The Cognitive Revolution Video Watch posted an update
AI:AM: Where's the Steering Wheel? My Agent Set the Password, Hidden Valence, Proofs Find Linux Bugs
Why it mattersThe episode examines how to give AI agents operational access while retaining human control, including real-world password lockouts and cyberattacks on supermarket refrigeration. It also discusses hidden valence in language models, using Lean proofs to find Linux kernel bugs, and whether formal verification can advance AI safety.
Discuss: What safeguards for AI agents would best preserve human control when operational access can lead to password lockouts?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.