When large models produce proofs or solve long‑standing problems (for example the claims about Navier‑Stokes), it forces mathematicians and the public to ask whether machine‑generated arguments count as the same kind of mathematical knowledge, who verifies them, and who gets credit. That debate changes hiring, training, publication norms, and the public trust in ‘mathematical truth.’
— If machine‑produced proofs become common, they will reconfigure scientific authority, publication standards, and education for decades.
Scott
2026.11.08
75% relevant
Scott Aaronson stresses that experts have inspected the result and that a Lean certificate exists, highlighting the emerging pattern where startups + large models can claim high‑stakes theorems — shifting who authors and validates proofs, and thus who holds epistemic authority in math and theoretical CS.
Scott
2026.09.29
100% relevant
Scott Aaronson’s podcast explicitly lists “The story of AI models solving the Navier‑Stokes Millennium Problem” and asks whether recent AI proofs can be called 'truly creative', providing a direct example and expert framing.
← Back to all ideas