Technology

Mathematician Levent Alpöge claims Claude helped solve Riemann hypothesis

Anthropic mathematician Levent Alpöge claims the AI Claude assisted him in developing a solution to the 170-year-old Riemann hypothesis using Lean code. While other mathematicians work to verify the findings, some express skepticism regarding the reliability of AI in solving complex math problems.

Mathematician Levent Alpöge claims Claude helped solve Riemann hypothesis

Alpoge and Claude

Levent Alpoge, a Turkish American mathematician at Anthropic, claims Claude helped him solve the 170-year-old Riemann hypothesis. He posted a note on his personal website explaining that Claude helped him develop a solution over the course of about six weeks. He used computer code written in Lean, which he says can verify the findings. I find the timing of his announcement strange because Anthropic went public on the Nasdaq stock market index just one day before he posted the note. Alpoge wrote: "Claude and I have had a great time together with such breadth in recent months, but this time Claude far outpaced me near the finish line." He added: "No, Claude flew, and danced, and sparkled near that finish line." The Riemann hypothesis, which Bernhard Riemann proposed in 1859, deals with the appearance and distribution of prime numbers along the number line. This problem concerns prime numbers, which are whole numbers greater than 1 that can only be divided by themselves and 1. Riemann also developed a formula for estimating how many prime numbers exist under a certain number. The Clay Mathematics Institute, which selected seven Millennium prize math problems in 2000, provides $1m for a solution to this problem.

Math competition and AI skepticism

Other mathematicians are working to verify Alpoge’s proof, but many express skepticism regarding AI. Alessandro Masullo, an independent mathematician in Italy, released an incomplete proof on Friday that he worked on for almost 12 years. Masullo said he did not use AI because he did not fully trust the system. He preferred human-led verification. I see a clear tension between traditional research and the massive spending by tech companies. In September, OpenAI claimed a non-public version of its tools helped solve part of another Millennium prize problem after the company spent $15m to employ 10,000 AI systems. The decision by a near-trillion-dollar private company to spend $15m on 10,000 AI systems struck many observers as strange. James Robinson, a mathematician at the University of Warwick, called the use of massive resources for these claims "immature playground boasting" that could cause significant environmental damage in an age when climate change is a major challenge. Alpoge was working on that separate problem for Anthropic when OpenAI made its announcement.