Increasingly, economic growth in the United States relies on investments in artificial intelligence.
Today, many consumers interact directly with AI through apps like ChatGPT or through summarized search results on Google or Apple. They may also encounter AI-generated images on social media.
But that’s not what drives growth. Rather, it’s about investing in building the raw computing power and electrical infrastructure needed to power these applications, as well as those that may arise from them in the future. Think about data center construction, computer processing chips, information processing equipment, and power transmission equipment.
“Where AI has had a direct impact on the economy is in the capital expenditures of (technology companies) and other companies on hardware and software needed to expand their cloud computing capacity so that it can meet the larger demand for AI computing,” said Ed. Yardeni, president of Yardeni Research, an economics and market consulting firm.
So far, the gains from this increased spending on AI technology and infrastructure have been relatively small, both in terms of jobs and financial returns.
And yet, the stock market gains have been massive. According to a new estimate from Skanda Amarnath, executive director of Employ America, spending on the development of AI technologies accounted for between 16% and 20% of real gross domestic product growth in the third quarter of 2024 alone, and only should only increase. In terms of share of total spending, AI investments are on track to surpass the share of GDP attributed to the dot-com boom of the late 1990s and become as important as real estate was during the bubble of the 2000s, Amarnath wrote in a Bloomberg News article published Monday.
“We’re starting to see it show up in the data,” Amarnath told NBC News. “This is something that means it’s more macroeconomically relevant and will likely be a driver of growth in 2025.”
Technologists pushing for AI development are touting AI not only as having the potential to transform business productivity, but as one of the most important developments in human history. At the moment, it’s a ‘FOMO’ mentality, not immediate profits, that appears to be driving much of the current cycle.
“Superintelligent tools could massively accelerate scientific discovery and innovation far beyond what we are capable of doing on our own, and in turn massively increase abundance and prosperity,” Sam Altman, director of ‘OpenAI, which built the original ChatGPT, wrote in a recent blog.
At the same time, a few scattered voices have already called the current round of investment in AI “bubble” which, like those in the Internet and housing sectors, could burst and lead to a slowdown.
“Overbuilding things that the world doesn’t need or isn’t ready for usually ends badly,” said Jim Covello, head of equity research at Goldman Sachs. said in a July report by the investment banking giant. “The NASDAQ fell about 70% between the height of the dot-com boom and the creation of Uber.”
In the meantime, however, high-profile AI-related announcements continue to pile up. On Tuesday, President-elect Donald Trump announcement a $20 billion deal to build new data centers in partnership with a billionaire UAE developer. Later today, Amazon’s AWS cloud computing group announcement it would spend $11 billion on AI-related investments in the state of Georgia. Earlier this month, Microsoft announced it was spending a total of $80 billion on AI-powered data centers for its 2025 fiscal year. And in December, Japanese conglomerate Softbank held a joint press conference with Trump to announce $100 billion in AI spending in the United States
The gains are also trickling down to the stock market, which had a stellar 2024 thanks in large part to the performance of the so-called Magnificent Seven. These seven tech companies – Amazon, Apple, Google parent Alphabet, Facebook and Instagram parent Meta Platforms, Microsoft, Nvidia and Tesla – gained an average of 63% last year, with Nvidia jumping 171%. Together, these companies now represent a third of the total value of the S&P 500 index and are responsible for more than half of its gains.
“Almost all of them are considered an AI game in one way or another,” Yardeni said.
Last year, about a third of all startup investments went into AI-related companies, the highest percentage ever, according to data from Crunchbase, which tracks venture capital data .
In addition to technology companies, power companies and infrastructure providers have also seen their share prices soar. Among the companies benefiting from this dynamic are Constellation Energy and Vistra, both considered to specialize in nuclear energy. Constellation announced last year it was partnering with Microsoft to restart one of the reactors at the Three Mile Island nuclear site in Pennsylvania.
What is missing at the moment is a wave of new jobs… although some AI-focused investment announcements promise that tens of thousands of them will eventually happen.
In fact, one of the essential premises of AI is its ability to automate human-centered roles, which could lead to job losses. And even if all new technologies end up creating new jobs, they can also cause entire professions to disappear. Surveys have found that many tasks, from writing and computer coding to illustration and translation, that were once done by people are likely to be taken over by robots, if not is already done.
So far, the most direct beneficiary of the current investment boost has been the construction sector, which continues to experience healthy annual employment growth of over 2.5%. Spending on data center construction increased 43.1% from the previous year. according to surveys monitored by the Associated General Contractors of America.
Employment in the public services sector also is now at levels not seen in more than 20 years.
But other sectors that had benefited from previous technological surges, particularly professional and business services – namely white-collar work – have stalled. Even traditional software engineering jobs are rare: Job postings for these positions on Indeed have fallen below pre-pandemic levels.
“It hasn’t created a huge job boom,” Yardeni said. He continued: “AI increases the productivity of programmers, so maybe overall we’re not going to see a big increase in the number of programmers hired to produce AI,” Yardei said. “The reward should be productivity.”
Yet, for better or worse, the U.S. economy, not to mention the stock market, increasingly relies on expectations of a profit from AI. A financial manager recently summed up with irony the disproportionate weight of AI on the economy. “I sometimes joke that we leveraged the entire U.S. retreat on Nvidia’s performance,” Marc Rowan, chief executive of Apollo Global Management, said at a corporate event this fall. according to the Financial Timesreferring to the chipmaker’s role in the stock’s rise.
Indeed, some academics argue that while there is broad consensus that bets on AI will eventually pay off in terms of increased productivity, no one knows when or how the actual benefits promised to public by increasingly sensitive computers will occur.
“These investments represent costs,” said Tania Babina, an associate professor of finance at Columbia Business School, meaning tech companies are making investments that they hope will pay off in the long run. “So I hope the benefits will be on a broader scale, not just for these technology leaders.”