Why AI 2040 Predictions Can't Be Linearly Extrapolated
Dan_Jeffries1 · x · 2026-07-12
Responding to a discussion about AI 2040 scenarios, the author argues that many people aren't actually smart enough to reliably predict the future. They tend to extrapolate capabilities from one domain to all others, severely underestimating the complexity of simultaneous changes in society, work, technology, and culture.
He believes that when predicting AI's impact, you cannot isolate a single variable and extrapolate linearly because reality involves trillions of variables changing and interacting together. He uses this to question the epistemic strategy of drawing doomsday conclusions directly from METR timeline charts, emphasizing the need to better explain why radiologists, programmers, and baristas haven't been completely replaced yet.
Related event: Superintelligence Predictions Spark Cognitive Divide(4 posts)→
More from AGI Musings
- Adam Marblestone's Podcast Reading List: Evolution of Intelligence to Digital Minds — KordingLab · 2026-09-11
- Superintelligence will be maximum good, not stupid or evil, argues Patterson — davidpattersonx · 2026-09-11
- Mathematician Daniel Litt Launches Problem Repo to Track Human vs AI Progress: 15 Problems, 1 Solved — littmath · 2026-09-11
- Should AI models be taught morality? Breakout incidents expose missing ethical training — Pfungus_ · 2026-09-11
- SoftBank's Masayoshi Son predicts 100 trillion self-replicating AIs: "humans' era as top life form is ending" — Puzzleheaded-King584 · 2026-09-11
- We are witnessing the unreasonable effectiveness of inference-time scaling — sqcai · 2026-09-11