| Welcome to Bw Reads, our weekend newsletter featuring one great magazine story from Bloomberg Businessweek. Today Shawn Donnan, Brooke Sutherland, Mark Niquette and Alexandre Tanzi have a story about an economy in transition: Spending on factory construction is down 2.5% so far this year, while that for data centers is up almost 18%. That poses a challenge to a president promising a manufacturing revolution. You can find the whole story online here. If you like what you see, tell your friends! Sign up here. If you’re searching for a symbol of the wild ride the US economy has been on over the past decade and where it’s going next, the former General Motors Co. plant in Lordstown, Ohio, checks a lot of boxes. President Donald Trump held a rally in nearby Youngstown in July 2017 at which he told the crowd, “Don’t sell your house,” because “we’re going to get those jobs coming back, and we’re going to fill up those factories or rip them down and build brand-new ones.” GM shuttered the facility less than two years later, and then a startup moved in, planning to make electric cars, only for it to go bankrupt by 2023. Now some of the biggest names in tech—Foxconn, OpenAI and SoftBank—have hatched a plan to repurpose the 6.2 million-square-foot facility into a manufacturing hub for artificial intelligence equipment. A demonstration data center will also be built on the site. The AI boom is coming to America’s industrial heartland, which speaks to how the economy is being transformed—even propped up—by a surge in investment in software, chipmaking and data centers. But the Lordstown plant’s reinvention raises the question: Is the exuberance around the technology that Trump has enthusiastically endorsed getting in the way of his goal of making America great again at churning out things like gasoline-powered cars, steel and furniture? Access to capital, power and people is key to the success of any construction or manufacturing project—and right now AI is gobbling up all three. “AI is consuming a lot of the oxygen in the room,” says John Engel, chief executive officer of Wesco International Inc., a distributor of electrical gear that’s getting a growing share of its revenue from supplying data centers. “That’s where the major growth is now. And if you don’t have, in some form or fashion, a solution or play in that, it’s unfortunate. You’re missing out,” he says. The AI investment surge invites comparisons to other historical buildouts (and bubbles), such as the construction of the railways in the 1870s and the late 1990s flood of capital into fiber-optic and wireless networks. Altogether the largest participants are expected to pour as much as $4 trillion into AI infrastructure through 2030, according to Bloomberg Economics. The massive splurge is masking softness in other parts of the economy. Manufacturing shows no signs of breaking out of its extended funk; the payroll numbers are troubling (even though unemployment remains low overall); and consumer sentiment has plunged to Great Recession levels. Pantheon Macroeconomics, a research firm, estimates that, had it not been for spending on AI-related infrastructure and the boost to household wealth provided by soaring valuations for AI-linked stocks, US gross domestic product would have grown at a pace of 1% in the first half of 2025, rather than 1.6%. Bloomberg Economics expects AI’s contribution to GDP growth could rise to as much as 1.5 percentage points next year as the so-called hyperscalers—Alphabet, Amazon.com, Meta Platforms, Microsoft and others—dial up their investments from almost $400 billion this year to as much as $600 billion. “If it weren’t for AI, we’d probably be looking at significantly slower growth or even a modest recession,” says Rebecca Patterson, former chief investment strategist at hedge fund Bridgewater Associates who’s now a senior fellow at the Council on Foreign Relations. Trump has embraced AI for its potential to boost America’s economic competitiveness and strengthen national security. The White House released an AI Action Plan in July, and the president hosted a dinner in September attended by the CEOs of Alphabet, Meta, Microsoft and OpenAI. The president has also been an energetic promoter of the Stargate Project, a $500 billion collaboration involving OpenAI, Oracle and SoftBank to build data centers and other AI infrastructure across the US, including in Lordstown.  Trump hosting Oracle’s Larry Ellison, SoftBank’s Masayoshi Son and OpenAI’s Sam Altman at the White House in January. Photographer: Jabin Botsford/The Washington Post/Getty Images But while AI is in the spotlight, signs are few as of yet that Trump is delivering on his promise to engineer a manufacturing renaissance. Factory activity has contracted for seven consecutive months through September, according to the Institute for Supply Management. Surveys show the administration’s ever-shifting tariffs, restrictions on immigration and withdrawal of government support for electric vehicles and other cleantech are sowing uncertainty that’s deterring investment. That’s showing up in the data too: Spending on construction of new factories is down 2.5% so far this year, while that for data centers is up almost 18%. Paul Kedrosky, a fellow at the MIT Initiative on the Digital Economy, who was among the first to put the AI capital expenditure, or capex, boom in historical context, says he worries that the euphoria over the technology has blinded Trump and his advisers to the negative impact their policies are having on the rest of the economy. “This thing is so large that it’s breaking people’s mental models of how the economy works, and they’re making mistakes as a result,” he says. The hyperscalers have managed to secure broad tariff exemptions from the White House on imported servers and other hardware for data centers. But the administration has been less open to manufacturers’ pleas for waivers on imports of the equipment needed to either expand or build US factories. In September the Department of Commerce opened an investigation into imports of robots and industrial machinery that will likely lead to yet more tariffs, creating further obstacles to reshoring production. Indeed, for many companies, Trump’s tariffs appear to be canceling out the expected lift of the pro-business provisions in the One Big Beautiful Bill Act, which Congress passed this summer, including improved tax treatment for purchases of equipment and investments in research and development. The National Association of Manufacturers publicly campaigned in favor of the legislation, but when it surveyed its members, almost 80% identified the burden of tariffs as their biggest concern. Based on the responses, NAM predicts that capital expenditures will rise only 1% not adjusted for inflation in the coming year, which is in line with recent years. By some calculations, Trump’s import levies amount to the largest tax increase on American companies since the early 1990s. Tariffs will cost the heavy machinery maker Caterpillar Inc. as much as $1.8 billion this year and GM faces a gross bill of $4.5 billion. The duties haven’t halted the decades-long shrinkage of the US manufacturing workforce; 42,000 jobs have disappeared since April, when Trump unveiled his “Liberation Day” tariffs, in the longest losing streak since the Covid-19 pandemic. The “industrial recovery looks suspect,” Barclays Plc analyst Julian Mitchell wrote in a note in August in which he called AI “the only game in town.” Keep reading: AI Data Center Boom Threatens Trump’s Manufacturing Revival |
