-05:00 2019-11-18T16:31:21-05:00

<p>Despite its reprehensible refusal to fact-check ads, Facebook has gone the farthest in ad transparency: A couple of clicks reveals the date an ad launched and all the active ads from that advertiser. But self-regulation is not enough. Better would be to require this through regulation, aimed at both platform and client. Platforms must build these advertiser archives, and keep and display the accompanying data. They must go to reasonable lengths to ensure that advertisers are who they say they are; anyone evading these restrictions (e.g., buying ads under different names) should no longer be allowed to advertise on the platform. Advertisers misrepresenting themselves should face penalties. An “Honest Ads Act” for all advertising, across the digital ecosystem—for several reasons.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2558942 2019-11-18T16:27:17-05:00 2019-11-18T16:27:17-05:00

<p>Intel isn’t alone. Within the last few years, a number of tech companies, from Google to Huawei, have launched their own programs under the AI for Good banner. They deploy technologies like machine-learning algorithms to address critical issues like crime, poverty, hunger, and disease. In May, French president Emmanuel Macron invited about 60 leaders of AI-driven companies, like Facebook’s Mark Zuckerberg, to a Tech for Good Summit in Paris. The same month, the United Nations in Geneva hosted its third annual AI for Global Good Summit sponsored by XPrize. (Disclosure: I have spoken at it twice.) A recent McKinsey report on AI for Social Good provides an analysis of 160 current cases claiming to use AI to address the world’s most pressing and intractable problems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2556602 2019-11-14T16:11:23-05:00 2019-11-14T16:11:23-05:00

<p>But the assessment tool she was using, a mouthful called the Vulnerability Index — Service Prioritization Decision Assistance Tool, or VI-SPDAT for short, regularly prioritized white clients. Post, a community health manager at the nonprofit Neighborhood House, and 11 other homeless service providers signed a letter to the board of All Home, King County’s coordinating agency for homelessness services, this past spring and put their concerns in no uncertain terms.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2556479 2019-11-14T13:30:37-05:00 2019-11-14T13:30:37-05:00

<p><span>Throughout our internal and external scan, </span><span>our </span><span>working group was struck by 1) the enormous </span><span>potential and need to aggressively address and incorporate research and teaching on the societal </span><span>implications of computing and 2) the profound cultural and structura</span><span>l inertia and</span><span>, </span><span>at times</span><span>, </span><span>barriers </span><span>preventing precisely this kind of work at MIT and elsewhere.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2555906 2019-11-13T16:38:21-05:00 2019-11-13T16:38:21-05:00

<p>BERT is one of a number of A.I. systems that learn from lots and lots of digitized information, as varied as old books, Wikipedia entries and news articles. Decades and even centuries of biases — along with a few new ones — are probably baked into all that material. BERT and its peers are more likely to associate men with computer programming, for example, and generally don’t give women enough credit. One program decided almost everything written about President Trump was negative, even if the actual content was flattering.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2555783 2019-11-13T13:57:16-05:00 2019-11-13T13:57:16-05:00

<p>The point of regulating social media is to create incentives for social media companies to become responsible and trustworthy institutions that will help foster a healthy and vibrant digital public sphere. It is equally important to ensure that there are a large number of different kinds of social media companies, with diverse affordances, value systems, and innovations. Treating social media companies as state actors or as public utilities does not solve the problems of the digital public sphere. One might create a public option for social media services, but this, too, cannot serve as a general solution to the problems that social media create. Instead, this essay describes three policy levers that might create better incentives for privately-owned companies: (1) antitrust and competition law; (2) privacy and consumer protection law; and (3) a careful balance of intermediary liability and intermediary immunity rules.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2555733 2019-11-13T10:55:52-05:00 2019-11-13T10:55:52-05:00

<p><span>United Nations Activities on Artificial Intelligence (AI) 2019</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2554802 2019-11-12T10:18:04-05:00 2019-11-12T10:18:04-05:00

<p>In these and other cases, judges must understand the role that AI and machine learning play in the legal system itself. Lawyers, as well as judges and their staff, use machine learning to improve case law searches for relevant legal authority to cite in briefs and decisions. Document production and technology-assisted reviews use AI to search for relevant documents to produce and to mine those documents for the information most important to a party’s claims without attorneys’ having to review every document. Some scholars and practitioners are already using AI to predict the outcome of cases based on algorithms based on tens of thousands of prior cases. Recent research suggests that such outcome predictions may have around a 70% accuracy rate. AI is ushering in a new era of quantitative legal decision forecasting.[2]</p>

tag:tagteam.harvard.edu,2005:FeedItem/2554079 2019-11-11T09:58:59-05:00 2019-12-23T11:58:56-05:00

<p>Similarly, Facebook workers have called on CEO Mark Zuckerberg to fact-check or ban political advertising, while employees at Microsoft and GitHub called for the end of a contract with ICE, the Department of Homeland Security agency that carries out Trump deportation and detention policies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2551981 2019-11-07T16:59:42-05:00 2019-11-07T16:59:42-05:00

<p>In the 20 years that I’ve spent studying how communities in the global South use technology, II've found that these are the real innovators, not just passive users of Western or Chinese tech. They are also makers, creators, and hackers. And they are guided by values far removed from the self-serving corporate strategies that govern the Facebooks, Googles, Facebooks, and Amazons of the world, which ensure that users sit at the bottom of the tech production chain with outstretched, hopeful hands. While Western companies focus on keeping our attention at all costs, gathering our personal data, and taking us down behavioral rabbit holes, in the global South the users are seizing control.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2551784 2019-11-07T14:14:12-05:00 2019-11-07T14:14:12-05:00

<p>We’ve worked on identifying a gap in the evaluation metrics that are widely used to measure performance, and go on to propose a human-algorithmic interaction metric called we call a barrier-to-exit. It aims to be a proxy for quantifying the ability of the AI model to recognize and allow for a change in user preference. See the full paper here — https://arxiv.org/abs/1909.06713</p>

tag:tagteam.harvard.edu,2005:FeedItem/2550914 2019-11-06T11:22:37-05:00 2019-11-06T11:22:37-05:00

<p>This is the first in a series of papers about delivery-driven policy, which builds on the overarching concept of delivery-driven government with a focus on policymaking and advocacy. In it, we explain the key disconnect between policymakers and government delivery, introduce delivery-driven policy as a concept, and offer examples of delivery-driven policymaking—informed both by our own work and that of others in the field. In collaboration with New America and the Delivering Better Outcomes Working Group hosted by the Beeck Center for Social Impact and Innovation at Georgetown University, with whom we’ve developed these ideas, we invite all readers to share their own examples of delivery-driven policy at resources@codeforamerica.org. In subsequent papers, we will go deeper into some of the examples introduced here, highlight new examples submitted by our wider community, and expand on these ideas as they relate to adjacent fields—including advocacy, procurement, and digital talent in government. You’ll find these on the delivery-driven government page of Code for America’s website.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2550745 2019-11-06T08:13:55-05:00 2019-11-06T11:14:09-05:00

<p>Steinmeier warned that German and American companies are likely to employ the same practice in the future, making it necessary to examine the morality of the digital age. “That brings me to the question of whether we really need a more intensive exchange between the tech community and the political scientists about the philosophy of the individual. I would like to leave Boston today having received the confirmation from all of you that I need not be afraid, that the debate is taking place.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2550108 2019-11-05T10:25:29-05:00 2019-11-05T10:25:29-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2548213 2019-11-01T16:19:41-04:00 2019-11-01T16:19:41-04:00

<p>The RSA’s Forum for Ethical AI ran a citizens’ jury to explore the use of AI in decision-making. We convened participants to grapple with the ethical issues raised by this application of AI under different circumstances and enter into a deliberative dialogue about how companies, organisations, and public institutions should respond. This report presents a toolkit for organisations seeking to deploy their own ethical processes around the proliferation of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2548212 2019-11-01T15:50:06-04:00 2019-11-01T15:50:06-04:00

<p>I will have more to say about Macron’s initiative and what it means for Canada, but this post focuses on the Global Forum itself. Global discussions on AI policy are absolutely essential. However, as the title to this post suggests, this Global Forum was not global. Rather, the overwhelmingly European and North American crowd was strikingly unrepresentative. The most obvious example occurred after the panel on AI and bias, which featured several presentations on bias in data sets that are often used in the development of AI systems. These included Françoise Soulié-Fogelman, who cited a well-known study that facial recognition programs work well with white males (1% error rate), but not for minority women (35% error rate); Kate Crawford, who emphasized the diversity crisis in AI, noting that women are poorly represented in major labs at companies such as Google and Facebook; and Laurence Devillers, who discussed how AI systems can reinforce stereotypes with 80% of coders being male and 80% of chatbots female.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2546923 2019-10-30T17:06:09-04:00 2019-10-30T17:06:09-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2545489 2019-10-28T16:57:27-04:00 2019-10-28T16:57:27-04:00

<p>But over and above this practical consideration, I contend that there are principled reasons for abandoning anonymity. In this era, thousands of organizations—tiny and gigantic—are attempting to survive by raising funds, and many individuals of wealth have motives for supporting such organizations. Sometimes those motives are idealistic (I believe in a free college education, and I want to support deserving youth who can’t afford tuition); sometimes those motives are dubious (I am involved with fossil fuels, and I want to support individuals who cast doubt on the consensus around climate change for personal benefit); but almost always, courtesy of “the human condition,” motives are mixed—as, indeed, are the motives for anonymity.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2545487 2019-10-28T16:38:21-04:00 2019-10-28T16:38:21-04:00

<p>Importantly, the principles also recommend a range of other actions by policy makers that seek to move governance upstream, including governments using experimentation to provide controlled environments for the testing of AI systems. Such environments could include regulatory sandboxes, innovation centres and policy labs. Policy experiments can operate in “start-up mode.” In this case, experiments are deployed, evaluated and modified, and then scaled up or down, or abandoned quickly. Another option to spur faster and more effective decisions is the use of digital tools to design policy, including innovation policy, and to monitor policy targets. For instance, some governments use “agent-based modelling” to anticipate the impact of policy variants on different types of businesses. Governments can also encourage AI actors to develop self-regulatory mechanisms, such as codes of conduct, voluntary standards and best practices. These can help guide AI actors through the AI life cycle, including for monitoring, reporting, assessing and addressing harmful effects or misuse of AI systems. Finally, governments can also establish and encourage public and private sector oversight mechanisms of AI systems, as appropriate. These could include compliance reviews, audits, conformity assessments and certification schemes.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2545461 2019-10-28T16:01:58-04:00 2019-10-28T22:25:39-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2545459 2019-10-28T15:48:15-04:00 2019-10-28T15:48:15-04:00

<p>For geeks (like me) who LIVE for understanding the ways different groups define different words, this glossary by Sara Jordan &amp; team @IEEEorg is FANTASTIC. All the AI words as defined by engineers, lawyers, philosophers and more https://standards.ieee.org/content/dam/ie</p>

tag:tagteam.harvard.edu,2005:FeedItem/2545404 2019-10-28T15:19:23-04:00 2019-10-28T15:19:23-04:00

<p>The companies will also work on "data trusts," a new, proposed technological solution born from AI to measure and maintain data control, which may become key as AI works its way into data collection solutions. Data trusts are third-party stewardship models based on "common law trust." These tools, as documented in an Element AI whitepaper, are proposed as a way to give individuals more control over their personal information; to balance power and data rights between companies, governments, and individuals; to enhance privacy, and to give the public the opportunity to "share in the value of data and artificial intelligence."</p>

tag:tagteam.harvard.edu,2005:FeedItem/2545322 2019-10-28T14:18:59-04:00 2019-11-07T09:53:05-05:00

<p>Joan Donovan and Jon Penney</p> <p>Google, Facebook and Amazon serve billions of users around the globe and increasingly perform core functions in society. The private gains are obvious — these are among the most profitable companies in history. But they come with a cost: platforms threaten our social fabric, our economy and our democracy. To begin to address this, CIGI has convened leading thinkers to explore new models for platform governance. Read the series below or watch the video.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2543774 2019-10-25T11:44:35-04:00 2019-10-25T11:44:35-04:00

<p>The charter, which is open for public consultation, sets out 10 points that agencies would have to adhere to. These include pledging to explain how significant decisions are informed by algorithms or, where it cannot – for national security reasons, for example – explain the reason; taking into account the perspectives of communities, such as LGBTQI+, Pacific islanders and people with disabilities; and identifying and consulting with groups or stakeholders with an interest in algorithm development.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2543114 2019-10-24T15:42:58-04:00 2019-10-24T15:42:58-04:00

<p>“There are tons of cases. In many of them, the government has to abandon the system, like in the Medicaid allocation system cases, because those were class actions. They’re now in the process of trying to figure out how to fix the system, because now they’ve like spent so much money in it not working, then getting sued, and now like they’re obliged by court order to fix it,” Richardson said. “And they’re kind of like, ‘We don’t know how to fix it,’ or they’re realizing that these may not be problems that are best solved using some type of automated decision system, and it’s you only find that out after millions of dollars are spent and a few people die.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2542932 2019-10-24T12:54:28-04:00 2019-10-28T15:01:45-04:00 Rebecca Crootof

<p>It is heartening that AI researchers are working to better understand the range of harms. Some researchers are delineating different kinds of potential accidents and their associated risks. Some are identifying risks from malicious actors, ranging from individuals engaging in criminal activity and harassment, to industry exploiting users, to states and others engaged in social and political disruption. Still others are focused on how AI may create long-term, less obvious “structural risks”—shifts to social, political and economic structures that have negative effects—such as destabilizing the nuclear deterrence regime. Many AI governance principles suggest that researchers attempt to minimize and mitigate these risks throughout the AI lifecycle. Meanwhile, in the wake of AI advances, multistate and supranational bodies, individual states, industry actors, professional organizations and civil society groups have been churning out ethical principles for AI governance.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2541354 2019-10-22T15:49:02-04:00 2019-10-22T15:49:02-04:00

<p>In the second part of the paper, we challenge the mainstream legal position that the use of a protected characteristic always violates the Equal Protection Clause. We first develop a statistical framework that formalizes exactly how the direct and proxy effects of race can lead to algorithmic predictions that disadvantage minorities relative to non-minorities. While an overly formalistic legal solution requires exclusion of race and all potential non-race correlates, we show that this type of algorithm is unlikely to work in practice because nearly all algorithmic inputs are correlated with race. We then show that there are two simple statistical solutions that can eliminate the direct and proxy effects of race, and which are implementable even when all inputs are correlated with race. We argue that our proposed algorithms uphold the principles of the Equal Protection doctrine because they ensure that individuals are not treated differently on the basis of membership in a protected class, in stark contrast to commonly-used algorithms that unfairly disadvantage minorities relative to non-minorities despite the exclusion of race.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2541198 2019-10-22T12:35:49-04:00 2019-10-22T12:35:49-04:00

<p>For example, IDDP researchers will use our university’s expanding data infrastructure to track the spread of disinformation—stories, memes, videos, photos—across the internet, including on platforms such as Twitter, Facebook, Reddit and the Russian social network VK. Researchers will be able to detect the origins and patterns of propaganda and learn how it is amplified, informing strategies to stop it. With this and other research and scholarship, such as timely publications about the role of online content related to current events, IDDP will contribute to public policy debates. And it will help train journalists to better understand the misinformation landscape this election season, which we plan to do in collaboration with the Poynter Institute and its PolitiFact initiative.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2537422 2019-10-17T11:37:51-04:00 2019-10-17T11:37:51-04:00

<p>Abu Dhabi on Wednesday announced the opening of the world’s first dedicated AI university – Mohammad Bin Zayed University of Artificial Intelligence (MBZUAI). Located in Masdar City with the latest state-of-the-art facilities and equipment, the university will offer both masters (two years) and PhD programmes (four years) for local and international graduate students across three main specialised fields – machine learning, computer vision and natural language processing – as the UAE looks to equip the next generation of students with the latest expertise in the field of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2536595 2019-10-16T13:04:52-04:00 2019-10-16T13:04:52-04:00

<p><span>The </span><a href="https://www.newamerica.org/public-interest-technology/university-network/">Public Interest Technology University Network</a><span> today announced the award of $3.1 million in grants to 21 colleges and universities to support new and existing programs designed to train engineers, policymakers, and social justice advocates to use technology for the public good. Awarded through the network’s inaugural “Network Challenge,” the grants will support the development of new public interest technology initiatives and institutions in academia, and foster collaboration among the network’s partner institutions.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2535680 2019-10-15T13:13:06-04:00 2019-10-15T13:13:06-04:00

<p>“It’s (going to) help bring all of these efforts and initiatives to more cohesion,” Mangold said. “The interdisciplinary is what’s needed, agnostic of discipline, agnostic of technology.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2532531 2019-10-10T13:54:36-04:00 2019-10-10T13:54:36-04:00

<p>Although it is very difficult to peer into the inner workings of a neural network’s processes, the team could easily audit the data it was generating. And with a little experimentation, they found that the CycleGAN had indeed pulled a fast one. The intention was for the agent to be able to interpret the features of either type of map and match them to the correct features of the other. But what the agent was actually being graded on (among other things) was how close an aerial map was to the original, and the clarity of the street map.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2531021 2019-10-08T15:41:58-04:00 2019-10-08T15:41:58-04:00

<p>The rise of social media platforms has produced novel security threats and vulnerabilities. Malicious actors can now exploit entanglements of once disparate technical and social systems to target exposed communities. These exploits pose a challenge to legacy security frameworks drawn from technical and state-based conceptions of referent objects and bounded information systems. In this paper we propose a new framework of analysis to meet this challenge, Sociotechnical Security (STsec), which acknowledges how the interplay between actors produces emergent threats to participant communities. This exploratory paper offers an overview of sociotechnical systems, explains why these threats and vulnerabilities require us to expand our understanding of security with regards to participatory technology, and how sociotechnical security can be operationalized as a framework for analysis.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2527132 2019-10-03T10:24:25-04:00 2019-10-03T10:24:25-04:00

<p>The workshop brought together a multidisciplinary group of scholars to consider (a) barriers to legitimate and effective algorithmic governance and (b) the research methods needed to address the nature and impact of specific barriers. An interactive management workshop technique was used to harness the collective intelligence of this multidisciplinary group. This method enabled participants to produce a framework and research agenda for those who are concerned about algorithmic governance. We outline this research agenda below, providing a detailed map of key research themes, questions and methods that our workshop felt ought to be pursued. This builds upon existing work on research agendas for critical algorithm studies in a unique way through the method of collective intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2527103 2019-10-03T10:01:27-04:00 2019-10-03T10:01:27-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2526305 2019-10-02T10:33:28-04:00 2019-10-02T10:33:28-04:00

<p>For their part, Franks and Citron are working on a federal bill to criminalize malicious deepfakes, which they call “digital forgeries.” To their thinking, a digital forgery is something that a reasonable person would think is real. It must also be likely to cause harm to a particular person or to public order—for example, if a video falsely showed a Muslim person committing a crime.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2525667 2019-10-01T13:33:13-04:00 2019-10-01T13:33:13-04:00

<p>A Benchmark Dataset for Learning to Intervene in Online Hate Speech</p>

tag:tagteam.harvard.edu,2005:FeedItem/2525665 2019-10-01T13:32:58-04:00 2019-10-01T13:32:58-04:00

<p>Although policymakers, commentators, and the mass media have paid growing attention to algorithmic bias and the shortcomings of machine learning and artificial intelligence, the algorithmic divide has yet to attract much policy and scholarly attention. To fill this lacuna, this article draws on the digital divide literature to systematically analyze this new inequitable gap between the technology haves and have-nots. Utilizing the analytical framework that the Author developed in the early 2000s, the article discusses the five attributes of the algorithmic divide: awareness, access, affordability, availability, and adaptability.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2519311 2019-09-23T15:00:29-04:00 2019-09-23T15:00:29-04:00

<p><strong>A group of researchers is calling for a way to measure and standardize how energy-intensive AI systems are.</strong><span><span> </span>In<span> </span></span><a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__u5080173.ct.sendgrid.net_wf_click-3Fupn-3Dvh9lyFdXNdtD4GLMrKHUIWJ-2D2BoGaZNxucLfoCqKe-2D2BJBymBg5jwKfEUaAD5XoflUu0-5FEvue9s2vKaD3k5UA2ksz2Wxxl3uP-2D2BSc31NPXsBgZj-2D2BaFmUwAeroLyHvfHrrjaK-2D2FaPbMMLPR4SXkYM4NXohEgbD-2D2B9P8xbqAZq-2D2BWjF6I0W6s13b-2D2FRhMI9WTB3v-2D2B40dzOpvUT7jK7HCtj0rqaXo1GUBoSPFp5dWbIoeTK51K3djC7bpTgiTXK9jW-2D2FBYEJv1P8qp7E9EFS9VRLn9vwtfTKUlifDdzco0z4seMho4vp3OwUZOspTaK7PbWJ2-2D2Bf7HWNurxtilVIfTRrsYsE18MmJvinE6dmXQJx0MkJEyt0H-2D2FfxKTOwnEk9nN6371HKCXbvUjS7QaTQF6wBgkzGlw0tU9WD5lewckyG4OliRRgXO1QvosYkKhvoKyiXvbeas-2D2FnjwTa6yVlQMxycfkXK7KyX8Db8Q-2D3D-2D3D&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=Qjk2L3tVGmU460B73RIT8bWVgfbkFCt2qrNH4AHdfMo&amp;s=96QGuM40Dbm8WBb2kEf2qBEUOuzH3ugsTrAvZc7gIB8&amp;e=">a paper</a><span>, the researchers — Roy Schwartz from the Allen Institute for Artificial Intelligence, Jesse Dodge from Carnegie Mellon University, Noah Smith from the University of Washington, and Oren Etzioni from the Allen Institute—make a case that future AI research should include measurements about energy use, which could lead to innovations that make AI less carbon-intensive. Schwartz spoke with Slate's April Glaser about the need to focus on energy efficiency in addition to accuracy, as people "don’t want to reach a state where AI will become a significant contributor to global warming,” he said. -<span> </span></span><strong><a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__u5080173.ct.sendgrid.net_wf_click-3Fupn-3Dvh9lyFdXNdtD4GLMrKHUIR8OmExqLnkY0Ztywo-2D2BuBYtF4HuajVr0gHukVwnXzhKa3x5gZ1VKDoWlrlEVdAyxDDsxbxlsHPo1DMLVu9JbqI46fjlWkgGAY8Q3NvmyTzH3SAZ6oFuFsWH7xy-2D2F8R1vqsUBPGhhodLiDsvp9lg7Lw2k-2D3D-5FEvue9s2vKaD3k5UA2ksz2Wxxl3uP-2D2BSc31NPXsBgZj-2D2BaFmUwAeroLyHvfHrrjaK-2D2FaPbMMLPR4SXkYM4NXohEgbD-2D2B9P8xbqAZq-2D2BWjF6I0W6s13b-2D2FRhMI9WTB3v-2D2B40dzOpvUT7jK7HCtj0rqaXo1GUBoSPFp5dWbIoeTK51K3djC7bpTgiTXK9jW-2D2FBYEJv1P8qp7E9EFS9VRLn9vwtfTKUlifDdzco0z4seMho4vp3OwUZOspTaK7PbWJ2-2D2Bf7HWNurxpPL-2D2BKw8fuMZ8UEPRulQO61tuCphoo6fdT6dPZkFv7PCTj0m2pBMuw2PNMJMhA6eUSawOyNSxFgWSgf-2D2FvpFpfUB8VYq-2D2BJqQdczVnCa5XOD0FEjSDVj0qnxI5cQEnGJATBqLRSpReyxR7bMGGjoup56g-2D3D-2D3D&amp;d=DwMFaQ&amp;c=WO-RGvefibhHBZq3fL85hQ&amp;r=zR5UBNtKTrqE3nNRLMXAH5MBO806eezpozU6LX9kmbQ&amp;m=Qjk2L3tVGmU460B73RIT8bWVgfbkFCt2qrNH4AHdfMo&amp;s=70ZgzwSQGQjIC4DOwdaHn3gp80g2623pSKUrBmD1xds&amp;e=">SLATE</a></strong></p>

tag:tagteam.harvard.edu,2005:FeedItem/2516089 2019-09-19T09:53:32-04:00 2019-09-19T09:53:32-04:00

<p>Young people in Africa are rising above infrastructure and resource constraints to create ingenious processes to adopt and apply the fourth industrial revolution technologies like artificial intelligence (AI) to generate more value for their localities. Since we founded the Alliance for Africa’s Intelligence (Alliance4ai) one year ago, we have interacted with more than 100 AI startups to learn about their work and support them with a platform where they exchange knowledge and opportunities with students, schools and other players in the ecosystem.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2516082 2019-09-19T09:34:11-04:00 2019-09-19T09:34:11-04:00

<p>The continuous interaction between intelligent devices, sensors and people points to the increasing number of data being produced, stored and processed, changing, in various aspects and increasingly, our daily life. This increasing connectivity and symbiotic interaction among humans and intelligent machines brings significant challenges for the rule of law and contemporary ethics. Do machines have morality? What legal liability regime should we adopt for damages arising from increasingly advanced artificial intelligence? Which ethical guideline should we adopt to orient its advancement? In this paper we will discuss the main normative and ethical challenges imposed by the advancement of artificial intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2516080 2019-09-19T09:32:35-04:00 2019-09-19T09:32:35-04:00

<p>There is further risk of inflating the number of apparent studies in the field of artificial intelligence and knowledge work. By publishing large numbers of journal articles which draw data or insights from popular sources, there is an illusion of a bigger body of original scholarship in the field than actually exists, despite a low volume of original surveys, interviews, or similar.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2515381 2019-09-18T10:53:24-04:00 2019-09-18T10:53:24-04:00

<p>I don’t know exactly what I was expecting the machine to tell me about myself, but I wasn’t expecting what I got: a new version of my official Guardian headshot, labeled in neon green print: “gook, slant-eye”. Below the photo, my label was helpfully defined as “a disparaging term for an Asian person (especially for North Vietnamese soldiers in the Vietnam War)”.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2515378 2019-09-18T10:46:18-04:00 2019-09-18T10:46:18-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2515302 2019-09-18T10:07:43-04:00 2019-09-18T10:07:43-04:00

<p>This handbook covers 20 evidence-based tools, designed to help you bring bold ideas to life to change lives from within government. This includes tools that uncover new insights and generate ideas, to tools for developing and testing, or tools that support system change.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2514575 2019-09-17T10:44:14-04:00 2019-09-17T10:44:14-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2512264 2019-09-13T14:51:53-04:00 2019-09-13T14:51:53-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2512190 2019-09-13T13:29:15-04:00 2019-09-13T13:29:15-04:00

<p>We believe the most effective way to promote systems-level social change is to place the voices of people who use social services—the people for whom change matters most—at the center of change processes. But while many philanthropic organizations tout the importance of listening to the “end beneficiaries” or “service users,” the practice nevertheless remains an underutilized methodology for countering systemic obstacles to change and, ultimately, reforming complex systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2512189 2019-09-13T13:26:09-04:00 2019-09-13T13:26:09-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2512022 2019-09-13T10:13:39-04:00 2019-09-13T10:13:39-04:00

<p>To that end, on Thursday researchers at Harvard Kennedy School’s Belfer Center for Science and International Affairs published a detailed examination of this congressional expertise shortfall. Drawing on more than 140 interviews with current and former members of Congress, their staff, lobbyists, civil-society groups, and think-tank experts,the researchers proposed various options to close the knowledge gap, such as hiring and developing more staffers with the necessary backgrounds and producing a more regular flow of accessible briefing material.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2510222 2019-09-11T09:52:43-04:00 2019-09-11T09:52:43-04:00

<p>The final version of Russia’s national artificial-intelligence strategy isn’t due until next month, but a draft prepared by the country’s largest bank appears to focus on science, academia, and healthcare. It also outlines a regulatory framework for AI development.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2506819 2019-09-06T13:44:46-04:00 2019-09-06T13:44:46-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2505922 2019-09-05T13:18:06-04:00 2019-09-05T13:18:06-04:00

<p>This Steering Committee builds upon the essential work of an Expert Group we launched last year focused on early investigation of these topics. Co-Chaired by WITNESS and Facebook, the group is comprised of a broad coalition of organizations convened from across the Partnership, and is focused on pressing issues at the intersection of AI and the media, including on mis- and dis-information and synthetically-generated media. Together, our community has been working to promote a more global conversation on these issues; to help news organizations and platforms better understand how to authenticate content and enhance detection capabilities; to understand and bolster the research landscape related to new technical capabilities in media production and synthetic media detection; to build coordination capacity across newsrooms and the broader media landscape; and to amplify potential implications of media preparedness issues for under-represented or vulnerable communities. The new Steering Committee aims to further galvanize institutions to make commitments to share more information, resources, and expertise toward these goals.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2505921 2019-09-05T13:09:09-04:00 2019-09-05T13:09:09-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2505881 2019-09-05T11:54:13-04:00 2019-09-05T11:54:13-04:00

<p>Facebook Inc. FB +1.81% has promised regulators that it will protect users’ privacy and give them the right to take their data where they choose. Now the company is trying to figure out how it can do both. In a white paper issued Wednesday, the social-media giant outlines how those two big mandates are potentially in conflict. The company said it is seeking outside input on how to deliver on both objectives and get a consensus on how the industry will handle such matters going forward.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2504911 2019-09-04T11:29:07-04:00 2019-09-04T11:29:07-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2499988 2019-08-28T14:14:42-04:00 2019-08-28T14:14:42-04:00

<p>To perform the membership inference attacks, we leverage the existing inference methods that exploit model predictions. We also propose two new inference methods that exploit structural properties of robust models on adversarially perturbed data. Our experimental evaluation demonstrates that compared with the natural training (undefended) approach, adversarial defense methods can indeed increase the target model’s risk against membership inference attacks. When using adversarial defenses to train the robust models, the membership inference advantage increases by up to 4.5 times compared to the naturally undefended models. Beyond revealing the privacy risks of adversarial defenses, we further investigate the factors, such as model capacity, that influence the membership information leakage</p>

tag:tagteam.harvard.edu,2005:FeedItem/2499987 2019-08-28T13:53:52-04:00 2019-08-28T13:53:52-04:00

<p>On Wednesday, Pinterest announced a new step in its efforts to combat health misinformation on its platform: users will be able to search for 200 terms related to vaccines, but the results displayed will come from major public health organizations, including the World Health Organization (WHO), Centers for Disease Control, American Academy of Pediatrics (AAP) and Vaccine Safety Net. <span>The platform will also bar advertisements, recommendations and comments on those pages.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2499936 2019-08-28T12:42:39-04:00 2019-08-28T12:42:39-04:00

<p>Given the rising concern and criticism surrounding the lack of diversity in the tech sector, how do prestigious and influential STEM establishments like MIT compare when we dig into the actual numbers?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2498845 2019-08-27T09:57:23-04:00 2019-08-27T09:57:23-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2498844 2019-08-27T09:50:35-04:00 2019-08-27T09:50:35-04:00

<p>Experts from the Future of Privacy Forum and the data management platform Immuta say the industry has yet to come up with a standard way to assess AI risks.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2498199 2019-08-26T12:39:17-04:00 2019-08-26T12:39:17-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2494633 2019-08-21T08:55:35-04:00 2019-08-26T16:36:17-04:00

<p>France and Canada are establishing an international committee to advise on the ethics of artificial intelligence. The group should be supported and shielded from undue influence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2489077 2019-08-20T14:02:57-04:00 2019-08-20T14:02:57-04:00

<p>Texas Attorney General Ken Paxton, a Republican, put out a statement after last month’s meeting saying that he and other attorneys general discussed “the real concerns consumers across the country have with big tech companies stifling competition on the internet.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488862 2019-08-16T15:18:09-04:00 2019-08-16T15:18:09-04:00

<p>Earlier this year, I negotiated a look behind the curtain that Silicon Valley’s wizards rarely grant. I made a meandering trip across India and stopped at a facility across the street from the Superdome in downtown New Orleans. In all, I visited five offices where people are doing the endlessly repetitive work needed to teach A.I. systems, all run by a company called iMerit. Editors’ Picks She Wanted a Man With a Good Job Who Is Nice to Animals He Gave a Bike to a Refugee Girl. 24 Years Later, She Got to Thank Him. When Did Self-Help Become Self-Care?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488861 2019-08-16T15:10:11-04:00 2019-08-16T15:10:11-04:00

<p>In reality, algorithms applied to malleable humans can have drastically different and pernicious side effects on a global scale. They modify our tastes to make us ever more predictable, edge us toward extremes, and ultimately erode civility and trust in society. It is time we stop blithely allowing this and create the digital equivalent of drug trials.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488710 2019-08-15T09:11:24-04:00 2019-08-15T09:11:24-04:00

<p>Canadian disinfo challenge grantees</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488665 2019-08-14T17:18:25-04:00 2019-08-14T17:18:25-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2488602 2019-08-14T09:24:18-04:00 2019-08-14T09:38:59-04:00

<p>"As art installations go, it is low key: a filing cabinet filled with meticulously labelled hanging folders. Visitors are welcome to browse under any heading that sparks their interest: publicly available gun trace data; the Nanjing massacre death toll; English language rules internalised by native speakers; how much Spotify pays each artist per play of song. The folders are all empty.</p> <p>The work, titled “<a href="http://mimionuoha.com/the-library-of-missing-datasets">The Library of Missing Datasets</a>”, is by Mimi Onuoha, an artist and adjunct professor at New York University. The aim, she says, is to expose the “blank spots in spaces that are otherwise saturated with data”. The blanks can reveal hidden biases in a society...."</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488414 2019-08-12T16:17:32-04:00 2019-08-12T16:17:32-04:00

<p>Machine learning is neither a panacea nor as technically daunting as some believe. When it comes to global development, the key is to ask the right questions, and then see if and how it can help.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488415 2019-08-12T16:17:45-04:00 2019-08-12T16:17:45-04:00

<p>TREND #2 – INCLUSIVENESS</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488412 2019-08-12T15:42:44-04:00 2019-08-12T15:42:44-04:00

<p>The Federal government should “commit to deeper, consistent, long-term engagement” in AI development and deployment, according to the plan, which focuses on nine identified areas of focus for AI standards, including:</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488396 2019-08-12T13:28:08-04:00 2019-08-12T13:28:08-04:00

<p>Support for this workshop is provided by the Omidyar Network - Tech and Society Solutions Lab</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488389 2019-08-12T12:49:24-04:00 2019-08-12T12:49:24-04:00

<p>Data Governance: Platforms play a central role in ADM because of the wealth of data they capture from users. Data is at the heart of monetization and governance processes. Since data is both input, and output for ADM, the way in which control over, and access to data is regulated forms a critical element of the question how to govern platforms, and how to realize rights and values in an ADM context.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488098 2019-08-08T11:00:15-04:00 2019-08-08T11:00:15-04:00

<p>The ethical implications and socialimpacts of artificial intelligence have become topics of compelling interest to industry, researchers in academia, and the public. However, current analyses of AI in a global context are biased toward perspectives held in the U.S., and limited by a lack of research, especially outside the U.S. and Western Europe.This article summarizes the key findings of a literature review of recent social science scholarship on the socialimpacts of AI and related technologies in five global regions. Our team of social science researchers reviewed more than 800 academic journal articles and monographsin over a dozen languages.Our review of the literature suggests that AI is likely to have markedly different socialimpacts depending on geographical setting. Likewise, perceptions and understandings of AI are likely to be profoundly shaped by local cultural and social context.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2488097 2019-08-08T10:51:50-04:00 2019-08-08T10:51:50-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2487784 2019-08-02T15:35:08-04:00 2019-08-02T15:35:08-04:00

<p>The initial report provided a summary of the group’s activities and its plans for the future but did not offer any recommendations to Congress.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2487736 2019-08-02T10:09:03-04:00 2019-08-02T10:09:03-04:00

<p>This project examined the potential that Artificial Intelligence (AI) technologies have in enhancing Australia’s wellbeing, lifting the economy, improving environmental sustainability and creating a more equitable, inclusive and fair society. Placing society at the core of AI development, the report analyses the opportunities, challenges and prospects that AI technologies present, and explores considerations such as workforce, education, human rights and our regulatory environment.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2487737 2019-08-02T10:09:48-04:00 2019-08-02T10:09:48-04:00

<p>As the world becomes more dependent on artificial intelligence, it has become clear that Turkey is in need of a national strategy to develop more advanced AI tech, which will be undertaken by the government, private sector and NGOs</p>

tag:tagteam.harvard.edu,2005:FeedItem/2487601 2019-07-31T16:23:10-04:00 2019-07-31T16:23:10-04:00

<p><span>Many have noted that a failure to center historically marginalized and underserved </span><span>communities risks defaulting to groups that have already been afforded significant </span><span>amounts of privilege. All too often, disadvantaged communities are an afterthought, </span><span>and ensuring that AI-driven solutions account for their needs is treated as an add-on to </span><span>or extension of the “main” problem. Communities that bear the burden of historical </span><span>discrimination are, unsurprisingly, also those most acutely harmed by the shortcom-</span><span>ings of AI. Without centering such communities, we risk repeating the same mistakes </span><span>of the past</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2487482 2019-07-30T14:15:39-04:00 2019-08-06T08:58:39-04:00

<p>One of the reasons philanthropists sometimes fail to measure what really matters is that the global political economy primarily seeks what is efficient and scalable. Unfortunately, efficiency and scalability are not the same as a healthy system. In fact, many things that grow quickly and without constraints are far from healthy—consider cancer. Because of our belief in markets, we tend to accept that an economy has to be growing for society to be healthy—but this notion is misguided, particularly when it comes to things we consider social goods. If we examine a complex system like the environment, for instance, we can see that healthy rainforests don’t grow in overall size but rather are extremely resilient, always changing and adapting.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2487388 2019-07-29T12:47:36-04:00 2019-07-29T12:47:36-04:00

<p>So what should we, as computer scientists, do instead? How can we avoid centering the algorithm and instead focus on helping people flourish, while at the same time allowing ourselves to be solution-driven? One idea that I’m becoming more and more convinced of is that, as Mitchell and Hutchinson argue in their FAT* 2019 paper, we should make the shift from thinking about fairness to thinking about (un)fairness.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486921 2019-07-22T15:11:16-04:00 2019-07-22T15:11:16-04:00

<p>My first publication as a grad student was a field experiment using Twitter “bots” to socially sanction users engaged in racist harassment. The ascendant paradigm in quantitative social science emphasizes the need for research to be “internally valid,” with a preference for randomized control trials like the one I conducted. That research project was well received, both by the political science discipline and the public, but I no longer believe that one-off field experiments are as valuable a tool for studying online behavior as I once did. The problem is that the knowledge they generate decays too rapidly (alternatively, that the realms in which it can be applied are too few) because the object of study changes too rapidly. I have been developing the concept of “temporal validity” as form of “external validity” that is particularly relevant to the study of social media. I suggest two avenues for producing more temporally valid research: (1) faster, more transparent publication (adapting the CS model of conference proceedings); (2) a “hollowing out” of empirical research, replacing medium-scale experimentation like my own work with either purely descriptive work (ethnographic or large-scale) or with massive, collaborative, replicable experimentation.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486909 2019-07-22T12:23:40-04:00 2019-07-22T12:23:40-04:00

<p>Mr. Nadella said Microsoft would not necessarily invest that billion dollars all at once. It could be doled out over the course of a decade or more. Microsoft is investing dollars that will be fed back into its own business, as OpenAI purchases computing power from the software giant, and the collaboration between the two companies could yield a wide array of technologies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486784 2019-07-19T15:55:07-04:00 2019-07-19T15:55:07-04:00

<p>The surprising commonality in principles expressed across governments, professional associations, research labs and private corporations suggests that we have the basis for an international treaty on AI. The recent interest and rapid progress suggest that we need to move swiftly from vague principles to concrete action.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486495 2019-07-16T13:08:50-04:00 2019-07-16T13:08:50-04:00

<p>When an organization announces to move from a non-profit to a capped-profit model, following the withholding of a contentious product, we should examine their motives.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486493 2019-07-16T12:26:29-04:00 2019-07-16T12:26:29-04:00

<p><span>In this paper, we focus on AI for </span><span>the SDGs. AI extends and amplifies </span><span>the capacity of human beings to </span><span>understand and solve complex, </span><span>dynamic, and interconnected </span><span>systems challenges like the SDGs. </span><span>Our main objective was to survey </span><span>the landscape of research and </span><span>initiatives on AI and the SDGs to </span><span>identify key themes and questions </span><span>in need of further exploration. We </span><span>also reviewed the state of AI and </span><span>the SDGs in two sectors – food </span><span>and agriculture and healthcare – to </span><span>understand if and how AI is being </span><span>deployed to address the SDGs and </span><span>the challenges and opportunities in </span><span>doing so.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2486486 2019-07-16T11:19:58-04:00 2019-07-16T11:19:58-04:00

<p>Evaluating whether a dataset is biased or a methodology is adequate depends on the context in which research takes place, and fundamentally on the goals of the researcher(s). To better grasp how data and methodological issues might affect or shape research outcomes, we first describe the prototypical goals (section 2.1) and classes of validity threats (section 2.2) to social data research. We then overview our framework to describe biases and pitfalls in social data analysis along an illustrative vignette (section 2.3), showing how they can compromise research validity and goals.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486485 2019-07-16T11:18:45-04:00 2019-07-16T11:18:45-04:00

<p>AI Ethics is now a global topic of discussion in academic and policy circles. At least 63 public-private initiatives have produced statements describing high-level principles, values, and other tenets to guide the ethical development, deployment, and governance of AI. According to recent meta-analyses, AI Ethics has seemingly converged on a set of principles that closely resemble the four classic principles of medical ethics. Despite the initial credibility granted to a principled approach to AI Ethics by the connection to principles in medical ethics, there are reasons to be concerned about its future impact on AI development and governance. Significant differences exist between medicine and AI development that suggest a principled approach in the latter may not enjoy success comparable to the former. Compared to medicine, AI development lacks (1) common aims and fiduciary duties, (2) professional history and norms, (3) proven methods to translate principles into practice, and (4) robust legal and professional accountability mechanisms. These differences suggest we should not yet celebrate consensus around high-level principles that hide deep political and normative disagreement.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486482 2019-07-16T10:38:40-04:00 2019-07-16T10:38:40-04:00

<p>Starting this month, Videgaray, an expert in geopolitics and AI policy, will serve as director of the MIT Artificial Intelligence Policy for the World Project (MIT AIPW), a collaboration between the MIT Sloan School of Management and the new MIT Stephen A. Schwarzman College of Computing. Videgaray will also serve as a senior lecturer at the MIT Sloan and as a distinguished fellow at the MIT Internet Policy Research Initiative.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2486144 2019-07-11T11:27:09-04:00 2019-07-11T11:27:09-04:00

<p>Nonetheless, given the countless technology related disasters and cautionary tales that the global tech-community is waking up to, there are numerous crucial lessons that African developers, start-ups and policy makers can learn from. The African continent need not go through its own disastrous cautionary tales to discover the dark side of digitisation and technologization of every aspect of life.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485927 2019-07-08T15:19:58-04:00 2019-07-08T15:19:58-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2485925 2019-07-08T14:27:52-04:00 2019-07-08T14:27:52-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2485865 2019-07-08T09:56:43-04:00 2019-07-08T10:55:37-04:00

<p>However, some feel that this overstates the threat of fake text. According to Yochai Benkler, co-head of the Berkman Klein Center for Internet &amp; Society at Harvard, the most damaging instances of fake news are written by political extremists and trolls, and tend to be about controversial topics that “trigger deep-seated hatred”, like election fraud or immigration. While a system like GPT-2 can produce semi-coherent articles at scale, it is a long way from being able to replicate this type of psychological manipulation. “The simple ability to generate false text at scale is not likely to affect most forms of disinformation,” he told me.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485875 2019-07-08T10:38:53-04:00 2019-07-08T10:38:53-04:00

<p>Guidance on building and using artificial intelligence in the public sector.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485873 2019-07-08T10:36:04-04:00 2019-07-08T10:36:04-04:00

<p>We have always been subject to social structures and our environment that have shaped who we are, what we want, and how we make decisions. What we need to pay heed to is how AI is destabilising not only our institutional-human relations and arrangements but also forming a critical part of our social environment (sometimes very sublimely, sometimes forcefully and pervasively). We need to make sure these profound changes do not happen without us.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485872 2019-07-08T10:33:26-04:00 2019-07-08T10:33:26-04:00

<p>Policymakers struggle to assess the ethical, legal and human rights impacts of IT systems in research, industry, and at home. At the same time, research needs to be useful for industry, academia, and society to have impact on policy. Right now, three European projects: PANELFIT, SHERPA and SIENNA, are working together with stakeholders to improve ethical, human rights and legal frameworks for information and communication technologies (ICT), big data analytics, artificial intelligence (AI) and robotics. Stakeholder involvement is key, and the outputs will support the European Union’s vision of Responsible Research and Innovation (RRI), as a means to foster the design of inclusive research and innovation. Here, we provide a short introduction to the projects and outline plans for collaboration with the aim to maximiseour joint policy impact.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485870 2019-07-08T10:26:11-04:00 2019-07-08T10:26:11-04:00

<p>In other words, this is a debate that we have been having, almost unchanged, for more than a century, whether we have been talking about chess, about facial image recognition, about neural networks, about spatial navigation, you name it: the argument over whether or not something thinks, what things should be classified as thought, what we mean by thought, and so on.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485541 2019-07-02T16:01:28-04:00 2019-07-02T16:01:28-04:00

<p>We expanded our work in 2014 following the Snowden revelations from a focus on open data to a dual focus on open data and privacy. Two years later, in the run-up to the Brexit referendum and US presidential election, we broadened our strategy to include work on the public scrutiny of automated decision-making systems. Our current strategy takes account of these changes by focusing explicitly on data rights and digital rights. We work to uphold data rights by focusing on three complementary areas: open data, privacy, and artificial intelligence. Our digital rights work focuses on freedom of expression issues such as hate speech, online surveillance of journalists, and net neutrality. Our overall goal is to support people and communities to exercise their human rights in the design and use of data and technology.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485453 2019-07-01T14:27:34-04:00 2019-07-01T14:27:34-04:00

<p>The goal of the research initiative he heads is to develop a global perspective on tech ethics. In his previous job, O’Brien observed that different societies around the world have very different perspectives on privacy and ethics. Just within Europe, for example, U.K. citizens are willing to tolerate video camera monitoring on London’s central High Street, perhaps because of IRA bombings of the past, while Germans are much more privacy oriented, influenced by the former intrusions of East German Stasi spies. O’Brien suggested that in China, the public is tolerant of AI-driven applications like facial recognition and social credit scores at least in part because social order is a key tenet of Confucian moral philosophy. Microsoft’s AI ethics research project involves ethnographic analysis of different cultures, gathered through close observation of behaviors, and advice from external academics such as Erin Meyer of INSEAD, who wrote The Culture Map. It may result in a number of publications, depending on the findings. My feeling, and O’Brien’s as well, is that there aren’t a lot of sources out there on global tech ethics. This research could start to fill that gap. Within Microsoft, O’Brien is trying to extend the community of people who are focused on the ethics topic. He sent me a collection of materials suggesting that lots of people are already engaged. The materials include, for example, a blog post on Microsoft guidelines about developing responsible conversation bots (no doubt occasioned by the off-the-rails comments from the Tay AI chatbot research project in 2016). They also include a presentation and article on intelligible models in health care, and an internal project for a “learning door” that recognizes (with opt-in) who is coming in and out of Microsoft buildings. O’Brien said he works closely with Smith’s legal team and also has a matrixed reporting relationship to Eric Horvitz, technical fellow and director at Microsoft Research Labs. Of course, Microsoft is a big company, but awareness of AI ethics seems to be catching on rapidly and broadly there.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485412 2019-07-01T12:01:26-04:00 2019-07-01T12:05:20-04:00

<p>In the last five years, private companies, research institutions as well as public sector organisations have issued principles and guidelines for ethical AI, yet there is debate about both what constitutes "ethical AI" and which ethical requirements, technical standards and best practices are needed for its realization. To investigate whether a global agreement on these questions is emerging, we mapped and analyzed the current corpus of principles and guidelines on ethical AI. Our results reveal a global convergence emerging around five ethical principles (transparency, justice and fairness, non-maleficence, responsibility and privacy), with substantive divergence in relation to how these principles are interpreted; why they are deemed important; what issue, domain or actors they pertain to; and how they should be implemented. Our findings highlight the importance of integrating guideline-development efforts with substantive ethical analysis and adequate implementation strategies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485413 2019-07-01T12:05:01-04:00 2019-07-01T12:05:01-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2485408 2019-07-01T10:44:07-04:00 2019-07-01T10:44:07-04:00

<p>In this report, we set out to capture a snapshot of the exponential progress in AI with a focus on developments in the past 12 months. Consider this report as a compilation of the most interesting things we’ve seen that seeks to trigger an informed conversation about the state of AI and its implication for the future. This edition builds on the inaugural State of AI Report 2018, which can be found here.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485247 2019-06-28T11:30:16-04:00 2019-06-28T11:30:16-04:00

<p>The is a Comment on the National Institute of Standards and Technology (NIST) Notice: Artificial Intelligence Standards</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485222 2019-06-28T09:28:36-04:00 2019-07-01T11:01:06-04:00

<p>“A lot of decisions that we make about how we feel about civil rights or gay marriage, whatever it is, those aren’t decisions that are optimizations,” he said. “Those are these cultural transformations that occur in society. And I think what’s missing right now in the conversation in AI is that that’s kind of the point of life, in my own personal opinion.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485151 2019-06-27T14:50:26-04:00 2019-06-27T14:50:26-04:00

<p>We should educate all—not just engineers—in how to become active technological citizens. This does not mean learning how to code. It means knowing enough to ask questions about how the code affects work, life, and community. It means understanding how people can hold institutions and firms accountable through literacy, advocacy, and social movements.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485073 2019-06-26T13:52:37-04:00 2019-06-26T13:52:37-04:00

<p>Fluxus Landscape: An Expansive View of AI Ethics and Governance is an art and research project by Şerife Wong, created in partnership with the Center for Advanced Study in the Behavioral Sciences (CASBS) at Stanford University with support from The Stanford Institute for Human-Centered Artificial Intelligence (HAI). The project maps and categorizes about 500 AI ethics and governance stakeholders and actors. Its goals are both practical and artistic: to help the global community interested in AI ethics and governance discover new organizations, and encourage a broader, more nuanced perspective on the AI ethics and governance landscape.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485034 2019-06-26T11:45:36-04:00 2019-06-26T11:45:36-04:00

<p>I provide nonpartisan advice on artificial intelligence on Capitol Hill. My job is to help policymakers ensure that technological progress is in the public interest. If someone were to make deepfake pornographic content of me, it would undermine public trust and derail my career. I do not have the resources to salvage my reputation, to place my attack in a larger social context and to recoup the lost income. This is why I need the government to regulate the spread of deepfakes — and so do the rest of us.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2485031 2019-06-26T10:02:52-04:00 2019-06-26T10:02:52-04:00

<p>So, how do you build “digital trust” and what does it look like? At the World Economic Forum, our new report provides a framework for a more efficient and effective global dialogue on digital trust built on two main components: mechanical trust and relational trust.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484896 2019-06-24T16:56:33-04:00 2019-06-24T16:56:33-04:00

<p>Clark and Hadfield’s model above only works if the regulators are both competitive and independent. Competitiveness — or the fact that no single party can have too large a market share — seems the most feasible. An oversight committee or a law can address this.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484867 2019-06-24T09:59:32-04:00 2019-06-24T09:59:32-04:00

<p>When the University of Oxford created my department, a multidisciplinary environment devoted to research and scholarship on the relationship between society and the internet — and now AI — it was prescient and brave. There could be an economic as well as a moral imperative for other bold moves. The US and China have outstripped all other nations in the AI technology race, however many unicorns we claim. But the giants of Silicon Valley have not been so keen to welcome the philosophers. US president Donald Trump signed an executive order on “Maintaining Leadership in AI” in February, which barely mentions ethics at all. As for China, even the Trump administration has blacklisted Chinese AI companies over ethical concerns. As post-Brexit Britain searches for a comparative advantage, ethical AI could be part of the answer.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484755 2019-06-21T13:10:00-04:00 2019-06-21T13:10:00-04:00

<p>The platform encourages broad participation in the policy-making process of the European Commission. Members of the Alliance will be able to engage in discussions with each other and the experts of the AI HLEG in a dedicated forum. Feedback will also be sought on specific (closed or open) questions, as well as on draft documents prepared by the AI HLEG under a section dedicated thereto. A specific events section will keep members informed of the upcoming relevant meetings and events which the Commission is organising, or in which it is taking part, and members will also be able to register their own event on the platform. In addition, Members will be able to access official documents on AI, and contribute reports and papers to an open library.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484571 2019-06-20T13:28:33-04:00 2019-06-20T13:28:33-04:00

<p>Yet companies have also seized on the idea of fairness in machine learning in a way that I will call a third type of co-opting, one in which the work done by socially conscious computer scientists working in the service of traditional civil rights goals, which was really meant to be empowering, suddenly becomes something that potentially fits in quite nicely with the existing interests of companies. If your main concern is that some of these systems don’t work as well for certain populations; if it turns out that the models are less likely to accurately assess, for instance, female applicants; well, one possible response is to go and collect more information about that population, so that the accuracy of your model can be improved. And in the end, this is potentially a very attractive problem for industry: it justifies investing more in data collection and seeing the problem of fairness or bias as one that is overcome by simply maximizing accuracy, which would have been a concern all along. Thus critical work done in computer science has quickly turned into something that is much more aligned with the traditional concerns of industry and government.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484545 2019-06-20T12:04:47-04:00 2019-06-20T12:04:47-04:00

<p>When Cicurel and her team looked more closely at the assessment technology, they discovered that it hadn’t been properly validated by any scientific group or judicial organization. Its previous review had come from an unpublished graduate-student thesis. Cicurel realized that for more than a decade, juvenile defendants in Washington, D.C., had been judged, and even committed to detention facilities, because the courts had relied on a tool whose only validation in the previous 20 years had come from a college paper.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484456 2019-06-19T11:02:36-04:00 2019-06-19T11:02:36-04:00

<p>The study of the ethics of AI at Oxford will be in a new humanities centre, bringing together subjects from languages to philosophy.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484455 2019-06-19T11:00:39-04:00 2019-06-19T11:00:39-04:00

<p>The Belfer and Shorenstein Center’s “Big Tech and Democracy” workshop fostered a space for congressional staffers to express their most-pressing concerns, inquiries and insights regarding tech policy issues. It added depth to the ongoing discussion of how to bring both policymakers and technical experts to the same table, undertake problem identification, and build a unified bipartisan front. The Belfer and Shorenstein Centers will continue to carry this momentum by providing additional resources to support congressional policy solutions. While an examination of past case studies built a foundation for discussion and policy frameworks, policymakers will need to consider a menu of technical, market-based, social, and regulatory solutions to adapt these historical observations into a blueprint for addressing today’s tech marketplace.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484413 2019-06-18T14:41:54-04:00 2019-06-18T14:41:54-04:00

<p>As states nationwide increasingly turn to risk assessment algorithms as tools for criminal justice reform, scholars and civil society actors alike are increasingly warning that this technological turn comes with complications. Research to date tends to focus on fairness, accountability, and transparency within algorithmic tools. Although attention to whether these instruments are fair or biased is normatively essential, this Article contends that this inquiry cannot be the whole conversation. Looking at issues such as fairness or bias in a tool in isolation elides vital bigger-picture considerations about the institutions and political systems within which tools are developed and deployed. Using California’s Money Bail Reform Act of 2017 (SB 10) as an example, this Article analyzes how risk assessment statutes create frameworks within which policymakers and technical actors are constrained and empowered when it comes to the design and implementation of a particular instrument. Specifically, it focuses on the tension between, on one hand, a top-down, global understanding of fairness, accuracy, and lack of bias and, on the other, a tool that is well-tailored to local considerations. It explores three potential technical and associated policy consequences of SB 10’s framework: proxies, Simpson’s paradox, and thresholding. And it calls for greater attention to the design of risk assessment statutes and their allocation of global and local authority.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484092 2019-06-13T12:03:20-04:00 2019-06-14T16:10:19-04:00

<p>Future research advances in machine learning are focused on two areas. The first is in enhancing how these systems distinguish between variations of an algorithm. As different versions of an algorithm are run over the training data, there needs to be some way of deciding which version is “better.” These evaluation functions need to balance the recognition of an improvement with not over-fitting to the particular training data. Getting functions that can automatically and accurately distinguish between two algorithms based on minor differences in the outputs is an art form that no amount of data can improve.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484102 2019-06-13T12:59:07-04:00 2019-06-13T12:59:07-04:00

<p>The point is that we have little conceptual clarity in the great AI debate because there is little agreement on what to include or exclude. If we cannot find the lines of inclusion and exclusion, then having any serious discussion about the possible benefits and risks of AI is almost impossible. Indeed, even discussing “AI governance,” including hard law, professional standards, and norms, presupposes having a finer-grained level of detail than currently exists, despite the many approaches proffered to date.17</p>

tag:tagteam.harvard.edu,2005:FeedItem/2484101 2019-06-13T12:58:34-04:00 2019-06-13T12:58:34-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2483962 2019-06-12T06:23:08-04:00 2019-08-12T13:51:44-04:00

<p>Alongside existing models, new models of governance and cooperation may be needed. They will need to be multi-stakeholder, including the private sector, civil society and users. Their debates should be transparent and open to citizens, as modelled by Mexico’s National Institute for Transparency, Access to Information and Personal Data Protection.93</p>

tag:tagteam.harvard.edu,2005:FeedItem/2483640 2019-06-07T11:57:25-04:00 2019-06-07T11:57:25-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2483639 2019-06-07T11:49:02-04:00 2019-06-07T11:49:02-04:00

<p>“Suddenly there was this categorical sentiment by various industry representatives and others that this phrase ‘red line’ cannot appear in that document anymore,” he recalls. The final version of the text, which has now been shared with the European Commission, instead referred to these issues as “critical concerns”.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2483372 2019-06-04T15:31:58-04:00 2019-06-04T15:31:58-04:00

<p>Levine targeted EFF, but his point redounds to a tough challenge for all of digital civil society. Two forces threaten individual liberties: corporate practices and government policies. The notion of corporate capture typically focuses on financial ties, but tech companies have other, more insidious means of exerting influence on nonprofits—the pervasive reach of their products.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2483331 2019-06-04T11:17:29-04:00 2019-06-04T11:17:29-04:00

<p>In the race to achieve AI mastery, the United States’ AI strategy has so far been led by corporations such as Google and Microsoft, who are focusing on self-regulation and rapid technological development, while others including China, the UK and the European Union have taken a government-led approach to defining AI principles.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2483227 2019-06-03T10:24:17-04:00 2019-06-03T10:24:17-04:00

<p>The Beijing AI Principles was jointly unveiled Saturday by the Beijing Academy of Artificial Intelligence (BAAI), Peking University, Tsinghua University, Institute of Automation and Institute of Computing Technology in Chinese Academy of Sciences, and an AI industrial league involving firms like Baidu, Alibaba and Tencent.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2483000 2019-05-30T15:37:18-04:00 2019-05-30T15:37:18-04:00

<p>The Government AI Readiness Index provides an assessment of the level of preparedness of governments to benefit from the potential of AI. Indicators and weighting criteria — for example, availability of open government data, level of digital skills and education, private sector capacity to benefit from AI, the existence of a national AI-related vision, and data protection and privacy laws —are applied to measure the capacity for wide-scale AI adoption in 194 countries and territories.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482997 2019-05-30T15:19:38-04:00 2019-05-30T15:19:38-04:00

<p>Microsoft unveiled a curriculum in its Artificial Intelligence Business School this week that is specifically tailored for government decision-makers. “The fact is that government workers across the board—and especially decision-makers—don’t necessarily have that familiarity or depth on AI,” Anthony Salcito, Microsoft’s vice president for government, said in a statement. “This new learning path is a way to get them introduced to the concept and to understand why it’s important in the context of government work.” The tech giant launched its AI Business School in March to help executives and other business leaders better understand how to implement the technology through a free online master class. Now that more than 140,000 people have gained practical guidance from the AI course material to date, the company is offering a new path that is designed especially for government agencies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482996 2019-05-30T14:24:45-04:00 2019-05-30T14:24:45-04:00

<p>Inspired by Shneiderman’s vision, Hosanagar advocates for the creation of an independent Algorithmic Safety Board, modeled on the Federal Reserve Board. Each country would have its own board at the federal level, and these boards would talk to one another and ensure there is some consistency across countries. Right now, some places have aggressive regulatory legislation (like the EU’s General Data Protection Regulation), while in other countries there’s almost no regulation at all. Hosanagar wants to see international coordination handled by a body modeled on the International Telecom Union, which governs cellphone communication. He argues for a mix of government regulation and self-regulation from tech companies. We need both, in his opinion, because tech companies can’t always be trusted to police themselves, and government regulators can’t always understand the complexities of fast-developing AI technology by themselves.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482938 2019-05-30T09:43:57-04:00 2019-05-30T09:43:57-04:00

<p>How can the multilateral system better understand and anticipate risks as AI convergence with dual-use technologies intrudes increasingly into the political, social, economic, and security spheres, creating new potential for systemic vulnerabilities and distributive inequalities? How can actors within the multilateral system build better anticipation and prevention capacities in the face of these risks?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482807 2019-05-28T16:11:20-04:00 2019-05-28T16:11:20-04:00

<p>Facebook has many failings, but the misconstrual of fakeness on its service must be attributed to us, its users and critics. We’ve fooled ourselves into believing we have some common ground with Facebook, one rooted in a responsibility to something larger than ourselves. Anderson Cooper looks like an earnest, hardline journalist when he challenges Monika Bickert, who seems to take the mantle of the sneaky tech executive, on Facebook’s latest failing. But that posture is only noble among a community who agrees about that the purpose of content like the Pelosi video is to tell the truth. That is not the case here.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482775 2019-05-28T14:03:41-04:00 2019-05-28T14:03:41-04:00

<p>Ethics shirking, like ethics dumping, has historical roots and often follows geopolitical outlines. Actors are more likely to engage in ethics dumping and shirking in contexts where disadvantage populations, weaker institutions, legal uncertainties, corrupted regimes, unfair power distributions, and other economic, legal, political, or social ills prevail. It is not unusual to map, correctly, both malpractices along the divide between Global North and Global South, or to see both as affecting above all Low- and Middle-Income Countries. The colonial past still exerts a disgraceful role. It is also important to recall that, in digital contexts, these malpractices can affect segments of a population within the Global North. The Gig Economy may be seen as a case of ethics shirking within developed countries. And the development of self-driving cars may be interpreted as an instance of research dumping in some states of the USA. In this case, the 1968 Vienna Convention on Road Traffic, which establishes international principles to govern traffic laws, requires that a driver is always fully in control and responsible for the behaviour of a vehicle in traffic. However, the USA is not a signatory country and the requirement does not apply, meaning state vehicle codes do not prohibit automated vehicles, and several states have enacted laws for automated vehicles. This is also why research on self-driving cars happens mostly in the USA—as well as the related incidents and human suffering.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482760 2019-05-28T11:27:41-04:00 2019-05-28T11:27:41-04:00

<p>The WEF will also announce the creation of an “AI Council” designed to find common ground on policy between nations that increasingly seem at odds over the power and the potential of AI and other emerging technologies (see “Trump’s feud with Huawei and China could lead to the Balkanization of tech”). The issue is of paramount importance given the current geopolitical winds. AI is widely viewed as critical to national competitiveness and geopolitical advantage. The effort to find common ground is also important considering the way technology is driving a wedge between countries, especially the United States and its big economic rival, China.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482507 2019-05-24T15:16:22-04:00 2019-05-24T15:16:22-04:00

<p>With algorithms appearing in a variety of applications, we argue that operators and other concerned stakeholders must be diligent in proactively addressing factors which contribute to bias. Surfacing and responding to algorithmic bias upfront can potentially avert harmful impacts to users and heavy liabilities against the operators and creators of algorithms, including computer programmers, government, and industry leaders. These actors comprise the audience for the series of mitigation proposals to be presented in this paper because they either build, license, distribute, or are tasked with regulating or legislating algorithmic decision-making to reduce discriminatory intent or effects.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482502 2019-05-24T14:22:39-04:00 2019-05-24T14:22:39-04:00

<p>Problems of bias and fairness are central to data justice, as they speak directly to the threat that ‘big data’ and algorithmic decision-making may worsen already existing injustices. In the United States, grappling with these problems has found clearest expression through liberal discourses of rights, due process, and antidiscrimination. Work in this area, however, has tended to overlook certain established limits of antidiscrimination discourses for bringing about the change demanded by social justice. In this paper, I engage three of these limits: 1) an overemphasis on discrete ‘bad actors’, 2) single-axis thinking that centers disadvantage, and 3) an inordinate focus on a limited set of goods. I show that, in mirroring some of antidiscrimination discourse’s most problematic tendencies, efforts to achieve fairness and combat algorithmic discrimination fail to address the very hierarchical logic that produces advantaged and disadvantaged subjects in the first place. Finally, I conclude by sketching three paths for future work to better account for the structural conditions against which we come to understand problems of data and unjust discrimination in the first place.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482494 2019-05-24T11:21:25-04:00 2019-05-24T11:21:25-04:00

<p>"Information operations are rarely about changing the things people believe, but changing the way they feel. Anger and fear are not things we can correct with better facts"</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482478 2019-05-24T10:18:16-04:00 2019-05-24T10:18:16-04:00

<p><span>In response, PeaceTech Lab partnered with South Africa’s </span><a href="https://www.mediamonitoringafrica.org/"><span>Media Monitoring Africa</span></a><span> to examine the origins, context, and influence of hateful speech in the country. </span><a href="https://www.peacetechlab.org/south-africa-lexicon"><span>Social Media, Discrimination and Intolerance in South Africa: A Lexicon of Hateful Terms</span></a><span> pairs social media analysis with in-depth qualitative research to help everyone - from civil society activists to government officials - in their efforts to address the surging problem of hate speech.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2482262 2019-05-22T15:48:29-04:00 2019-05-22T15:48:29-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2482260 2019-05-22T15:40:16-04:00 2019-05-22T15:40:16-04:00

<p>I hope that what follows can help explain why so many people feel angry, put-upon, and powerless. And why so many, in expressing their anger, refer to something called “the establishment,” that shadowy and pervasive entity. In this essay I will be critiquing both algorithmic governance and (more controversially) the tenets of pro­gressive governance that make these digital innovations attractive to managers, bureaucrats, and visionaries. What is at stake is the qual­itative character of institutional authority—how we experience it.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482231 2019-05-22T11:05:21-04:00 2019-05-22T11:05:21-04:00

<p>The problem is that AI adoption is already happening and some of its risks are playing out in the private sector. In the face of rapid technological change, governments need to move quickly to address social, political and ethical risks. When they move slowly, private sector actors often fill the void with their own principles and procedures. But self-regulation is always shaped by existing distributions of resources and power, frequently prioritizes profit over people and often amounts to ethics washing rather than sincere, committed action to address social, political and ethical risks of business activities and technologies. The government is right to highlight the need for a “human-centric” approach to AI, grounded in respect for “human rights, inclusion and diversity.” But recent developments still seem to prioritize innovation over risk management, industry over civil society and rhetoric over concrete efforts to develop regulatory mechanisms to protect the rights and interests of Canadians in the age of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482230 2019-05-22T10:32:47-04:00 2019-05-22T10:32:47-04:00

<p>AI Ethics is now a global topic of discussion in academic and policy circles. At least 63 public-private initiatives have produced statements describing high-level principles, values, and other tenets to guide the ethical development, deployment, and governance of AI. According to recent meta-analyses, AI Ethics has seemingly converged on a set of principles that closely resemble the four classic principles of medical ethics. Despite the initial credibility granted to a principled approach to AI Ethics by the connection to principles in medical ethics, there are reasons to be concerned about its future impact on AI development and governance. Significant differences exist between medicine and AI development that suggest a principled approach in the latter may not enjoy success comparable to the former. Compared to medicine, AI development lacks (1) common aims and fiduciary duties, (2) professional history and norms, (3) proven methods to translate principles into practice, and (4) robust legal and professional accountability mechanisms. These differences suggest we should not yet celebrate consensus around high-level principles that hide deep political and normative disagreement.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482229 2019-05-22T09:04:21-04:00 2019-05-22T09:47:52-04:00

<p>At an annual meeting on Wednesday, the 36 countries in the Organization for Economic Cooperation and Development (OECD) plus a handful of other nations are set to adopt a list of guidelines for the development and use of artificial intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482052 2019-05-20T10:30:38-04:00 2019-05-20T10:30:38-04:00

<p>The Ministry of Economic Development, Job Creation and Trade cut $20 million from the Vector Institute for Artificial Intelligence and $4 million annually from the Canadian Institute for Advanced Research (CIFAR), which supports a hub of AI-focused computer scientists. Both draw funding from the federal government and other sources and say they will adjust programming or operations.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2482049 2019-05-20T10:18:50-04:00 2019-05-20T10:18:50-04:00

<p>She's one of tens of thousands of non-technology experts who are taking part in a grand experiment aimed at repurposing the country's economy toward high-end applications of artificial intelligence. The idea has a simple, Nordic ring to it: Start by teaching 1 percent of the country's population, or about 55,000 people, the basic concepts at the root of artificial technology, and gradually build on the number over the next few years.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481934 2019-05-17T10:43:42-04:00 2019-05-17T10:43:42-04:00

<p>Across the world, artificial intelligence (AI) is showing its potential for abetting repressive regimes and upending the relationship between citizen and state, thereby exacerbating a global resurgence of authoritarianism. AI is a component in a broader ecosystem of digital repression, but it is relevant to several different techniques, including surveillance, censorship, disinformation, and cyber attacks. AI offers three distinct advantages to autocratic leaders: it helps solve principal-agent loyalty problems, it offers substantial cost-efficiencies over traditional means of surveillance, and it is particularly effective against external regime challenges. China is a key proliferator of AI technology to authoritarian and illiberal regimes; such proliferation is an important component of Chinese geopolitical strategy. To counter the spread of high-tech repression abroad, as well as potential abuses at home, policy makers in democratic states must think seriously about how to mitigate harms and to shape better practices.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481846 2019-05-16T14:55:50-04:00 2019-05-16T14:55:50-04:00

<p>The Trump Administration Turns Its Back on a Chance to Fight Online Extremism The White House rejected an international effort to stop the spread of hate and violence on the internet. Its reasons are specious and counterproductive.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481752 2019-05-15T14:37:30-04:00 2019-05-15T16:44:17-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2481614 2019-05-14T10:03:06-04:00 2019-05-14T10:03:06-04:00

<p>human rights impact assessment; public consultations; human rights standards in the private sector; information and transparency; independent oversight; non-discrimination and equality; data protection and privacy; freedom of expression, freedom of assembly and association, and the right to work; access to remedies; and the promotion of artificial intelligence literacy.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481529 2019-05-13T09:56:22-04:00 2019-05-22T15:01:39-04:00

<p>People, Power and Technology: The Tech Workers’ View is the first in-depth research into the attitudes of the people who design and build digital technologies in the UK. It shows that workers are calling for an end to the era of moving fast and breaking things.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481347 2019-05-10T10:43:50-04:00 2019-05-10T10:43:50-04:00

<p>Instead of thinking about machine intelligence in terms of humans vs machines, we should consider the system that integrates humans and machines – not artificial intelligence but extended intelligence. Instead of trying to control or design or even understand systems, it is more important to design systems that participate as responsible, aware and robust elements of even more complex systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481052 2019-05-07T11:25:11-04:00 2019-05-07T11:25:11-04:00

<p>Artificial Intelligence (AI) presents novel policy challenges that require coordinated global responses.2Standards, particularly those developed by existing international standards bodies, can support the globalgovernance of AI development. International standards bodies have a track record of governing a range ofsocio-technical issues: they have spread cybersecurity practices to nearly 160 countries, they have seen firmsaround the world incur significant costs in order to improve their environmental sustainability, and they havedeveloped safety standards used in numerous industries including autonomous vehicles and nuclear energy.These bodies have the institutional capacity to achieve expert consensus and then promulgate standards acrossthe world. Other existing institutions can then enforce these nominally voluntary standards through both ​defacto​ and ​dejure​ methods</p>

tag:tagteam.harvard.edu,2005:FeedItem/2481051 2019-05-07T11:12:05-04:00 2019-05-07T11:12:05-04:00

<p>A proposed law would make big companies determine whether their algorithms discriminate, but it’s lacking in some big ways.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2480995 2019-05-06T13:25:12-04:00 2019-05-06T13:25:12-04:00

<p>Given this trend, it is especially disappointing to learn that some of the clearest statements about ethical boundaries for AI were removed from the report. To effectively shape the development and use of AI systems, policy guidelines need to provide enough detail and certainty to allow for their application in the real world and to allow others to verify efforts to comply with them. To its credit, the guidelines acknowledge the necessity of applying different ethical principles depending on the application of any given AI system. However, the guidelines do not clarify which principles should apply to which applications or how one should evaluate applications according to any set of principles (beyond the four high-level tenets above).</p>

tag:tagteam.harvard.edu,2005:FeedItem/2480687 2019-05-02T09:11:16-04:00 2019-08-30T16:29:09-04:00

<p>The dominant academic publishers are busy positioning themselves to monetize not only on content, but increasingly on data analytics and predictive products on research assessment and funding trends. Their growing investment and control over the entire knowledge production workflow, from article submissions, to metrics to reputation management and global rankings means that researchers and their institutions are increasingly locked into the publishers’ “value chain”. I will discuss some of the implications of this growing form of “surveillance capitalism” in the higher education sector and what it means in terms of the autonomy of the researchers and the academy. The intent is to call attention to the need to support community-governed infrastructure and to rethink our understanding of “openness” in terms of consent and social values.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2480749 2019-05-02T14:18:47-04:00 2019-05-02T14:18:47-04:00

<p>Industry has the data and expertise necessary to design fairness into AI systems. It cannot be excluded from the processes by which we investigate which worries are real and which safeguards work, but it must not be allowed to direct them. Organizations working to ensure that AI is fair and beneficial must be publicly funded, subject to peer review and transparent to civil society. And society must demand increased public investment in independent research rather than hoping that industry funding will fill the gap without corrupting the process.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2479431 2019-04-17T14:35:17-04:00 2019-08-08T10:42:15-04:00

<p>Counterintuitively, the Article contends that the framing of algorithmic bias as a technical problem is misguided. Rather, the Article’s central claim is that bias is introduced in the hiring process, in large part, due to an American legal tradition of deference to employers, especially allowing for such nebulous hiring criterion as “cultural fit.” The Article observes the lack of legal frameworks that take into account the emerging technological capabilities of hiring tools which make it difficult to detect disparate impact. The Article thus argues for a re-thinking of legal frameworks that take into account both the liability of employers and those of the makers of algorithmic hiring systems who, as brokers, owe a fiduciary duty of care. Particularly related to Title VII, the Article proposes that in legal reasoning corollary to extant tort doctrines, an employer’s failure to audit and correct its automated hiring platforms for disparate impact could serve as prima facie evidence of discriminatory intent, leading to the development of the doctrine of discrimination per se. The article also considers other approaches separate from employment law such as establishing consumer legal protections for job applicants that would mandate their access to the dossier of information consulted by automated hiring systems in making the employment decision.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2479237 2019-04-15T12:20:43-04:00 2019-04-15T12:20:43-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2478169 2019-04-02T12:12:33-04:00 2019-04-02T12:12:33-04:00

<p>Singapore introduced draft legislation on Monday that it said would combat false or misleading information on the internet, but critics said the measure could be used as a cudgel against the government’s critics.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2478124 2019-04-02T09:30:11-04:00 2019-04-02T09:30:11-04:00

<p><span>Calls</span><span>for</span><span>sector</span><span>-</span><span>specific regulation</span><span>or</span><span>the creation of a</span><span>federal agency</span><span>or commission</span><span>to guide </span><span>and constrain artificial intelligence, or AI, development are </span><span>increasing.</span><span>This turn to administrative </span><span>law is understandable because AI’s regulatory challenges</span><span>seem</span><span>similar to those in other </span><span>technocratic domains, such as the pharmaceutical industry or environmental law.</span><span>But an “FDA </span><span>for algorithms” or federal robotics </span><span>commission is not a cross</span><span>-</span><span>cutting AI solution. AI is</span><span>unique, </span><span>even if it is not entirely different.</span><span>AI’s distinctiveness comes from technical</span><span>attributes (speed, </span><span>complexity, and unpredictability) that</span><span>strain traditional administrative law tactics, in combina</span><span>tion </span><span>with</span><span>institutional</span><span>settings</span><span>and incentives, or </span><span>strategic context</span><span>, that affect its development path. </span><span>This Article puts American AI governance in strategic context. Today, there is an imbalance of </span><span>state and non</span><span>-</span><span>state AI authority. Commercial actors do</span><span>minate research and development and </span><span>private resources outstrip public investments. Even if we could redress this baseline, a </span><span>fundamental, yet under</span><span>-</span><span>recognized problem remains. Any governance strategy must contend with </span><span>the ways in which algorithmic applicat</span><span>ions permit seemingly technical decisions</span><span>to de facto </span><span>regulate human behavior, with a greater potential for physical and social impact than ever before. </span><span>When coding choices functionally operate as policy in this manner, the current trajectory of AI </span><span>develop</span><span>ment augurs an era of private governance. </span><span>Without rethinking our regulatory strategies, </span><span>we risk losing the democratic accountability that is at the heart of public governance.</span></p>

tag:tagteam.harvard.edu,2005:FeedItem/2478036 2019-04-01T11:21:33-04:00 2019-04-01T11:21:33-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2477849 2019-03-29T10:55:24-04:00 2019-03-29T10:55:24-04:00

<p>These issues are examined in a new book, Knowledge and Global Power: Making new sciences in the South (published by Wits University Press), which I co-authored with Fran Collyer, Raewyn Connell and Joao Maia. The focus is especially on those areas where old patterns are not being replicated, so the study chooses climate change, gender and HIV and AIDS as three new areas of knowledge production in which new voices from the South might be prominent.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2477848 2019-03-29T10:55:00-04:00 2019-03-29T10:55:00-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2477804 2019-03-28T21:25:09-04:00 2019-03-28T21:25:09-04:00

<p>The leaked emails offer a window into the intense internal discussions and disagreements Facebook executives have over harmful content and concern over how its actions will be portrayed in the press.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2477170 2019-03-21T15:02:52-04:00 2019-03-21T15:02:52-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2477073 2019-03-20T16:33:48-04:00 2019-03-20T16:33:48-04:00

<p>In their new Nature Human Behaviour paper, “The Wisdom of Polarized Crowds,” Evans and Teplitskiy concluded that polarization doesn’t poison the wells of information. On the contrary, they showed politically diverse editor teams on Wikipedia put out better entries—articles with higher accuracy or completeness—than uniformly liberal or conservative or moderate teams. It’s a surprising result and so I caught up with Evans and Teplitskiy to offer their interpretations.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2476427 2019-03-15T10:14:23-04:00 2019-03-15T10:14:23-04:00

<p>The papers in this special issue focus on various aspects of machine ethics, including the central question of how autonomous systems can be imbued with ethical values.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2476426 2019-03-15T09:50:19-04:00 2019-03-15T09:50:19-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2476425 2019-03-15T09:50:02-04:00 2019-03-15T09:50:02-04:00

<p>Regulating disinformation with artificial intelligence</p>

tag:tagteam.harvard.edu,2005:FeedItem/2476406 2019-03-15T09:45:50-04:00 2019-03-15T09:45:50-04:00

<p>This paper is a response to the Singapore Personal Data Protection Commission consultation on a draft AI Governance Framework. It analyses the five data trust models proposed by the UK Open Data Institute and identifies that only the contractual and corporate models are likely to be legally suitable for achieving the aims of a data trust.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2476105 2019-03-12T09:46:38-04:00 2019-03-12T09:46:38-04:00

<p>This report focuses on domestically generated disinformation in the U.S.: the nature and scope of the problem, what the social media platforms have done about it, and what more they need to do.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2476074 2019-03-11T14:02:34-04:00 2019-03-11T14:02:34-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2476064 2019-03-11T12:34:19-04:00 2019-03-11T12:34:19-04:00

<p>The best way to get away with something is to make it feel new and surprising. Using a computer is hardly enough anymore; today’s machines offer all kinds of ways to generate images that can be output, framed, displayed, and sold—from digital photography to artificial intelligence. Recently, the fashionable choice has become generative adversarial networks, or GANs, the technology that created Portrait of Edmond de Belamy. Like other machine-learning methods, GANs use a sample set—in this case, art, or at least images of it—to deduce patterns, and then they use that knowledge to create new pieces. A typical Renaissance portrait, for example, might be composed as a bust or three-quarter view of a subject. The computer may have no idea what a bust is, but if it sees enough of them, it might learn the pattern and try to replicate it in an image.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2476033 2019-03-11T09:35:05-04:00 2019-03-11T09:35:05-04:00

<p>Sandbox beta phase discussion paper</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475688 2019-03-06T10:54:34-05:00 2019-03-06T10:54:34-05:00

<p>“Men Are Scum”: Inside Facebook’s War on Hate Speech The company blew it on privacy and fake news. Can it do better against trolls and racists? An exclusive embed with Facebook’s shadow government.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475596 2019-03-05T13:43:16-05:00 2019-03-05T13:43:16-05:00

<p>What if we had a “social media” space where information was organized based on its context, legal significance and cultural meaning? How might we see the present time, and the past, differently?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475563 2019-03-05T10:23:55-05:00 2019-03-05T10:23:55-05:00

<p>One study from North Carolina State University has suggested that reading ACM’s code of ethics, on its own, doesn’t have much of an effect on individual decision-making. Nevertheless, I believe studying such a code arms readers with the language needed to describe types of ethical infractions. Establishing a code of ethics gives potential whistleblowers a leg to stand on so that an engineer could say, “We cannot publish this until it complies with our company’s ethical standards.” Indeed, other research has suggested that a code, complemented with a more holistic approach to building an ethical culture, can effectively change employee behavior. As major CS schools such as Stanford and MIT start to build ethics training into their curricula, hopefully the next generation of engineers can bring those lessons into the workplace.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475482 2019-03-04T13:09:58-05:00 2019-03-04T13:09:58-05:00

<p>Baidu’s Robin Li Yanhong and Tencent’s Pony Ma Huateng submitted separate proposals on the need for ethical guides on the development of new technology on Sunday</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475347 2019-03-01T13:59:51-05:00 2019-03-01T13:59:51-05:00

<p>About 50 experts in the use of artificial intelligence and other digital health care technologies gathered at Loeb House for an all-day symposium on “Hype vs. Reality: The Role of AI in Global Health,” sponsored by HGHI. It featured talks by representatives from academia, health care, and industry, including Google AI and the pharmaceutical giant Novartis. Principals also spoke for startups Aindra Systems, which uses AI to rapidly analyze pap smears for evidence of cervical cancer, and Ubenwa, which has developed an AI-based system to analyze infant cries to detect birth asphyxia, a leading killer of children under 5 that’s traditionally diagnosed through time-consuming blood tests and lab analyses.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475348 2019-03-01T14:02:03-05:00 2019-04-01T11:45:26-04:00

<p>How data trusts, collaboratives and coops can help govern data for the maximum public benefit</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475198 2019-02-28T11:02:52-05:00 2019-02-28T11:02:52-05:00

<p>In this work, we investigate whether state-of-the-art object detection systems haveequitable pre-dictive performanceon pedestrians with differentskin tones. This work is motivated by many recentexamples of ML and vision systems displayinghigher error rates for certain demographic groupsthan others. We annotate an existing large scaledataset which contains pedestrians, BDD100K,with Fitzpatrick skin tones in ranges [1-3] or [4-6].We then provide an in depth comparative analy-sis of performance between these two skin tonegroupings, finding that neither time of day norocclusion explain this behavior, suggesting thisdisparity is not merely the result of pedestrians inthe 4-6 range appearing in more difficult scenesfor detection. We investigate to what extent timeof day, occlusion, and reweighting the supervisedloss during training affect this predictive bias. cd</p>

tag:tagteam.harvard.edu,2005:FeedItem/2475060 2019-02-27T08:52:34-05:00 2019-02-27T08:52:34-05:00

<p>De Kleijn-Lloyd was presenting the results of an Elsevier study of trends in the publication and use of academic studies on AI. According to the report, published in December, only around 0.4 per cent of search keywords had any real bearing on ethics, and their use was limited to teaching materials and media sources.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2474464 2019-02-19T11:25:36-05:00 2019-02-19T11:25:36-05:00

<p>World leaders, CEOs, and academics have suggested that a revolution in artificial intelligence is upon us. Are they right, and what will advances in artificial intelligence mean for international competition and the balance of power? This article evaluates how developments in artificial intelligence (AI) — advanced, narrow applications in particular — are poised to influence military power and international politics. It describes how AI more closely resembles “enabling” technologies such as the combustion engine or electricity than a specific weapon. AI’s still-emerging developments make it harder to assess than many technological changes, especially since many of the organizational decisions about the adoption and uses of new technology that generally shape the impact of that technology are in their infancy. The article then explores the possibility that key drivers of AI development in the private sector could cause the rapid diffusion of military applications of AI, limiting first-mover advantages for innovators. Alternatively, given uncertainty about the technological trajectory of AI, it is also possible that military uses of AI will be harder to develop based on private-sector AI technologies than many expect, generating more potential first-mover advantages for existing powers such as China and the United States, as well as larger consequences for relative power if a country fails to adapt. Finally, the article discusses the extent to which U.S. military rhetoric about the importance of AI matches the reality of U.S. investments.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2474455 2019-02-19T10:43:44-05:00 2019-02-19T10:43:44-05:00

<p>In the absence of thinking about both technical feasibility and social viability, there is a strong risk that AI-based technology gains are likely to benefit only a select few Indians. In this context, the ‘AI for All’ narrative obscures rather than answers many of the fundamental challenges that India faces.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2474432 2019-02-19T09:54:06-05:00 2019-02-19T09:54:06-05:00

<p>Debiasing Word Embeddings</p>

tag:tagteam.harvard.edu,2005:FeedItem/2474050 2019-02-13T14:41:26-05:00 2019-02-13T14:41:26-05:00

<p>Interesting landscape analysis</p> <p> </p>

tag:tagteam.harvard.edu,2005:FeedItem/2473962 2019-02-12T14:14:02-05:00 2019-02-12T14:14:02-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2473842 2019-02-11T13:08:24-05:00 2019-02-11T13:08:24-05:00

<p>President Trump is expected to sign an executive order on Monday meant to spur the development and regulation of artificial intelligence, technology that many experts believe will define the future of everything from consumer products to health care to warfare.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473843 2019-02-11T13:11:45-05:00 2019-02-19T11:25:07-05:00 Remco Zwetsloot, Allan Dafoe

<p>The idea that many of the risks from AI have structural causes is a sobering one: It implies that solving these problems will require collective action both domestically and internationally, which has always been a difficult problem—especially on the international stage. Yet at several points in history, even tense ones, nations managed to find ways to stave off (at least for a while) the unintended and destabilizing effects of emerging technologies, from the Anti-Ballistic Missile Treaty to the Montreal Protocol. Such cooperation becomes possible when leaders realize that structural risks are also collective risks, and that there are therefore mutual gains to be had from working hard to understand and address them—even if those involved otherwise see each other as competitors. The fact that nations’ fates are fundamentally interlocked is a source of complexity in the governance of AI, but also a source of hope.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473830 2019-02-11T10:53:21-05:00 2019-02-11T10:53:21-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2473829 2019-02-11T10:52:48-05:00 2019-02-11T10:52:48-05:00

<p>etter, Nicer, Clearer, Fairer: A Critical Assessment of theMovement for Ethical Artificial Intelligence and Machine Learning</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473828 2019-02-11T10:47:24-05:00 2019-02-11T10:47:24-05:00

<p>The importance of public legitimacy – which refers to the broad base of public support that allows companies, designers, public servants and others to design and develop AI to deliver beneficial outcomes – was illustrated by a series of public, highly controversial events that took place in 2018</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473826 2019-02-11T10:37:45-05:00 2019-02-11T10:37:45-05:00

<p>Debate about the impacts of AI is often split into two camps, one associated with the near term and the other with the long term. This divide is a mistake — the connections between the two perspectives deserve more attention</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473645 2019-02-08T09:23:55-05:00 2019-02-08T09:23:55-05:00

<p>Artificial Intelligence and National Security</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473644 2019-02-08T09:22:57-05:00 2019-02-08T09:22:57-05:00

<p>Demystifying Machine Learning for Regulators On Wednesday, January 16, Professor Hal Abelson and I led a Machine Learning (ML) workshop for members of the Organisation for Economic Co-operation and Development (OECD) Artificial Intelligence Expert Group (AIGO), where we aimed to demystify ML and discuss challenges to developing effective public policy in the space. As a PhD student working on ML research and education, I was looking forward to hearing the perspectives of people who would be in charge of crafting the AI Policy Principles, which will represent consensus views of over 30 countries.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473399 2019-02-06T13:56:31-05:00 2019-02-06T13:56:31-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2473313 2019-02-05T12:00:00-05:00 2019-02-05T12:00:00-05:00

<p>Mappingthe challengesand opportunitiesof artificial intelligencefor the conductof diplomacy</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473268 2019-02-04T19:42:30-05:00 2019-02-04T19:42:30-05:00

<p>One type of progress is on what is often referred to as AI safety research. As our profile on this problem explains, this subfield within computer science addresses the technical question of how to ensure advanced AI systems do what we want them to do without unwanted side-effects. It’s being studied by groups like Prof. Stuart Russell’s Centre on Human-compatible AI at University of Berkeley, OpenAI’s safety team, DeepMind’s safety team, Oxford University’s Future of Humanity Institute’s safety group,2 and MIRI. The DeepMind safety team give a good high-level overview of the work required in this field in this post. An alternative landscape overview is given in ‘Concrete Problems in AI safety‘ (summary). We also have a podcast with OpenAI safety resea</p>

tag:tagteam.harvard.edu,2005:FeedItem/2473234 2019-02-04T14:17:06-05:00 2019-02-04T14:17:06-05:00

<div>Specifically, we highlight five areas where government, in collaboration with wider civil</div> <div>society and AI practitioners, has a crucial role to play in clarifying expectations about AI’s</div> <div>application on a context-specific basis. These include explainability standards, approaches</div> <div>to appraising fairness, safety considerations, requirements for human-AI collaboration,</div> <div>and general liability frameworks.</div>

tag:tagteam.harvard.edu,2005:FeedItem/2473233 2019-02-04T14:16:21-05:00 2019-02-04T14:16:21-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2470670 2018-12-21T10:30:54-05:00 2018-12-21T10:30:54-05:00

<p>This working document constitutes a draft of the AI Ethics Guidelines produced by the European Commission’s High-Level Expert Group on Artificial Intelligence (AI HLEG), of which a final version is due in March 2019.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2470633 2018-12-20T16:13:30-05:00 2018-12-20T16:13:30-05:00

<p>Report on National and Regional AI Strategies</p>

tag:tagteam.harvard.edu,2005:FeedItem/2470551 2018-12-19T14:55:47-05:00 2018-12-19T14:55:47-05:00

<p>Kent Walker, senior vice president of global affairs at Google, said in a blog post that a formal review structure to assess new “projects, products and deals” has been established, and that more than 100 reviews have been completed so far. Some have resulted in decisions to modify research in visual speech recognition and to hold off on commercial offerings of technology like general-purpose facial recognition.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2470130 2018-12-14T09:52:19-05:00 2018-12-14T09:52:19-05:00

<p>U.S. cybersecurity law is largely an outgrowth of the early-aughts concerns over identity theft and financial fraud. Cybersecurity laws focus on protecting identifiers such as driver’s licenses and social security numbers, and financial data such as credit card numbers. Federal and state laws require companies to protect this data and notify individuals when it is breached, and impose civil and criminal liability on hackers who steal or damage this data. In this paper, I argue that our current cybersecurity laws are too narrowly focused on financial harms. While such concerns remain valid, they are only one part of the cybersecurity challenge that our nation faces. Too often overlooked by the cybersecurity profession are the harms to individuals, such as revenge pornography and online harassment. Our legal system typically addresses these harms through retrospective criminal prosecution and civil litigation, both of which face significant limits. Accounting for such harms in our conception of cybersecurity will help to better align our laws with these threats and reduce the likelihood of the harms occurring.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2470048 2018-12-13T13:06:40-05:00 2018-12-13T13:06:40-05:00

<p>In this series of digital interactives, we will take you through the main building blocks for creating and deploying AI, illuminating unique aspects of the Chinese ecosystem for each one. Before doing that, we need a rudimentary definition and model for AI today.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469968 2018-12-12T13:10:03-05:00 2018-12-12T13:10:03-05:00

<p>Delivering on its strategy on artificial intelligence (AI) adopted in April 2018 today the Commission presents a coordinated plan prepared with Member States to foster the development and use of AI in Europe. This plan proposes joint actions for closer and more efficient cooperation between Member States, Norway, Switzerland and the Commission in four key areas: increasing investment, making more data available, fostering talent and ensuring trust. Stronger coordination is essential for Europe to become the world-leading region for developing and deploying cutting-edge, ethical and secure AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469965 2018-12-12T12:22:47-05:00 2018-12-12T12:22:47-05:00

<p>The central prediction I want to make and defend in this post is that continued rapid progress in machine learning will drive the emergence of a new kind of geopolitics; I have been calling it AI Nationalism. Machine learning is an omni-use technology that will come to touch all sectors and parts of society. The transformation of both the economy and the military by machine learning will create instability at the national and international level forcing governments to act. AI policy will become the single most important area of government policy. An accelerated arms race will emerge between key countries and we will see increased protectionist state action to support national champions, block takeovers by foreign firms and attract talent. I use the example of Google, DeepMind and the UK as a specific example of this issue. This arms race will potentially speed up the pace of AI development and shorten the timescale for getting to AGI. Although there will be many common aspects to this techno-nationalist agenda, there will also be important state specific policies. There is a difference between predicting that something will happen and believing this is a good thing. Nationalism is a dangerous path, particular when the international order and international norms will be in flux as a result and in the concluding section I discuss how a period of AI Nationalism might transition to one of global cooperation where AI is treated as a global public good.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469883 2018-12-11T14:00:25-05:00 2018-12-11T14:00:25-05:00

<p>The impact of AI is likely to disproportionately affect populations in the developing world. Compared with the advanced economies, emerging economies will have gaps that will prove problematic for AI deployment, such as in information and communications technology (ICT) infrastructure, digital skills and literacy, capital market maturity, funding for research and development (R&amp;D), and legal and institutional frameworks to regulate new technologies. As a result, they face uphill challenges to nurture local innovation ecosystems, develop new digital industries, and prepare people for new kinds of employment. Jobs in many sectors are already undergoing a transformation. For instance, in electronics manufacturing, Foxconn, a major Apple and Samsung supplier, has reportedly replaced 60,000 workers with robots and now aims for full automation of its Chinese factories.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469834 2018-12-11T08:45:39-05:00 2018-12-12T13:50:39-05:00

<p>A year after YouTube’s chief executive promised to curb “problematic” videos, it continues to harbor and even recommend hateful, conspiratorial videos, allowing racists, anti-Semites and proponents of other extremist views to use the platform as an online library for spreading their ideas.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469800 2018-12-10T17:04:56-05:00 2018-12-10T17:04:56-05:00

<p>Describes opportunities for harnessing the value of big data and artificial intelligence (AI) for social good and how new families of AI algorithms now make it possible to obtain actionable insights automatically and at scale. Beyond internet business or commercial applications, multiple examples already exist of how big data and AI can help achieve shared development objectives, such as the 2030 Agenda for Sustainable Development and the Sustainable Development Goals (SDGs). But ethical frameworks in line with increased uptake of these new technologies remain necessary—not only concerning data privacy but also relating to the impact and consequences of using data and algorithms. Public recognition has grown concerning AI’s potential to create both opportunities for societal benefit and risks to human rights. Development calls for seizing the opportunity to shape future use as a force for good, while at the same time ensuring the technologies address inequalities and avoid widening the digital divide.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469799 2018-12-10T17:04:39-05:00 2018-12-10T17:04:39-05:00

<p>This Memorandum of Understanding between the PPPs follows the European Commission's approach to AI presented in April 2018 and the Declaration of Cooperation on Artificial Intelligence signed by all 28 Member States and Norway. This Friday 7 December the Commission will present its EU coordinated plan.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469707 2018-12-09T13:25:15-05:00 2018-12-11T13:54:38-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2469603 2018-12-07T14:15:03-05:00 2018-12-07T14:15:03-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2469490 2018-12-06T13:54:45-05:00 2018-12-06T13:54:45-05:00

<p>Black box machine learning models are currently being used for high stakes decision-making throughout society, causing problems throughout healthcare, criminal justice, and in other domains. People have hoped that creating methods for explaining these black box models will alleviate some of these problems, but trying to explain black box models, rather than creating models that are interpretable in the first place, is likely to perpetuate bad practices and can potentially cause catastrophic harm to society. There is a way forward - it is to design models that are inherently interpretable.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469393 2018-12-05T15:46:44-05:00 2018-12-05T15:46:44-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2469383 2018-12-05T15:38:37-05:00 2018-12-06T13:59:01-05:00

<p>The United States is the global leader in developing and using artificial intelligence (AI), but it may not be for long. Succeeding in AI requires more than just having leading companies make investments. It requires a healthy ecoystem of AI companies, robust AI inputs—including skills, research, and data—and organizations that are motivated and free to use AI. And that requires the federal government to support the development and adoption of AI. Many other countries, including China, France, and the United Kingdom, are developing significant initiatives to gain global market share in AI. While the U.S. government has taken some steps, it lacks a comprehensive strategy to proactively spur the development and adoption of AI. This report explains why a national AI strategy is necessary to bolster U.S. competitiveness, strengthen national security, and maximize the societal benefits that the country could derive from AI. It then lays out six overarching goals and 40 specific recommendations for Congress and the administration to support AI development and adoption.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2469286 2018-12-04T17:05:33-05:00 2018-12-04T17:05:33-05:00

<p>The difficulty of finding a middle path is why negotiations on topics like artificial intelligence and big data were so fraught in Dubai: as much as developing countries hoped ITU could provide relatively quick fixes for the deluge of ICT issues they increasingly face, many developed countries were concerned about mission creep, duplicating existing processes, and ITU entering fields in which it lacks expertise. The Dubai resolutions show this tension, with many references to “in collaboration with relevant bodies,” “information sharing” as well as to ITU-specific studies and activities.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468821 2018-12-03T13:19:31-05:00 2018-12-03T13:19:31-05:00

<p>More than half of the world’s roads lack adequate street addressing systems. Lack of addresses is even more visible in daily lives of people in developing countries. We would like to object to the assumption that having an address is a luxury, by proposing a generative address design that maps the world in accordance with streets. The addressing scheme is designed considering several traditional street addressing methodologies employed in the urban development scenarios around the world. Our algorithm applies deep learning to extract roads from satellite images, converts the road pixel confidences into a road network, partitions the road network to find neighborhoods, and labels the regions, roads, and address units using graph- and proximity-based algorithms. We present our results on a sample US city, and several developing cities, compare travel times of users using current ad hoc and new complete addresses, and contrast our addressing solution to current industrial and open geocoding alternatives</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468820 2018-12-03T13:18:51-05:00 2018-12-03T13:18:51-05:00

<p>The results: Hosted in partnership with the Neural Information Processing Systems (NeurIPS) conference, one of the largest annual gatherings for AI research, the competition received submissions from over 100 participants. Google Brain researcher Pallavi Baljekar noted at a conference talk on Sunday, December 2, that the first-year competition winners were able to make small steps toward more inclusive systems. But only one of the top five approaches successfully recognized an Indian bride. It’s clear that more work needs to be done.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468806 2018-12-03T11:21:36-05:00 2018-12-03T11:21:36-05:00

<p>More The Commission pledges to spend up to €100 million on efforts "to facilitate access to, connectivity, interoperability and aggregation of public data including the development of relevant Application Programming Interfaces (APIs) for accessing data of dynamic nature." /9</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468563 2018-11-30T09:32:10-05:00 2018-11-30T16:31:16-05:00

<p><img src="http://clinic.cyber.harvard.edu/files/2018/11/podcast-3332163_640-300x197.png"></p> <p>We found that podcast platforms varied widely in the language of their content policies. Some were short and broad — for example, Apple Podcasts’ terms merely say that they will prohibit content that may be “construed as racist, misogynist, or homophobic” or that depicts “hate themes.” But even within Apple there are significantly different approaches to content regulation. Apple’s App Store Guidelines, unlike their podcast equivalents, include seven paragraphs defining “offensive content.”</p> <p><a href="http://clinic.cyber.harvard.edu/2018/11/29/how-podcast-platforms-respond-to-hate-speech-clinic-releases-new-memo/">Read more at the Cyberlaw Clinic...</a></p>

tag:tagteam.harvard.edu,2005:FeedItem/2468568 2018-11-30T09:40:47-05:00 2018-11-30T09:40:47-05:00

<p>Software development remains the province of males — only about one-quarter of computer scientists in the United States are women3 — and minority racial groups, including blacks and Hispanics, are underrepresented in tech work, too.4 Groups like Girls Who Code and AI4ALL have been founded to help close those gaps. Girls Who Code has reached almost 90,000 girls from various backgrounds in all 50 states,5 and AI4ALL specifically targets girls in minority communities.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468499 2018-11-29T15:50:38-05:00 2018-11-29T15:50:38-05:00

<p>A key goal of the FAT* community is to develop machine-learning based systems that, once introduced into a social context, can achieve social and legal outcomes such as fairness, justice, and due process. Bedrock concepts in computer science—such as abstraction and modular design—are used to define notions of fairness and discrimination, to produce fairness-aware learning algorithms, and to intervene at different stages of a decision-making pipeline to produce "fair" outcomes. In this paper, however, we contend that these concepts render technical interventions ineffective, inaccurate, and sometimes dangerously misguided when they enter the societal context that surrounds decision-making systems. We outline this mismatch with five "traps" that fair-ML work can fall into even as it attempts to be more context-aware in comparison to traditional data science. We draw on studies of sociotechnical systems in Science and Technology Studies to explain why such traps occur and how to avoid them. Finally, we suggest ways in which technical designers can mitigate the traps through a refocusing of design in terms of process rather than solutions, and by drawing abstraction boundaries to include social actors rather than purely technical ones.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468501 2018-11-29T16:02:54-05:00 2018-12-03T14:53:22-05:00

<p>For many reasons, it matters who is conducting AI research. The researcher gets to set the question, decide what datasets to use, how to conduct analyses and how to present results. There are many points at which the researcher makes subjective decisions, which can impact what community will benefit from the research. Greater diversity within research communities can control for bias in this process. Diverse teams can think creatively about solving problems to which we can apply techniques and insights from AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468451 2018-11-29T11:08:46-05:00 2018-11-29T11:08:46-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2468450 2018-11-29T11:03:32-05:00 2018-11-29T11:03:32-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2468440 2018-11-29T10:05:19-05:00 2018-11-30T16:47:31-05:00

<p><img src="https://cdn-images-1.medium.com/max/800/1*Y3Gi8fxzSg7Q1Mn3EJG5fg.png"></p> <p>On October 12th, we brought together specialists, academics and activists to investigate the main points arising from the “Data, Artificial intelligence and the Global South” conversation. The purpose of the workshop was to explore economic fairness and the Global South. This post shares the questions we asked, some of our findings, and what we need to consider next.</p> <p><a href="https://medium.com/digitalhks/big-data-meager-returns-c7e7beceb3a7">Read more at <em>Digital HKS</em>...</a></p>

tag:tagteam.harvard.edu,2005:FeedItem/2468369 2018-11-28T15:38:37-05:00 2018-11-28T15:38:37-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2468279 2018-11-27T13:16:02-05:00 2018-11-27T13:16:02-05:00

<p>The Governance of Decision - Making Algorithms</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468278 2018-11-27T13:12:34-05:00 2018-11-30T17:07:41-05:00

<p>Artificial intelligence is everywhere but it is considered in a wholly ahistorical way. To understand the impact AI will have on our lives, it is vital to appreciate the context in which the field was established. After all, statistics and state control have evolved hand in hand for hundreds of years.</p> <p><a href="https://www.economist.com/open-future/2018/11/26/ai-thinks-like-a-corporation-and-thats-worrying">Read more at <em>The Economist</em>...</a></p>

tag:tagteam.harvard.edu,2005:FeedItem/2468149 2018-11-26T11:43:37-05:00 2018-11-26T11:43:37-05:00

<p>"Now, with the technology that Waycare has brought to Southern Nevada, the Highway Patrol can proactively deploy units into an area when the system identifies a high likelihood of something happening and prevent crashes before they even happen. "This translates into significantly faster treatment for those injured in crashes, quicker clearance and restoration of normal traffic flow, and untold millions of dollars in commerce related delays that are prevented."</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468146 2018-11-26T11:04:07-05:00 2018-11-26T11:04:07-05:00

<p>A facial recognition system in the city of Ningbo caught Dong Mingzhu, the chair of appliance-making giant Gree Electric, running a red light. Only it turned out not to be Dong, but rather an advertisement featuring her face on the side of a bus, local police said on Weibo Wednesday.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2468145 2018-11-26T10:46:13-05:00 2018-11-26T10:46:13-05:00

<p>Voice recognition and artificial intelligence are being paired to point out potential inaccuracies in real time</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467891 2018-11-21T18:20:44-05:00 2018-11-21T18:20:44-05:00

<p>In 2001 Google adopted the tagline “Don’t Be Evil” to describe its corporate code of conduct. In 2015 when Google was restructured within its parent conglomerate, Alphabet, the motto was replaced by “Do the Right Thing,” which remains the corporation’s tagline today. Though particular to one company, these mottos reflect and help shape attitudes across the technology industry more broadly—even gesturing toward something like a Hippocratic Oath. Although brevity is their virtue, both mottos also raise essential questions, starting with this one: Do they set a high bar or a low bar? And, more to the point, have technologists given enough thought to the world they’re making, to the dislocations they’re enabling, and to the lives and livelihoods they are often unknowingly changing, both here at home and across the world? This quarter, three senior Stanford faculty (a computer scientist, a political scientist, and a philosopher, all award-winning teachers) have joined forces along with invited entrepreneurs, engineers, policy-makers, and investors throughout the tech community in order to ask, and perhaps to answer, some of the profound ethical questions posed by the rapidly expanding and unpredictably evolving technology sector. Among other issues, we will explore data privacy, artificial intelligence, autonomous systems, algorithmic bias, the attention economy, and the power of private platforms. With each topic, we will encounter a range of perspectives, allowing us genuinely to wrestle with the tensions and trade-offs that accompany technological change. With every new innovation, we need to ask: What does this technology enable others to do? What responsibilities does this imply for me as an innovator, a citizen, and a human being? The goal of the course is to bring about a fundamental shift in how we think about our roles as enablers and shapers of technological change in society. The course will challenge all of us to internalize a commitment to our responsibilities as innovators, designers, coders, engineers, policy-makers, citizens, and consumers.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467797 2018-11-20T09:52:10-05:00 2018-11-20T09:52:10-05:00

<p>There have been numerous calls for algorithmic accountability in the government space, including from AI Now, which has developed Algorithmic Impact Assessments as a practical framework for public agency accountability. A 2018 report by Harvard University’s Berkman Klein Center, which studies the internet and society, made recommendations for how government could be more transparent and accountable for the algorithms it uses. Step one involves establishing technical standards that encourage transparency and eliminate bias, and step two introduces procurement guidelines to ensure that government software meets those standards. And between those two steps is algorithmic auditing. Setting technical and ethical standards is an opportunity to rethink some of our social processes, and build a more inclusive vision of the future.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467635 2018-11-19T10:18:12-05:00 2018-11-19T10:18:12-05:00

<p>Given the foreseeable pervasiveness of artificial intelligence (AI) in modern societies, it is legitimate and necessary to ask the question how this new technology must be shaped to support the maintenance and strengthening of constitutional democracy. This paper first describes the four core elements of today's digital power concentration, which need to be seen in cumulation and which, seen together, are both a threat to democracy and to functioning markets. It then recalls the experience with the lawless Internet and the relationship between technology and the law as it has developed in the Internet economy and the experience with GDPR before it moves on to the key question for AI in democracy, namely which of the challenges of AI can be safely and with good conscience left to ethics, and which challenges of AI need to be addressed by rules which are enforceable and encompass the legitimacy of democratic process, thus laws. The paper closes with a call for a new culture of incorporating the principles of democracy, rule of law and human rights by design in AI and a three-level technological impact assessment for new technologies like AI as a practical way forward for this purpose.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467639 2018-11-19T10:18:40-05:00 2018-11-19T10:18:40-05:00

<p>The way forward requires greater attention to these risks at the nation-al level, and attendant regulation. In its absence, technology giants, all of whom are heavily investing in and profiting from AI, will dominate not only the public discourse, but also the future of our core values and democratic institutions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467634 2018-11-19T10:17:44-05:00 2018-11-19T10:17:44-05:00

<p>The Global Indicators of Regulatory Governance project explores how governments interact with the public when shaping regulations that affect their business community. Concerned stakeholders could be professional associations, civic groups or foreign investors. The project charts how interested groups learn about new regulations being considered, and the extent to which they are able to engage with officials on the content. It also measures whether or not governments assess the possible impact of new regulations in their countries (including economic, social and environmental considerations) and whether those calculations form part of the public consultation. Finally, Global Indicators of Regulatory Governance capture two additional components of a predictable regulatory environment: the ability of stakeholders to challenge regulations, and the ability of people to access all the laws and regulations currently in force in one, consolidated place.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467415 2018-11-15T14:22:49-05:00 2018-11-15T14:23:00-05:00

<p>Facebook doesn’t want to be the arbiter of decency when it comes to content policy decisions, similar to how it looked to third-party fact checkers rather than becoming an arbiter of truth. Today on a press call with journalists, Mark Zuckerberg announced that a new external oversight committee would be created in 2019 to handle some of Facebook’s content policy decisions. The body will take appeals and make final decisions. The hope is that beyond the influence of Facebook’s business imperatives or the public’s skepticism about the company’s internal choices, the oversight body can come to the proper conclusions about how to handle false information, calls to violence, hate speech, harassment, and other problems that flow through Facebook’s user generate content network. Users will be able to appeal decisions about content they report or when their content is reported, and Facebook will direct these appeals to the independent body. Zuckerberg said Facebook will be working to get the oversight body up and running over the next year. For now, there are plenty of unanswered questions about who will be on the committee, which of the many appeals it will review, and what ensures it’s truly independent from Facebook’s power.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467407 2018-11-15T12:41:13-05:00 2018-11-15T12:41:13-05:00

<p>“This new project will bring together a mini-think tank of real-world AI experts, tech policy wonks, and sci-fi visionaries to help us imagine an even broader range of possible futures for AI—both the dark futures that we want to avoid, and the brighter tomorrows that we should all work toward.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467402 2018-11-15T11:59:27-05:00 2018-11-15T11:59:27-05:00

<p>Intelligence has always been used as fig-leaf to justify domination and destruction. No wonder we fear super-smart robots</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467401 2018-11-15T11:40:15-05:00 2018-11-15T11:40:15-05:00

<p>Much of the media coverage about the walkout has focused on sexual harassment of women. But the issues go even further. The Google workers argued that sexual harassment and abuse in the workplace are not only about gender and sexuality, but are interconnected with other forms of racialized discrimination that point to entrenched abuses of power.² In targeting the structural practices that serve to maintain inequality, these workers are echoing the demands of social justice movements throughout history: to call out the abuse of power, and to unite the many people who are exploited and discriminated against in demanding an end to these practices, be they contract workers, junior employees, or the many others who experience marginalization.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467397 2018-11-15T11:14:41-05:00 2018-11-15T11:14:41-05:00

<p>A tech boot camp will teach US politicians and policymakers about the potential, and the risks, of artificial intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467396 2018-11-15T11:14:29-05:00 2018-11-15T11:14:29-05:00

<p>The sum to be spent up to 2025, outlined in a draft paper seen by Reuters on “AI made in Germany”, is a measure of the concern felt in Berlin at the challenge that digital technologies and AI pose to Germany’s traditional export-focused manufacturing industries. The paper also stresses the “social policy and labor” aspects of AI, reflecting nervousness in traditionally privacy-conscious Germany at the way rampant technological change might disrupt existing social models.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467389 2018-11-15T11:08:35-05:00 2018-11-15T11:08:35-05:00

<p>The Fairness Toolkit has been developed for UnBias by Giles Lane and his team at Proboscis, with the input of young people and stakeholders. It is one of our project outputs aiming to promote awareness and stimulate a public civic dialogue about how algorithms shape online experiences and to reflect on possible changes to address issues of online unfairness. The tools are not just for critical thinking, but for civic thinking – supporting a more collective approach to imagining the future as a contrast to the individual atomising effect that such technologies often cause.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467280 2018-11-14T09:48:48-05:00 2018-11-20T13:34:43-05:00

<p>It is no secret that technology companies have greased the wheels for human rights abuses around the world — backed by a global web of private sector support and investment that has yielded significant financial returns. For example, the University of Toronto's Citizen Lab recently published research analyzing the use of Internet filtering technology developed by Canadian company Netsweeper in ten countries globally — Afghanistan, Bahrain, India, Kuwait, Pakistan, Qatar, Somalia, Sudan, United Arab Emirates, and Yemen — and concluded these uses likely violated international human rights law. Products like Netsweeper’s Internet filtering systems are often referred to as "dual use" technologies: though they may serve legitimate societal objectives in some cases, they also used to undermine human rights like freedom of expression and privacy. Yet Netsweeper is but one example among a growing number of such dual-use tech companies, within a wider and complex cyber security industry, prepared to facilitate mass censorship and surveillance — and increasingly doing so with the financial backing of specialized and powerful investment firms. This paper cites this and other examples to help document this now billion dollar worldwide market and offers ideas and proposals to help clean it up via stronger human rights norms — including human rights-by-design for dual use technologies — among all stakeholders in the cyber security industry: from governments, to businesses, to their employees and shareholders, to industry associations, and the private investment firms funding it all.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467298 2018-11-14T11:42:29-05:00 2018-11-14T11:42:29-05:00

<p>The restructure, critics argue, breaks a pledge DeepMind made when it started working with the NHS that “data will never be connected to Google accounts or services”. The change has also resulted in the dismantling of an independent review board, created to oversee the company’s work with the healthcare sector, with Google arguing that the board was too focused on Britain to provide effective oversight for a newly global body.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467290 2018-11-14T10:21:14-05:00 2018-11-14T10:21:14-05:00

<p>“HabLatam: Digital skills and misinformation among youth in Latin America” by Fabro Boaz Steibel (Institute for Technology and Society of Rio de Janeiro), Andrés Lombana (Harvard University), Debora Albu (ITS), Diego Cerqueira (ITS), Ezequiel Passerón (Faro Digital), Lionel Brossi (University of Chile), Marco Konopacki (ITS) and Thayane Guimarães (ITS)</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467289 2018-11-14T10:19:43-05:00 2018-11-14T10:19:43-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2467243 2018-11-13T15:48:49-05:00 2018-11-13T15:48:49-05:00

<p>Under new policies in California, New Jersey, New York and beyond, “risk assessment” algorithms recommend to judges whether a person who’s been arrested should be released. These advanced mathematical models — or “weapons of math destruction” as data scientist Cathy O’Neil calls them — appear colorblind on the surface but they are based on factors that are not only highly correlated with race and class, but are also significantly influenced by pervasive bias in the criminal justice system.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467242 2018-11-13T15:46:03-05:00 2018-11-13T15:46:03-05:00

<p>Citizens and lawmakers need to be more proactive about setting the AI agenda—and doing so in a manner that includes the voices of the marginalized.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467191 2018-11-13T11:48:55-05:00 2018-11-13T11:48:55-05:00

<p>As drafted, the bill establishes a five-year federal grant program at the Department of Commerce that would provide $50m to $100m to 50 two- and four-year universities and colleges “in rural and urban areas that have been left behind by the digital divide to create a United States Technology Institute”.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467120 2018-11-12T10:26:01-05:00 2018-11-19T11:32:23-05:00

<p>In response to recent controversies, the field of data science has rushed to adopt codes of ethics. Such professional codes, however, are ill-equipped to address broad matters of social justice. Instead of ethics codes, I argue, the field must embrace politics. Data scientists must recognize themselves as political actors engaged in normative constructions of society and, as befits political work, evaluate their work according to its downstream material impacts on people's lives. I justify this notion in two parts: first, by articulating why data scientists must recognize themselves as political actors, and second, by describing how the field can evolve toward a deliberative and rigorous grounding in a politics of social justice. Part 1 responds to three common arguments that have been invoked by data scientists when they are challenged to take political positions regarding their work. In confronting these arguments, I will demonstrate why attempting to remain apolitical is itself a political stance--a fundamentally conservative one--and why the field's current attempts to promote "social good" dangerously rely on vague and unarticulated political assumptions. Part 2 proposes a framework for what a politically-engaged data science could look like and how to achieve it, recognizing the challenge of reforming the field in this manner. I conceptualize the process of incorporating politics into data science in four stages: becoming interested in directly addressing social issues, recognizing the politics underlying these issues, redirecting existing methods toward new applications, and, finally, developing new practices and methods that orient data science around a mission of social justice. The path ahead does not require data scientists to abandon their technical expertise, but it does entail expanding their notions of what problems to work on and how to engage with society.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2467055 2018-11-09T15:58:18-05:00 2018-11-09T15:58:23-05:00

<p> </p> <p> </p>

tag:tagteam.harvard.edu,2005:FeedItem/2466646 2018-11-05T12:07:19-05:00 2018-11-05T12:07:19-05:00

<p>This month, Access Now launched a report that maps and analyses strategies and proposals for regulation on artificial intelligence (AI) in Europe. The report covers regional strategies from the European Union and the Council of Europe as well as national plans from several member states including France, Finland, Germany, and Italy. Access Now lays out a criteria to assess AI strategies to make sure that the development and deployment of AI is individual-centric and human rights-respecting. Read the full report here.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2466645 2018-11-05T11:25:44-05:00 2018-11-05T11:25:44-05:00

<p>Those who have been following my work know that I’m a big fan of metrics, stats, and comparisons. One question I always get is which Chinese companies have the top AI labs, so this week’s translation features a relatively new project from Leiphone (a leading Chinese new media platform that covers S&amp;T, which we’ve used for ChinAI translations in the past). Earlier in the summer, they launched a database called AI Impact Factors (AI 影响因子):</p>

tag:tagteam.harvard.edu,2005:FeedItem/2466638 2018-11-05T10:24:24-05:00 2018-11-05T10:24:24-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2466639 2018-11-05T10:24:55-05:00 2018-11-05T10:24:55-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2465882 2018-10-26T11:12:47-04:00 2018-10-26T11:12:47-04:00

<p>A.I. presents the challenge of reckoning with our skewed histories, while working to counterbalance our biases, and genuinely recognizing ourselves in each other. This is an opportunity to expand — rather than further homogenize — what it means to be human through and alongside A.I. technologies. This implies changes in many systems: education, government, labor, and protest, to name a few. All are opportunities if we, the people, demand them and our leaders are brave enough to take them on.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2465871 2018-10-26T10:59:30-04:00 2018-10-26T10:59:30-04:00

<p>The pervasive gridlock affecting the traditional global governance approach calls into question the idea of broadening its scope beyond its core remit, and it calls for alternatives, either as substitutes for obsolete arrangements or to address emerging collective action problems in new, inadequately covered fields.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2465870 2018-10-26T10:53:10-04:00 2018-10-26T10:53:10-04:00

<p><span>The use of algorithms in modern data processing techniques, as well as data-intensive technological trends, suggests the adoption of a broader view of the data protection impact assessment. This will force data controllers to go beyond the traditional focus on data quality and security, and consider the impact of data processing on fundamental rights and collective social and ethical values.</span></p> <p> </p> <p><span>Building on studies of the collective dimension of data protection, this article sets out to embed this new perspective in an assessment model centred on human rights (Human Rights, Ethical and Social Impact Assessment-HRESIA). This self-assessment model intends to overcome the limitations of the existing assessment models, which are either too closely focused on data processing or have an extent and granularity that make them too complicated to evaluate the consequences of a given use of data.</span></p> <p>In terms of architecture, the HRESIA has two main elements: a self-assessment questionnaire and an ad hoc expert committee. As a blueprint, this contribution focuses mainly on the nature of the proposed model, its architecture and its challenges; a more detailed description of the model and the content of the questionnaire will be discussed in a future publication drawing on the ongoing research.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2465857 2018-10-26T08:39:56-04:00 2018-10-26T08:39:56-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2465676 2018-10-24T10:43:09-04:00 2018-10-24T10:43:09-04:00

<p>This document presents the final results of UNESCO’s project “Defining Internet Universality Indicators” for the consideration of the 31st session of the IPDC Council. The project describes the background and process to develop the indicators. The results contribute to reinforcing UNESCO’s priority program areas related to promote freedom of expression and media development in the digital age and universal access to information and knowledge, enriching UNESCO’s leadership contribution to global Internet governance and achieving the 2030 Sustainable Development Agenda.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2465628 2018-10-23T14:50:10-04:00 2018-10-23T14:50:10-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2465581 2018-10-23T10:11:19-04:00 2018-10-23T10:11:19-04:00

<p>Spurred on by a mandate in the Secretary-General’s Strategy on New Technologies, the platform is curated by the Centre for Policy Research at United Nations University. The project is led by Eleonore Pauwels, Research Fellow on Emerging Cybertechnologies and a global expert on AI, its ethics, and its convergence with other frontier technologies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2465510 2018-10-22T12:38:16-04:00 2018-10-23T11:33:53-04:00

<p>We propose these Universal Guidelines to inform and improve the design and use of AI. The Guidelines are intended to maximize the benefits of AI, to minimize the risk, and to ensure the protection of human rights. These Guidelines should be incorporated into ethical standards, adopted in national law and international agreements, and built into the design of systems. We state clearly that the primary responsibility for AI systems must reside with those institutions that fund, develop, and deploy these systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2465150 2018-10-17T15:12:47-04:00 2018-10-17T16:44:52-04:00

<p>Spurred on by a mandate given to the United Nations University (UNU) in the Secretary-General’s Strategy on New Technologies, The Centre for Policy Research at the UNU has created an “AI and Global Governance Platform” as an inclusive space for researchers, policy actors, corporate and thought leaders to explore the global policy challenges raised by artificial intelligence.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2464721 2018-10-12T14:30:00-04:00 2018-10-12T14:30:00-04:00

<p>BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE benchmark to 80.4% (7.6% absolute improvement), MultiNLI accuracy to 86.7 (5.6% absolute improvement) and the SQuAD v1.1 question answering Test F1 to 93.2 (1.5% absolute improvement), outperforming human performance by 2.0%.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2464390 2018-10-09T15:35:27-04:00 2018-10-09T15:35:27-04:00

<p>This paper presents the PDPC's preliminary analysis of some of the issues pertinent to the commercial development and adoption of AI solutions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463718 2018-10-01T14:28:06-04:00 2018-10-01T14:28:06-04:00

<p>In this inaugural post, we discuss three areas of technical AI safety: specification, robustness, and assurance. Future posts will broadly fit within the framework outlined here. While our views will inevitably evolve over time, we feel these three areas cover a sufficiently wide spectrum to provide a useful categorisation for ongoing and future research.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463255 2018-09-25T10:51:36-04:00 2018-09-25T10:51:36-04:00

<p>A Facebook user logs in on a smartphone. A new report is calling for stronger regulation of technology companies. A Facebook user logs in on a smartphone. A new report is calling for stronger regulation of technology companies. Photo: Wilfredo Lee/Associated Press 24 Comments By Deepa Seetharaman Sept. 24, 2018 6:00 p.m. ET Silicon Valley tech giants can’t be trusted to police themselves and should be subject to tougher regulation, including around their pattern of acquiring competitors to accumulate ever-larger stores of user data, according to a critical new report released Monday. The business models powering digital advertising platforms like Facebook Inc. and Alphabet Inc.’s Google still undermine user privacy and incentivize disinformation campaigns despite recent efforts by tech companies to prevent abuse, says the report from Harvard’s Shorenstein Center on Media, Politics and Public Policy and New America, a left-leaning Washington-based think tank. “We need to completely reorganize the way that industry works,” said Dipayan Ghosh, who previously worked on privacy and policy issues at Facebook and is now a fellow at the Shorenstein Center. The report lands amid a broad discussion in the industry and in Washington, D.C., about how aggressively lawmakers should move on Silicon Valley’s biggest firms, and comes just days before the Senate Commerce Committee is due to hold a high-profile hearing on the privacy practices of several large tech companies. Google, Amazon.com Inc., Twitter Inc. and Apple Inc. will all send privacy executives to testify at the Senate hearing, along with telecommunications firms AT&amp;T Inc. and Charter Communications Inc. A Facebook representative isn’t scheduled to attend. The authors argue that the federal government should take the lead on regulations rather than states like California, which recently passed a tough new privacy law that the industry broadly opposes. Most big tech companies have said they are open to sensible regulation as long as it doesn’t overly curtail their ability to offer personalized products and services to consumers.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463254 2018-09-25T10:30:56-04:00 2018-09-25T10:30:56-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2463166 2018-09-24T13:13:03-04:00 2018-09-24T13:13:03-04:00

<p>Selena Scola was a content moderator at Facebook’s Menlo Park, California headquarters from June 2017 through March of this year, according to the lawsuit. She worked for a contractor called Pro Unlimited, Inc., which helps Facebook delete content that violates its Community Standards. Facebook has roughly 7,500 content moderators worldwide, who are tasked with deleting hate speech, graphic violence and self harm images and video, nudity and sexual content, bullying, and a host of other content that violates its policies.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463147 2018-09-24T10:36:48-04:00 2018-09-24T10:36:48-04:00

<p>So how do we begin to understand the changes taking place? One method is to look to the courtroom where evidence, expert testimony, and judicial scrutiny often reveal new insights into the current state of systems. Recently, AI Now partnered with NYU Law’s Center on Race, Inequality and the Law and the Electronic Frontier Foundation to conduct an examination of current United States courtroom litigation where the use of algorithms by government was central to the rights and liberties at issue in the case.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463020 2018-09-21T11:03:50-04:00 2018-09-21T19:21:55-04:00

<p>0.38801985560121732 This is analogous to what I saw when I experimented with analyzing restaurant reviews using word embeddings, and found out that all the Mexican restaurants were ending up with lower sentiment for no good reason.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463021 2018-09-21T12:18:42-04:00 2018-09-21T15:29:18-04:00

<p>The MIT algorithm had done all the city could reasonably ask. It had sorted through more possibilities than any human being could possibly contemplate. And it had come up with a solution no bureaucrat had ever mustered. But it was people who made the final call. People with competing interests and a mish-mash of motivations. This was a fundamentally human conflict, and all the computing power in the world couldn’t solve it. If anything, the algorithm fueled the conflict and made the choices stark. Before the district commissioned the formula, few parents had thought about the interplay between high school start times and teenage sleep deprivation. And even fewer understood that starting high school later would mean sending younger kids to school earlier. But even if the algorithm flopped, says Goldsmith, the former Indianapolis mayor who now runs Data-Smart City Solutions at Harvard, it was worth pursuing. “We live in an inherently political world,” he says, “and sometimes, politics are going to trump science. But if the science can illuminate the disparities, that’s better than continuing in ignorance.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463056 2018-09-21T15:06:18-04:00 2018-09-21T15:06:18-04:00

<p>In this paper, I examine five AI ethical dilemmas: weapons and military-related applications, law and border enforcement, government surveillance, issues of racial bias, and social credit systems. I discuss how technology companies are handling these issues and the importance of having principles and processes for addressing these concerns. I close by noting ways to strengthen ethics in AI-related corporate decisions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463053 2018-09-21T15:05:32-04:00 2018-09-21T15:05:32-04:00

<p>More than 2,000 researchers join forces to urge EU to help continent build ‘Google-style’ infrastructure as counterweight to the two leading AI players</p>

tag:tagteam.harvard.edu,2005:FeedItem/2463052 2018-09-21T14:56:56-04:00 2018-09-21T14:56:56-04:00

<p>This article argues that the technical limitations of AI systems should be reckoned with at the time of developing policy, and the societal and ethical concerns that arise due to such limitations should be used to inform what policy processes aspire to achieve. It proposes a framework for such deliberation to occur, by analysing the three main stages of bringing machine learning (the most popular subset of AI techniques) to deployment - the data, model, and application stage. It is written against the backdrop of India’s current AI policy landscape, and applies the proposed framework to ongoing sectoral challenges in India. With a view to influence existing policy deliberation in the country, it focuses on potential risks that arise from data-driven decisions in general, and in the Indian context in particular.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2462857 2018-09-19T15:06:07-04:00 2018-09-19T15:06:07-04:00

<p>This extensible open source toolkit can help you examine, report, and mitigate discrimination and bias in machine learning models throughout the AI application lifecycle. Containing over 30 fairness metrics and 9 state-of-the-art bias mitigation algorithms developed by the research community, it is designed to translate algorithmic research from the lab into the actual practice of domains as wide-ranging as finance, human capital management, healthcare, and education. We invite you to use it and improve it.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2462673 2018-09-17T15:40:03-04:00 2018-09-21T15:56:47-04:00

<p>Researchers and designers at Google's PAIR (People and AI Research) initiative created the What-If visualization tool as a pragmatic resource for developers of machine learning systems. Using the What-If tool reveals, however, one of the hardest, most complex, and most utterly human, questions raised by artificial intelligence systems: What do users want to count as fair?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2461256 2018-08-29T23:30:30-04:00 2018-08-29T23:30:30-04:00

<p>In fact, as technology evolves from voice to data services, and over-the-top platforms, IoT and artificial intelligence become increasingly pervasive, the central policy paradox is that as we increase ICT access and use, digital inequality is also amplified. Not only is the gap between the connected and the unconnected greater, the divide between those minimally connected and those with the resources to use the internet optimally to meet their needs expands.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2461257 2018-08-29T23:55:32-04:00 2018-09-05T11:36:44-04:00

<p>We have published three papers describing a human rights-based blueprint for responsible business practice with regard to AI, both within and beyond the technology sector.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2460793 2018-08-24T02:03:01-04:00 2018-10-03T10:09:30-04:00

<p>This has happened because traditional academic disciplines, as they currently operate, often aren't designed to help students study and apply technical expertise to advance the public interest (as distinct from advancing commercial interests). Students doggedly find their own public interest paths; in fact, the digital generation now in college and graduate school craves meaningful work that will change the world. As long as they can make ends meet, they'll happily work for less money in public interest jobs in government and nonprofits. But most universities haven't provided pathways for these digital natives to cross-train in policy and computer science by working on real problems, or to combine expertise in data science with the capacity to think deeply about the ethical and social implications of the use of digital technology.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2460664 2018-08-23T00:34:20-04:00 2018-09-24T09:51:14-04:00

<p>Wikimedia/Yale Law School Initiative on Intermediaries and Information (WIII) at Yale Law School has released a report detailing the most critical issues necessary for understanding the role of information platforms, such as Facebook and Google, in law and society today. The report synthesizes discussions held at “Beyond Intermediary Liability: The Future of Information Platforms,” an invitation-only workshop convened by WIII, featuring leading experts from industry, civil society, and academia.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2460770 2018-08-23T17:07:46-04:00 2018-08-24T00:13:55-04:00 Alex Stamos

<p>In the swirl of news this week, it would be easy to miss recent announcements from two of America's largest and most influential technology companies that have implications for our democracy as a whole. First, on Tuesday morning, Microsoft <a href="https://www.nytimes.com/2018/08/21/us/politics/russia-cyber-hack.html">revealed</a> that it had detected continued attempts at spear-phishing by APT 28/Fancy Bear, the hacking group tied to Russia’s Main Intelligence Directorate (known as the GRU). Later that day, my friends and former colleagues at Facebook <a href="https://newsroom.fb.com/news/2018/08/more-coordinated-inauthentic-behavior/">unveiled details</a> on more than 600 accounts that were being used by Russian and Iranian groups to distort the information environment worldwide.</p> <p>The revelations are evidence that Russia has not been deterred and that Iran is following in its footsteps. This underlines a sobering reality: America’s adversaries believe that it is still both safe and effective to attack U.S. democracy using American technologies and the freedoms we cherish.</p> <p>And why wouldn’t they believe that? In some ways, the United States has broadcast to the world that it doesn’t take these issues seriously and that any perpetrators of information warfare against the West will get, at most, a slap on the wrist. While this failure has left the U.S. unprepared to protect the 2018 elections, there is still a chance to defend American democracy in 2020.</p> <p>From 2014 until very recently, I worked on security and safety at Yahoo and then at Facebook, both companies on the front line of Russia’s information and cyber-warfare campaign. From that vantage point, the facts are indisputable: There was a multiyear effort by a coalition of Russian agents to harm the likely presidency of Hillary Rodham Clinton and sow deep division in America’s political discourse. The uniformed officers of the GRU and the jeans-wearing millennial trolls of the private Internet Research Agency turned American technology, media and this country’s culture of discourse back against the United States. Stymied by a lack of shared understanding of what happened, the government’s sclerotic response has left the United States profoundly vulnerable to future attacks. As a security leader in my former role at Facebook, my personal responsibility for the failures of 2016 continues to weigh on me, and I hope that I can help elucidate and amplify some hard-learned lessons so that the same mistakes will not be made again and again.</p> <p>The fundamental flaws in the collective American reaction date to summer 2016, when much of the information being reported today was in the hands of the executive branch. Well before Americans went to the polls, U.S. law enforcement was in possession of forensics from the hacks against the Democratic National Committee; important metadata from the GRU’s spear-phishing of John Podesta and other high-profile individuals; and proactive reports from technology companies. Following an acrimonious debate inside the White House, as <a href="https://www.washingtonpost.com/outlook/the-united-states-a-rich-cyber-target-for-hackers-that-fails-to-protect-itself/2018/08/02/a73215e0-8b97-11e8-8aea-86e88ae760d8_story.html">reported by the New York Times’s David Sanger</a>, President Obama rejected several retaliatory measures in response to Russian interference—and U.S. intelligence agencies did not emerge with a full-throated description of Russia’s meddling until after the election.</p> <p>If the weak response of the Obama White House indicated to America’s adversaries that the U.S. government would not respond forcefully, then the subsequent actions of House Republicans and President Trump have signaled that our adversaries can expect powerful elected officials to help a hostile foreign power cover up attacks against their domestic opposition. The bizarre behavior of the chairman of the House Permanent Select Committee on Intelligence, Rep. Devin Nunes, has destroyed that body’s ability to come to any credible consensus, and the relative comity of the Senate Select Committee on Intelligence has not yet produced the detailed analysis and recommendations our country needs. Although by now Americans are likely inured to chronic gridlock in Congress, they should be alarmed and unmoored that their elected representatives have passed no legislation to address the fundamental issues exposed in 2016.</p> <p>Republican efforts to downplay Russia’s role constitute a dangerous gamble: It is highly unlikely that future election meddling will continue to have such an unbalanced and positive impact for the GOP. The Russians are currently the United States’ most visible information-warfare adversaries, but they are not alone. Their proven playbook is now “in the wild” for anyone to use. Recent history has shown that once a large, powerful nation-state actor demonstrates the effectiveness of a technique, many other groups <a href="https://www.nbcnews.com/tech/security/do-we-need-cyber-warfare-treaties-legacy-stuxnet-n8736">rush to build che

tag:tagteam.harvard.edu,2005:FeedItem/2460160 2018-08-19T23:39:51-04:00 2018-08-19T23:39:51-04:00

<p>Abstract That AI will have a major impact on society is no longer in question. Current debate turns instead on how far this impact will be positive or negative, for whom, in which ways, in which places, and on what timescale. In order to frame these questions in a more substantive way, in this prolegomena we introduce what we consider the four core opportunities for society offered by the use of AI, four associated risks which could emerge from its overuse or misuse, and the opportunity costs associated with its under use. We then offer a high-level view of the emerging advantages for organisations of taking an ethical approach to developing and deploying AI. Finally, we introduce a set of five principles which should guide the development and deployment of AI technologies. The development of laws, policies and best practices for seizing the opportunities and minimizing the risks posed by AI technologies would benefit from building on ethical frameworks such as the one offered here.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2459834 2018-08-16T15:28:31-04:00 2018-08-16T15:28:31-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2459796 2018-08-16T09:54:59-04:00 2018-08-16T09:54:59-04:00

<p>The rate at which bad content is reported in Burmese, whether it’s hate speech or misinformation, is low. This is due to challenges with our reporting tools, technical issues with font display and a lack of familiarity with our policies. So we’re investing heavily in artificial intelligence that can proactively flag posts that break our rules.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2459537 2018-08-13T16:03:23-04:00 2018-08-13T16:03:23-04:00

<p>Election meddling, state-sponsored disinformation campaigns, and the potential manipulation of platform users is provoking intense reactions to technology around the world.<span> </span>The outrage following news reports that the data of millions of people were used without their knowledge to train sophisticated targeting tools that may have manipulated voters suggests that consumers expectations of how their data are collected and used do not correspond with the reality of the business models of many data-driven platforms.The allegations, if true, underscore the power of increasingly sophisticated predictive technology and the limitations of the United States’ largely self-regulatory approach to consumer data rights, privacy, and security. These allegations also raise the possibility that regulators, policymakers, consumers, and even the platforms themselves<span> </span>may be significantly underestimating the risks of data-fueled analytics and automated technology. In the last year new risks emerged including the use of technology to: manipulate perceptions and emotion; rapidly disseminate “fake news;” and deceive people through AI-driven systems that can create “deepfakes” (fake video and audio in which a person’s face is substituted). Technology is also being deployed to undermine democratic processes. For example, millions of fake comments were filed with the Federal Communications Commission during its proceeding revising the Open Internet rules, and bad actors used social network platforms to engage in widespread propaganda and disinformation campaigns.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2459532 2018-08-13T15:01:36-04:00 2018-08-13T15:01:36-04:00

<p>New research by scientists from UNICEF’s Office of Innovation — published today in the journal Nature Human Behaviour — applies methods from network science and machine learning to constitutional law. UNICEF Innovation Data Scientists Alex Rutherford and Manuel Garcia-Herranz collaborated with computer scientists and political scientists at MIT, George Washington University, and UC Merced to apply data analysis to the world’s constitutions over the last 300 years. This work sheds new light on how to better understand why countries’ laws change and incorporate social rights.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2459530 2018-08-13T14:41:57-04:00 2018-08-13T14:41:57-04:00

<p>In the country reports by Tactical Tech's partners as part of our research on the global Influence Industry, it is clear that since 2015, WhatsApp is increasingly becoming the main tool for political campaigns in the Global South, and its influence is continuing to grow. According to Brendan Tobin, head of the political-campaign software company Ecanvasser, "...in South America and Central America [...] some campaigns just run on WhatsApp groups."</p>

tag:tagteam.harvard.edu,2005:FeedItem/2459526 2018-08-13T13:44:14-04:00 2018-08-13T13:44:14-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2459491 2018-08-13T10:37:43-04:00 2018-08-13T10:37:43-04:00

<p>Data colonialism combines the predatory extractive practices of historical colonialism with the abstract quantification methods of computing. Understanding Big Data from the Global South means understanding capitalism’s current dependence on this new type of appropriation that works at every point in space where people or things are attached to today’s infrastructures of connection. The scale of this transformation means that it is premature to map the forms of capitalism that will emerge from it on a global scale.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2453830 2018-07-23T12:12:06-04:00 2018-07-23T12:12:06-04:00

<div>This report summarizes findings from a three-year</div> <div>collaboration between the World Economic Forum and</div> <div>The Boston Consulting Group (BCG) to explore how</div> <div>autonomous vehicles could reshape the future of urban</div> <div>mobility. The project built on the collective insights</div> <div>generated from the Autonomous and Urban Mobility</div> <div>Working Group (Working Group) of the System Initiative</div> <div>on Shaping the Future of Mobility, composed of roughly</div> <div>35 business executives from diverse industries (including</div> <div>automotive, technology, logistics, insurance, utilities and</div> <div>infrastructure) that convened for 10 full-day workshops and</div> <div>numerous conference calls.</div>

tag:tagteam.harvard.edu,2005:FeedItem/2453540 2018-07-19T09:48:15-04:00 2018-07-19T09:48:15-04:00

<p>However, the workplace in Japan will not immediately be taken over by machines. Technology and society are mutually related, so we must closely examine how each research area is interacting with ethical, legal, and social issues. In addition, experts in domains such as healthcare and agriculture are also carefully reorganizing their tasks through the use of AI and robots and negotiating numerous variables in the process, such as institutions, organizational cultures, economic efficiency, and human values. These practices can be seen all over the world, not only in Japan; however, there is not as much information on Japan’s practices in this regard due to the language barrier. We hope that this report will contribute to an understanding of Japanese technological, institutional, cultural, and social aspects on AI and robotics.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2453482 2018-07-18T14:48:39-04:00 2018-07-18T14:48:39-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2453481 2018-07-18T14:43:30-04:00 2018-07-18T14:43:30-04:00

<p>Praising the high quality of the paper’s analysis, the Committee expressed particular appreciation for its focus on social equity and the 2030 Agenda’s pledge to leave no one behind. While recogni zing the political complexity of the spheres in which AI operates, members highlighted the potential role the UN system could play in fostering a social compact approach and closing the “AI divide” to be better aligned with universal values and standards</p>

tag:tagteam.harvard.edu,2005:FeedItem/2453366 2018-07-17T13:50:12-04:00 2018-07-17T13:50:12-04:00

<p>To address this challenge, a team of scientists at IBM Research, in collaboration with MIT Media Lab, has developed an AI recommendation technique that, while optimizing its results for the user’s preferences, also makes sure it stays conformant to other constraints, such as ethical and behavioral guidelines. Led by Francesca Rossi, AI Ethics Global Leader at IBM Research, the team of scientists has demonstrated the functionality of the AI in a movie recommendation system that allows parents to set moral constraints for their children.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2453331 2018-07-17T09:30:55-04:00 2018-07-17T12:56:09-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2452702 2018-07-09T12:50:33-04:00 2018-07-17T09:31:06-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2452008 2018-06-28T15:54:20-04:00 2018-06-28T15:54:20-04:00

<p>Fernando Diaz: One of the reasons we’re concerned about bias in data is that the trained model will be biased when it’s deployed. And so, step one is to be able to detect whether or not the actions of an artificial intelligence are biased themselves and, if they are, how do I go back and retrain that algorithm, or add constraints to the algorithm, so it doesn’t learn the biases from the data? And so, my work to date has focused primarily on the measurement side of things. On the measurement side of things, it has more to do with understanding the users that are coming into the system, what they’re asking for and whether or not the system, by virtue of the fact of who the user is or what population they’re coming from, is behaving in a way that you would consider biased. And that requires a lot of the expertise from the information retrieval community who have been thinking a lot about measurement and evaluation for almost since the beginning of the research agenda of the community in the 50s. And so, this is what makes it a good natural fit between auditing and measurement and information retrieval.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2452003 2018-06-28T15:27:45-04:00 2018-06-28T15:27:45-04:00

<p>After fact-checkers repeatedly flagged Amor a México’s content as problematic, Facebook this month punished the page with its highest-level “demotion,” dramatically reducing its likes, shares and other interactions to 17,000 on June 3 from 121,000 four days prior. Still, Amor a México has doubled its followers to 300,000 in the past few months. Facebook said it was investigating the page but declined to share any information about it.</p> <p>In interviews, executives conceded that determining the origin and motivation of many page operators is too great an effort for a private company to manage. Instead, the focus is on limiting the reach of serial offenders, punishing behaviors without often being able to get to the source. The brunt of Facebook’s news vetting in Mexico falls to a small group of third-party fact-checkers, whose job is to play whack-a-mole — debunking one story at a time, with each taking several days to disprove.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2451759 2018-06-26T09:50:35-04:00 2018-06-26T09:50:35-04:00

<p>Terms of service of on-line platforms too often contain clauses that are potentially unfair to the consumer. We present an experimental study where machine learning is employed to automatically detect such potentially unfair clauses. Results show that the proposed system could provide a valuable tool for lawyers and consumers alike.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2451727 2018-06-25T18:10:07-04:00 2018-06-25T18:10:07-04:00

<p>Angry users bring up the issue of political bias largely because the platforms themselves have so consistently and nervously shied away from anything that could be construed as censorship. But in 2018, ideas of political bias and censorship feel almost irrelevant, given the reality of how these platforms actually operate. The discussion of complete neutrality is now a quaint notion when nearly every discussion that takes place on these networks is weaponized toward a political end. It’s a luxurious philosophical debate that the platforms have little time for. Instead, they should be focused on preventing harm.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2451703 2018-06-25T12:58:16-04:00 2018-06-25T12:58:16-04:00

<p>Many modern Artificial Intelligence (AI) systems make use of data embeddings, particularly in the domain of Natural Language Processing (NLP). These embeddings are learnt from data that has been gathered "from the wild" and have been found to contain unwanted biases. In this paper we make three contributions towards measuring, understanding and removing this problem. We present a rigorous way to measure some of these biases, based on the use of word lists created for social psychology applications; we observe how gender bias in occupations reflects actual gender bias in the same occupations in the real world; and finally we demonstrate how a simple projection can significantly reduce the effects of embedding bias. All this is part of an ongoing effort to understand how trust can be built into AI systems.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2451352 2018-06-20T15:45:56-04:00 2018-06-20T15:45:56-04:00

<p>One big obstacle to accountability is that agencies often do not collect data on how the tools are used or their performance, says Goel. “A lot of times there’s no transparency because there’s nothing to share.” The California legislature, for instance, has a draft bill that calls for risk-assessment tools to help reduce how often defendants must pay bail — a practice that has been criticized for penalizing lower-income defendants. Goel wants the bill to mandate that data are collected on instances when judges disagree with the tool and on specific details, including outcomes, of every case. “The goal is fundamentally to decrease incarceration while maintaining public safety,” he says, “so we have to know — is that working?”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2450788 2018-06-13T13:25:06-04:00 2018-06-13T13:25:06-04:00

<p>Today, we’re announcing a Google AI research center in Africa, which will open later this year in Accra, Ghana. We’ll bring together top machine learning researchers and engineers in this new center dedicated to AI research and its applications.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2450642 2018-06-12T10:38:11-04:00 2018-06-13T14:00:59-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2450123 2018-06-05T20:24:35-04:00 2018-06-05T20:24:35-04:00

<p>Passing judgement on a series of legal questions referred to it, the CJEU has held that the administrator of a fan page on Facebook is jointly responsible with Facebook for the processing of the data of visitors to the page — aligning with the the Advocate General’s opinion to the court, which we covered back in October. In practical terms the ruling means tech giants could face more challenges from European data protection authorities. While anyone piggybacking on or plugging into platform services in Europe shouldn’t imagine they can just pass responsibility to the platforms for ensuring they are compliant with privacy rules.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2450085 2018-06-05T13:27:55-04:00 2018-06-05T13:27:55-04:00

<p>The award was announced today by Mr S Iswaran, Minister for Communications and Information, and Minister-in-charge of Trade Relations, at the Opening of innovfest unbound, the anchor event of Smart Nation Innovations Week. The grant call by NRF and IMDA was made with the aim of collaborating with an IHL to set up the aforementioned Programme, so as to achieve various strategic objectives, including to (i) promote cutting edge thinking and practices in AI and data policies and regulations; (ii) inform AI and data policy and regulation formulation in Singapore through research publications and stakeholder engagement events and activities; and (iii) establish Singapore as a global thought leader in AI and data policies and regulations.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2449958 2018-06-04T11:56:48-04:00 2018-06-04T11:56:48-04:00

<p>The internet search company — widely regarded as having the most advanced AI — is planning to publish an ethical framework to guide its use of the technology this week. The move follows a decision not to renew a drone contract with the US Department of Defense. The Pentagon has been using Google’s vision technology to help drones interpret objects on the ground, prompting fierce opposition inside the company to its co-operation with the military.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2449951 2018-06-04T11:27:44-04:00 2018-06-04T11:27:44-04:00

<p>From a technical point of view, we’ve found that even when explicit information about race, gender, age and socioeconomic status is withheld from models, part of the remaining data often continues to correlate with these categories, serving as a proxy for them. A person’s postal code, for instance, tends to reveal much about their protected characteristics. Directly removing information about protected attributes therefore does little to shield people from discrimination — and may even make things worse. Commenting on this problem, Silvia Chiappa, a research scientist here at DeepMind, observes that ‘information about group membership is often needed to disentangle complex patterns of causation and to protect people from indirect discrimination.’</p>

tag:tagteam.harvard.edu,2005:FeedItem/2449939 2018-06-04T10:57:59-04:00 2018-06-04T10:57:59-04:00

<p>Our new report, launched today, argues that the public needs to be engaged early and more deeply in the use of AI if it is to be ethical. One reason why is because there is a real risk that if people feel like decisions about how technology is used are increasingly beyond their control, they may resist innovation, even if this means they could lose out on benefits.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2449927 2018-06-04T09:18:18-04:00 2018-06-04T09:18:18-04:00

<p>Mr Tessier-Lavigne said that the university would begin earmarking funds this autumn for research and faculty appointments linked to the new strategy, ahead of fresh fundraising and broader announcements next year. He said there was scope for companies and philanthropists to contribute but support “has to respect academic values, freedom to publish and objectivity”.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2449928 2018-06-04T09:18:48-04:00 2018-06-04T09:18:48-04:00

<p>Citing the enormous power of artificial intelligence to benefit but also disrupt society, Microsoft President Brad Smith on Thursday, March 1, called for standards of accountability and a “Hippocratic oath” among technologists to do no harm with the emerging tools.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2448948 2018-05-22T15:05:46-04:00 2018-05-22T15:05:46-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2448925 2018-05-22T11:53:57-04:00 2018-05-22T11:53:57-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2445312 2018-05-16T17:44:51-04:00 2018-05-16T17:44:51-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2443441 2018-05-14T16:14:18-04:00 2018-05-14T16:14:18-04:00

<p>But given the wide-ranging implications of AI on societies and individuals, it is certain that internationally protected human rights will be affected by developments in this field. The goal of this blog is not so much to dive into concrete examples of human rights harm, but rather to have a discussion about what ‘value’ human rights may ‘add’ to the AI debate. We will then briefly go into the peripheral role of human rights in the AI debate to date and indicate what the global human rights community, from UN bodies to domestic human rights institutions and human rights NGOs, can do to ensure that human rights become part and parcel of discussions on the future of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2433266 2018-05-01T13:01:16-04:00 2018-05-01T13:01:16-04:00

<p>Mrs May responded by asking Educational Testing Services, the US-based company that ran the system, to analyse voice files to work out if students were using proxies to sit the tests for them. Conversations with those affected have convinced me that many were entirely innocent. Their treatment has been shameful Stephen Timms, MP for East Ham After analysing the files, ETS said it was confident 33,725 test takers had used someone else to sit the test for them. It added that it had “limited confidence” that there had been fraud in another 22,694 cases. By the end of 2016, when it stopped reporting the figures, the Home Office said it had revoked the visas of 35,870 people as a result. Most were told they had no right to appeal and should leave the country.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2433136 2018-05-01T10:33:37-04:00 2018-05-01T10:33:37-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2427346 2018-04-23T12:59:21-04:00 2018-04-23T12:59:21-04:00

<p>Given a trained text embedding model, we can directly measure the associations the model has between words or phrases. Many of these associations are expected and are helpful for natural language tasks. However, some associations may be problematic or hurtful. For example, the ground-breaking paper by Bolukbasi et al. [4] found that the vector-relationship between "man" and "woman" was similar to the relationship between "physician" and "registered nurse" or "shopkeeper" and "housewife" in the popular publicly-available word2vec embedding trained on Google News text.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2427345 2018-04-23T12:52:55-04:00 2018-04-23T12:52:55-04:00

<p>McKinsey Global Institute’s latest research on artificial intelligence focuses on how AI will affect business, analyzing 400 use cases — from airlines avoiding flight cancellations to online retailers recommending items — across 19 industries.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2425666 2018-04-20T15:19:43-04:00 2018-04-20T15:19:43-04:00

<p>This confluence of ideas and technology trends has been rebranded as “AI” over the past few years. This rebranding is worthy of some scrutiny.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2424692 2018-04-19T13:27:13-04:00 2018-04-19T13:27:13-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2423070 2018-04-17T14:35:53-04:00 2018-04-17T14:35:53-04:00

<p>On 27 March 2018, the European Political Strategy Centre hosted six leading international experts for a High-Level Hearing on ‘A European Union Strategy for Artificial Intelligence’ to intellectually accompany the European Commission’s ongoing public consultation on Artificial Intelligence. During the Hearing, the experts were asked to address a set of predetermined questions with the knowledge that a full transcript of the Hearing would be submitted as a public contribution to the consultation.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2422067 2018-04-16T12:10:45-04:00 2018-04-16T12:10:45-04:00

<p>Despite the fact that sinuous news cycles and crisis-prone economics drive much of the interest in ethics in CS, there’s hope that remaining consistent in teaching what’s important to consider in an ethical computer science career will be enough in the long run. As Jasmine puts it, “I think that interest in ethics ebbs and flows a lot, just based on the world, but I think that focusing on incorporating ethics into the undergraduate education is something that can help nudge that ebb and flow in the right direction.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2419487 2018-04-12T15:28:01-04:00 2018-04-12T15:28:01-04:00

<p>Ultimately, we’re talking about choice architecture, a term for the way that information or products are presented to people in a manner that takes into account individual or societal welfare while preserving consumer choice. The presentation of choices has an impact on what people choose, and social networks' recommender systems are a key component of that presentation; they are already curating the set of options. This is the idea behind the “nudge”—do you put the apples or the potato chips front and center on the school lunch line?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2419421 2018-04-12T14:07:14-04:00 2018-04-12T14:07:14-04:00

<p>The tension between advancing engineering and science while respecting benefits to individuals and to society has been at the heart of decades of careful thinking about how best to define ethical principles, and at what granularity or generality. Both in the research community and in information security, a few ethical principles set out to guide how we think about impact of research and technology. As codified by the Belmont5 and Menlo6 reports, these may be summarized briefly as:</p>

tag:tagteam.harvard.edu,2005:FeedItem/2416375 2018-04-09T13:30:42-04:00 2018-04-09T13:30:42-04:00

<p>Trustworthiness goes hand in hand with ethics, and if trust is to be regained, lots of work needs to be done to close gaping ethics gaps. Here, I’ll talk about three areas that need improvement: ethics in design, ethics training and procedures at work, and ethics education and policy at school. Throughout the conversation, I’ll explain why it’s so damned hard for tech companies to live up to high ethical expectations, and I’ll offer suggestions for improving upon the status quo.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2412054 2018-04-03T17:22:32-04:00 2018-04-03T17:22:32-04:00

<p>Fake news that you can “see” will be politically explosive in many ways. Human rights groups will have to pour limited resources into sorting real evidence from bot-driven simulacra. How and at what cost will investigators demonstrate that satellite images of mass graves or chemical weapons attacks are real instead of fabricated — or vice versa? Denials and counternarrative campaigns will become increasingly sophisticated. Investigators and non-govermental organizations will struggle to keep up.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2411724 2018-04-03T11:50:38-04:00 2018-04-03T11:50:38-04:00

<p>As part of continuing reflections on the legal and ethical issues raised by autonomous weapons systems, the ICRC convened a round-table meeting in Geneva on 28-29 August 2017 to explore the ethical aspects.</p> <p> This report - "Ethics and autonomous weapon systems: An ethical basis for human control?" - summarizes discussions and highlights the ICRC's main conclusions:</p>

tag:tagteam.harvard.edu,2005:FeedItem/2411568 2018-04-03T10:07:34-04:00 2018-04-03T10:07:34-04:00

<p>This new discipline is concerned with the scientific study of machines, not as engineering artifacts, but as a new class of actors with their unique behavioral patterns and ecology. Crucially, this field overlaps with, but is distinct from computer science and robotics, as it treats machine behavior observationally and experimentally, without necessarily appealing to the machine’s internal mechanisms. Machine Behavior is akin to how the fields of animal behavior—also known as ethology—and behavioral ecology study the behavior of animals without necessarily focusing on physiology or biochemistry.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2410777 2018-04-02T10:10:45-04:00 2018-05-07T15:34:01-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2411081 2018-04-02T16:08:25-04:00 2018-04-02T16:08:25-04:00

<p>The point of making AI more ethical is so it won’t reproduce the prejudices of random jerks on the internet. Community participation throughout the design process of new AI tools is a good idea, but let’s not do it by having trolls decide ethical questions. Instead, representatives from the populations affected by technological change should be consulted about what outcomes they value most, what needs the technology should address and whether proposed designs would be usable given the resources available. Input from residents of heavily policed neighbourhoods would have revealed that a predictive policing system trained on historical data would exacerbate racial profiling. Having a person of colour on the design team for that soap dispenser should have made it obvious that a peachy skin tone detector wouldn’t work for everyone. Anyone who has had a stalker is sure to notice the potential abuses of selfie drones. Diversifying the pool of talent in AI is part of the solution, but AI also needs outside help from experts in other fields, more public consultation and stronger government oversight.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2410776 2018-04-02T10:10:30-04:00 2018-04-02T10:10:30-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2410327 2018-04-01T16:33:00-04:00 2018-04-01T16:33:00-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2409170 2018-03-30T14:29:28-04:00 2018-03-30T14:29:28-04:00

<p>These scenarios tell us that while machine learning can do incredibly good things for this world, those benefits are not inevitable. We need to look closely at the ways discrimination can creep into these systems, and the ways that companies can act proactively to secure a bright future for machine learning. To that end, we’ve recommend eight steps all companies involved in machine learning can and should do to maximize the shared benefit of this game-changing technology while minimizing real risks to human rights:</p>

tag:tagteam.harvard.edu,2005:FeedItem/2408946 2018-03-30T09:39:43-04:00 2018-03-30T09:39:43-04:00

<p>Currently there is no standard way to identify how a dataset was created, and what characteristics, motivations, and potential skews it represents. To begin to address this issue, we propose the concept of a datasheet for datasets, a short document to accompany public datasets, commercial APIs, and pretrained models. The goal of this proposal is to enable better communication between dataset creators and users, and help the AI community move toward greater transparency and accountability. By analogy, in computer hardware, it has become industry standard to accompany everything from the simplest components (e.g., resistors), to the most complex microprocessor chips, with datasheets detailing standard operating characteristics, test results, recommended usage, and other information. We outline some of the questions a datasheet for datasets should answer. These questions focus on when, where, and how the training data was gathered, its recommended use cases, and, in the case of human-centric datasets, information regarding the subjects’ demographics and consent as applicable. We develop prototypes of datasheets for two well-known datasets: Labeled Faces in The Wild [33] and the Pang &amp; Lee Polarity Dataset</p>

tag:tagteam.harvard.edu,2005:FeedItem/2407362 2018-03-28T11:44:13-04:00 2018-03-28T11:44:13-04:00

<p>Deployed wisely, AI holds the promise of addressing some of the world’s most intractable challenges. But the significance of its positive impact is mirrored by its likely destabilising effects on some aspects of economic and social life. Our paper explores the opportunities and ethical challenges that come with AI and focuses on how Europe can sharpen its competitive edge vis-à-vis other leading economies, such as the United States and China.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2407318 2018-03-28T10:48:53-04:00 2018-03-28T10:48:53-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2406812 2018-03-27T16:58:42-04:00 2018-03-27T16:58:42-04:00

<p>Like Sulston, we should look for ways of making data available for the common interest. Intellectual-property rights expire after a fixed time period: what if, similarly, technology companies were allowed to use the data that they gather only for a limited period, say, five years? The data could then revert to a national charitable corporation that could provide access to certified researchers, who would both be held to account and be subject to scrutiny that ensure the data are used for the common good.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2405547 2018-03-26T10:35:14-04:00 2018-03-26T10:35:14-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2403943 2018-03-23T17:24:27-04:00 2018-03-23T17:24:27-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2400762 2018-03-20T12:03:49-04:00 2018-03-20T12:03:49-04:00

<p>The ethical questions surrounding self-driving cars—and especially a conundrum known as the “trolley problem,” which requires that a car choose between two potential victims in an accident—have confused the issue, he adds: “I don’t think the trolley-problem conversation has been at all helpful.” Of the accident that killed the woman in Arizona, he says, “The sensors probably didn’t pick her up, or the algorithm didn’t understand what it’s seeing.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2398280 2018-03-16T15:23:28-04:00 2018-03-16T15:23:28-04:00

<p>In order for any research area to benefit all of society, it must foster the inclusion of a diverse group of researchers, especially those who have been historically excluded. About a year ago, to address the lack of representation of Black and African researchers in AI, along with Timnit Gebru, Moustapha Cisse, and others, I co-founded the Black in AI group. Black in AI is a place for sharing ideas, fostering collaborations, and discussing initiatives to increase the presence of Black researchers in the field. What started as a small Facebook group has now grown to include over 600 members from Africa, Latin America, Asia, Europe, and North America. Over 200 members are women, and we are deliberate about prioritizing the participation of African women. Members of the group span a wide range of career stages, from undergraduates to senior professors, engineers, and entrepreneurs. We work on applications of AI in health-care, algorithmic fairness, development, art and many many other fields.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2398212 2018-03-16T14:31:12-04:00 2018-03-16T14:31:12-04:00

<p>This brings me to another interesting point about computerized AI, as opposed to corporatized AI: AI algorithms tend to embody the prejudices and beliefs of the programmers. A couple of years ago I ran across an account of a webcam developed by mostly-pale-skinned silicon valley engineers that have difficulty focusing or achieving correct colour balance when pointing at dark-skinned faces. That's an example of human-programmer-induced bias. But with today's deep learning, bias can creep in via the data sets the neural networks are trained on. Microsoft's first foray into a conversational chatbot driven by machine learning, Tay, was yanked offline within days because when 4chan and Reddit based trolls discovered they could train it towards racism and sexism for shits and giggles.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2397870 2018-03-16T08:50:57-04:00 2018-03-16T08:50:57-04:00

<p>Most of the discourse on algorithmic decision-making, whether it comes in the form of praise or warning, assumes that algorithms apply to a static world. But automated decision-making is a dynamic process. Algorithms attempt to estimate some difficult-to-measure quality about a subject using proxies, and the subjects in turn change their behavior in order to game the system and get a better treatment for themselves (or, in some cases, to protest the system.) These behavioral changes can then prompt the algorithm to make corrections. The moves and counter-moves create a dance that has great import to the fairness and efficiency of a decision-making process. And this dance can be structured through law. Yet existing law lacks a clear policy vision or even a coherent language to foster productive debate.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2396196 2018-03-14T11:24:59-04:00 2018-03-14T11:24:59-04:00

<p>The report is organized in four parts: (1) Context - I place China’s AI strategy in the context of its past science and technology plans, as well as other countries’ AI plans; (2) Components - I relate the key features of China’s AI strategy to the drivers of AI development (e.g. data, talented scientists); (3) Capabilities - I assess China’s current AI capabilities by constructing a novel index to measure a country’s AI potential; (4) Consequences - I highlight the potential implications of China’s AI dream for issues of AI safety, national security, economic development, and social governance. In each of these four parts, I dispel a common misconception about China’s approach to AI (Table 1). Then, using the deconstruction of these myths as a starting point, I derive my own findings. What follows is a summary of the key findings in each section.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2395528 2018-03-13T15:16:47-04:00 2018-03-13T15:16:47-04:00

<p>This statement calls for the launch of a process that would pave the way towards a common, internationally recognised ethical and legal framework for the design, production, use and governance of artificial intelligence, robotics, and ‘autonomous’ systems. The statement also proposes a set of fundamental ethical principles, based on the values laid down in the EU Treaties and the EU Charter of Fundamental Rights, that can guide its development.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2394199 2018-03-12T11:21:14-04:00 2018-03-13T17:54:59-04:00

<p>Sir Tim Berners-Lee, the inventor of the world wide web, has called for large technology firms to be regulated to prevent the web from being “weaponised at scale”.</p> <p>Berners-Lee, in an open letter to mark the 29th anniversary of his invention, said: “In recent years, we’ve seen conspiracy theories trend on social media platforms, fake Twitter and Facebook accounts stoke social tensions, external actors interfere in elections, and criminals steal troves of personal data.”</p> <p>These problems have proliferated because of the concentration of power in the hands of a few platforms – including Facebook, Google, and Twitter – which “control which ideas and opinions are seen and shared”.</p> <p>“What was once a rich selection of blogs and websites has been compressed under the powerful weight of a few dominant platforms,” said the 62-year-old British computer scientist.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2394200 2018-03-12T11:22:02-04:00 2018-03-13T17:32:40-04:00

<p>The impact of extralegal coercion will be far reaching. Unlike national laws that are limited by geographic borders, terms-of-service agreements apply to platforms’ services on a global scale. Whereas local courts can order platforms only to block material viewed in their jurisdictions, a blacklist database raises the risk of global censorship. Companies should counter the serious potential for censorship creep with definitional clarity, robust accountability, detailed transparency, and ombudsman oversight.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2394176 2018-03-12T11:04:14-04:00 2018-03-12T11:04:14-04:00

<p>Adequately facing these challenges will require commitments from many of our largest institutions. Universities are uniquely positioned to foster connections between computer science and traditionally unrelated departments like the social sciences and even humanities, through interdisciplinary projects, courses and seminars. Governments can make a greater effort to encourage computer science education, especially among young girls, racial minorities and other groups whose perspectives have been underrepresented in A.I. And corporations should combine their aggressive investment in intelligent algorithms with ethical A.I. policies that temper ambition with responsibility.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2394131 2018-03-12T09:50:04-04:00 2018-03-12T09:50:04-04:00

<p>Sheldon County is a podcast that will never sound the same twice. Every time someone listens to it, they’ll begin by typing a random number into a website. This “seed” will set in motion a Rube Goldberg machine of calculation that will create characters, relationships, jealousies, betrayals, and maybe even a murder or two. These plot points will be turned into a text narrative, read aloud by a voice synthesizer, and then zipped up into an audio file. Each time it will be a unique version of Sheldon County’s story. A podcast made just for you.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2392527 2018-03-09T15:09:55-05:00 2018-03-09T15:09:55-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2392128 2018-03-09T08:24:04-05:00 2018-03-09T08:24:04-05:00

<p>Artificial intelligence and machine learning can help the news media with its three core problems: The overabundance of information and sources that leave the public confused The credibility of journalism in a world of disinformation and falling trust and literacy The Business model crisis – how can journalism become more efficient – avoiding duplication; be more engaged, add value and be relevant to the individual’s and communities’ need for quality, accurate information and informed, useful debate.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2390887 2018-03-07T17:25:56-05:00 2018-03-07T17:25:56-05:00

<p>However, legal automation can also elide or exclude important human values, necessary improvisations, and irreducibly deliberative governance. Due process, appeals, and narratively intelligible explanation from persons, for persons, depend on forms of communication that are not reducible to software. Language is constitutive of these aspects of law. To preserve accountability and a humane legal order, these reasons must be expressed in language by a responsible person. This basic requirement for legitimacy limits legal automation in several contexts, including corporate compliance, property recordation, and contracting. A robust and ethical legal profession respects the flexibility and subtlety of legal language as a prerequisite for a just and accountable social order. It ensures a rule of persons, not machines.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2388760 2018-03-05T13:42:06-05:00 2018-03-05T13:42:06-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2385738 2018-03-01T11:28:11-05:00 2018-03-01T11:28:11-05:00

<p>This experience led us to the idea that perhaps we could measure aspects of the health of the public sphere—in terms of communication exchanges between groups or tribes—grounded in data from public social media and other public media sources. As a starting point, we are developing a set of health indicators for the U.S. (with the potential to expand to other nations) aligned with four principles of a healthy public sphere: Shared Attention: Is there overlap in what we are talking about? Shared Reality: Are we using the same facts? Variety: Are we exposed to different opinions grounded in shared reality? Receptivity: Are we open, civil, and listening to different opinions?</p>

tag:tagteam.harvard.edu,2005:FeedItem/2385164 2018-02-28T17:22:07-05:00 2018-02-28T17:27:32-05:00

<p>ipation about robotics and automation and how they will impact the future of work. Critics are talking about the coming disruption of machine learning, algorithms, chatbots, voice recognition and intelligent automation. What do you make of that pessimistic view? Frank: Based on our research at MIT, I think that the automation trend is not so alarming—in terms of current technologies and technologies in the pipeline today. But there are certain jobs and certain industries that should be concerned about automation.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2385163 2018-02-28T17:18:40-05:00 2018-02-28T17:18:40-05:00

<p>To me, then, this highlights an important path forward. Clearly, machine learning is incredibly useful—and, in particular, machine learning is useful for social science. But we must treat machine learning for social science very differently from the way we treat machine learning for, say, handwriting recognition or playing chess. We cannot just apply machine learning methods in a black-box fashion, as if computational social science were simply computer science plus social data. We need transparency. We need to prioritize interpretability—even in predictive contexts. We need to conduct rigorous, detailed error analyses. We need to represent uncertainty. But, most importantly, we need to work with social scientists in order to understand the ethical implications and consequences of our modeling decisions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2383042 2018-02-26T13:05:07-05:00 2018-02-26T15:32:05-05:00

<p>A predictive policing algorithm might perfectly forecast where crime will occur (hence satisfying notions of fairness), but if those predictions are used by police departments to harass and oppress communities of color, then the algorithm is unjust.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2375074 2018-02-15T19:25:36-05:00 2018-02-15T19:25:36-05:00

<p>To get at whether there are any guiding ethical principles when it comes to the kinds of ambiguous scenarios a car might face, Rahwan and others in his lab created an online poll of sorts called the Moral Machine, in which participants are presented with morally charged setups similar to the one in the illustration. The Moral Machine went viral. Almost three million people from all over the world have clicked through the scenarios. Rahwan’s team translated the experience into nine languages including Arabic, Chinese, and Russian. They’re still working up the results but a couple of things are clear. “Almost every person has something to say about this,” observes postdo</p>

tag:tagteam.harvard.edu,2005:FeedItem/2374112 2018-02-14T13:04:10-05:00 2018-02-14T13:04:10-05:00

<p>Yet these systems can only be understood as the flawed extensions of human creation. They act on our biases by replicating them and distributing them into the background of everyday life, thereby reinforcing and even exacerbating existing structural inequalities.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2374111 2018-02-14T12:53:39-05:00 2018-02-14T12:53:39-05:00

<p>But as AI grows in importance, some believe the scope of the problems the industry faces will extend beyond the ability of legacy agencies to keep up. Some prominent voices, including Elon Musk, a founder of research group OpenAI, and former Department of Justice attorney-adviser Andrew Tutt, have proposed going so far as to establish a standalone federal agency that would oversee the development of AI, much like the Federal Communications Commission and Food and Drug Administration do for their respective industries.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2373601 2018-02-13T16:32:07-05:00 2018-02-13T16:32:07-05:00

<p>But Africa need not dread the age of robotics and automation. Across the continent, from Ghana to Zimbabwe, this technology has the potential to bring myriad positive changes in sectors such as health care and finance, bridging the gap between physical infrastructure inadequacies and consumer demands, while freeing up more time for skilled labor and increased labor productivity. For Africans to reap these benefits, African governments, investors, and NGOs must prepare for the fourth industrial revolution’s transformation of the modern workplace by training workers for complex tasks, and reforming laws and education to meet the demands of tomorrow.       </p>

tag:tagteam.harvard.edu,2005:FeedItem/2373414 2018-02-13T12:21:59-05:00 2018-02-13T12:21:59-05:00

<p>Given the significant economic and social impact of these developments, policymakers should be leading the charge to enable data-driven innovation in their states. This effort should encompass not just the development and growth of data-driven companies, but also the use of data to address their states’ most important priorities, such as improving health care, reducing crime, developing sustainable communities, and improving schools.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2373223 2018-02-13T08:50:49-05:00 2018-02-13T08:50:49-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2372634 2018-02-12T11:55:23-05:00 2018-02-12T11:55:23-05:00

<p>“Unilever will not invest in platforms or environments that do not protect our children or which create division in society, and promote anger or hate,” Unilever Chief Marketing Officer Keith Weed is expected to say Monday during the Interactive Advertising Bureau’s annual leadership meeting in Palm Desert, Calif.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2371620 2018-02-10T12:53:36-05:00 2018-02-10T12:53:36-05:00

<p>This paper seeks to provide a foundation for planning a foreign policy strategy that responds effectively to the emerging power of AI in international affairs. The developments in AI are so dynamic and the implications so wide-ranging that ministries need to begin engaging immediately. That means starting with the assets and resources at hand while planning for more significant changes in the future. Many of the tools of traditional diplomacy can be adapted to this new field. While the existing toolkit can get us started, this pragmatic approach does not preclude thinking about more drastic changes that the technological changes might require for our foreign policy institutions and instruments.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2371211 2018-02-09T16:42:29-05:00 2018-02-09T16:42:29-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2370473 2018-02-08T18:27:28-05:00 2018-02-08T18:27:28-05:00

<p>The ease of transformation from ethical to unethical robot is hardly surprising. It is a straightforward consequence of the fact that both ethical and unethical behaviors require the same cognitive machinery with—in our implementation—only a subtle difference in the way a single value is calculated. In fact, the difference between an ethical (i.e. seeking the most desirable outcomes for the human) robot and an aggressive (i.e. seeking the least desirable outcomes for the human) robot is a simple negation of this value.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2369590 2018-02-07T16:29:27-05:00 2018-02-07T16:29:27-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2369540 2018-02-07T16:00:30-05:00 2018-02-07T16:00:30-05:00

<p>The skewed accuracy appears to be due to underrepresentation of darker skin tones in the training data used to create the face-analysis algorithms.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2367974 2018-02-05T16:01:33-05:00 2018-02-05T16:01:33-05:00

<p>This article argues that once ML diagnosticians, such as those based on neural networks, are shown to be superior, existing medical malpractice law will require superior ML-generated medical diagnostics as the standard of care in clinical settings. Further, unless implemented carefully, a physician’s duty to use ML systems in medical diagnostics could, paradoxically, undermine the very safety standard that malpractice law set out to achieve.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2367776 2018-02-05T11:28:18-05:00 2018-02-05T11:28:18-05:00

<p>But some experts said Facebook’s AI teams have, like at other tech giants, struggled with the natural tension between the teams in research, with their long timelines and academic pursuits, and application, whose mandate involves building features that actually work.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2367768 2018-02-05T11:23:34-05:00 2018-02-05T11:23:34-05:00

<p>In the context of machine learning, disparate impact refers to a form of systematic discrimination whereby the output distribution of a model depends on the value of a sensitive attribute (e.g., race or gender). In this paper, we present an information-theoretic framework to analyze the disparate impact of a binary classification model. We view the model as a fixed channel, and quantify disparate impact as the divergence in output distributions over two groups. We then aim to find a \textit{correction function} that can be used to perturb the input distributions of each group in order to align their output distributions. We present an optimization problem that can be solved to obtain a correction function that will make the output distributions statistically indistinguishable. We derive closed-form expression for the correction function that can be used to compute it efficiently. We illustrate the use of the correction function for a recidivism prediction application derived from the ProPublica COMPAS dataset.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2365908 2018-02-02T09:19:08-05:00 2018-02-02T09:43:19-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2363845 2018-01-30T14:08:24-05:00 2018-01-30T16:41:39-05:00

<p>In contrast to automated decision-making, profiling is a relatively novel concept in European data protection law. It is now explicitly defined in Article 4(4) of the EU General Data Protection Regulation (GDPR), and refers to the automated processing of data (personal and not) to derive, infer, predict or evaluate information about an individual (or group), in particular to analyse or predict an individual’s identity, their attributes, interests or behaviour. Through profiling, highly sensitive details can be inferred or predicted from seemingly uninteresting data, leading to detailed and comprehensive profiles that may or may not be accurate or fair. Increasingly, profiles are being used to make or inform consequential decisions, from credit scoring, to hiring, policing and national security. Ever since the approval of the regulation in 2016, debates have focussed on the GDPR’s potential to limit or offer protection against increasingly sophisticated means of processing data, in particular with regard to profiling and automated decision-making. While the GDPR offers new rights and protection, their scope and limits are open to debate, partly due to the clumsy syntax of the relevant articles and the lack of authoritative guidance concerning their interpretation. The European Data Protection Board that will replace the Working Party on the Protection of Individuals with regard to the Processing of Personal Data is specifically tasked with publishing ‘guidelines, recommendations and best practices’ on the GDPR. In October 2017, the Working Party 29 has published draft guidance on profiling and automated decisionmaking. In this article we propose our suggestions to contribute to the development of guidelines which provide the strongest protections for data subjects.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2363155 2018-01-29T14:34:15-05:00 2018-01-29T14:34:15-05:00

<p>Academic institutions also need to adapt. University machine learning curricula need more policy-based instruction so technologists are aware of these concerns. And the public-policy profession needs a new sub-specialty: the data and machine learning ethics analyst. These analysts will have a solid grounding in both public-policy and technical fields. They will speak the language of machine learning developers, data scientists, and policy professionals. They will work with technologists to help them understand the value judgments inherent in their systems, assist them in removing bias from data-sets, and advise them how to encode values to remove bias. Universities such as Harvard,[22] Cornell,[23] and the University of Edinburgh have introduced courses on AI ethics.[24] These should be required for any institution providing AI instruction and mandatory for policy and technology students studying AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2355908 2018-01-17T16:19:28-05:00 2018-01-17T16:19:28-05:00

<p>While there’s no comprehensive checklist one can go through to systematically address issues pertaining to fairness, transparency, and accountability, the good news is that the machine learning research community has started to offer suggestions and some initial steps model builders can take. Let me go through a couple of simple examples.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2350049 2018-01-08T15:18:44-05:00 2018-02-23T14:51:47-05:00

<p>Artificial intelligence and big data present new challenges to protecting our right to privacy both online and offline. Predictive algorithms and online social networks add complex nuances to the idea of informed consent. A global, cross-disciplinary effort is underway to make predictive algorithms fair, accountable, and transparent as these algorithms come to mediate more and more the areas of civic life and public discourse. But these tools and ideas operate inside of existing socio-technical systems that need to balance conflicting goals and tensions.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2309613 2017-11-06T10:16:57-05:00 2017-11-06T10:16:57-05:00

tag:tagteam.harvard.edu,2005:FeedItem/2301582 2017-10-24T21:48:54-04:00 2017-10-24T21:48:54-04:00

<p>Until now, perhaps. A new study from MIT’s Media Lab posits that the smaller the city, the greater the impact it faces from automation. The finding, they say, could encourage legislators to pay special attention to workers in smaller cities and offer them support services.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2298761 2017-10-20T04:36:17-04:00 2017-10-20T04:36:17-04:00

tag:tagteam.harvard.edu,2005:FeedItem/2298762 2017-10-20T04:36:51-04:00 2017-10-20T04:36:51-04:00

<p>The review of artificial intelligence argues a new AI council should be created but it wouldn't be in charge of regulating systems</p>

tag:tagteam.harvard.edu,2005:FeedItem/2298759 2017-10-20T04:35:18-04:00 2017-10-20T04:35:18-04:00

<p>Crowdsourced morality “doesn't make the AI ethical,” said James Grimmelmann, a professor at Cornell Law School who studies the relationships between software, wealth, and power. “It makes the AI ethical or unethical in the same way that large numbers of people are ethical or unethical.”</p>

tag:tagteam.harvard.edu,2005:FeedItem/2298760 2017-10-20T04:35:49-04:00 2017-10-20T04:35:49-04:00

<p>We present a general approach to automating ethical decisions, drawing on machine learning and computational social choice. In a nutshell, we propose to learn a model of societal preferences, and, when faced with a specific ethical dilemma at runtime, efficiently aggregate those preferences to identify a desirable choice. We provide a concrete algorithm that instantiates our approach; some of its crucial steps are informed by a new theory of swap-dominance efficient voting rules. Finally, we implement and evaluate a system for ethical decision making in the autonomous vehicle domain, using preference data collected from 1.3 million people through the Moral Machine website.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2298299 2017-10-19T11:32:38-04:00 2017-10-20T09:44:13-04:00 Elsa Kania

<p>Looking to the future, it is clear that becoming a world “leader” in AI will require not only technological innovation but also policy choices. Since the release of its new plan, it is clear that China is actively formulating a national strategy to leverage AI, while planning to actively participate in the future “global governance” of AI.</p>

tag:tagteam.harvard.edu,2005:FeedItem/2483466 2019-06-05T15:33:35-04:00 2019-06-05T15:33:35-04:00