Google AI Overview invents "weak prisoners dilemma" concept with no real source
jd_pressman · x · 2026-11-21
Following up his decision-theory argument, jdpressman searched for the claimed "weak prisoners dilemma" (a version where cooperation and unilateral defection both get zero punishment). Google Search's AI Summary asserted the distinction exists, yet no search result actually surfaces any author ever outlining such a concept — a suspected AI-summary hallucination of an academic term, and a vivid example of AI search overviews being unreliable.
Related event: CDT Prisoner's Dilemma Debate Sparks Suspected Google AI Hallucination(2 posts)→
More from Fun
- levelsio publishes 3,728-account AI reply blacklist; listed user slams 'no evidence' entry — PierceLilholt · 2026-11-21
- ChatGPT sides with friend in CDT prisoners-dilemma argument, author 'losing his mind' — jd_pressman · 2026-11-21
- Viral reply: 'Why is everything so much?' Answered with cosmic entropy comfort — CatAstro_Piyush · 2026-11-21
- Claude's spinner verbs deliver again: 'Whatchamacalliting' for 16 seconds — Angaisb_ · 2026-11-21
- User finds an ad on his iPhone lock screen and posts the receipts — TheMoonMidas · 2026-11-21
- Inkling 2 Mini builds a flight simulator fully offline on a Mac mid-flight — miramurati · 2026-11-21