OpenAI model cracks century-old Siegel zeros problem; Ben Green condenses proof to 5 pages in 12 days
新智元 · wechat · 2026-10-25
The century-old Landau–Siegel zeros problem is solved. An unreleased OpenAI internal model produced a consistent exclusion proof among 700+ math manuscripts on Oct 6 (Zhang Yitang's 2022 attack was a weakened 111-page version).
Human mathematicians then raced to shorten it: William Banks expanded it to 19 pages (constant computed as 0.003), Bryce Kerr cut it to 6 pages via pigeonhole (crediting ChatGPT), and Oxford's Ben Green — of Green–Tao fame — delivered a 5-page contradiction proof built on a determinant ceiling vs. divisibility floor, writing on page one "this paper claims no human originality" after a vibemathing session with ChatGPT.
About three-quarters of Green's own ERC-funded agenda was scooped; he says the shock removed pressure, while Tao laments nobody can lecture on or extend these AI proofs. From manuscript to classroom lectures at Max Planck took just 12 days.
More from AGI Musings
- Investor argues LLMs make specialists commoditized and generalists will capture the value — StewartalsopIII · 2026-10-26
- AI researcher calls out X users for dunking on people angry about AI-driven layoffs — mervenoyann · 2026-10-26
- Weight-tampering worm scenario for near-frontier models called 'somewhat scary' — teortaxesTex · 2026-10-26
- What if all human-written code goes in .gitignore? Redditors debate dev's future role — michal-rd · 2026-10-26
- Voice-to-voice needs more attention for less information density, OpenAI researcher argues — athyuttamre · 2026-10-26
- Senior devs are in their Kasparov era: chess history says the centaur phase expires — nixiewright · 2026-10-26