AI use leads Berkeley math professor UC to deem grading homework pointless
| Source: HN | Original article
A UC Berkeley math professor and Fields medalist has stopped grading homework, saying AI tools make the task pointless.
UC Berkeley mathematics professor and Fields Medalist Richard Borcherds has stopped grading his students’ homework assignments, saying the practice has become “pointless” in an era where AI tools can generate perfect solutions. Borcherds observed that a growing number of his students were turning to generative‑AI programs to complete problem sets, consistently earning top marks without demonstrating genuine understanding. Faced with this reality, he concluded that allocating staff time to verify correctness no longer serves an educational purpose.
The shift underscores a broader dilemma confronting higher‑education instructors: how to assess learning when AI can produce flawless answers on demand. If homework scores no longer reflect mastery, traditional grading may fail to identify gaps in reasoning, potentially widening the “math deficiency” many educators have warned about. Moreover, the move raises questions about academic integrity policies, the design of assessments that resist automation, and the role of AI literacy in curricula.
What to watch next includes how UC Berkeley and other institutions adapt their evaluation frameworks. Faculty committees may experiment with oral examinations, project‑based assessments, or AI‑augmented feedback that emphasizes process over final answer. Policy makers could also tighten guidelines on AI use in coursework, while researchers are likely to study the impact of ungraded homework on student performance and motivation. The conversation is already spilling into the public sphere, with op‑eds and campus debates probing whether AI will reshape the very purpose of homework in mathematics education.
Sources
Back to AIPULSEN