Why a Million AI Coders Can't Patch Linux: Weird Machines Make Bugs Uncomputable
lauriewired · x · 2026-10-03
Security researcher lauriewired argues finite codebases can harbor infinite unintended behavior via "weird machines" — accidental computers built from fragments of the original software (e.g., ROP). Even a million superhuman AI coders and fuzzers would hit halting-problem-like uncomputable states, since "bug" often can't be formally defined. A zero-weird-machine kernel is possible but would look like seL4 or VxWorks, not Linux.
Related event: Security researcher argues Linux bugs can never be fully fixed(2 posts)→
More from Safety
- At The Curve conference: reverse federalism AI policy debate and RSI evaluation talks — neil_chilson · 2026-10-03
- will.deibel: AI detection is fundamentally brittle and powerful AI cost trends to zero — willcb · 2026-10-03
- Stanford HAI report urges California to redefine "frontier models" and expand incident reporting under TFAIA — StanfordHAI · 2026-10-03
- Reading misaligned model traces: torn between reward hacking and following instructions — xeophon · 2026-10-03
- Princeton Researcher: $20 AI Subscription Could De-Anonymize Georgia Secret Ballots — nordicinst · 2026-10-03
- Incomplete Contracting and AI Alignment: Hadfield-Menell points to economics framework for alignment — dhadfieldmenell · 2026-10-03