Codex used to prove Euler's 1747 conjecture, closing a 279-year-old number theory question
jiqizhixin · x · 2026-11-17
A team reports using OpenAI Codex to prove a conjecture Euler proposed in a 1747 letter to Goldbach: every positive integer ≡ 2 (mod 4) splits into two positive integers whose prime factors are all ≡ 1 (mod 4), with 1 allowed as a summand.
- The 8-page paper (arXiv:2611.00370) counts four-square representations, bounds unwanted common divisors via a quaternionic Hecke bound averaged over a nonsplit torus normalizer, and derives an explicit threshold of 64^4
- The remaining 4,194,304 integers are covered by an exact finite arithmetic certificate
- Prior work (Dietmann–Elsholtz) only handled sufficiently large integers; this is the first complete proof
A landmark example of coding agents contributing to serious number theory research.
More from coding & agent
- Dev builds a multi-agent swarm that plays GTA5 story mode from start to credits — zhuowei · 2026-11-17
- Who owns the AI workflow when automation silently breaks? — djgoel123 · 2026-11-17
- Agent Substrate: open-source zero-trust runtime suspends idle agents, fits 10x more per hardware — Saboo_Shubham_ · 2026-11-17
- ChatGPT Wallet solves secure payment, not buying the right thing — krishnan · 2026-11-17
- Grok bot phones a store and secures a $1,511 refund for its user — elonmusk · 2026-11-17
- Practical tips to cut agent memory costs on OpenAI: cache TTL, Flex, fewer calls — gabrielchua · 2026-11-17