How the AI spending boom is reshaping an otherwise slowing economy - The Washington Post

As revisions to U.S. jobs numbers call into question economic growth, Big Tech is becoming an even more important part of American prosperity.

An Amazon data center in Boardman, Oregon. (Jenny Kane/AP)

SAN FRANCISCO — Big Tech’s unprecedented spending spree on artificial intelligence is getting so big that it’s starting to reshape the U.S. economy.

Google, Meta, Amazon and Microsoft reaffirmed this past week that they are on track to spend more than $350 billion this year building and equipping AI data centers — a massive influx of money that economists and analysts say could be a countervailing force to what appears to be a decelerating economy.

Revised job numbers released Friday by the Labor Department suggest the U.S. job market appears weaker than previously understood, with employers pulling back sharply on new hiring in May and June. At the same time, Big Tech’s infusion of cash will go toward building and expanding data centers, which could create more infrastructure jobs as well as higher demand for computer chips, servers and other network equipment that power them. AI investments could grow the economy by as much as 0.7 percent in 2025, according to a calculation by Jens Nordvig, an economist and the founder of economic data platform Exante Data. That would represent half of the 1.4 percent growth projection for the U.S. economy this year from the Federal Reserve.

Hefty AI spending is creating a situation where the U.S. economy is becoming even more dependent on Big Tech, raising concerns from some analysts and economists that should the AI boom slow down, it could further harm the economy.

“The AI complex seems to be carrying the economy on its back now,” said Callie Cox, a market strategist with investment firm Ritholtz Wealth Management. “In a healthy economy, consumers and businesses from all backgrounds and industries should be participating meaningfully. That’s not the case right now.”

AI executives argue the spending boom will create more jobs and bring about scientific breakthroughs with advancements in the technology. OpenAI has said that once its AI data centers are built, the resulting economic boom will create “hundreds of thousands of American jobs.”

Meta CEO Mark Zuckerberg has been on an AI hiring spree that began in June, offering hundreds of millions of dollars in salaries as the company launches a new team dedicated to creating “superintelligence,” a term for machines hoped to one day outperform humans at every possible task.

“AI will improve all our existing systems and enable the creation and discovery of new things that aren’t imaginable today,” Zuckerberg wrote in a Wednesday manifesto laying out the company’s AI plans, predicting the tech would usher in a “new era” of productivity and economic development.

A handful of regions across the United States, including Texas and Northern Virginia, may benefit from increased construction jobs as the data centers are built, but once they’re up and running, they will require fewer people to maintain them. Nvidia, which is the biggest AI chip supplier in the world, brought in $44 billion in revenue in the most recent quarter alone. That company’s stock has skyrocketed over the past two years, turning many of its individual employees with stock options into millionaires and some of its executives into billionaires.

Wall Street has so far rewarded Big Tech companies for their spending on AI, increasing their market value and making them an even bigger part of the overall stock market. Microsoft, an investor in OpenAI, last week hit a $4 trillion market valuation after reporting strong earnings growth, the second company to hit that threshold after Nvidia. Meta’s stock jumped 11 percent after reporting strong earnings and telling investors its AI spending would be on the high end of its previous predictions.

Investor enthusiasm for AI has boosted the stock portfolios of many Americans, but if the economy sours, Big Tech stocks could be hit hard, affecting 401(k)s and investment portfolios, market strategists and investors say.

“Recessions can snap your focus back to reality quickly, and what gets hit the worst in recession-fueled market sell-offs is high-valuation growth stocks,” Cox said.

Investment bank Raymond James said it expects the AI spending boom to continue for the “next couple years,” but only if tech companies can show they’re able to make real money from AI services.

AI boosters argue that demand for the tech will increase exponentially as more people use it, but skeptics point out that the costs of AI services have already fallen as companies compete to drive them down and AI algorithms become more efficient.

“This build-out presumes something that isn’t true, and that is you can make a reasonable rate of return well into the future,” said Paul Kedrosky, a tech investor and research affiliate at MIT’s Initiative on the Digital Economy.

There are also questions about whether all of the promised funding from AI companies will actually materialize. On the first day of Trump’s term, he hosted OpenAI and Japanese investment giant SoftBank to announce the companies’ plans to build $500 billion worth of data centers during Trump’s presidency. But the companies also said not all of the funding for the project had been secured yet.

Other big-ticket tech projects have failed to materialize in the past. During Trump’s first term, Taiwanese tech manufacturer Foxconn unveiled plans to build a $10 billion facility in Wisconsin. It never opened.

The tech industry has been feverishly investing in AI since OpenAI kicked off an arms race in late 2022 with the release of ChatGPT. The tech behind ChatGPT and other tools, such as Google’s AI search answers, demands more computational power than regular apps or internet searches. That means the companies must buy expensive computer chips manufactured abroad, and build energy-intensive data centers to keep them running around-the-clock.

There aren’t any signs Big Tech will slow down its AI spending. In earnings calls over the past two weeks, company executives said capital expenditures would keep growing. Google increased its planned spending to $85 billon from $75 billion. Amazon said it probably would overshoot its previous expectation of investing $100 billion this year. Meta said its spending would be as high as $72 billion in 2025 and would increase in 2026.

The companies have surpassed previous forecasts they’ve made. Microsoft said last week it is on track to spend $30 billion on AI in the current quarter. If that rate continues, the company will surpass its previous prediction of spending $80 billion on AI this year.

The numbers are bigger than anything in the tech industry’s history. From 2023 through the end of this year, Google, Microsoft, Meta and Amazon will spend more on new investments than they did for all the years from 2010 to 2022 combined, according to an analysis of historical spending by The Washington Post. That period covers a time in which Big Tech was growing rapidly, building new offices all over the world, constructing data centers to run cloud services and laying thousands of miles of transoceanic internet cables.

The spending has also eclipsed other major tech projects in U.S. history. NASA’s Apollo program cost about $25 billion, the agency told Congress in 1973. In today’s dollars, that is about $180 billion, a fraction of what the tech industry will spend on AI this year alone.

In China, which has its own AI spending boom as the country seeks to compete with the U.S. for control over the future of the technology, President Xi Jinping recently criticized local officials for investing too much in AI data centers.

“The foundation of the [U.S.] economy is cracked, even if AI optimism is keeping the momentum going for now,” said Cox.

SOURCE: The Washington Post