Article · AI translation
IT Home reported on October 28 that a professor at the University of California, Berkeley (UC Berkeley) has announced that he will no longer assign graded homework because the misuse of artificial intelligence (AI) has rendered such assignments meaningless.
According to an October 26 report by the New York Post, renowned mathematician and Fields Medal winner Richard Borcherds said it is now almost impossible to tell whether students completed submitted assignments themselves or generated them with the help of AI. As a result, he now primarily assesses students’ learning through in-person exams that do not allow electronic devices.
IT Home noted that Borcherds told the outlet: “Online exams are now a waste of time, because some students simply use AI to answer the questions. I cannot tell which students have used AI, but I have noticed that average scores on online exams have risen sharply.”
He also pointed out: “Grading based on whether answers are correct gives students a huge incentive to use AI to do their homework instead of solving the problems themselves, while penalizing students who complete their work honestly. Spending time and money grading AI-generated answers is also a waste.”
The University of California, Berkeley had previously considered using AI detection software to address the issue, but ultimately concluded that such tools were not reliable enough. The university therefore decided to leave AI-use policies to professors, requiring only that they clearly state the relevant rules in their course syllabuses.
As for Borcherds, he said that if AI’s worst effect is merely to make it harder for teachers to grade students’ work, he would be “very relieved.”