Semiconductor rivalry is less a new Silicon-era order than a logistics-driven reinvention of U.S. dominance: control of supply-chain chokepoints and infrastructure sustains geopolitical power. The so-called 'Pax Silica' is fragile—hardware dependencies and networked control, not chip design alone, will shape technological governance and global stability.

Main Finding

The paper develops a theory of "logistical hegemony" and argues that contemporary struggles over semiconductor supply chains—often labeled “Pax Silica”—are not a wholly novel global order but a logistical adaptation of Pax Americana. Control over semiconductor chokepoints, export controls, friend‑shoring, and denial logistics extend long‑standing hegemonic logic (control of infrastructures that organize circulation) into the digital/semiconductor domain, creating new strengths and fragilities for global capitalism and geopolitical stability.

Key Points

• Logistical hegemony: Hegemonic orders are best understood as historically specific logistical regimes—configurations of foundational resources, logistical domains, and enforcement mechanisms that organize global circulation.
• Logistical Hegemony Triangle: Three interdependent dimensions
  • Foundational power resource (e.g., agrarian surplus, industrial capital, financial‑technological innovation, semiconductor/digital infrastructure)
  • Primary logistical domain (roads, sea lanes, financial networks, global value chains/digital infra)
  • Core enforcement mechanism (garrisoning/legal integration, naval blockade, institutional access/financial sanctions, export controls/technological denial)
• Typology of logistical power: integrative, circulatory, chokepoint, and denial logistics. Contemporary U.S. techno‑security strategy mixes chokepoint and denial logistics.
• Comparative historical analysis: Pax Romana (territorial/integrative), Pax Britannica (maritime/circulatory), Pax Americana (institutional/chokepoint), and the emerging Pax Silica (digital/chokepoint + denial).
• Weaponized interdependence and chokepoints: States occupying central nodes in networks can coerce others; semiconductor production (TSMC, ASML, foundry-capable fabs) is a concentrated chokepoint amenable to such leverage.
• Contemporary policy moves (CHIPS Act, export controls, friend‑shoring) reflect an exclusionary shift—denial logistics—aimed at constraining rivals’ access to advanced nodes and equipment.
• Fragilities identified: the Innovation Diffusion Paradox (denial can slow adversary capabilities but also reduces overall innovation diffusion), supply chain fragility, and potential for escalation and instability.

Data & Methods

• Methodological approach: comparative historical analysis (CHA) combined with infrastructural process tracing.
  • CHA justifies long-duration, path-dependent comparison across Pax systems.
  • Infrastructural process tracing examines mobility, integration, surveillance, coercion, and innovation mechanisms of logistical infrastructures.
• Case selection: Pax Romana, Pax Britannica, Pax Americana, and the emerging U.S. techno‑security architecture (Pax Silica hypothesis).
• Data sources: historical scholarship, International Political Economy literature (structural power, GVCs, infrastructure studies), policy documents (e.g., CHIPS Act), think‑tank and industry reports, and analyses of key firms (e.g., TSMC, ASML, SEMI).
• Scope and limits: Framework applies when a dominant power has disproportionate capabilities and when logistical infrastructures are central; logistics are treated as necessary but not solely sufficient to explain hegemonic outcomes.

Implications for AI Economics

• Compute bottlenecks map to geopolitical risk: Semiconductors (advanced logic nodes, GPUs/accelerators) are critical inputs to large AI models; chokepoints in fabrication and equipment supply increase strategic exposure for AI development and deployment.
• Higher and risk‑volatile training costs: Denial logistics, export controls, and friend‑shoring are likely to raise costs and unpredictability of advanced compute, increasing barriers to large‑scale model training and potentially concentrating AI capability among firms/nations that secure privileged access.
• Concentration of AI capability: Geographic and firm concentration in semiconductor production (and cloud datacenter capacity) amplifies incumbency advantages—few producers/cloud providers can control access and pricing, affecting market structure and competition.
• Innovation diffusion paradox: Policies that deliberately slow diffusion (exports/denial) may retard adversary AI capabilities but also impede global innovation spillovers, reducing the overall rate of technological progress and cross‑border knowledge diffusion relevant to AI.
• Strategic industrial policy matters: CHIPS‑style subsidies, targeted investment in fabs, and supply‑chain diversification reshape comparative advantage in AI hardware and thus national AI competitiveness.
• Standards, protocols, and platform governance: Logistical control extends beyond chips to standards, software toolchains, and cloud APIs—control over these can serve as non‑hardware chokepoints affecting model interoperability and diffusion.
• Redistributional and inequality effects: Bottlenecks and friend‑shoring may raise compute costs for developing countries and smaller firms, exacerbating global inequality in AI capabilities.
• Risk of escalation and instability: Weaponized interdependence can incentivize redundancy and decoupling, which fragment markets and raise inefficiencies—outcomes that have direct macroeconomic effects on AI investment and growth trajectories.

Suggested empirical research directions for AI economists: - Quantify how chip supply shocks (capacity constraints, export controls) translate into increases in training and inference costs and hence into slower model scaling. - Model market concentration effects: simulate how chokepoint control impacts rents, entry barriers, and welfare in the AI ecosystem. - Measure the diffusion effects of denial policies on global innovation (patents, publications, open‑source contributions) and firm‑level R&D investment. - Assess policy trade‑offs: evaluate welfare implications of friend‑shoring/subsidies versus open global supply chains for AI growth and security. - Explore heterogeneity across countries and firms: which actors can absorb higher compute costs, and how does that shape comparative AI advantage?

Limitations and caveats from the paper: - Logistics is framed as necessary but not sole determinant of hegemonic stability—political, military, institutional, and ideational factors remain important. - The Pax Silica thesis is presented as an emergent configuration with considerable fragility; outcomes depend on policy choices, technological diffusion, and counterstrategies by other states and firms.

If you’d like, I can (a) produce a one‑page policy brief for AI firms/policymakers summarizing recommended strategic responses, or (b) draft research designs for any of the empirical directions above.