Fields Medallist Tao Calls for 'Maths 2.0' as AI Floods Field With Proofs

Rivka Galchen's New Yorker profile casts Fields medallist Terence Tao, once an enthusiastic early user of AI assistants in mathematics, as the field's most prominent brake-puller. The occasion is OpenAI's release of answers to 722 problems, across more than 350 findings in nearly every subfield, including a claimed proof of the Navier-Stokes problem, one of the seven Millennium Prize problems.

Tao's response was blunt. At a Caltech talk days after the Navier-Stokes announcement, he told the audience, "We have to slow down. It's insane, the pace," per his own public aggregator of his AI remarks. He has called the OpenAI release "toxic waste," later softening to "maybe treated waste," and said of the 165-page Navier-Stokes write-up — produced by roughly 10,000 agents over 88 hours — that it is "technically readable," but that "no expert I've talked to has actually understood the whole argument."

Tao's proposal, developed in talks and the profile, is what he calls "Maths 2.0": a reorganization of the discipline in which, as Nature summarized the piece, "proof scarcity" is no longer the field's defining feature. The frame is less about machines being wrong than about mathematicians losing the slow, deliberative process from which new problems emerge.

Underneath the technical worry sits a governance one. "If we don't ask these questions ourselves," Tao says in the profile, "then they will get answered for us by a technology company." That is the sentence that explains why a mathematician who built one of the field's most-read research blogs is now spending his time on norms rather than theorems.

It lands in a week when OpenAI is already everywhere in our tracker — the Abu Dhabi ADIPEC deals and the new per-device ChatGPT caps both hit the same day — and the mathematical community is the latest profession being asked to decide, on OpenAI's timetable, what its own work is for.