Four years of AI timeline forecasting may have hurt more than helped
JacquesThibs · x · 2026-11-15
Jacques Thibault questions the value of AI forecasting work: the past four years of attaching numbers and probabilities to far-future events may have delivered little counterfactual benefit compared to other efforts — possibly even net negative, by focusing attention and creating feedback loops on the wrong things. He also notes the AI Futures Project's median prediction of at least 2032 made at the start of the year, arguing this at least warrants reflection on how one should update on such forecasts, though not to zero.
Related event: AI safety community debates whether forecasting work is overrated(4 posts)→
More from AGI Musings
- DeepMind researcher: AGI will likely decentralize power, not centralize it — sebkrier · 2026-11-15
- Skeptic asks how RSI can start when LLMs fail at far out-of-distribution tasks — mariofilhoml · 2026-11-15
- Ex-Anthropic researcher Jacob Coxon: AI near recursive self-improvement within a year — rohanpaul_ai · 2026-11-15
- Prediction: Superintelligence kills profits and riches; UHI worth 100x salaries by 2035 — davidpattersonx · 2026-11-15
- a_rohan_ dreams of a sun-powered GPU constellation training in space after SpaceX rocket catch — _arohan_ · 2026-11-15
- tszzl: automated research can crack mechanistic alignment — 'it's more like math than cell bio' — tszzl · 2026-11-15