teortaxesTex: RL may give coding models 'full fluency' with no superhuman headroom
teortaxesTex · x · 2026-11-11
AI commentator teortaxesTex offers a notable mental model: current midtraining and RL approaches may enable "full fluency" in coding — perfect alignment with the human code distribution — without any need or potential for "superfluency" beyond it. He stresses this isn't a prediction, but says such an outcome would no longer shock him, unlike in the past. He considered writing a longpost on the idea but likely won't.
More from AGI Musings
- "Skipping your 401k over fantasy x-risks means you're innumerate" — inductionheads · 2026-11-11
- Beff Jeffzos: data centers face nuclear-power-style stigma from decel mind virus — beffjezos · 2026-11-11
- Founders quit profitable YC startup, went salary-less for 2 years to warn on AI risk — jeremiecharris · 2026-11-11
- Human edge over frontier AI: the willingness to commit heresy — JeremyNguyenPhD · 2026-11-11
- François Fleuret to biologists: look how cocksure other fields were before AI — francoisfleuret · 2026-11-11
- Could Recursive AI Improvement Compress Millennia of Progress into One Decade? — VraserX · 2026-11-11