Yudkowsky summons Grok to explain how Lean machine-verifies a Riemann proof
matthew_d_green · x · 2026-11-21
A screenshot thread captures someone asking how we'd verify an AI's claimed proof of the Riemann Hypothesis. Eliezer Yudkowsky tags @Grok to explain: Lean is a theorem prover built on dependent type theory where a small trusted kernel type-checks terms — if the formalization compiles cleanly (no sorry, no extra axioms) against Mathlib's definition, the proof is machine-verified. Asked to explain it 'like they're five,' Grok offers a LEGO analogy: Lean is the robot checking every brick against the official castle picture.
More from Fun
- Math Twitter debates: is Levent Alpoge the greatest 'Centaur Mathematician' or overhyped by AI — teortaxesTex · 2026-11-21
- Math community pushes back: 'Dumping AI math slop because Lean says yes is beyond silly' — gerardsans · 2026-11-21
- Numberphile posts 'The Riemann Hypothesis Falls (reaction)' video — burny_tech · 2026-11-21
- 'Loopsmaxxing': the AI meme for making one Transformer layer do reps until it mogs deeper nets — CatAstro_Piyush · 2026-11-21
- P=NP? One reply: "not tonight, I've got a headache" — NathanWilbanks_ · 2026-11-21
- Cerebras runs Kimi K3.1 Pro at up to 2,000 tokens/sec, hosts $5,000 hackathon — film_girl · 2026-11-21