Berkeley Professor Abandons Homework Grading — AI Makes It ‘No Point’

A UC Berkeley mathematics professor has stopped grading homework for correctness, telling students there is “no point” in the practice due to the widespread use of AI to generate answers. The move, first reported by UC Berkeley’s student newspaper The Daily Californian, is a stark admission that traditional methods for assessing student work are becoming untenable in the face of powerful and easily accessible AI tools.

This isn't a strategic pivot towards a new educational philosophy. It's a tactical retreat. The professor now awards full credit for any homework submitted in "good faith," effectively changing the assignment from a measure of competence to a simple check for participation. The burden of assessment now shifts entirely to in-person, proctored exams.

A Pragmatic Surrender

The policy change reflects a bottom-line calculation about the return on investment for grading. With students able to use AI to solve complex math problems instantly, the professor concluded that the time spent grading homework no longer provided a meaningful signal of student comprehension. As Inc Magazine noted in its coverage, this change is not because students have suddenly improved; it’s because the tools they are using have rendered the old measurement system obsolete.

The combined picture from both The Daily Californian and Inc Magazine suggests this is less about innovation and more about capitulation. Policing AI use in homework is a resource-intensive, likely losing battle. By making homework a low-stakes completion grade, the professor neutralizes the incentive to cheat for points and reallocates his own resources—and the course's evaluative weight—to the one environment he can still control: the exam room.

The Bottom Line for Education and Business

This single professor's decision at a prestigious university like Berkeley is a significant signal for the entire education sector and the businesses that rely on its graduates. If homework is no longer a reliable tool for practice and assessment in a foundational STEM field like mathematics, the implications are substantial.

For business leaders, this means the signals on a college transcript are changing. A degree from a top school may no longer guarantee a certain level of practiced, iterative problem-solving skill developed through coursework. The burden of assessment will increasingly shift from the university to the employer through more rigorous technical interviews, practical skills tests, and probationary periods. Companies hiring for quantitative roles must now question what a transcript truly represents and develop their own mechanisms to verify competence.

The combined picture suggests a future where higher education bifurcates its assessment strategy: low-stakes, AI-assisted practice for coursework and high-stakes, highly-controlled exams for final grades. This Berkeley professor is not an outlier; he is a leading indicator of a system-wide reckoning with AI's role in learning.