

My favorite chart from this report is the one on page ~230ish where they show science benchmark test scores for students who use AI to study, and it basically just makes you perform much worse at any level (except one data point where the AI users perform only slightly worse). I’ve kept a few print outs of the page to hand out to the AI-pilled students I work with. In one of their words “wow. This is pretty damning”.











They did this in my city, but the furthest point from school anyone lives is, like a bit more than a mile and there is reasonably functional bus transit and bike lanes. It’s been a bit of a non-issue here. I could imagine in many other cities it would be a problem.