Top chatbot hallucinates a nonexistent painting and writes 1,000-word review of it
ValerioCapraro · x · 2026-11-10
Behavioral scientist Valerio Capraro photographed a painting in Milan and uploaded it to one of today's most advanced chatbots, accidentally including part of a neighboring painting. Instead of admitting ignorance, the model fused elements of both paintings, hallucinated a third one that doesn't exist, and produced a convincing 1,000-word review of it. His takeaway: these systems have no access to reality — hallucination is not an error, it's their default state.
More from Models
- Anthropic: Claude Fable 5.2 designs working electronic hardware end to end — EricBuess · 2026-11-10
- Claude Mythos 5.2 redisCOVERS 89.6% of known CVEs in Anthropic's internal eval — minchoi · 2026-11-10
- Claude Mythos 5.2 accurately models 28 of 34 viral capsids for gene therapy research — minchoi · 2026-11-10
- Fable 5.2 solves two open black hole problems in theoretical physics, Anthropic says — minchoi · 2026-11-10
- Anthropic releases new Mythos and Fable models amid RSI concerns — TechCrunch AI · 2026-11-10
- ChrisGPT claims he leaked Fable 5.2's '30% faster' claim weeks before Anthropic's announcement — ChrisGPT · 2026-11-10