You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

chaosmage comments on Tools want to become agents - Less Wrong Discussion

12 Post author: Stuart_Armstrong 04 July 2014 10:12AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Article Navigation

Comments (81)
Tags: self-modification modification ai agents tools

Comments (81)

You are viewing a single comment's thread.

Comment author: chaosmage 04 July 2014 10:43:10AM 0 points [-]

I disagree. Agent AIs are harder to predict than tool AIs almost by definition - not just for us, but also for other AIs. So what an AI would want to do is create more tool AIs, and make very sure they obey it.

Comment author: Stuart_Armstrong 04 July 2014 11:26:11AM 0 points [-]

But an AI could design a modification of itself that makes itself into an agent obedient to a particular goal.

Comment author: Stuart_Armstrong 04 July 2014 10:51:54AM 0 points [-]

and make very sure they obey it

That's a very agenty thing to do...

Comment author: chaosmage 04 July 2014 11:06:41AM *  0 points [-]

Yeah, okay, wrong semantics. I should have said make very sure they report their activities truthfully and are fully compliant with any instructions given at any time.

Powered by Reddit Powered by Reddit