{"slug": "a-mathematician-and-an-ai-say-they-solved-a-167-year-old-puzzle", "title": "A Mathematician and an AI Say They Solved a 167-Year-Old Puzzle", "summary": "Mathematician Levent Alpöge announced on Friday that he had found a proof of the Riemann hypothesis, a 167-year-old conjecture about prime numbers that carries a $1 million Clay Mathematics Institute Millennium Prize, with help from an AI called Claude. An independent verification log from MIT's Andrew Sutherland checked every step, and mathematicians cited by Scientific American called the proof almost certainly true, though the result is not yet settled. Bernhard Riemann first posed the guess in 1859.", "body_md": "# A Mathematician and an AI Say They Solved a 167-Year-Old Puzzle\n\nIn 1859, a mathematician named Bernhard Riemann made a guess about prime numbers that nobody could prove, even with a million-dollar prize waiting. On Friday, mathematician Levent Alpöge said he had found a proof with help from an AI called Claude. Leah and Sam explain what prime numbers are, why Riemann's guess matters, how a proof is different from checking examples, how computer programs checked every step, and why mathematicians say it looks promising but isn't settled yet.\n\n## Show notes\n\n### In this episode\n\n- 0:01 Opening\n- 0:26 A mathematician and an AI say they solved a 167-year-old puzzle\n- 4:55 Recap and outro\n\n### Sources and context\n\n- **A mathematician and an AI say they solved a 167-year-old puzzle:**[Levent Alpöge — Announcement of a proof of the Riemann hypothesis (PDF)](https://alpo.ge/remark.pdf) and[MIT, Andrew Sutherland — Independent verification log](https://math.mit.edu/~drew/grh/) and[Clay Mathematics Institute — The Riemann hypothesis](https://www.claymath.org/millennium/riemann-hypothesis/) and[Clay Mathematics Institute — Millennium Prize rules (PDF)](https://www.claymath.org/wp-content/uploads/2022/03/millennium_prize_rules_0.pdf) and[BBC News — Mathematician says AI helped prove the Riemann hypothesis](https://www.bbc.com/news/articles/c6pqr2g7w059o) and[Science — AI may have solved math's greatest mystery](https://www.science.org/content/article/ai-may-have-solved-math-s-greatest-mystery-can-humans-understand-answer) and[Nature — Reaction to the claimed Riemann hypothesis proof](https://www.nature.com/articles/d41586-026-03812-7) and[Scientific American — Independent mathematicians on the proof](https://www.scientificamerican.com/article/the-ai-riemann-hypothesis-proof-is-almost-certainly-true-independent-mathematicians-attest/)", "url": "https://wpnews.pro/news/a-mathematician-and-an-ai-say-they-solved-a-167-year-old-puzzle", "canonical_source": "https://myfirst.news/episodes/099-mathematician-ai-riemann-puzzle/", "published_at": "2026-11-21 13:10:00+00:00", "updated_at": "2026-11-21 13:13:45.610597+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-research", "large-language-models"], "entities": ["Levent Alpöge", "Claude", "Bernhard Riemann", "Clay Mathematics Institute", "Andrew Sutherland", "MIT"], "also_reported_by": [], "takeaway": "An independent verification log from MIT's Andrew Sutherland checked every step of the AI-assisted Riemann hypothesis proof, but mathematicians say the result is promising rather than settled.", "recommendation": {"advice": "When an AI system produces a mathematical or logical result, have an independent party check every step, as MIT's verification log did for the Riemann proof, before treating it as settled.", "playbook": "output-verification", "playbook_url": "https://wpnews.pro/playbooks/output-verification"}, "alternates": {"html": "https://wpnews.pro/news/a-mathematician-and-an-ai-say-they-solved-a-167-year-old-puzzle", "markdown": "https://wpnews.pro/news/a-mathematician-and-an-ai-say-they-solved-a-167-year-old-puzzle.md", "text": "https://wpnews.pro/news/a-mathematician-and-an-ai-say-they-solved-a-167-year-old-puzzle.txt", "jsonld": "https://wpnews.pro/news/a-mathematician-and-an-ai-say-they-solved-a-167-year-old-puzzle.jsonld"}}