Daijiworld Media Network - New York
New York, Nov 21: A researcher at Anthropic has claimed that he may have used the company’s artificial intelligence software to solve the famous Riemann hypothesis, one of the most challenging mathematical problems that has remained unresolved for more than 150 years.
Mathematician Levent Alpöge published a note on his personal website on Friday, claiming to have developed a proof of the hypothesis. He also shared a link to code written in the programming language Lean, which he said provided evidence supporting his breakthrough.
According to the note, Alpöge contacted Anthropic chief executive officer Dario Amodei on Sunday night to seek advice on making the findings public. Describing the timing of the announcement as “horrible”, he said he could not wait any longer. Anthropic declined to comment, while Alpöge did not respond to a request for comment.

The Riemann hypothesis concerns the distribution of prime numbers and is widely regarded as one of the most difficult unsolved problems in mathematics.
Artificial intelligence companies, including OpenAI and Anthropic, have been working to demonstrate the advanced capabilities of their latest models by tackling complex mathematical and life sciences problems.
If confirmed, Alpöge’s work would represent a possible second breakthrough involving a Millennium Prize Problem achieved with AI assistance. OpenAI had said in September that its models generated a solution to the Navier-Stokes problem, which has remained unsolved for around 90 years.
Andrew Sutherland, a senior research scientist at the Massachusetts Institute of Technology, said he had independently verified Alpöge’s work on his computer. He had also seen at least three other verifications carried out on different systems by professional mathematicians.
Sutherland said Alpöge’s efforts using Claude built on the contributions of generations of mathematicians who had studied the Riemann hypothesis, as well as advances in the formalisation of mathematical proofs.
Expressing both excitement and apprehension about the development, Sutherland said mathematicians would need years to examine the proof and understand its possible applications. He described the development as a day for celebration, while acknowledging concerns within the mathematical community about AI’s potential impact on the profession and other jobs.
Noah Giansiracusa, an associate professor of mathematics at Bentley University and faculty associate at Harvard University, said that if the solution proved valid, it would be a significant achievement for artificial intelligence.
“It sounds like there’s a pretty good chance this is airtight and valid,” he said.
Alpöge described his attempt to prove the hypothesis as a “stream-of-consciousness jaunt”, adding that while he and Claude had worked together extensively in recent months, the AI system had outpaced him towards the final stages of the proof.