This and all episodes at: https://aiandyou.net/ .

We are talking about international governance of AI again today, a field that is just growing and growing as governments across the globe grapple with the seemingly intractable idea of regulating something they don’t understand. Helping them understand that is Markus Anderljung, Director of Policy and Research at the Centre for the Governance of AI in the UK. He aims to produce rigorous recommendations for governments and AI companies, researching frontier AI regulation, responsible cutting-edge development, national security implications of AI, and compute governance. He is an Adjunct Fellow at the Center for a New American Security, and a member of the OECD AI Policy Observatory’s Expert Group on AI Futures. He was previously seconded to the UK Cabinet Office as a Senior Policy Specialist.

I know “governance” sounds really dry and a million miles away from the drama of existential threats, and jobs going away, and loss of privacy on a global scale; but governance is exactly the mechanism by which we can hope to do something about all of those things. Whenever you say, or you hear someone say, “Someone ought to do something about that,” governance is what answers that call.

We talk about just what the Centre is, what it does and how it does it, and definitions of artificial general intelligence insofar as they affect governance – just what is the difference between training a system with 1025 and 1026 flops, for instance? And also in this part Markus will talk about how monitoring and verification might specifically work.

All this plus our usual look at today's AI headlines.

Transcript and URLs referenced at HumanCusp Blog.