Staff Reports
A UC Berkeley math professor has quietly done what many teachers only joke about: he stopped grading homework for correctness. Professor Richard Borcherds told the campus paper that students are using generative AI so often and so well that checking answers is mostly wasted effort. That is the new reality in a top university math class — and it should make parents, taxpayers, and employers sit up straight.
Berkeley Professor Stops Grading for Correctness
Professor Richard Borcherds, a Fields Medal–winning mathematician and a member of the UC Berkeley math faculty, says student use of AI on homework is “widespread.” He told reporters he now gives credit for completion only, because AI solutions are hard to tell apart from real student work. He also said the campus tried AI‑detection tools but found them unreliable, leaving instructors to decide their own fixes. That’s not a classroom tweak. It’s an admission that a fundamental tool of education — homework as proof of learning — is breaking down.
The Tool Problem: Turnitin, Rumi, and False Hopes
Berkeley tested detectors like Turnitin and Rumi and concluded they aren’t ready for prime time. That tracks with what many educators have found: detectors flag style, not truth, and clever students learn to game them. So instead of fixing the problem, some schools are nudging instructors to change syllabi or move to more in‑person and oral assessments. Those are good ideas, but they are slow and uneven. Meanwhile, AI keeps getting better at solving the same math problems students were supposed to learn how to do.
Why This Is a Bigger Problem
This isn’t just about grades. It’s about value and trust. If students graduate with shiny transcripts but hollow knowledge, employers lose, too. If a mathematician says he can’t advise students because AI may outperform human mathematicians, that should be a flashing red light for anyone who cares about careers and competence. Universities should not treat this as an abstract technology issue. It is a real test of whether higher education can set standards and enforce them.
What Conservatives Should Push For
Conservatives should demand accountability, not excuses. Push for tougher assessment rules: in‑person exams, oral defenses for proofs, random checks, and real consequences for cheating. Audit those AI‑detectors before relying on them. Fund instructor training and small‑class pilots that preserve honest learning. And stop pretending homework exists to be graded by a machine. If universities won’t defend the value of a degree, the market will — employers will hire differently, and parents will choose different schools. That would be the clean outcome. Until then, don’t be surprised when top professors admit there’s “no point” in grading the way we used to.
