Abstract: Advanced machine intelligence is likely to arrive in the coming decades. When it does it will transform employment, wealth and power, and world order. The potential benefits are tremendous, but so are the potential risks. We urgently need AI Strategy: the study of how humanity can best navigate the transition to advanced AI systems. This document surveys this research domain, dividing it into four complementary research clusters. (1) Understanding the technical landscape: foreseeing potentially transformative capabilities, modeling and forecasting the development of AI, and appraising the viability and character of scalably safe systems. (2) Understanding AI politics: who are the relevant actors, what do they want, how will they interact, how will wealth and power be transformed, what is the risk of a great power arms race, what routes exist for safe (international) governance of AI? (3) Understanding ideal global AI governance: given the opportunity to build institutions to globally govern AI, how should we do so in a way that remains effective, safe, legitimate, accountable, inclusive, and incentive compatible? (4) Formulating and advising near-term policy for relevant actors.

 

Speaker Bio: Allan Dafoe is an Assistant Professor of Political Science at Yale University and a Research Associate at the Future of Humanity Institute, University of Oxford. His research seeks to understand the causes of world peace and stability. Specifically, his research has examined the causes of the liberal peace, and the role of reputation and honor as motives for war. He develops methodological tools and approaches to enable more transparent, credible causal inference. Allan is beginning research on the international politics of transformative artificial intelligence.