As America celebrates 250 years of freedom, the question is whether we can win the global race against China in artificial intelligence, the newest technology. The country that generates the most reliable electricity fastest will train the most powerful models and set technological terms for the rest of the century.
China is building dozens of new coal-fired power plants a year and has no permitting or environmental constraints. The question is whether America can keep pace.
Recommended Stories
In addition to an energy problem, America has a shortage of technical workers. June’s labor force numbers, released Thursday, suggest that the problem is getting harder, not easier, at a time when AI and data center demand is rising.
President Donald Trump’s energy dominance agenda created the conditions for the energy race to be winnable. America holds 46 billion barrels of oil and nearly 584 trillion cubic feet of natural gas, and the president is increasing permits for development. He is keeping coal-fired power plants open to alleviate shortages.
Trump has called for faster nuclear power plant construction, modernized safety regulations, new domestic fuel sources so America is not dependent on foreign uranium enrichment, and a resolved long-term waste disposal site.
This energy agenda is vital for AI. The combined capital expenditure of five large American technology companies exceeded $400 billion last year and is projected to rise by another 75% in 2026, according to the International Energy Agency. Every dollar of technology investment depends on reliable power, delivered on schedule.
But permitting is still a problem. Every new gas plant, transmission line, and reactor requires a sometimes decade-long fight in court and at the local zoning board.
China, unfettered by Western environmental regulations, has built industrial capacity to dominate solar panels, batteries, and rare earth processing, and is now applying the same strategic patience to grid buildout and nuclear deployment. China connected more nuclear reactor capacity to its grid last year than the United States has added in three decades.
The Chinese Communist Party is racing to ensure that when the AI era matures, the infrastructure underneath it runs on Chinese terms.
America has the resources to win this race, yet we are squandering the time it has to do so. Despite natural resources, we lack the regulatory speed to turn them into electrons on the grid before the AI buildout stalls or moves elsewhere.
Regional grid operators from PJM to ERCOT are sounding alarms about capacity shortfalls. PJM’s own capacity auction prices rose more than 800% in a single year, a market signal that America is not building fast enough.
The fixes are not mysterious. Permitting reform should be accelerated so new generation and transmission projects move from proposal to operation in years rather than decades. Every month a power plant waits in an interconnection queue to come online is a month China does not have to wait.
America also increasingly lacks the labor force, as the disappointing June jobs numbers from the Labor Department showed. At a time when the electricity and technology centers are desperate for workers, June’s labor force shrank by 720,000. The share of Americans who are employed or looking for work declined to 61.5%, the lowest since the pandemic and almost 6 percentage points lower than its 2000 peak.
America faces an energy workforce shortage, with 81,000 electrician openings annually, a 30% near-retirement rate among union electricians, and a 320,000-welder shortfall by 2029. The International Energy Agency reports that most surveyed energy employers report critical hiring bottlenecks.
Meta’s new America’s Workforce Academy, a new free five-week training program for positions in AI infrastructure, power generation, grid modernization, and data center construction, offers one solution that could be replicated. Launched in June, trainees are promised jobs before they start.
I’M A PROUD DEMOCRAT, BUT AMERICA’S ENERGY PARALYSIS ISN’T TRUMP’S FAULT — STOP FIGHTING REFORM
China has low-wage labor to build excess capacity with no regard for environmental or human cost, also propped up by state subsidies. America’s advantage has always been that its innovation is powered by private capital operating under the rule of law, not command-economy fiat. But that advantage disappears if private capital cannot get a permit or if people leave the labor force.
The next decade of the AI race will be decided by which country is allowed the labor and energy to run the best algorithms. America still has the resources, the capital, and the technical lead to win. But we have no time to waste.
Diana Furchtgott-Roth, former deputy assistant secretary for research and technology at the Transportation Department, is a distinguished fellow at the Energy Policy Research Foundation and an adjunct professor at George Washington University.
