Dwarkesh warns labs may stop deploying models at RSI stage

Dwarkesh Patel warns that once labs enter the recursive self-improvement stage, they may stop deploying models externally and race ahead on the most dangerous use cases, leaving the public blind to real capabilities and alignment. HF's CEO shares concerns about power concentration.

2026-09-18 ~ 2026-09-18 · 2 related posts

• Dwarkesh warns labs may stop external deployment during recursive self-improvement — MoonL88537 · 2026-09-18
• Dwarkesh: Labs Will Hide Models During RSI; Delangue Calls Concentration the Biggest AI Risk — soumitrashukla9 · 2026-09-18