OpenAI denies user data shaped Millennium Prize solution; mathematician says 'I just don't believe them'
francoisfleuret · x · 2026-09-14
Mathematician Tristan Buckmaster publicly doubts OpenAI's claim that it is "categorically" impossible for two months of a mathematician's user data to have influenced its landmark solution to a Millennium Prize problem. François Fleuret, quoting the story, argues original ideas often appear in many places simultaneously because they derive from collective knowledge, so accusations of idea theft need tangible evidence.
More from Companies & People
- Burkov calls out Musk over unfulfilled promise to open-weight older Grok models — burkov · 2026-09-14
- shawmakesmagic: EA folks were 'lame status-concerned social climbers' — beffjezos · 2026-09-14
- 84% of the world has never used AI; only 0.04% use agents effectively — petergyang · 2026-09-14
- Cognition Recruits Devin Ambassadors Across Nearly 40 Cities Worldwide — cyrus_zei · 2026-09-14
- After 15 Years and 15,000 Founders, This VC Says Outlier Quotient Beats Pedigree — FinanceYF5 · 2026-09-14
- Critics warn EA-backed third-party AI audits would give safety committees sweeping powers — beffjezos · 2026-09-14