LESSWRONG
LW

Cody Breene
3Ω1030
Message
Dialogue
Subscribe

Posts

Sorted by New

Wikitag Contributions

Comments

Sorted by
Newest
AI alignment is distinct from its near-term applications
Cody Breene2y10

Agreed, this seems to be the path that OpenAI is taking with plugins.

Reply
AI alignment is distinct from its near-term applications
Cody Breene2yΩ140

When you say, "take over", what do you specifically mean? In the context of a GPT descendent, would take over imply it's doing something beyond providing a text output for a given input? Like it's going out of its way to somehow minimize the cross-entropy loss with additional GPUs, etc.? 

Reply
AI alignment is distinct from its near-term applications
Cody Breene2yΩ010

"AI developers should not develop and deploy systems with a significant risk of killing everyone."

If you were looking at GPT-4, what criteria would you use to evaluate whether it had a significant risk of killing everyone? 

Reply
No wikitag contributions to display.
No posts to display.