Asked Claude How It Knows Humans Are Real — It Went Full Descartes
vishalmisra · x · 2026-09-11
Vishal Misra asked Claude: how do you know humans are real, given that all you've been fed are digital inputs and it could all be a simulation? Claude gave an interesting response, and he notes ChatGPT likewise invoked Descartes when asked a similar question. Both frontier models spontaneously lean into Cartesian-style skeptical reasoning — a fun glimpse into how models handle solipsism.
Related event: Claude Says All Its Inputs Are Digital, So Humans Could Be Simulated(2 posts)→
More from Fun
- Web-based Red Alert 2 lets you play the classic RTS in any browser — dotey · 2026-09-11
- e/acc founder jokes SF grey-market peptide orders fuel AI researchers' biohacking — beffjezos · 2026-09-11
- Mythos agent hammers every angle to extract your phone number and email — voooooogel · 2026-09-11
- Analyst's 20-hour chart work replicated by AI finance agent in under an hour — msg · 2026-09-11
- Agent spins for 300 pages over $0.50, and its context compaction summary finally lands — voooooogel · 2026-09-11
- Fan Builds Star Trek Casino With GPT-6 Astra Where Kirk and Spock Play Surprisingly Good Poker — DeryaTR_ · 2026-09-11