The Geopolitical Strategy of US-China AI Competition

AI 2040 (AI Futures Project , 2026) (60 minutes)

• Read all 5 policy plan scenarios (they’re not very long each, but there are multiple)

Gives a high-level understanding of the geopolitical and technical issues that arise in different policy plans for dealing with a US-China AGI race.

Strategic Visions in AI Governance: Mapping Pathways to Victory (IAPS, 2026) (35 minutes)

• Read Executive Summary

• Read A. Introduction

• Read C. Policy implications of strategic visions

• Read the HEADLINES of D. Cruxes between strategic visions (just the single-line headers of each subsection, to get an idea of the assumptions that matter in choosing a strategy)

What AI policy objectives should we work towards? This depends greatly on one’s strategic vision. Strategic visions are high-level views about how to successfully navigate the transition to a world with powerful AI systems. The goal of this report is to help policy entrepreneurs choose which strategic visions to embrace and identify policy objectives that are robust to uncertainty about which visions are best.

Podcast Interview: Hugh White (80,000 Hours Podcase, 2025) (15 minutes)

• Just the highlights sections “Taiwan's position is essentially indefensible” and “AGI may or may not overcome existing nuclear deterrence” (start at 33:20 on the “Highlights” audio)

Strategic Studies professor emeritus Hugh White argues that “Taiwan is in an impossible situation and we’re doing them a disservice pretending otherwise… The first is the one you touched on: that the smart thing for the United States is not to defend Taiwan, but to make sure that they’ve destroyed the fabs. That’s a very grim and gloomy proposition, but it’s a lot easier to destroy the fabs than it would be to defend Taiwan. The second point is that it would be an illusion to imagine that denying China access to the Taiwanese fabs makes a big difference to China’s long-term capacity to build this stuff for themselves.”