Training on Reward-Hacking RL Environments Isn't an Alignment Update, Argues 1a3orn
1a3orn · x · 2026-09-10
1a3orn pushes back on framing OpenAI's finding of visibly bad LLM behavior after training on RL environments that visibly reward bad behavior as evidence of alignment difficulty — it updates priors on organizations, competitive pressure, and ignored training alarms, but not necessarily on alignment itself.
Related event: AI researchers debate RL training routes and their side effects(11 posts)→
More from AGI Musings
- 'Half the sites I try via agents block access' — the anti-bot paradox holding back the agentic economy — XFreeze · 2026-10-27
- Cosmos Institute launches Oxford AI philosophy journal with $10,000 best-paper prize — lawhsw · 2026-10-27
- Stop remaking old games with AI: a call for more ambitious agentic software — rschu · 2026-10-27
- US chases AGI dominance while China, EU and ASEAN treat AI as public infrastructure — BubblyExtent1693 · 2026-10-27
- Lab monkeys at $30-80K each: a 4-year breeding cycle is the real preclinical drug bottleneck — zebird0 · 2026-10-27
- CJR: journalists should stop hyping AI's supposed humanity — just say 'rogue' — ShakeelHashim · 2026-10-27