Gemini 4 reportedly "proved" nearly half of the false theorems it was given
RishabJainK · x · 2026-10-29
Citing Adam Holtererer's finding, Rishab Jain notes that Gemini 4 "proved" almost half of the false theorems it was given in a test — meaning the model constructed plausible-looking proofs for wrong propositions instead of catching their errors.
More from Fun
- Are we ready to admit React hooks were a mistake? — tlakomy · 2026-10-29
- AI-generated Batman F1 helmet concept image wows with carbon-fiber realism — Delahuntagram · 2026-10-29
- Google touts Carbon beating Opus 5.5, but good luck actually trying it — QuixiAI · 2026-10-29
- AI-Generated Fake: Korn's 'Freak on a Leash' as a 1973 Spanish TV Performance — Mother-Pear-2392 · 2026-10-29
- Jonathan Chang posts Atlanta photo of three AI figures, crowdsources memes — MatthewChang · 2026-10-29
- "Babysit this for me" is now the number one request people make to AI agents — RichardsonDx · 2026-10-29