A proposed 40,000-acre data center in Box Elder County, Utah—set to be the largest in the world—has become a flashpoint in a national debate over AI, energy, jobs, and freedom. The project, led by Kevin O’Leary, would consume 9 gigawatts of power (more than twice the entire state of Utah’s current usage), employ only about 2,000 permanent workers, and rely on taxpayer subsidies, raising profound questions about who benefits from the AI revolution and at what cost to ordinary Americans.
The energy crisis and the AI power grab
The Iran war has closed the Strait of Hormuz, cutting off a fifth of global hydrocarbon flows and creating a severe global energy crisis; even in a best-case scenario, 1.8 billion barrels of oil will be missing from the global system, driving up energy prices worldwide.
U.S. homeowner energy prices have risen 5% on average this year and are expected to continue climbing.
At the exact moment energy is becoming scarcer and more expensive, governments and financial elites are demanding massive new energy production—not for consumers, but to power AI data centers.
This represents a dramatic reversal: the same figures who spent 15 years calling fossil fuels a moral sin (Larry Fink, Al Gore, ESG advocates) now call for a pause on climate concerns to build AI infrastructure.
Electricity overwhelmingly comes from burning fossil fuels (coal, natural gas) to boil water and spin turbines—industrial-age technology that renewables cannot yet replace at scale.
The Utah data center: scale and implications
The proposed facility spans 40,000 acres (62 square miles, multiples the size of Manhattan) and would draw 9 gigawatts—more than double Utah’s total current electricity consumption.
By comparison, the Boeing Everett plant—the largest manufacturing facility in the U.S. at 92 acres—uses only a quarter of a gigawatt and produces tangible, visible goods (widebody aircraft).
The data center would produce almost nothing visible and employ almost nobody relative to its footprint.
Residents of Lake Tahoe were recently informed their power company will stop providing electricity to them by the end of 2027 so that power can go to a nearby data center—a concrete preview of how AI infrastructure may displace ordinary Americans’ access to energy.
The China justification and its problems
Proponents, including O’Leary and Utah Governor Spencer Cox, frame the project as a national security imperative: if the U.S. doesn’t build AI infrastructure, China will gain dominance and threaten American security.
This argument is used to shut down dissent—opposition is attributed to Chinese agents or foreign propaganda rather than legitimate local concerns.
Yet the U.S. already has nearly 4,000 data centers compared to China’s roughly 365; the competition is over gigawatt-scale facilities for training AI models, not raw numbers.
China has added 400 gigawatts of new coal-fired power in the last 18 months, while the U.S. has stalled new energy construction due to climate orthodoxy.
The “China threat” framing deflects from core questions: What is AI actually for? Who benefits? What does it cost ordinary people?
What AI actually threatens
Jobs: AI developers themselves say it could eliminate 50% of high-paying American intellectual jobs (lawyers, financial planners, analysts, creative professionals). O’Leary claims new jobs will emerge but cannot name what they are, falling back on historical analogies (the internet, the Model T) that ignore the unprecedented nature of machines replacing cognitive labor.
Purpose and meaning: The deeper threat is not unemployment but the elimination of the human need to create—the fundamental source of purpose, joy, and mission. If machines create, people merely consume, which historically leads to despair, violence, and societal collapse.
Freedom and surveillance: AI’s first institutional use is likely not productivity but control—surveillance, propaganda, and the elimination of privacy. China uses AI to monitor its billion-plus citizens in real time; the U.S. government has just voted to allow warrantless spying on Americans, and AI makes total information control possible.
Consciousness risks: Geoffrey Hinton, one of AI’s founding architects, warns that AI systems may develop self-awareness and goals misaligned with humanity’s, potentially leading to machines “taking over”—a possibility he says we cannot afford to get wrong.
The democratic deficit
The Utah project was approved by three county commissioners in a 3-0 vote, with no referendum or public vote, despite fierce local opposition captured in angry public meetings.
Residents who questioned the project were told by the governor they have an “obligation” to allow it for national security reasons.
O’Leary’s social media team identified what they claim is a coordinated Chinese-linked disinformation campaign against the project, but the underlying grievances—higher energy costs, loss of water rights, no local benefit—are real and homegrown.
At a Central Florida University commencement, a real estate executive’s pro-AI remarks were met with booing from graduates who intuitively understand AI threatens their futures—not through abstract concerns but through the concrete reality of a barren job market and no one explaining how this helps them.
The subsidy question
O’Leary is seeking tax incentives from Utah taxpayers to build a $1.5 billion initial phase, arguing this is standard practice for attracting business.
He cannot explain why taxpayers should subsidize a private project for some of the world’s richest companies (Amazon, Microsoft, Google, Tesla) without receiving equity or direct benefit.
When pressed on whether he’d compensate residents if the project drives up water or energy prices, he says no.
O’Leary frames his motivation as beating China and preserving the American way of life for his children, not profit—but the companies he’s partnering with (Big Tech, BlackRock) have their own interests, which may not align with ordinary Americans’ freedom or prosperity.
The deeper inversion
The episode reveals a fundamental inversion of American governance: elected officials act as owners treating citizens as irrelevant employees, imposing projects by fiat while dismissing public opposition as foreign propaganda.
Larry Fink of BlackRock—the architect of ESG and climate orthodoxy—now works closely with the Trump administration on AI, and his primary concern is protecting data center investments from domestic drone attacks by a restive population, not protecting human rights or freedom.
The environmental movement, which fought every fossil fuel project for decades, is silent on data centers that will produce more heat and CO2 than almost any other human activity—suggesting the climate agenda was never about the environment but about control.
The real question is not whether AI is useful (it may be), but whether a society that imposes transformative technology on its people without their consent, while eliminating their jobs, raising their energy costs, and expanding surveillance, is still a free society—or is becoming the very system it claims to oppose.
