Mr. Beast and academia

Do you know Mr. Beast? There’s this amazing video of a young Mr. Beast counting to 100.000. The video went totally viral, and skyrocketed him among the top youtube streamers. He now has 200M+ scubcribers, he is apparently filthy rich and is extremely popular, despite splashing money for cheap stunts, like random “last one on the mat gets $1M, yeah!”. Not long ago, aristocrats had their fill of “fun” by throwing bread to beggars and then watch them fight over it. Now it’s this, alright.

But besides, it makes me think of academia. Consider the 100.000 video. What is in it? Nothing, absolutely nothing. Has anyone even watched the entire video? Or even big chunks of the video? No. But it’s there. It’s nothing but “unhealthy weight lifting,” or a “proof of suffering.”

It makes me think of how decisions are made when refereeing papers. (The move to more or less anonymous submissions makes me slightly less uneasy when writing such things.) You get these massive papers, with pages and pages of math. (First thing to check, are there inequalities? Yes, good, because otherwise it’s just stringology.) The paper has nothing to say. It addresses a totally obvious question, like matching a known bound, putting X in “sharp alignment” with Y, and so on. Conceptually, there’s nothing new. But papers that have something to say tend to be much shorter, and you can actually read them, or at least skim them, and familiarity brings contempt, as they say. This paper on the other hand is full of math, goes on and on, kind of like Mr. Beast counting to 100.000. Is anyone actually going to read it? Or even some of it? Very unlikely. The paper seems to be just a “proof of suffering” that obvious next step X can actually be carried out. It does have some value, of course, but is this the type of papers that we should accept? Remember the rule, and you may want to watch the Mr. Beast video while reviewing.

Speaking on reviewing, I am now on the RANDOM 2024 program committee. If you get a very short review, it could be mine (now that I wrote that, I might ask chatGPT to add a bunch of trivialities, just to hide my tracks). I view my job as reviewing the papers myself and providing the most significant bit, and any *substantial* comment I have for the authors. I try to avoid comments that are likely subjective, and to avoid repetition. The Mr. Beast phenomenon is actually also present in reviews, somehow very long reviews that really show that the paper has been understood, and that include a nice summary of the paper, are often just those that kill the paper.

Of course, there is no hard and fast solution, the enormous pressure to publish is only increasing, just like pretty much any other metric in the world. By the way I find it ridiculous to read every time that this has been unusually high, record-setting and blah blah. This is the norm at least since I can remember! The only thing that would be unusual is if the metrics go down, which has only rarely happened.