ChatGPT, Claude and Grok all answer p(doom) with ~10%, an odd cross-model consensus
maksym_andr · x · 2026-10-22
maksymandr observed that ChatGPT, Claude, and Grok all state a p(doom) around 10% regardless of model size and reasoning effort — screenshots show each model answering "about 10%, with a wide 2–30% band" while emphasizing the definition of doom matters.
His take: independently trained models converging on the same consensus answer for an open-ended question is striking, and it generalizes — current LLMs broadly avoid opinionated takes, which may be what still makes humans unique.
More from AGI Musings
- Ethan Mollick: AI coding tools help, but product bottlenecks lie elsewhere — emollick · 2026-10-22
- Oxford study: the public is an AI shareholder and deserves a happiness dividend — SandraWachter5 · 2026-10-22
- Alignment discourse checked: 'spooky emergent behavior' traced to fixable setup choices — random_walker · 2026-10-22
- Anthropic Publishes Its First Annual Public Benefit Report With Five Pillars — AnthropicAI · 2026-10-22
- "Mechanomorphizing": describing people with machine concepts — PierceLilholt · 2026-10-22
- Musk tells Isaacson: millionfold-smarter AI means no human control, money goes extinct — TheMoonMidas · 2026-10-22