11 - Attainable Utility and Power with Alex Turner
AXRP - the AI X-risk Research Podcast
00:00
What Kinds of Symmetries Do You Need for Instrumental Conversion?
There are two ways that a this gets formalized in the paper. They can be in comparing two different sets of states and saying, you know, there's some relation between them, and one's just kind of bigger somehow. What kinds of symmetries do you need for this to be true? And like, how often in reality do you expect those symmetries to show up? So with the symmetry argument, we want to be thinking, well, what parts of the environment can make instrumental conversion true or false? Ar like, hold or not hold, in a given situation? The other way is by looking at long term reward maximizing agents who tend to seek power seeking over short
Transcript
Play full episode
