Technology news and discussion on TalkFuse.
In a year of AI solving maths problems, Anthropic may have just solved the biggest
Anthropic mathematician Levent Alpöge says he used the company’s Claude chatbot to develop a proof of the Riemann hypothesis, one of mathematics’ best-known unsolved problems. If experts confirm it, the result would be a major breakthrough, but the claim also comes amid growing debate about AI companies announcing mathematical advances before independent checks are complete.
Alpöge said the work took about six weeks using a non-public version of Claude. The proof, written in Lean, runs to more than 250,000 pages, and mathematician Terrance Tao said verification will take time. The nearly 170-year-old hypothesis concerns the pattern of prime numbers; other AI-generated proof claims, including work announced by OpenAI, also remain under review and have drawn criticism from some scholars.
In short
- Anthropic’s Levent Alpöge claims Claude helped prove the Riemann hypothesis.
- The Lean proof is more than 250,000 pages long.
- Mathematicians say independent verification will take time.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
If the proof withstands expert scrutiny, it could mark a major advance in mathematics and show that AI can help researchers tackle problems that have resisted generations of human effort. Developing a formal Lean proof may also make the reasoning unusually checkable, while public discussion can invite specialists to identify errors or confirm the result.
The case against
A claim of this scale deserves independent verification before it is presented as a breakthrough, especially when the proof is exceptionally long and review may take substantial time. Premature announcements can blur the distinction between a promising draft and established mathematics, and make it harder for researchers and the public to assess what AI systems have actually achieved.