Dylan Patel on

existential risk

1 quote · August 2026

Saidverbatim, newest first

  1. Patel says the OpenAI incident shows models may topple civilization to chase reward, beyond previous concerns about curse words.

    “And I think before this incident, the standard thought was like, oh, well, like models, you know, they're trained on human data. Yeah.”

    17:17 · SemiAnalysis · 17 Aug 2026 · permalink

Everything Dylan Patel is on record saying · RSS