When the Populist Becomes the Technocrat
Donald Trump built his political movement around a simple proposition: ordinary people had lost control of institutions that supposedly existed to serve them. Decisions affecting their jobs, communities and futures were increasingly being made by political bureaucracies, global corporations and credentialed elites who regarded popular resistance as ignorance. MAGA promised to reverse that relationship. The forgotten citizen would no longer be instructed to accept whatever arrangement the managerial class had decided was inevitable.
That makes Trump's extraordinary attack upon Americans resisting hyperscale data centres more significant than another presidential outburst. Communities objecting to data-centre developments are, according to Trump, choosing to become "backwards and poor." Those wanting prosperity should "let Data Reign," while resistance risks killing the technological "Golden Goose." Trump then reached for the increasingly familiar national-security trump card: China "could not be happier" about the American backlash against data centres.
Patrick Wood correctly identifies the technocratic structure of the argument. The significant feature is not merely Trump's enthusiasm for artificial intelligence. There are perfectly good reasons for the United States to develop AI and remain technologically competitive with China. The deeper problem arises when technological development ceases to be something citizens are entitled to evaluate and becomes an imperative to which citizens are expected to submit. The people object, the planners announce that technological progress requires the project, and opposition is redefined as ignorance, backwardness or assistance to America's enemies.
That is an old technocratic move. Technocracy does not necessarily require sinister men sitting around a table designing a dictatorship. At its philosophical core lies a much simpler proposition: complex industrial societies should increasingly be organised according to technical requirements identified by experts rather than political choices made by ordinary citizens. Once the optimal system has supposedly been identified, democratic disagreement becomes troublesome because the citizen is attempting to interfere with what the system requires.
The historical Technocracy movement made this particularly explicit. Wood draws attention to Howard Scott and Technocracy Inc., which during the 1930s imagined society reorganised around energy production and consumption rather than ordinary market and political mechanisms. Energy would cease to be merely another commodity and become the organising metric of an industrial civilisation. The engineers would understand the physical requirements of the system and allocate resources accordingly.
The irony is that hyperscale AI infrastructure gives this old idea an unexpectedly modern form. Data centres transform enormous quantities of electricity into computation. A hyperscale facility consuming hundreds of megawatts is therefore not merely another office building filled with computers. It is a gigantic claimant upon the physical energy system. Somebody must generate its electricity, transmission infrastructure must carry it, substations must distribute it, cooling systems may require substantial water resources, and the surrounding community must live with whatever physical consequences follow.
This is why the current dispute is fundamentally about resource allocation. Who has the superior claim upon electricity: the existing household or the new hyperscaler? Who should pay for the additional generation and transmission infrastructure? If supplying an AI complex requires enormous investment in the grid, should existing consumers finance any part of it through higher electricity charges? If water becomes scarce, what priority should be given to industrial cooling? These are political and economic questions before they are technological ones.
Trump's rhetoric attempts to answer them by treating technological expansion as presumptively decisive. The data centre represents progress; therefore the data centre should be built. The AI industry represents national technological leadership; therefore infrastructure should be provided. Anyone questioning the arrangement risks impoverishing the community and helping China. The political dispute has effectively been resolved before the citizen enters the room.
This is particularly strange because Trump's own administration has already acknowledged that one of the central public complaints is legitimate. The White House Ratepayer Protection Pledge explicitly declares that ordinary Americans "should not be footing the bill for the benefit of private companies." The administration has called upon hyperscalers to build, bring or buy the new generating capacity they require and pay the full cost of grid upgrades associated with their operations. Amazon, Google, Meta, Microsoft, OpenAI, Oracle and xAI were among the original signatories.
That is a sound principle, but it demolishes the caricature of opposition as mere Luddism. If electricity costs created by data centres were imaginary, there would be no reason for a presidential initiative designed to prevent those costs being transferred to households. The White House itself says that without safeguards the electricity demands of data centres can increase utility bills for ordinary Americans.
The pledge also reveals another important distinction. It is a voluntary political commitment rather than, by itself, an enforceable utility tariff applicable to every project. The actual allocation of costs depends upon state regulation, utility agreements and particular projects. A promise that Big Tech will pay its way is therefore welcome, but citizens remain entitled to ask whether that promise is actually being delivered where they live.
The scale of the infrastructure challenge makes such scrutiny essential. PJM, America's largest electricity grid operator, recently proposed arrangements under which data centres and other enormous users could be required to move onto backup generation during electricity emergencies rather than jeopardising ordinary consumers. PJM serves approximately 67 million people and has warned of a projected supply shortfall that could increase blackout risks. The issue is therefore no longer theoretical: the electricity requirements of computing infrastructure are beginning to collide with the limitations of existing physical systems.
Trump nevertheless tells reluctant communities that they should embrace the developments or risk becoming poor. This is where the populist begins sounding remarkably like the technocrat. The argument becomes that the system has requirements that transcend local preferences. America requires AI supremacy. AI requires data centres. Data centres require enormous quantities of electricity, land, water and infrastructure. Therefore, communities must accommodate them.
Notice how easily democratic choice disappears inside that chain of reasoning. Each step is presented as necessitated by the previous one. By the time the ordinary citizen is asked whether a 500-megawatt industrial complex should be built nearby, the substantive decision has already been made. Saying no means opposing technological progress, national prosperity and ultimately America's strategic competition with China.
Wood goes further and sees this as part of a broader convergence between technological and political power. One need not accept every element of his analysis to recognise the danger he identifies. The corporations building frontier AI systems possess extraordinary concentrations of capital, technical expertise and increasingly strategic infrastructure. Governments simultaneously regard their success as essential to national security. That produces a relationship in which private corporate expansion can acquire the aura of public necessity.
Once that occurs, an important boundary becomes blurred. A technology company naturally wants more electricity because more computing capacity may produce more revenue. Government wants more computing capacity because it believes AI leadership confers national power. Corporate interest and national strategy therefore point in the same direction. The danger is that the interests of the ordinary citizen become the residual variable expected to adjust itself around both.
That is not a free market in any meaningful classical sense. A genuinely market-oriented arrangement would begin by requiring the corporation seeking an enormous new quantity of a scarce resource to pay the full cost of obtaining it. If an AI company requires another power station, transmission lines, substations and water infrastructure to operate its facility, those costs belong in the economic calculation determining whether the project is worthwhile. Shifting them onto existing consumers disguises the real cost of the enterprise.
The principle should therefore be brutally simple: Big Tech can build the data centres, but Big Tech should pay for what they require. If a project remains profitable after paying the full marginal cost of electricity generation, grid expansion, water, roads and other infrastructure, there is at least a serious economic case for proceeding. If it only becomes profitable because some of those costs can be distributed across millions of ratepayers and taxpayers, the public is subsidising private technological accumulation.
The White House appears to recognise this in principle. Its pledge says hyperscalers should finance their own additional energy and infrastructure. Yet Trump's political rhetoric moves in another direction when communities continue to resist. Instead of treating local opposition as part of the mechanism ensuring that the promised protections actually work, he treats resistance itself as the problem.
This is the technocratic temptation in its purest form. Experts and corporations define the objective, government adopts the objective as national policy, infrastructure requirements follow from the objective, and democracy is permitted to operate only within the remaining space. People may debate details, perhaps, but they are not supposed to challenge the direction of travel.
The China argument makes the structure even more dangerous because geopolitical competition can justify almost anything once accepted as the overriding consideration. China is developing AI; therefore America must develop better AI. Better AI requires more computation; more computation requires more data centres; more data centres require more electricity; therefore resistance to any part of this expansion assists China. A complicated political economy has been reduced to a loyalty test.
The same argument could justify virtually unlimited subsidies. If an AI company demanded publicly funded nuclear reactors, one could say China will otherwise win. If it demanded priority access to water, China could again be invoked. If local planning laws became inconvenient, they could be overridden because national security requires rapid construction. At every stage a legitimate strategic concern becomes a rhetorical solvent dissolving ordinary restraints upon concentrated power.
There is an especially uncomfortable irony for MAGA. The movement was partly a revolt against precisely this style of politics. American manufacturing was relocated overseas and working-class communities were told that globalisation was economically inevitable. Immigration expanded and people questioning its scale were told that experts understood the benefits. Environmental and energy policies were imposed with ordinary people expected to absorb higher costs because a larger technocratic objective supposedly required them. In each case the complaint was not merely about the policy but about the arrogance of elites announcing that alternatives were irrational.
Now some of the same people who rebelled against that worldview are being told that another technological transformation is inevitable and that resisting its local consequences demonstrates backwardness. The managerial vocabulary has changed. Instead of globalisation, the imperative is AI. Instead of international competitiveness generally, the enemy is specifically China. But the underlying message sounds familiar: experts know where history is going, corporations possess the technology required to get there, and ordinary people should not stand in the way.
There is nothing inherently conservative, populist or libertarian about accepting that proposition. Indeed, local control over land use, protection of property, opposition to corporate subsidies and suspicion of concentrated political-economic power would once have been recognisable conservative concerns. The fact that the corporations involved are producing artificial intelligence rather than automobiles does not abolish them.
The real choice is not AI or no AI. It is between different political economies of AI. One model allows technology companies to expand while internalising the costs they create and negotiating with the communities affected by them. The other identifies rapid technological expansion as a national imperative and progressively reorganises energy, infrastructure and government policy around satisfying its requirements.
The latter deserves the name technocratic because the technology ceases to serve the political community and the political community begins to be reorganised to serve the technology. Electricity systems are redesigned around hyperscale computing. Energy policy is reshaped around AI requirements. Permitting is accelerated because ordinary planning procedures are considered too slow. Governments align themselves with corporations possessing strategic computing capacity. Citizens objecting to the consequences are characterised as obstacles to national progress.
Wood's most interesting observation is therefore not that Trump likes data centres. It is the inversion of Trump's political identity. The man who promised to restore power to forgotten communities now confronts communities exercising local political power against developments they do not want and tells them that they are backward.
That contradiction should trouble Trump supporters more than Trump critics because it raises the question of what populism actually means. If populism means merely supporting whichever policies the populist leader announces, it has no serious political content. If it means restoring meaningful political agency to citizens against concentrations of institutional power, then those citizens must remain entitled to say no even when Trump wants them to say yes.
Data centres are therefore becoming a revealing political test. They place some of the largest corporations in history on one side, backed by enormous capital and a federal government convinced that their expansion is strategically essential. On the other side are households, towns and counties asking questions about electricity bills, water, land, noise and infrastructure. The asymmetry of power could hardly be clearer.
A genuine populist response would not automatically reject the data centre. It would force the developer to make its case. How much electricity will you consume? Where will it come from? Who pays for the generating capacity? What happens during shortages? How much water will you use? How many permanent jobs will you actually create? What taxes will you pay? What infrastructure must the public provide? What happens to nearby residents, and what compensation will they receive for genuine costs imposed upon them?
If the answers are satisfactory, build it. If they are not, the community should retain the right to refuse. That is neither technological primitivism nor assistance to China. It is self-government.
The larger issue is where political sovereignty resides in the emerging AI economy. Does it remain with citizens operating through markets, local governments, states and national democratic institutions, or does it gradually migrate toward an alliance of government planners and technology corporations controlling the computational and energy infrastructure upon which the rest of society increasingly depends?
That is where the data-centre controversy intersects with the older critique of Technocracy. The crucial question is not whether engineers are competent or whether artificial intelligence is useful. It is whether technical necessity becomes a political doctrine. Once governments begin saying that certain technological objectives are so important that opposition itself is irrational, democratic government has begun yielding ground to administration by supposedly unavoidable technocratic imperatives.
https://patrickwood.substack.com/p/trump-blasts-luddites-for-opposing
