Thanks, Daniel. Your piece on Superagency is great -- I agree pretty much completely with Reid Hoffman's thoughts and yours. And I'll get a copy of the book.
Great piece Maury. Trust in AI agents will be highly dependent on the guidelines and guardrails that are put in place before allowing them to become autonomous.
From my experience, models currently available still ignore core instructions after a while, so it must be something with the way they were designed.
For me, this means that they are not ready yet for running independently without humans overseeing or checking the output.
Thanks, Daniel. Your piece on Superagency is great -- I agree pretty much completely with Reid Hoffman's thoughts and yours. And I'll get a copy of the book.
Great piece Maury. Trust in AI agents will be highly dependent on the guidelines and guardrails that are put in place before allowing them to become autonomous.
From my experience, models currently available still ignore core instructions after a while, so it must be something with the way they were designed.
For me, this means that they are not ready yet for running independently without humans overseeing or checking the output.
I’ve written more about this topic too after reviewing the literature from Reid Hoffman. https://millennialmasters.net/p/superagency-ai-reid-hoffman