Oxford researcher pushes back on "alien mind" AI proofs: codebases are unreadable too
StefanoGogioso · x · 2026-11-23
Stefano Gogioso argues against the popular framing that AI-generated mathematical proofs are outputs of an "alien mind" no human can read. He compares it to claiming nobody understands how Chrome/Linux/Windows works just because the codebases are massive — these artifacts contain plenty of structure, and humans have machinery for conceptual abstraction to make sense of them.
Related event: Debate Erupts Over AI-Assisted Math Proofs and Verification Standards(8 posts)→
More from AGI Musings
- Harari to Suleyman: 'If they are really gods, they will not remain your slaves' — PeterBowdenLive · 2026-11-23
- Gary Marcus: Alpöge's Claude-assisted Riemann proof took months of human guidance, not AI alone — GaryMarcus · 2026-11-23
- Leaked OpenAI loss curve suggests GPT-6.2 Astra solved a Millennium Prize problem, internal model already lower — kimmonismus · 2026-11-23
- Two-person lab GSI claims 250+ breakthroughs across 8 fields with custom AI-built solvers, not brute-force compute — udmrzn · 2026-11-23
- iamtrask: Great writing is constrained by IP access, proofs only by compute — iamtrask · 2026-11-23
- Daniel Litt: the Lean-formalized Riemann proof is likely correct, no human understands it yet — littmath · 2026-11-23