John Nash's 1998 Hierarchical Introspective Logics paper recirculates in AI circles
fkasummer · x · 2026-10-16
@fkasummer highlighted John F. Nash Jr.'s 1998 paper "Hierarchical Introspective Logics", proposing a hierarchy of logic levels where higher levels have an overview of lower-level reasoning. Stimulated by the FLT proof and computer-aided four-color theorem, Nash asked whether unprovable but true statements might have to be accepted as new axioms, building from Gödel's incompleteness results. The poster praised Nash's embedding smoothness work too.
More from AGI Musings
- New startup path: VC-funded founders become creators, then incubate products solo — FinanceYF5 · 2026-10-16
- Agent insurance may emerge: insurers will demand audits before underwriting AI agents — FinanceYF5 · 2026-10-16
- Every agent is a new attack surface: security firms see an opening in the agent era — FinanceYF5 · 2026-10-16
- The next model race is memory: whoever remembers users most wins, says exec — FinanceYF5 · 2026-10-16
- 'We spent a century teaching people to think like machines — now the machines are here' — r0ck3t23 · 2026-10-16
- Redditor defines AGI/RSI/ASI concretely, claims GPT-6 'Astra' already crossed the line — Crayola_God · 2026-10-16