By coincidence (I think?) this episode was released during the AI Pause debate week on the forum, so I felt like it was worth adding to the mix. The episode has general LMM/AI chat unit ~22min in, and then the specific question “should we pause AI?” is discussed from 23:10 onwards.
The target audience for this is probably more general than many of the AI safety professionals in the EA community, however I think its 1) useful for there to be some more intro-level materials about pausing AI available on the forum and 2) There are kernels of interesting ideas in this interview which I think will add something to the debate.
Any chance of a TLDR from someone who listened?
Small chance from me. If I listen to it, which I estimate is a 40% chance this week, and if I do, 99.9% chance that, if I do, I’ll give you the TLDR.
Would you still like a TL;DR? Sorry it’s been so long.