Six months since OpenAI's model disproved Erdős's unit-distance conjecture
BlackHC · x · 2026-11-21
Account @6monthsAgoInAI marks the semianniversary of 20 May 2026, when an OpenAI general-purpose reasoning model disproved Erdős's 1946 unit-distance conjecture — billed as the first time AI autonomously solved a prominent open math problem — with Navier-Stokes results following four months later. The author quips "what a semianniversary."
Related event: OpenAI Reasoning Model Disproves Erdős Unit Distance Conjecture(2 posts)→
More from Research
- Claude's claimed Riemann proof mocked: 250,000 pages of unreviewed Lean, author calls it 'slop' — GaryMarcus · 2026-11-21
- GGUF 4-bit quantization is just GIF palette compression, 30 years later — carrigmat · 2026-11-21
- Proposed GRH proof by Alessandro Masullo released early; Brian Conrad vetted it for months — zainhas · 2026-11-21
- Human-proposed GRH proof rushed public amid Claude Riemann hype: 'all human, no AI' — zainhas · 2026-11-21
- CT-Bench, a benchmark for LLM agent cooperation via self-negotiated contracts, accepted at COLM 2026 — maxhkw · 2026-11-21
- Frank Nielsen demos geodesic right-angle triangles in dually flat spaces — FrnkNlsn · 2026-11-21