The Real Economy Always Decides the AI Endgame

During my trip to San Francisco for one of the world's largest tech conferences, I brought along two extraordinary books to read. Although they are quite different, both share a strong focus on innovation and its impact on the economy.

Carlota Perez and Peter Thiel are rarely read together, yet they converge on a powerful idea: technological revolutions do not reward intelligence alone—they reward those who control how intelligence becomes productive.

San Francisco, Fisherman’s Wharf - Perez and Thiel on the windowsill, the Golden Gate in the distance. This city is exceptionally good at imagining the future, while also revealing the tension between financial narratives and real economic power.

Perez and Thiel on the windowsill, the Golden Gate in the distance.

Both writers arrive at the same conclusion from different directions: wealth is not created in markets, narratives, or abstractions. It is created in the real economy—where energy is transformed, goods are produced, health is maintained, security is enforced, and governments function.

Perez shows this repeatedly across technological revolutions. Financial capital races ahead of productive deployment, inflating bubbles and rewarding speculation. Only later—often after crisis—does technology become embedded into production, raising productivity and reshaping institutions. That is when durable wealth is created.

Thiel, from the firm-level perspective, makes the same argument more bluntly. Real progress comes from building something that fundamentally improves how the world works and defending it through proprietary technology and scale. Companies do not win by participating in markets alone; they win by reshaping production itself.

This shared insight is critical for understanding the AI endgame.

The current AI cycle is dominated by markets rather than production. Valuations, benchmarks, and model capabilities move faster than measurable economic output. Intelligence is treated as a standalone asset, as if cognition alone could compound into wealth. History suggests otherwise.

AI does not create value by being intelligent. It creates value when it is attached to a production function.

The decisive question is NOT how capable models become, but if they are deployed deeply enough to alter cost structures, productivity curves, and institutional capacity. Energy systems optimized by AI reduce marginal costs across the entire economy. Manufacturing systems augmented by AI compress design-to-production cycles. Healthcare systems that embed AI into diagnostics, operations, and prevention change population-level outcomes. Defense and state capacity strengthened by AI redefine power itself.

These domains are capital-intensive, regulated, slow-moving—and decisive. Perez would recognize them immediately as the terrain where technological revolutions either mature or fail.

Thiel’s emphasis on proprietary technology and economies of scale sharpens the point further. In AI, advantage accrues not to those who merely access intelligence, but to those who control how it is operationalized at scale. Compute infrastructure, specialized data, deployment rights, and institutional integration matter more than raw model performance. Open access may spread capability, but structural value concentrates where AI becomes indispensable to production.

Conclusion

The AI endgame will therefore not be decided by who builds the smartest systems first, but by who succeeds in rewiring the economic core: energy, manufacturing, healthcare, defense, and the administrative machinery of the state. That is where productivity compounds, where monopolistic advantage becomes defensible, and where financial capital finally finds something solid to stand on.

Perez teaches us that the success of a technological revolution is measured by how much it improves overall productivity, not just by the new inventions it creates. Thiel emphasizes that true wealth comes to those who can transform technological knowledge into a lasting competitive edge.

Seen through both lenses, the message is clear: markets may lead the conversation, but the real economy always decides the outcome. AI will be no exception.