No Wall Between 'Approximating Text' and 'Understanding Language': AI Timelines Debate Flares Up
teortaxesTex · x · 2026-09-20
Responding to phl43's claim that anyone confident in short AI timelines is overconfident, teortaxesTex pushes back in a long post: demanding first-principles proof suits mathematics and safety engineering but not understanding historical transitions that emerge from fuzzy probabilistic parts under pressure of capital and interest. His core argument: there is no real boundary between "merely approximating the text corpus" and "understanding language," nor between "better kernel engineering" and "RSI" — AI improvement is largely about better kernels, and in a sufficiently multidimensional space there are no walls; keep pouring in resources and progress continues.
More from AGI Musings
- AI labs' monitoring and actual research work is now 'basically all AI', says Jeff Ladish — JeffLadish · 2026-09-20
- Luiza Jarovsky: AI won't kill all humans by 2030, but may destabilize key global systems — LuizaJarovsky · 2026-09-20
- EA critics blast 'insects matter more than people' essay: keep them away from AI control — wen_ragnarok · 2026-09-20
- Helen Toner, who voted to oust Altman, on AI safety and overseeing OpenAI — trevposts · 2026-09-20
- AI Agent Switches Auto Insurance in 5 Minutes, Saving $3,500 a Year — rohanpaul_ai · 2026-09-20
- If continual learning is nearly solved, why do we still pick reasoning effort tiers? — akbirthko · 2026-09-20