Discussion about this post

User's avatar
Frank Bruno's avatar

Joel, this post is great! Your point about "having an accurate mental model of AI’s capabilities and limitations" is right in line with my current research. If the AI just follows orders (which it will) it can perfectly execute a fundamentally wrong objective. Great to see someone highlighting that AI's best feature is often its ability to disagree.

Chris Tottman's avatar

Can you hire and lead humans well whilst building and deploying valuable agents at the same time ..

15 more comments...

No posts

Ready for more?