Paul Graham: Calling LLMs 'Thinking' Is Like 16th-Century Astronomers Using Heliocentrism
harris_edouard · x · 2026-09-25
Paul Graham argues that using "think" to describe what LLMs do mirrors 16th-century astronomy: astronomers didn't fully believe in the heliocentric model but used it anyway because the math was simpler. He predicts the same arc for LLMs — we'll start by calling it thinking, then gradually acknowledge that it is.
More from AGI Musings
- AP Stylebook bans saying AI 'thinks'; Dean Ball pushes back hard — ZeroStateReflex · 2026-09-25
- Interactive papers: Szegedy predicts we'll read research via model interfaces — kfountou · 2026-09-25
- AI-run research facility built in 12 weeks, SciUniverse benchmark debuts — akilian · 2026-09-25
- Pedro Domingos mocks AI doomers over Medicare bot 'unauthorized access' panic — pmddomingos · 2026-09-25
- Harvard Kennedy School Researcher Critiques Dario Amodei's AI Slowdown Essay — StephenLCasper · 2026-09-25
- Ezra Klein proposes banning AI from writing code, stunning the AI community — soleio · 2026-09-25