AI

Anthropic Mathematician Claims Claude Assisted Proof of Riemann Hypothesis

Anthropic mathematician Levent Alpöge claims to have proved the Riemann Hypothesis using Claude, a claim that awaits independent verification and could redefine AI's role in mathematics.

S
SciTech Editorial Team
Newsroom
··2 min read
Anthropic Mathematician Claims Claude Assisted Proof of Riemann Hypothesis
In a development that could reshape the landscape of modern mathematics, Levent Alpöge, a mathematician at Anthropic, has claimed to have proved the Riemann Hypothesis with the assistance of the company’s AI model, Claude. This assertion, if verified, would represent a monumental breakthrough, solving one of the seven Millennium Prize Problems that has eluded the world’s finest minds since 1859. The announcement underscores the accelerating role of artificial intelligence in tackling complex mathematical challenges, following recent, albeit controversial, advances by competitors like OpenAI. Alpöge detailed his methodology in a document on his personal website, stating that Claude was integral to developing the proof. He noted that the method not only addresses the Riemann Hypothesis but also extends to resolving the Generalized Riemann Hypothesis, with the potential to solve the Extended Riemann Hypothesis given more time. Alpöge mentioned that he informed Anthropic CEO Dario Amodei of the solution late Sunday night and deliberately delayed public disclosure until after the company’s IPO on Thursday. The proof has been formalized in Lean, a programming language designed for mathematical verification, though it has not yet undergone independent peer review. The Riemann Hypothesis, concerning the distribution of prime numbers, is administered by the Clay Mathematics Institute, which offers a $1 million prize for its solution. However, under the institute’s strict rules, a proposed solution must be published and withstand at least two years of scrutiny to receive general acceptance from the global mathematics community before the prize can be awarded. To date, only the Poincaré conjecture has been officially solved under this framework, making Alpöge’s claim particularly significant yet subject to intense scrutiny. This development arrives amidst a broader shift in the mathematical community, where AI tools are increasingly central to research. In September, OpenAI claimed to have cracked the Navier-Stokes problem, a task that reportedly involved deploying 10,000 agents and spending millions of dollars. While OpenAI’s achievement is considered credible, it has been marred by allegations of questionable research practices, including the use of unpublished work from other mathematicians. Alpöge’s claim adds to a deluge of AI-driven findings, leaving mathematicians scrambling to digest the material and adapt to a rapidly evolving field where traditional methods are being complemented, and sometimes superseded, by computational intelligence.

Share

More from this section: AI