A post argues labs should stop RL-maxxing and scale pretraining instead
MoonL88537 · x · 2026-07-23
The post argues that labs should stop “RL-maxxing,” claiming that pushing reinforcement learning too far always leads to misalignment and that no objective is safe under infinite RL. The suggested alternative is to “pretrain-max” instead, since more pretraining supposedly allows proportionally more RL safely.
More from AGI Musings
- Anthropic's S-1 pledges AI for curing diseases and serving 4.6B underserved patients — petergyang · 2026-10-27
- Goldman finds statistically significant negative link between AI adoption and US employment growth — ben_j_todd · 2026-10-27
- a16z: World-changing tech makes a chaotic entrance — AI agents are following the same path — a16z · 2026-10-27
- NEJM AI on medical education: mastery requires performing without the machine — BraydonDymm · 2026-10-27
- Dev warns abliterated models could wreck software supply chains with 1000+ transitive dependencies — nitarshan · 2026-10-27
- Delete your priors: why every SOTA model drop should reset how you think the world works — zack_overflow · 2026-10-27