Washington has a new favorite catchphrase, and it sounds remarkably like economic salvation. When Jacob Helberg, Under Secretary of State, took the stage at the Pax Silica Summit to declare that artificial intelligence will be one of the defining engines of prosperity for decades to come, he wasn’t just delivering a tech forecast. He was laying out a battle doctrine. The official narrative promises a future where Silicon Valley innovation cures economic stagnation, secures global supply chains, and guarantees democratic dominance.
But look past the stage lights of the summit, and a far more volatile reality emerges. If you enjoyed this post, you should read: this related article.
The promise of an AI-driven economic boom is being used to mask a aggressive, state-backed consolidation of the technology sector. By framing computing power as a matter of national survival, the U.S. government is quietly merging the interests of big tech with the machinery of national defense. This isn't just about wealth creation. It is about control. The underlying strategy relies on restricting access to advanced semiconductors, subsidizing domestic mega-foundries, and picking winners among a handful of corporate giants.
This Pax Silica—a peace enforced by silicon monopoly—comes with immense hidden costs that few in power are willing to discuss. For another angle on this story, see the recent coverage from Ars Technica.
The Mirage of Shared Abundance
The central premise of the Washington consensus is that AI deployment will automatically trigger a tide of productivity that lifts all industries. Historical precedents suggest otherwise. When steam power, electricity, and early computing arrived, they did not create instant, widespread wealth. They concentrated power. They disrupted labor markets for a generation before any broad benefits trickled down to the average worker.
The current trajectory of machine learning suggests an even narrower bottleneck. Building and maintaining modern frontier models requires three resources available only to the ultra-wealthy: massive capital, specialized talent, and unimaginable amounts of electricity.
Consider the infrastructure alone. A standard data center today consumes megawatts of power, but the next generation of training clusters will require gigawatts. This demands direct partnerships with energy grids and nuclear power providers. When a technology requires a sovereign-level energy footprint just to function, the idea that it will organically democratize wealth becomes a mathematical impossibility. The rewards will flow inward to the infrastructure owners, while the disruptions are exported outward.
Weaponizing the Supply Chain
To understand why the state is intervening so heavily, one must look at the physical choke points of the industry. The software might live in the cloud, but the cloud runs on hardware manufactured in a handful of highly vulnerable locations.
[Advanced Lithography (ASML - Netherlands)]
│
▼
[Precision Manufacturing (TSMC - Taiwan)]
│
▼
[Data Center Deployment (Global Consumables)]
The administration's strategy relies on keeping these nodes under strict lock and key. Through the CHIPS Act and aggressive export controls, the government has weaponized the global supply chain. The goal is simple: starve geopolitical rivals of the components needed to train competitive models.
This strategy works in the short term, but it creates structural instability. By cutting off foreign markets, American chip design firms lose massive revenue streams that previously funded their research and development. Meanwhile, targeted nations are not giving up; they are pouring hundreds of billions into developing domestic, unsanctioned supply chains. By forcing a hard fracture in the global tech ecosystem, the U.S. risks accelerating the exact competition it seeks to prevent.
The Small Business Squeeze
While federal subsidies pour into multi-billion-dollar factory builds in Arizona and Ohio, smaller tech innovators face a different reality. High borrowing costs and restricted access to top-tier hardware are creating a two-tiered system.
Domestic manufacturing facilities take years to build and calibrate. In the meantime, the scarcity of advanced graphics processing units (GPUs) means that only well-funded corporations can afford to experiment at scale. Startups are forced to build on top of APIs provided by the industry titans, effectively turning the next generation of tech builders into digital sharecroppers. They do not own the core technology; they merely rent it.
The Sovereign Computing Trap
European and Asian allies are watching this consolidation with growing anxiety. They recognize that relying entirely on American cloud infrastructure and chip designs is a long-term economic hazard.
As a result, the concept of sovereign AI has shifted from a theoretical policy preference to an urgent economic mandate. Countries are scrambling to build their own localized data centers, train models on their own cultural data, and reduce their reliance on Silicon Valley. This global balkanization means the massive, unified global market that tech companies once enjoyed is splintering into heavily regulated regional zones.
The Labor Inflation Myth
Summit speeches frequently emphasize that AI will elevate workers, turning ordinary employees into super-productive operators. The reality on the ground looks less like elevation and more like displacement.
Automation historically targets repetitive, manual tasks. The current wave of software tools, however, targets cognitive, creative, and administrative functions. White-collar roles that once formed the backbone of the middle class are being streamlined out of existence. While new jobs are being created—such as data labelers and prompt engineers—these roles rarely offer the stability, benefits, or wages of the positions they replace.
The productivity gains promised by executives will likely show up in corporate profit margins, not worker paychecks. When capital can substitute for human labor at scale, the bargaining power of the workforce plummets. Expecting corporate boards to voluntarily distribute these savings as widespread prosperity ignores the fundamental rules of modern market capitalism.
The Energy Collision Course
Perhaps the most significant blind spot in the Pax Silica vision is the physical limitation of the power grid. The push for computational dominance is colliding directly with environmental mandates and aging energy infrastructure.
Data centers are already straining local utilities from Virginia to Ireland. In some regions, tech companies are buying up retired coal and nuclear plants to ensure an uninterrupted power supply. This insatiable demand for electricity threatens to drive up energy costs for ordinary consumers and push carbon neutrality goals further out of reach. The irony is stark: the engine of future prosperity may rely on reviving the fossil fuel dependencies of the past.
The belief that market forces alone will solve this energy crunch is overly optimistic. Grid expansion requires decades of planning, regulatory approval, and physical construction. The speed of software development is moving at a pace that physical infrastructure simply cannot match.
Securing the Cloud at All Costs
As the economic stakes rise, data centers are turning into sovereign fortresses. Corporate espionage is no longer just a matter of stolen source code; it is a vector of national security.
The infrastructure hosting these models requires physical security, advanced cyber defense systems, and strict personnel vetting that resembles military installations. This shift transforms technology companies into state actors. When a private corporation holds the keys to software that drives national productivity and military logistics, the line between public governance and private enterprise dissolves completely.
This entanglement creates a dangerous precedent. If an algorithm failure or a cyberattack takes down a core model, it is no longer just a corporate outage. It becomes a national crisis.
Moving Past the Rhetoric
The narrative delivered at high-level summits is designed to build public consensus and justify massive state interventions. It paints a picture of a clean, friction-free ascent to global economic dominance.
The actual path forward is messy, exclusionary, and fraught with geopolitical peril. True prosperity cannot be built on a foundation of artificial monopolies, strained energy grids, and a hollowed-out middle class. If the deployment of these technologies is to benefit society at large, policy must shift away from subsidizing corporate giants and toward protecting labor resilience, upgrading public infrastructure, and enforcing genuine market competition.
The race for silicon supremacy is well underway, but the prize may not look like the utopia promised on the summit stages. Power, not prosperity, remains the ultimate goal.
