“Another big AI math drop”: viral summary of AlphaProof Nexus
Dr Singularity @Dr_SingularityX
Why it matters
Highest-reach X post on the Science publication (~45.6k views, 1.16k likes at fetch time). It presented the May 2026 result as new news.
Summary
An aggregator account posted at 15:03 UTC on 10 Oct 2026 (read through api.fxtwitter.com). It summarises the paper’s headline numbers: 9 of 353 open Erdős problems, 2 of them open for 56 years; 44 of 492 OEIS conjectures; an algebraic-geometry question; a min-max optimization bound; and the LLM plus Lean verification loop. It does not say that the preprint dates from May 2026.
Archived text
Another big AI math drop. This time from Google DeepMind
Google DeepMind’s AlphaProof Nexus AI solved mathematical questions that had resisted progress for decades.
The system autonomously resolved 9 of 353 open Erdős problems, including 2 that had remained unsolved for 56 years, and also proved 44👀 of 492 open conjectures from the Online Encyclopedia of Integer Sequences.
It also resolved an open question in algebraic geometry and improved a known bound in min-max optimization.
The key: pairing an LLM’s ability to generate proofs with Lean’s formal verification. The AI proposes a proof, the checker tests every logical step, and failed attempts produce feedback for another try. Proofs must pass the checker to be accepted.
Researchers also tested a simpler approach. Repeatedly generating candidate proofs with an LLM and checking them in Lean, alongside their more sophisticated search framework.
