Competing with China on AI’s Social Impact

Summary

This opinion essay challenges the prevailing US and Silicon Valley goal of beating China in an AI race for technological supremacy. It argues that advances made in the United States are likely to diffuse to China through open-source adoption, capability extraction, proxy services that bypass restrictions, talent and academic links, and theft. In the author’s view, accelerating the most advanced US systems can therefore also accelerate Chinese capabilities. The essay says the more urgent risks are broader: inexpensive AI could give dangerous expertise to non-state actors, including for advanced weapons, while poorly controlled systems could cause harm through misalignment. It argues that the United States and China should cooperate on treaties and other measures addressing these risks rather than pursue an unrestricted arms race. The article also criticizes the domestic social effects of AI, claiming that cognitive offloading weakens memory, critical thinking, creativity, and independent judgment, while AI companions may deepen loneliness and reduce people’s ability to socialize. It concludes that the meaningful competition may be over which country has the least AI-damaged population, and says China is ahead on regulation in that comparison.