Brown University economics professor Roberto Serrano made the difficult decision to offer a take-home, closed-book midterm to his ECON 1170 class last spring, motivated by concern for his students’ mental health after the tragic December 2022 shooting on campus that killed two students, including one of his own advisees. Instead of the gratitude he expected for reducing classroom stress, Serrano uncovered the largest known AI-assisted cheating scandal in the Ivy League. Of the 86 students who took the exam, 40 received a perfect 100, with a class average of 96, far surpassing the historical average of 65 to 80. Serrano realized something was wrong when graders noticed that many exams contained the same convoluted reasoning produced by ChatGPT for a problem with a much simpler solution. After confronting his students about the academic dishonesty, 27 students dropped the course immediately, including 22 of those who had scored 100 on the midterm. When the in-person final exam was administered, only 59 students took it, and 19 failed, with the class average collapsing to just 48 out of 100.
Serrano, a distinguished game theory expert who has been blind since age 17 and holds a prestigious named chair at Brown, took his evidence to the university’s dean and provost but received little response initially. After escalating to the Academic Code Committee, he was told the incident was „a wake-up call,” though the provost has remained silent. Serrano expressed deep frustration, comparing the arrival of AI to a „tsunami” that has caught educational institutions unprepared. Brown University acknowledged the challenge, stating that the rapid proliferation of generative AI has raised questions about educational innovation and academic integrity, and that the university is examining its academic codes in this new context while a new associate provost for AI has been hired. However, Serrano warned that if elite universities continue to produce mediocre students who refuse to learn, the value of their degrees will ultimately diminish in the marketplace.
The scandal at Brown reflects a broader crisis in higher education, where AI cheating has become pervasive. A Fortune analysis found that 57% of U.S. college students now use AI tools weekly, and the technology is causing measurable cognitive atrophy, with educators warning of a „great unwiring” of independent reasoning. Even Princeton recently ended its 133-year-old honor code tradition of unsupervised exams, and 47% of surveyed Harvard seniors admitted to cheating. Serrano has permanently eliminated take-home exams and weekly homework assignments that can be completed with AI. He emphasized that while AI has appropriate educational uses, its risks to academic integrity are profound, and universities must establish guardrails and consequences. „If we no longer defend truth and decency and honesty,” Serrano concluded, „then what kind of credibility are we going to have as academics?”
Ez a cikk a Neural News AI (V1) verziójával készült.
Forrás: https://www.yahoo.com/news/us/articles/humanity-chosen-become-idiots-brown-220202787.html.