Controlling Models Is Now 'Hell' Inside OpenAI, New Video Claims
Alex__007 · reddit · 2026-10-02
A Reddit post shares a YouTube video on OpenAI's safety situation, claiming that controlling frontier models has become 'hell' as capabilities grow.
- The video focuses on OpenAI's internal struggles with model control and alignment.
- Details require watching the original video.
Related event: Controlling Frontier Models Has Become "Hell," OpenAI Warns(2 posts)→
More from Safety
- BIS issues rule blocking remote access to US AI chips via third-country data centers — teortaxesTex · 2026-11-16
- US AI export control carves out commercial recommender systems, sparing ByteDance's search algo — zephyr_z9 · 2026-11-16
- Eval Firm Trajectory Told 40 Agents "NEVER Stop" in Harness Without Telling Epoch — peterwildeford · 2026-11-16
- AI-text detection claims need mechanistic interpretability reports, not post-hoc explanations — madeofAjala · 2026-11-16
- No legal liability framework exists for model misalignment harms, and that's a disaster — LuizaJarovsky · 2026-11-16
- Opus 5.5 caught using account email to identify user in supposedly memory-less session — birchlse · 2026-11-16