#64 – Michael Aird on Strategies for Reducing AI Existential Risk
Hear This Idea
00:00
How to Monitor AI Development and Construct How It's Used
So we have this sort of like three layer thing where you have a portfolio of theories of victory and how much weight we put on each of them given each theory of victory. So for example the US led and the multi-lateral thing both share it's really important that we have an actual way of monitoring where the AI is and maybe constraining how it's used. But in contrast one concrete example where that doesn't happen is something like things likeThings like the US and October the US did quite intense export controls related to AI chips and chips in general on China. See AI can help for governance research. It might be useful because it might be overly involved, but ultimately any story will
Transcript
Play full episode
