Claude Code team keeps a framed 'Bitter Lesson' on the wall: 'never bet against the model'
Hesamation · x · 2026-11-03
@Hesamation notes the Claude Code team has a framed copy of Rich Sutton's 'The Bitter Lesson' on their wall, and resurfaces an interview with Boris Cherny from 8 months ago. Their philosophy: 'never bet against the model' — the general model always beats the specific one. Non-model code is 'scaffolding,' and the team constantly weighs building a feature to extend capabilities by 10-20% versus just waiting a few months for the next model to do it. Their rule: feel out the current model's boundary, then build for the model you expect in six months.
More from AGI Musings
- Old Vision Architecture Tricks Like ResNets Are Being Reinvented Weekly, Says Researcher — code_star · 2026-11-03
- Whoever Wins Today, AI Will Change Your Life More Than Any Politician — rand_longevity · 2026-11-03
- Chinese blogger's advice for CS students in the AI era: pick a niche, pay for top AI tokens, invest half your salary — vista8 · 2026-11-03
- "People who build things with AI will never exceed 10%": the metacognition argument sparks debate — IndraVahan · 2026-11-03
- AI safety researcher: technical 'AI Safety' work has been product work empowering big labs — j_foerst · 2026-11-03
- Alan's AGI Countdown Shows Just 1% Remaining, Sparking Reddit Debate — deles_dota · 2026-11-03