Samo Burja on Growth, Energy, and AI

What It Covers

Samo Burja argues with Theo Jaffe that AI's physical infrastructure demands — spanning steel, energy, chips, and construction — are triggering a new industrial revolution, while demographic decline, institutional decay, and fiscal policy determine which nations and organizations capture that growth.

Key Questions Answered

• •AI Demand as Industrial Catalyst: AI infrastructure requirements are large enough to force economies of scale across physical industries — steel, cement, natural gas, mirrors, and construction machinery. Investors and policymakers should track upstream industrial suppliers, not just software companies, as the demand shock propagates through supply chains over the next several years.
• •Industrial vs. Knowledge Economy Divergence: Countries with industrial economies tied to AI hardware — Taiwan, South Korea, The Netherlands via ASML — could see 10% year-on-year GDP growth in the near term. Knowledge-economy nations risk mass white-collar disruption as frontier AI firms scale elite services to millions, creating winner-take-all professional markets globally.
• •Government AI Investment Thesis: Rather than nationalizing AI companies, Burja proposes the US government purchasing equity stakes in frontier labs like OpenAI or Anthropic at market value. He argues this would function as productive industrial policy, directing printed capital toward data centers and compute rather than consumption spending with no long-term economic multiplier.
• •Functional Institutions as AI Winners: Organizations with coherent, functional processes will extract disproportionate value from AI, while broken bureaucracies will simply amplify existing bottlenecks. The actionable principle: audit and repair internal workflows before deploying AI at scale, since AI accelerates whatever process it enters — effective or dysfunctional — rather than correcting structural failures.
• •Fertility and Industrial Growth Correlation: Taiwan holds the world's lowest fertility rate at approximately 0.65, yet TSMC employees show notably higher fertility than the national average. Burja suggests this pattern — where workers at high-growth industrial anchors reproduce at higher rates — could signal long-term demographic stabilization if replicated in other technology-driven industrial clusters.