The bitter lesson, interpretability edition: humans steer only the weakest AI in the stack
burny_tech · x · 2026-11-21
A viral chat screenshot distills "the bitter lesson, interpretability edition": a tower of black-box AI systems, each steering the one slightly more capable than it, with humans steering the weakest one.
The reaction emojis say it all — as model capability stacks upward, the human oversight interface keeps getting pushed to the bottom of the chain.
More from AGI Musings
- Build for automation that serves humans: staying in control — gdechichi · 2026-11-21
- Multiplayer coding agents: agent attention is unbounded, human attention isn't — manosaie · 2026-11-21
- Konrad Kording doubles down: AI people should hold put options as a hedge — KordingLab · 2026-11-21
- Turing Award winner David Patterson claims superintelligence is here, all science solved by 2030 — davidpattersonx · 2026-11-21
- Lean has de-bundled discovery from explanation in math, argues economist — soumitrashukla9 · 2026-11-21
- Amazon's 2007 Subscribe & Save was arguably the first mainstream commerce agent, used by ~25% of customers — signulll · 2026-11-21