Report: OpenAI near verifying Hodge conjecture, racing Anthropic on BSD
新智元 · wechat · 2026-09-11
Two days after OpenAI claimed it cracked the Navier-Stokes problem with 10,000 agents in 88 hours (166-page paper plus Lean-verified proofs), an X leak says OpenAI is close to verifying the Hodge conjecture, and a New York Times feature confirms its internal team made substantial progress on another Millennium Prize Problem over five years.
Rumors add that OpenAI and Anthropic are quietly racing toward the BSD conjecture, with some work already in verification. If true, three decades-old Millennium Problems could fall to AI within a month.
- Status: Poincaré solved (2003); Navier-Stokes claimed pending Clay review; Hodge and BSD rumored; Riemann, P vs NP, Yang-Mills untouched.
- Hodge asks whether topological Hodge classes on algebraic varieties all come from algebraic cycles — a search-for-counterexample task well suited to brute force.
- BSD links the rank of elliptic curves to L-function behavior; proving it would also settle the congruent number problem.
The piece frames this as AGI benchmarks ratcheting upward: IMO gold, GPT-5.2 deriving a new particle-physics formula, GPT-6 Astra and Brockman's "Welcome to the AGI era." (Hodge/BSD claims remain unverified.)
More from Models
- MissionReport benchmark: only 3 of 540 AI-generated government reports pass all criteria — Exp_Mark · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27