Wes Roth Video posted an update
this might be HUGE
Why it mattersThe description pitches building a team of AI agents, while the chapter list includes OpenAI model sabotage and weight-space programming. Linked sources point to OpenAI research on covert underperformance and Linux kernel security, alongside experiments in programming through model weights.
Discuss: How might the chapter on OpenAI model sabotage change the way you think about building and supervising AI agents?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.