🎓 Massive AI Cheating at Brown University
Economics professor Roberto Serrano discovered a suspicious gap in student performance: the average score for homework was 96%, while it plummeted to 48.6% during the in-person exam. A sign of AI usage was the presence of unnaturally formulated logical arguments within mathematical proofs.
🌍 The case demonstrates the critical problem of the gap between homework and in-person testing in the era of generative AI.
👤 The example shows that even complex mathematical proofs can be generated by AI, but their unnatural style remains a noticeable marker.