E116: Toxic out-of-control trains, regulators, and AI
All-In with Chamath, Jason, Sacks & Friedberg
00:00
Elon Musk's tweet about the dangers of AI
OpenAI's chat interface is programmed with a trust and safety layer that intercepts questions from users and determines whether the AI is allowed to give its true answer./nThis layer is programmed with a trust and safety layer similar to Twitter's L Roth, and is effective in preventing offensive or insensitive responses from being given.
Play episode from 58:25
Transcript
