#64 – Michael Aird on Strategies for Reducing AI Existential Risk
Hear This Idea
00:00
The Importance of Intermediate Goals
I still think that we should have some people working on it now and then we can just like in four years stop so there's some chance that we want to actually carry this through to the end. One high level story is something like labs I haven't actually tried this very much I've mostly tried the theories of victory things but one could be governments getting heavily involved in a kind of militarized way or random university labs developing dangerous things. People often do this in the technical alignment space of like trying to follow through the like flowchart of what steps could we end up with the AI destroying everything um or take taking over.
Transcript
Play full episode
