On November 20, U.S. technology media reported that Levant Alpoglu, a mathematician at Anthropic, announced he had derived a proof of the Riemann Hypothesis using the chatbot Claude, with the same method extending to the generalized Riemann Hypothesis. He released a four-page explanation alongside formalized code written in Lean for peer review. First proposed in 1859, the Riemann Hypothesis concerns the distribution of prime numbers and has remained, for 167 years, one of the most renowned unsolved problems in mathematics.

The announcement quickly sparked debate within the mathematical community, not over the conclusion itself, but over how to define what constitutes a "proof." According to media reports, Alpoglu’s publicly shared Lean code amounts to approximately 17 million lines; when attempting to generate a human-readable version, he reached over 250,000 pages, which he described as “unreadable garbage.” He acknowledged the argument’s immense scale and admitted it requires a proper written presentation. Fields Medalist Terence Tao offered a cautious assessment on his blog: preliminary analysis suggests the proof draws upon advanced results from the Langlands program and arithmetic geometry, but comprehension by the broader academic community will take time. He added, somewhat wryly, that at least this time there is still a living author available for discussion.

Whether a proof so vast that even its author cannot fully read it counts as a genuine resolution of the problem lies at the heart of the controversy. Supporters emphasize the certainty brought by machine-assisted verification, while skeptics maintain that mathematics ultimately seeks to answer “why,” not merely “that.” Fields Medalist Shing-Tung Yau called the work a monumental contribution, one that could be said to have shaken the world. 2026 Fields Medalist Yu Deng was more direct: the number of pages is irrelevant; the key point is that 250,000 pages are fundamentally unreadable, and no evaluation will be made until a comprehensible version emerges. Mathematician Andrew Granville criticized the publicly released materials as incomplete and disorganized, and linked the timing of the release to Anthropic’s recent public listing. Alpoglu conceded the timing was awkward, coinciding with the company’s stock debut. As Tao put it most cautiously: verification and understanding will both require time.

Original source: toutiao.com/article/1882561622364169/

Disclaimer: The views expressed in this article are those of the author alone.