A mathematics professor at the University of California, Berkeley, has stopped grading students’ homework for correctness because so many students are using artificial intelligence to complete their assignments.
Richard Borcherds, a renowned mathematician and Fields Medal winner, said AI tools have made it difficult to determine whether students are solving problems themselves.
After noticing that students were consistently receiving high scores, he decided there was little reason to continue paying teaching assistants to check their answers.
Instead, Borcherds now awards marks based on whether students submit their homework, regardless of whether their answers are correct.
AI Is Changing How Students Are Graded
Borcherds admitted that his new grading method is not ideal but said he has not found a better alternative.
He believes AI has become so widely available that preventing students from using it for homework is increasingly difficult.
His comments came following OpenAI’s announcement that its AI systems had produced a claimed solution to the Navier-Stokes existence and smoothness problem, one of mathematics’ most difficult challenges.
The development has raised questions about how artificial intelligence could change mathematics education and research.
Borcherds warned that if AI continues improving at its current pace, machines could outperform human mathematicians across their work within a few years.
Professor Questions the Future of Mathematics Careers
Borcherds also expressed uncertainty about whether students should pursue careers in mathematics.
He explained that five years ago, mathematics appeared to be one of the professions least likely to face disruption from artificial intelligence.
However, recent developments have changed his outlook.
The professor said he even asked ChatGPT what advice he should give students considering mathematics careers, but the chatbot could not provide a clear answer.
Despite these concerns, Borcherds believes universities may take several years to respond to the changes.
He pointed out that academic institutions often continue following established practices long after new technologies make them less useful.
As an example, he mentioned printed academic journals, which he believes have remained unnecessary for around two decades.
He also compared AI’s potential impact on mathematics to earlier technological changes, such as machines replacing hand weaving and automobiles replacing horses.
However, he said mathematics has never experienced a technological disruption of this kind before.
University Struggles to Detect AI-Generated Homework
UC Berkeley is also reviewing how artificial intelligence should be handled in classrooms.
According to Anne Marie Richard, the university’s assistant vice provost for undergraduate education and chief academic technology officer, faculty members are asked to explain their expectations for AI use in course outlines.
However, individual professors can decide how students are allowed to use AI, depending on the subject and course level.
The university has tested AI detection tools, including Turnitin and Rumi, but found that they were not reliable enough to adopt as a standard system across campus.
UC Berkeley also maintains an online AI Hub where teachers can access sample policies, training materials, and guidance on using AI in education.
For now, the university has left decisions about AI use largely to individual faculty members as the technology continues changing how students learn and complete assignments.
The post Top University Professor Stops Grading Homework Because of ChatGPT appeared first on ProPakistani.
