Two Riemann proofs drop in one week: one built with Claude, one all human
Levent Alpöge, a mathematician at Anthropic, says he has a Lean-compilable solution to the Riemann Hypothesis, the Millennium Prize problem dating back to 1859. He went public the same week his employer became the first publicly-traded frontier AI lab. Meanwhile, Stanford's Ravi Vakil posted a rival General Riemann Hypothesis proof he stressed was entirely human.
- Original source: techcrunch.com.
- The source published it on Nov 20, 2026 at 7:32 PM.
- No second outlet has reported it yet. Treat it as a single-source claim.
Alpöge announced the finding in an essay, writing that he messaged Anthropic CEO Dario Amodei last Sunday night to ask how to break the news. His employer had just become the first publicly-traded frontier AI lab. He called the collision of events awkward. The Riemann Hypothesis dates back to 1859 and ranks among the Millennium Prize math problems.
Lean, an open-source language that helps verify mathematical proofs, is why much of the field seems confident the formalized result will hold. Mathematicians are working through the argument in a Zulip chat, where Chris Birbeck, a number theorist at the University of East Anglia, unpacked the cohomology bound and the automorphy-lifting reasoning behind it. Twenty readers answered his post with the exploding-head emoji. A full reckoning from the community will take a while.
Alpöge admitted some guilt about releasing the proof without a human-intelligible explanation. A childhood obsession with the problem convinced him he could not live if he waited, he wrote. He described a stream-of-consciousness jaunt through finding every piece of the argument amid incredible confusion with Claude, saying the model flew, danced and sparkled near the finish line. He has committed to pursuing a version humans can explain to each other.
The same week, Alpöge and NYU mathematician Tristan Buckmaster released another major proof, this one on the Navier-Stokes equations. Those problems helped spark the summer race between frontier AI labs and mathematicians.
Ravi Vakil, a Stanford professor and president of the American Mathematics Association, posted a proof of the General Riemann Hypothesis written by Stanford postdoc Alessandro Masullo. Vakil noted it is all human, with no AI. The original plan was to bring Masullo to Stanford for a year to polish the writing and run it past enough people before announcing; Vakil said it now seems best to put out the current draft.
Masullo said the proof represents twelve years of work. He aimed a message at any mathematician discouraged by the pace and tone of the current climate: we are still here, we are very much alive, and we are not going anywhere.
TopicsLevent Alpöge · Anthropic · Claude · Dario Amodei · Tristan Buckmaster · Ravi Vakil · Alessandro Masullo · Stanford University
Source: techcrunch.com
Written by AI from verified reporting.
More stories
- Tech & AI cryptobriefing.com 1 viewReport: Nvidia loses $2.8T in market value over 33 days after CEO exit
- Tech & AI cryptobriefing.com 3 sources 1 viewS&P 500 Falls Below 7,000 as SEC Reportedly Probes Nvidia
- Tech & AI blockchain.news 1 viewOpenAI Builds MedChemBench to Grade AI on Early Drug Discovery
- Tech & AI x.com 3 sources 0 viewsSEC opens probe into Nvidia over accounting questions, stock tanks

