Samo Burja on Growth, Energy, and AI

The conversation opens with a debate about whether American eschatological thinking around AI is rooted in Christianity. Burja pushes back, arguing that material conditions matter more than deep civilizational narratives, pointing to China's own history of apocalyptic cults like the Yellow Turban Rebellion as evidence that religious frameworks shift dramatically over centuries. He characterizes modern American nominal Christianity as a vague 'be a good person' ethic blended with karma beliefs, while arguing that Puritan work ethic persisted not because of theology but because capitalism materially rewarded hard work and reinforced it over generations.

Burja then pivots to fiscal policy, arguing that since the U.S. government will continue deficit spending regardless, directing that capital toward AI companies via equity stakes would be better stimulus than infrastructure bills or direct payments. He frames this partly as a cynical but pragmatic observation — money spent on AI companies will likely flow into data centers and compute, which is productive investment. He also interprets Sam Altman's nationalization proposal as a strategic PR move to position OpenAI as the 'least disfavored' AI company against Anthropic.

On labor automation, Burja argues capitalism can survive full automation as long as competing AI systems exist, but warns that the political economy will be transformed — drawing parallels to agricultural automation driving the New Deal era realignment. He suggests voters may lose political leverage as their value as conscripts, taxpayers, and rioters diminishes under full automation, referencing the concept of the 'intelligence curse.'

The macro economic argument Burja develops most extensively is that AI's physical infrastructure demands — energy, chips, data centers — are so large they require industrial revolutions in steel, cement, mirrors, and natural gas. He sees this as a 'demand shock from the future' rippling through supply chains. Countries with industrial economies (Taiwan, South Korea, Netherlands via ASML, Germany) stand to gain significantly, potentially seeing 3-10% GDP growth boosts, even as their knowledge-economy white-collar sectors face disruption from AI.

On demographics, Burja acknowledges aging populations and collapsing fertility rates (Taiwan at 0.65) as a serious headwind to sustained industrial growth. He notes the curious anomaly that TSMC employees have higher fertility than Taiwan's national average, suggesting concentrated high-growth companies may bucked demographic trends. He concludes by arguing that functional institutions — organizations with sound processes — will be the winners of the AI era, because fitting AI into broken bureaucratic processes merely amplifies existing bottlenecks rather than creating value.