Anthropic mathematician claims Claude-aided Riemann Hypothesis proof in Lean
Original title: [AINews] Riemann Hypothesis proved...twice independently
Elliot Glazer says the formalized statement is GRH and puts a Lean bug well under 1%, while compilation was unfinished.
Levent Alpöge of Anthropic released only Lean code for a claimed proof of the Riemann Hypothesis and the Generalized Riemann Hypothesis, described as roughly 250,000 pages if written out. Within hours, Alessandro Masullo released a separate, roughly 940-page human draft. The item says both AI-claimed Millennium Prize results remain unverified, including OpenAI’s September Navier-Stokes claim.
Unconfirmed: Compilation was unfinished, both Millennium claims remain unverified, and the model and human steering were not disclosed.
- Work began October 6; eight of 20 steps landed November 15, and the final entry reads GRH.
- Alpöge credits Claude for work on automorphic forms on higher-rank groups.
- The Claude model used and the degree of human steering were not disclosed.
- Masullo’s competing draft is reported as about 940 pages and purely human.
- ANT stock reportedly fell about 10% on the day of the claim.
- Organizations
- Artificial Analysis
- AI models
- Claude
From the original article
14 min read at latent.space
Within a few hours of each other ( timeline ), two mathematicians claimed independent proofs of the Riemann Hypothesis, the most famous open problem in mathematics ( Scientific American ). Levent Alpöge of Anthropic first made his Claude-driven proof in a note released with Lean code (1. 3GB worth, ~250,000 page equivalent) rather than a full paper.
By swyx .ai · latent.space
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.