'Online exams are now a waste of time': Top UC Berkeley professor calls time on homework grades as students are increasingly cheating using AI

University of California, Berkeley professor Richard Borcherds has discontinued grading homework assignments for his mathematics students. This decision stems from the increasing reliability of AI-generated schoolwork, making it nearly impossible for educators to distinguish between human-written and AI-generated submissions. As AI detection software proved unreliable at Berkeley, professors were given the autonomy to establish their own policies concerning AI usage within course syllabi.

Borcherds claims that online exams have become ineffective, as numerous students now utilize AI to answer questions. The professor argues that grading for correctness, a grading system that rewards students for submitting correct answers irrespective of their method, has incentivized the use of AI. He contends that this approach penalizes honest students who choose not to utilize AI tools.

Other educators have voiced comparable concerns about AI's impact on student performance. However, few have adopted Borcherds' more stringent stance on the matter. Some have discovered that permitting students to employ AI in their work has led to skewed results, and therefore, restricting access to such technology could yield more accurate grades.

Despite previous exploration of AI detection methods at Berkeley, the university ultimately concluded that allowing professors to dictate AI usage expectations in their respective departments proved more effective.

It appears that Professor Borcherds may be at the forefront of this shift in perspective, and if his stance garners attention, it could prompt more educators to reevaluate their policies regarding AI assistance in homework and assessments.