Capital, electricity, and talent cannot compete! American manufacturing is losing to AI

The revival of American manufacturing is facing an unexpected competitor— the surge of artificial intelligence data centers is fully squeezing traditional manufacturing in three key areas: capital, electricity, and labor, threatening the core policy goal of the Trump administration to revitalize American industry.

On October 24, Bloomberg reported that while the Trump administration held high the banner of "manufacturing resurgence," a more intense resource competition was unfolding behind the scenes. Tech giants are expected to invest up to $4 trillion in AI infrastructure by 2030, a wave of investment comparable to the railroad construction of the 19th century, is crazily siphoning capital, electricity, and labor from traditional factories to AI data centers.

So far this year, data center construction spending has surged by 18%, while new factory construction has shrunk by 2.5%; in areas dense with data centers, wholesale electricity prices have skyrocketed by 267% in five years; among the 439,000 construction worker shortages nationwide, one-fifth of contractors are busy with data center projects; ironically, tech companies easily obtain tariff exemptions, while manufacturers face the heaviest tax burden since the 1990s for the equipment needed to expand their factories.

The story of Lordstown, Ohio reflects the absurdity of this transformation. Trump promised in 2017 to "bring jobs back" here, but the General Motors plant still closed. Now SoftBank has taken over this facility, not to manufacture cars, but to produce data center equipment— this former automotive manufacturing base that once employed 12,000 workers is now being transformed by Foxconn, OpenAI, and SoftBank into an AI equipment manufacturing center and demonstration data center, expected to employ only about 1,600 people.

Research firm Pantheon Macroeconomics estimates that without AI-related infrastructure spending and the wealth effect brought by the soaring AI concept stocks, the growth rate of the U.S. GDP in the first half of 2025 would be only 1%, rather than the actual 1.6%. Bloomberg Economics predicts that as tech giants like Google, Meta, and Microsoft increase their AI capital spending from nearly $400 billion this year to $600 billion next year, AI's contribution to GDP growth could rise to 1.5 percentage points next year.

MIT researcher Paul Kedrosky warns that while manufacturing struggles with tariffs, electricity prices, and labor shortages, AI enjoys policy green lights, capital pursuit, and resource tilt. The frenzy for AI technology has "broken people's mental models of how the economy operates," causing decision-makers to overlook the negative impacts of policies.

Imbalance in Capital Flow: Surge in Data Center Investment, Decline in Factory Construction

The AI investment boom is causing a severe tilt in capital allocation. According to Bloomberg Economics, major players are expected to invest up to $4 trillion in AI infrastructure by 2030, a scale comparable to the railroad construction of the 1870s and the fiber optic network boom of the late 1990s.

This concentration of funds is reflected in the actual data. So far this year, data center construction spending has increased by nearly 18%, while new factory construction spending has decreased by 2.5%. Data from the Institute for Supply Management shows that factory activity has contracted for seven consecutive months through September Morten Wierod, CEO of Swiss manufacturer ABB, pointed out that due to tariffs raising costs and a shortage of construction workers, the return on investment for data center projects is currently much higher for builders and suppliers than for other projects.

ABB agreed in October to sell its industrial robotics business to SoftBank Group for over $5 billion to deepen its focus on the more profitable data center business.

John Engel, CEO of electrical equipment distributor Wesco International, stated:

"AI is consuming most of the oxygen in the room, which is where the main growth is right now. If you are not involved in some way, unfortunately, you are missing out on opportunities."

Power Struggle: Data Center Demand Drives Up Industrial Electricity Costs

Electricity supply is becoming a real limiting factor for the U.S. to support both the AI revolution and the revival of manufacturing simultaneously.

According to data from the International Energy Agency, a typical AI data center consumes as much electricity as 100,000 households, and the largest data centers currently under construction will consume 20 times that amount. Bloomberg Industry Research estimates that by 2032, data centers could account for 20% of U.S. electricity demand.

Bloomberg News analysis found that wholesale electricity costs in areas near data centers have risen by as much as 267% over the past five years. This surge in electricity demand is driving up utility bills for residential and industrial customers.

The Trump administration's determination to eliminate all federal support for renewable energy has further exacerbated the tight electricity supply.

To meet the enormous energy demands of AI technologies, utility companies are investing in gas turbines. The surge in data center construction is also driving a spike in demand for various electrical equipment—Eaton Corp. reported a 55% year-over-year increase in orders from data center customers in the second quarter.

Labor Shortage Intensifies: Construction Workers Siphoned by AI Projects

The U.S. was already facing a long-term shortage of skilled workers, and Trump's crackdown on illegal immigration has further worsened this gap. According to data from the American Contractors Association (ABC), the U.S. is short 439,000 construction workers this year.

The construction of AI infrastructure is exacerbating this tension. Among ABC's 23,000 member companies, one-fifth signed contracts for data center projects in September.

John Fish, CEO of Suffolk Construction, stated that the company is building over 30 data centers in eight states and is "struggling" nationwide to find enough plumbers, HVAC contractors, mechanics, and electricians. Fish warned:

"If we do not take action to address this issue, the U.S. workforce will face catastrophic problems. The economy will only need more and more of these skilled workers, and we are heading in the opposite direction."

Policy Contradictions: Tariff Exemptions Favor Some Over Others

The Trump administration's policy treatment of AI and traditional manufacturing is sharply contrasting. Tech giants have successfully secured broad tariff exemptions for imported servers and other data center hardware, but the government has been unresponsive to manufacturers' requests for exemptions on import tariffs for equipment needed to expand or build new factories in the U.S In September, the Ministry of Commerce launched an investigation into the import of robots and industrial machinery, which could lead to more tariffs and further obstacles for production repatriation.

According to estimates, Trump's import tariffs represent the largest tax increase on American businesses since the early 1990s. The tariffs will cost Caterpillar up to $1.8 billion this year, while General Motors faces a total bill of $4.5 billion.

Barclays analyst Julian Mitchell stated in an August report that the industrial recovery "looks questionable," and referred to AI as "the only game in town."

The National Association of Manufacturers publicly supported a "big and beautiful bill" passed by Congress this summer, which includes tax incentives for improving equipment procurement and R&D investments.

However, a survey conducted by the association revealed that nearly 80% of businesses listed the burden of tariffs as their biggest concern. Based on these responses, the association predicts that capital spending will grow only 1% in the coming year (not adjusted for inflation), remaining flat compared to recent years.

The Symbolic Significance of Lordstown: From Auto Manufacturing to AI Manufacturing

The transformation of the former General Motors plant in Lordstown has become a microcosm of this economic transition.

In July 2017, Trump told the crowd at a rally in nearby Youngstown, "Don't sell your house," because "we're bringing those jobs back, filling those factories, or tearing them down to build new ones." General Motors closed the plant less than two years later, and a startup moved in with plans to manufacture electric vehicles, but went bankrupt by 2023.

In August, Foxconn announced it would sell the 6.2 million square foot facility for $375 million to a mysterious buyer, later confirmed to be SoftBank. The two companies plan to start manufacturing data center equipment at the same site that once produced Chevrolet Impalas and other models. SoftBank will also collaborate with OpenAI to test "advanced data center designs" at the former auto plant site.

The project is set to begin operations in 2026, employing about 1,600 people, including 400 who already work there. For a region that has lost 40% of its manufacturing jobs over the past two decades, these job commitments are significant, but far from filling the gap.

Some local leaders are cautious. Foxconn had previously promised to build a $10 billion LCD factory complex in Wisconsin during Trump's first term, hailed by Trump as the "eighth wonder of the world," but it never materialized. Former Lordstown Mayor Arno Hill stated:

"They say this is the wave of the future, but will this market become saturated? People will always need cars."

Paul Kedrosky, a researcher at MIT's Digital Economy Initiative, is one of the first scholars to place the AI capital expenditure boom in historical context.

He worries that the frenzy over this technology has blinded Trump and his advisors to the negative impacts of policy on other parts of the economy. "This thing is so big that it has broken people's mental models of how the economy works, and as a result, they are making mistakes," he said