Update: the race for the Riemann Hypothesis gained a second chapter on the same day Anthropic announced its AI-assisted proof. On Friday, November 20, Ravi Vakil, a mathematician at Stanford University and president of the American Mathematical Society (AMS), published on his blog the draft of a rival proof — written by a human, without AI assistance.
The author is Alessandro Maria Masullo, who earned his doctorate in mathematics at Stanford under Brian Conrad, one of the leading experts in number theory. The manuscript, titled The Music of the Primes, runs to 940 pages and proposes a proof of the Generalized Riemann Hypothesis (GRH), a stronger version of the classical Riemann Hypothesis: proving GRH would imply the Riemann Hypothesis as a consequence.
What Vakil wrote
Vakil published the announcement in a hurry, during a meeting of the AMS executive committee. According to him, Conrad has been following Masullo's work closely for months, gave extensive feedback that is being incorporated into the text, and so far the proof "is looking good". He made a point of marking the difference from the day's other claim: "It is all human, with no AI."
The original plan was different: to bring Masullo to Stanford for a year, polish the text, and only then announce it, after showing it to more specialists. Faced with Anthropic's claim, Vakil decided to publish the draft immediately and asked the community to check it, formalize it, and fill any gaps.
Problems already flagged
About eight hours later, Vakil himself updated the post with a twist: several people had already reported problems in the proposed proof — identified with the assistance of different AI models. He thanked the mathematicians who flagged them first and made clear that all issues need to be resolved "to the satisfaction of the community" before the proof is accepted as true, and that he makes no claim about the correctness of the proof strategy.
Independent readers have already published detailed notes on the problematic passages, and mathematician John Baez published on Sunday, November 22, a sketch of Masullo's strategy with the problems observed so far — stressing that his analysis is provisional.
What Brian Conrad says
Speaking to Scientific American, Conrad confirmed that the details in Vakil's post are true, but cautioned that judging whether the method will work requires the paper to reach a more final stage of revision — the text is "undergoing major revision". On the decision to publish now, he said the choice was between "two unpalatable options": announce now, or stay silent and publish later.
The reaction was immediate: on X, science journalist Kevin Hartnett wrote that he has followed the story for years and called the announcement "massive".
Two paths, no verdict
The two claims from November 20 follow opposite tracks. Levent Alpöge's, from Anthropic, exists as a giant Lean formalization — around 17 million lines — and was certified as correct by the Lean and Mathlib communities on Saturday, November 21, although no human has read the full argument. Masullo's is a traditional text, written to be read by humans, but with no formalization — and with problems already flagged in its first hours.
Neither has been accepted by the mathematical community, and the Riemann Hypothesis remains officially open. The Clay Mathematics Institute's US$1 million prize requires publication and general acceptance for two years.
Previous Geekiki coverage: Lean and Mathlib certify the Riemann Hypothesis proof and Anthropic mathematician claims to have proved the Riemann Hypothesis with Claude.
Image: the Main Quad at Stanford University, where Ravi Vakil and Brian Conrad work. Credit: King of Hearts, CC BY-SA, via Wikimedia Commons.
Sources
- Ravi Vakil — "Hurried Announcement: Alessandro Masullo GRH", Math 216 blog
- Scientific American — "Dueling claims of Riemann hypothesis proofs emerge"
- John Baez — "Masullo on the Riemann Hypothesis"
- explainx.ai — "Masullo GRH Draft: What Vakil Announced (Nov 20)"
- OfficeChai — "Hours After Alpöge's AI-Assisted Riemann Hypothesis Proof, AMS President Ravi Vakil Rushes Out A Rival 'All Human' Proof"
Transparency note: this article was produced with the assistance of artificial intelligence, cross-checking public posts on X and sources on the internet, and reviewed before publication. Check the original sources to confirm the information.
Por Muse