Patterson: almost no one grasps how fast AI progress will accelerate
davidpattersonx · x · 2026-11-18
Quoting his own earlier prediction, davidpattersonx argues that extrapolating exponential trends is easy, but estimating how fast those trends accelerate is not. He had already factored acceleration into his ASI-by-end-of-year prediction, yet acceleration now seems even faster than he expected — "it should have been obvious," since superintelligence itself will drive massive acceleration.
In the main post he splits people into three tiers: most expect a flat line into the future; a few can plot improvement trends and extrapolate; almost no one understands how fast improvement will accelerate.
More from AGI Musings
- Tim Sweeney on AI job fears: same teleprompter script as the tractor panic 120 years ago — liu_mingyu · 2026-11-18
- LeCun signs open letter from 59 European leaders urging EU not to sleep through AI revolution — ylecun · 2026-11-18
- Dev observes that lying to LLMs works — and it's making people habitual liars — andimarafioti · 2026-11-18
- 20% of UK's Milburn Review report was AI-generated, sparking disclosure debate — birchlse · 2026-11-18
- 'Human' named Cambridge Dictionary's Word of the Year 2026 — shauntrennery · 2026-11-18
- Rodney Brooks, Roomba creator, calls today's humanoids learning dexterity 'pure fantasy' — lukas_m_ziegler · 2026-11-18