Tracking the AI Economy: Seven Charts Every Presidential Candidate Should Study

TEXT START: The next president will need a day one strategy for addressing the economic opportunity and political risk of a raging artificial intelligence debate.

The Dissection

This is a transition-management memo disguised as an empirical governing agenda. It correctly identifies where AI capital is concentrating: data centers, energy, skilled construction, supply chains, public capacity, and regional inequality. It also correctly notices that infrastructure can produce local winners, losers, political backlash, and temporary employment.

Then it performs the fatal substitution: it treats the construction and governance of AI infrastructure as a pathway to broad-based employment rather than as the physical plant for eliminating the need for broad-based employment.

The article’s governing fantasy is that Washington can convert AI’s ownership structure into an inclusive growth machine through better siting, redistribution, workforce investment, and administrative competence. That may redirect rents and soften local damage. It does not restore the mass employment-to-wage-to-consumption circuit.

The Core Fallacy

The text confuses capital formation with productive participation.

Data centers, chip facilities, energy projects, and construction booms can create jobs around AI deployment. But those jobs are ancillary to the machine economy, not evidence that humans remain necessary across the economy. Construction ends. Operations are highly automated and thinly staffed. Supply chains consolidate around owners of scarce infrastructure, compute, energy, and models.

The article also treats redistribution as if it were labor-market inclusion. Returning some AI-generated gains to communities can preserve purchasing power. It cannot preserve the social status, bargaining power, or economic necessity that came from having a job. Under the Discontinuity Thesis, transfers may keep the consumption circuit alive as an administered carcass, but they do not resurrect productive participation.

Its proposed solution assumes that policy can engineer durable human employment niches at scale. P1 makes that assumption unstable; P2 makes it politically unenforceable; P3 is the result. Once AI is cheaper and better across cognitive work, no administration can permanently command capital to hire humans merely to preserve the old social contract.

Hidden Assumptions

• That AI investment will create enough durable jobs to offset the work it destroys, rather than concentrating output while reducing labor demand.
• That short-term construction employment can be converted into permanent regional prosperity.
• That data-center “AI Hubs” will generate broad ecosystems rather than isolated capital enclaves surrounded by expensive infrastructure and weak local employment.
• That government can redistribute AI gains without confronting who owns the compute, energy, models, land, and financial claims producing those gains.
• That improved government expertise changes the competitive economics of automation. It does not.
• That workforce retraining remains meaningful when the displacement target is increasingly cognitive labor itself, including the administrative and professional strata that normally absorb displaced workers.
• That “jobs” remain the correct primary unit of economic inclusion after labor ceases to be the dominant production input.
• That political consent can be secured through local benefits indefinitely, even as residents experience resource strain, housing pressure, environmental costs, and declining labor leverage.
• That industrial policy can repeat the geographic employment effects of highways, railroads, NASA, or manufacturing. AI infrastructure is more capital-intensive and less labor-absorbing than those precedents.
• That private investment can be made socially inclusive without changing the underlying ownership hierarchy. The article gestures at redistribution while leaving sovereignty over AI capital intact.

Social Function

Primarily: transition management and ideological anesthetic, with a substantial layer of elite self-exoneration.

It is a partial truth because it accurately identifies infrastructure concentration, regional exposure, public-sector capacity gaps, and the political consequences of local buildouts. But it converts those truths into a reassuring policy narrative: if the next president deploys capital intelligently, staffs government better, and designs inclusive hubs, the AI revolution can “work for workers.”

That phrase is the sedative. It allows policymakers to discuss the distribution of AI rents without confronting the terminal question of ownership and necessity. The article asks how to make the new machine economy create opportunity for displaced people. It does not ask whether the new economy requires those people at all.

The Verdict

This is a competent map of the AI buildout and a failed theory of its social endpoint. It sees the factories of the new order but mistakes their construction crews, regulators, and local beneficiaries for a replacement mass employment system.

The proposed agenda may manage the transition, redistribute some gains, and delay political rupture. It cannot reverse the Discontinuity Thesis. The article is not describing a path to an economy that works for workers; it is drafting the administrative paperwork for a system in which workers become claimants on output produced by sovereign AI capital.