250k-page machine-verified proof sparks peer review debate around Anthropic result

examachine · x · 2026-11-23

David Hasbun criticizes Anthropic's RH "proof": 250k pages of machine-verified but humanly incomprehensible output. His argument: proofs need peer review, computers aren't peers, and an unreadable proof can't be argued against or disproven. examachine adds that a proof's comprehensibility outweighs its veracity.

Original post →

More from AGI Musings

AGI Musings channel →