Ethan Mollick has AI run a Fermi paradox study on public data: 73% odds we're alone in the Milky Way
emollick · x · 2026-11-11
Ethan Mollick tasked two models (Fable 5.2 and Gemini Carbon) with solving the Fermi paradox — marshal data, test hypotheses, pressure-test against critics, and give probabilistic answers. Fable 5.2 produced a full research site built by Claude over 2 days.
Key findings: a 73% chance we are alone in the Milky Way (no civilization ever arose, none inbound from other galaxies), 16% that civilizations arise but never spread between stars (some universal barrier), and 11% that someone has spread — 4% still en route, 7% already here unnoticed. Randomly varying assumptions across 400 reruns yields a typical 56%, with 90% of runs between 26% and 78%; other well-known stances imply 13%–94%. The evidence only firmly rules out a visible galactic settler.
Notably, the analysis, code, and text were all AI-generated, with only a separate Claude run as reviewer — no human expert checked anything, making it a public experiment in end-to-end AI research.
Related event: AI Models Tackle the Fermi Paradox: 73% Chance We're Alone(2 posts)→
More from AGI Musings
- Game devs shouldn't cave on AI use, argue founders: judge the output, not the tooling — petergyang · 2026-11-11
- Creon: algorithmic gains will eclipse hardware improvements — the reverse bitter lesson again — bradneuberg · 2026-11-11
- AI researcher Delip Rao opposes any AI bailout: families first, not recursive intelligences — deliprao · 2026-11-11
- LLM moral choices lack stable rankings; averaging fakes human-like moral hierarchy — rohanpaul_ai · 2026-11-11
- David Hoang revisits his 2026 focus areas amid high displacement variance — davidhoang · 2026-11-11
- OpenAI misalignment incident details drip out on X as analyst urges Anthropic to shorten risk reports — DKokotajlo · 2026-11-11