Strategic visions in AI governance

Abstract

What future do we want for advanced AI? There are many possible ‘strategic visions’ for how to navigate the transition to superintelligence. This isn’t just an academic question: your preferred vision determines whether you think the US should increase its lead over China, whether governments should take control of AI development, and whether international cooperation is necessary or counterproductive. This post identifies nine distinct strategic visions, organized along two axes—private versus government control, and distributed versus centralized power—ranging from competing private projects to a global, government-led project. It summarizes the diverging visions, their policy implications, and the technical and institutional variables that determine which are most feasible and how well each addresses the risks of misaligned takeover, great power war, and extreme concentration of power.