US-China Relations: History, Culture, and AI Competition, with Noah Smith, from Econ 102

Episode Summary

Executive Summary: The episode frames U.S.-China relations as a long-term strategic rivalry shaped less by ideology than by power, incentives, and China’s shift under Xi Jinping toward contesting the U.S.-led order. Noah Smith argues the U.S. is no longer fully in control of the relationship, that export controls are mainly about preserving conventional military deterrence, and that the most important near-term risk is Taiwan and autonomous warfare—not consumer AI models like DeepSeek.

Main Topics: Historical arc of U.S.-China relations (Priority: 5/5): The discussion traces a long period of cooperation, including alliance-like alignment against the Soviet Union, followed by rupture under Xi Jinping. Smith argues the relationship was not inevitably hostile and could be repaired in the future. China’s strategic goals under Xi Jinping (Priority: 5/5): Smith argues Xi shifted China toward rivalry with the U.S., seeking to weaken American influence, contest the U.S.-led global order, and increase China’s ability to act independently and coercively. What U.S. export controls are meant to do (Priority: 5/5): The core policy rationale discussed is preserving a military technological edge—especially in chips, tools, software, and talent—so the U.S. can maintain deterrence against China in future conflicts. AI, DeepSeek, and the limits of the AI arms-race narrative (Priority: 4/5): Both speakers question whether frontier LLM progress itself is a decisive geopolitical threat. Smith is skeptical that models like DeepSeek meaningfully harm the U.S., though he remains concerned about military AI applications. Taiwan, drones, and conventional deterrence (Priority: 5/5): The conversation emphasizes that the real danger is military conflict, especially around Taiwan, and that export controls matter insofar as they affect precision weapons, chips for drones, and autonomous systems. China’s pressure tactics and destabilization (Priority: 4/5): Smith describes influence operations, sharp power, cyber operations, cable-cutting, and economic leverage as ways China may weaken U.S. cohesion and reduce America’s ability to resist Chinese ambitions. Deterrence, escalation, and future stability (Priority: 4/5): Smith argues the goal should be to freeze the balance for roughly a decade, deter war, and eventually restore a stable relationship rather than pursue unilateral weakening or total decoupling.

Key Arguments: The U.S. and China were often cooperative in the 20th century, so permanent civilizational rivalry is not inevitable. Xi Jinping, not the U.S., was the first major actor to shift the relationship toward rivalry and decoupling. China’s ambitions are best understood in terms of power over trade routes, information infrastructure, and regional military dominance. The U.S. still has agency, but it no longer fully controls the bilateral relationship; China now has a decisive vote. Export controls are most defensible as a way to preserve conventional deterrence, especially for future autonomous drone and precision-weapons warfare. Banning or limiting chips alone will not stop China from building frontier AI models, and may mostly affect civilian access and some hardware progress. The biggest concern is not LLM competition per se, but whether AI contributes to military advantage in a Taiwan or broader regional conflict. China’s sharp power, cyber activity, and influence operations could weaken U.S. social cohesion and political resolve. A stable, deterred relationship is more likely to make future reconciliation possible than escalation or unilateral disarmament. DeepSeek appears to the speakers more like a competitive technical achievement than an existential geopolitical threat.

Data Points: Duration of rupture under Mao: 15–20 years - Smith references the period of rupture in U.S.-China relations under Mao before rapprochement. U.S. trade-control effectiveness window: Next 10 years - Smith argues the next decade is the peak risk period and the key horizon for deterrence. China’s chip manufacturing benchmark: 7-nanometer chip - Smith cites China’s attempt to manufacture a 7nm chip as evidence that equipment controls have slowed domestic capability. U.S. cloud customers cited in ad read: 42,000 businesses - Sponsor segment for NetSuite mentions scale of adoption, not the core argument. OCI cost claims in ad read: Up to 50% less compute, 70% less storage, 80% less networking - Sponsor segment for Oracle Cloud Infrastructure; not part of the policy discussion. U.S. trade deficit concern era: 1990s–2000s - Smith describes Trump’s trade anxieties as rooted in earlier fears of industrial decline. Relative power framing: China is bigger than the U.S. in some important ways - Smith argues the U.S. is no longer the sole dominant actor in the relationship.

Pivotal Quotes: "We have to get out of the mentality of thinking that America makes all the choices... We're not the big one anymore." — Noah Smith: Smith’s central framing of the bilateral power shift. "China would like to destabilize our society... the more we fight each other, the less we have the option value to stop China from doing whatever it feels like doing." — Noah Smith: On sharp power, social division, and strategic weakening. "The purpose of the export controls is to maintain a military technological edge over China." — Noah Smith: On the main rationale for semiconductor and equipment restrictions.

Implications: Listeners should expect a long period of U.S.-China tension centered on deterrence, chips, and Taiwan—not simple “AI race” headlines. Policy debates will hinge on what actually preserves security without unnecessarily slowing civilian AI growth or escalating conflict.