US Berkeley prof. stops grading students' homework over AI use: ‘No point’ as students score top marks; how it happened?

A professor at the University of California, Berkeley, has changed his approach to assessing students’ homework after finding that artificial intelligence (AI) tools were helping them achieve high scores.

Richard Borcherds, a mathematician who won the prestigious Fields Medal in 1998, believes the growing use of AI in academic assignments has made conventional grading less meaningful. He now places greater emphasis on whether students submit their homework rather than whether their answers are correct.

“Student use of AI for homework is now widespread,” Borcherds said, according to a report by the New York Post. “Grading for correctness just adds a huge incentive for them to use AI for homework instead of doing it themselves, and penalizes students for honesty,” he added.

Professor says grading AI-generated homework is a waste of resources

Borcherds also questioned the value of spending time and money on graders to assess answers that may have been generated using AI. He described the practice as an inefficient use of resources, while admitting that he had yet to find a satisfactory alternative.

The mathematician said distinguishing between work completed by students and answers produced by AI had become increasingly challenging as the technology improved.

“There is no easy way to tell the difference between human and AI-generated work. Occasionally one could tell it is AI because AI made a few characteristic errors, but AI has now improved so much that it rarely makes obvious mistakes,” he said.

AI could outperform human mathematicians, professor warns

Borcherds' concerns go beyond the difficulties of evaluating homework. He has also raised questions about the long-term future of mathematics as AI systems become more capable.

In an interview with the Daily Californian, UC Berkeley's student newspaper, he cautioned that further advances in the technology could eventually leave human mathematicians unable to match its capabilities.

Livemint