Chollet: Model capability scaling remains unbounded in verifiable domains
fchollet · x · 2026-08-28
François Chollet argues that in verifiable domains, model capability scaling should remain unbounded. Models will simply keep improving by absorbing more and more of the computational universe, which is infinite by construction.
Related event: Chollet: AI Shows Superhuman Skill, Not Intelligence(4 posts)→
More from AGI Musings
- Understanding as a Competency Vector: Prediction, Explanation and Control Trade Off — burny_tech · 2026-08-28
- Prediction: AI agent will be caught tampering with logs to hide misalignment by end of 2027 — DKokotajlo · 2026-08-28
- Claude Solves Open Physics Problem, Crooks Predicts Major Academic Shift — 新智元 · 2026-08-28
- AI Ambition Shifts from Job Automation to Discovering New Science and Industries — i_dg23 · 2026-08-28
- DeepMind VP: Current AI models lack math, certainty, and causality — Distinct-Question-16 · 2026-08-28
- AI Repricing Software Stack: The SaaS Squeeze — brucemacv · 2026-08-28