[VIDEO] Living with Data: Stories that Make Data More Personal

I had the privilege of speaking at the Berkman Lunch Series this week and talked about my ideas for telling more personal stories about our relationship to data to ground our understanding in more practical, everyday lived experience. The way I see the problem is that right now we don’t understand the causal relationship between our data and its uses in the world. My talk sets up a few examples that I’ve seen recently that both exposed and obscured what my data says about me and how it’s being used. I talk about why understanding data in our everyday lives matters more than ever, and I set up what personal stories can do to help us. I walk through a few canonical examples, and then end with a pitch for a column to tell these stories on a regular basis. Please send me your ideas, strange screengrabs, and questions—this is just the beginning of an effort to make data and its uses more legible to us. 

The video is embedded here, and I’ve also posted my crib notes with links below if you’d prefer to read or want to follow up on some of the examples.

CRIB NOTES

This talk is a reflection of a lot of the work I’ve been thinking about here as a fellow, but it’s also a kind of proposal for future work, so I’m very much interested in feedback from the braintrust here in the room and watching on the web.

The main idea is that we need more stories that ground data in personal, everyday experience. We need personal data stories make data uses intelligible and impacts personal.

I wanted to start off by talking about what I do and what I do not know about myself as other entities see me through my data.

Facebook advertising engine seems to think I like cheese boards. Even when they aren’t selling cheese, or boards, they are part of my advertisements.

But I don’t know if it is because I talked about my love of cheese boards, or if it is based on image recognition, or some combination of the two. I can’t tell if Facebook thinks I’m demographically bougie, or if it really knows I’m obsessed with cheese.

About the Data, Acxiom’s consumer portal into our data broker data tells me it thinks I am a truck owner and intending to purchase a vehicle. I am not. I’m assuming this is based on my Father’s truck registration (the last time he drove a truck was in the early 1990s).

But About the Data doesn’t tell me what Axciom thinks I’m a “Truckin’ and Stylin’” or “Outward Bound” consumer, one of the many consumer segmentation profiles that might link to that Truck data point. Acxiom, shows us the inferred demographic information of behavioral targeting, but it doesn’t show us how it is being used by its third party customers who very well could be insurance companies or loan underwriters, not just marketers.

When I start to worry about the traces of my connections to friends in my time abroad in the UK and in China, I can use Facebook’s graph search to query how many people in my network I know in China that show up in my “buddy list” as the PRISM documents.

But I don’t have any confidence that I don’t meet the threshold for confidence-based citizenship. I don’t know what it means to be a person on a “buddy list” “associated with a foreign power.” Nor do I know whether my use of VPN would contribute to my score. My algorithmically-determined citizenship is completely opaque to me.

These are just some personal encounters I’ve had recently in my daily life—from the trivial in the commercial, to the consequential in talking about my shifting sense citizenship. The concerns I raise point to an asymmetry that obscures what’s going on behind the scenes in interactions in my daily life.

The crux of the problem is that right now we don’t understand the causal relationship between our data and its uses in the world.

Joanne McNeil has described this as reading the algorithmic tea leaves—it’s a dark art. We don’t understand the how and the why of data’s uses, let alone what our data forecasts about us.

I like to think of it as a kind of uncanny valley of personalization. When we try to understand creepy ads that follow us around or are strangely personal, we can’t figure out if it’s just coarse demographics or hyper-targeted machine learning that generates the ads we see and that leaves us with this sense of the uncanny.

So while data is making our behaviors, habits, and interests more legible to firms and governments, as consumers we haven’t yet developed the critical literacies to understand what our data is saying about u