Levent Alpöge, a mathematician at Anthropic, says he used the company’s Claude AI model to discover a proof of the Riemann Hypothesis, a 167-year-old conjecture about the hidden order governing prime numbers. He has posted the argument in Lean, software that checks mathematical reasoning, and says it proves the stronger Generalized Riemann Hypothesis. If the proof is correct, it would resolve one of mathematics’ most important open problems, carry a $1-million Clay Mathematics Institute prize, and establish deep connections among branches of mathematics. Mathematicians quoted describe the potential result as transformative, but say the proof’s details still need to be understood and verified.
The claim arrives amid a year of striking AI-generated mathematical results, following earlier advances involving OpenAI and Alpöge’s work with Claude on the Jacobian conjecture. A verified proof would mark an extraordinary expansion of AI’s capabilities, although the Riemann Hypothesis has no immediate practical application such as improving banking or internet technology. The prospect has prompted mixed emotions among mathematicians: the end of a central unsolved problem may feel like a loss, while the new ideas could open research directions. MIT mathematician Andrew Sutherland, who independently compiled Alpöge’s Lean result, calls the moment historic and emphasizes the human contributions behind the tools and mathematical knowledge that made it possible. He also stresses that substantial work remains to examine the proof’s mathematics.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.