← November 21, 2026 briefing · added in the 12:00 KST update
Claude-assisted Riemann proof follow-up: 17.4 million lines of Lean; Scott Aaronson says it will likely stand (community report)
Announced November 21, 2026
Verified (fact)
According to a GeekNews summary of a post by Scott Aaronson, the generalized Riemann hypothesis proof that Levent Alpöge produced with an Anthropic model runs to 17.4 million lines of Lean code. Aaronson said he is not certain the proof is correct. But he has not yet seen a Lean-verified AI result be retracted, so he expects this proof to survive as well. Possible bugs in the Lean kernel and the soundness of the underlying type theory remain open verification issues, but he said he saw no sign that the proof exploited such gaps. He added that for mathematicians to accept it, they must go beyond checking the code: they need to understand the proof's key ideas and reconstruct it in shorter form.
Why it matters (interpretation)
This is the first time a well-known expert has said the proof is likely to stand. It also raises the next challenge: how humans can understand and accept a machine proof of more than 17 million lines.
Confidence medium · official source pending
Sources
- Community ‘확률적 앵무새’가 리만 가설을 증명한 모양이다 GeekNews
Story thread
- 2026-11-20 Anthropic mathematician Levent Alpöge claims proof of the Riemann hypothesis with Claude, extending to GRH; Lean formalization released (unverified)
- 2026-11-21 Claude-assisted Riemann proof follow-up: 17.4 million lines of Lean; Scott Aaronson says it will likely stand (community report)
- 2026-11-21 Riemann hypothesis follow-up: on the day of a Claude-assisted proof claim, a human-written GRH proof without AI is also released
More about Anthropic
-
Anthropic IPO follow-up: Shares fall 9% on second day; Nasdaq drops 5.3%, erasing 2026 gains; NVIDIA down about 14% (community report)
According to a market recap posted on GeekNews, Anthropic shares fell 9% on the trading day after the listing and led the decline. The...
-
Inkling 2 Mini follow-up: Ollama v0.40.17 runs it out of the box ('79.2% on SWE-Bench Verified'), adds concurrent requests on Apple silicon
Ollama's v0.40.17 release (tagged rc4) supports Thinking Machines Lab's Inkling 2 Mini. It can be run with a single command, `ollama run...
-
Riemann hypothesis follow-up: on the day of a Claude-assisted proof claim, a human-written GRH proof without AI is also released
Following Anthropic mathematician Levent Alpöge's claim of a Claude-assisted proof of the Riemann hypothesis, mathematician Ravi Vakil...
-
Anthropic IPO follow-up: 7 founders each cash out $280M, about $2B total; 'about 2,800 current and former staff are millionaires' (reports)
TechCrunch calculated that, at the IPO price of $83 per share, the seven founders each cashed out $280 million, nearly $2 billion...
-
Gemini 4 Neon follow-up: tops engineering benchmark 'Models Engineering Index' at 68.0, ahead of Fable 5.2 (65.3) and GPT-6.2 Astra (61.5) (per community posts)
Google's Gemini 4 Neon ranked first with 68.0 points on the Models Engineering Index (MEI), a public leaderboard comparing how well...