The U.S. economy remains relatively stable even as President Donald Trump’s tariff and immigration policies disrupt businesses. Experts say the country can thank the artificial intelligence (AI) industry.
“AI machines appear to be saving the U.S. economy right now, in a very literal sense,” George Saravelos of Deutsche Bank wrote to clients in late September. “Without tech spending, the U.S. will be close to or in recession this year.”
Recommended stories
list of 4 itemsend of list
Economist and Nobel Prize winner Paul Krugman expressed a similar opinion in his Substack newsletter. AI companies are investing hundreds of billions of dollars in AI infrastructure and development, and other U.S. companies are also spending billions on AI products.
Just last month, a data center in Aveline, Texas went online, the flagship site of the $500 billion Stargate program, a joint venture between Oracle, OpenAI, and Japan’s SoftBank to advance America’s AI infrastructure.
Around the same time, chipmaker Nvidia announced it would invest up to $100 billion in OpenAI to provide data center chips. It also became the first U.S. company to reach $4 trillion in market capitalization. This benchmark was quickly followed by Microsoft, whose stock soared as AI was one of the key factors driving business demand.
It’s not just Nvidia and Microsoft. Alphabet, Google’s parent company, and Meta Platforms, which owns Facebook, Instagram, and WhatsApp, are ramping up their AI ambitions and investment efforts.
While this enthusiasm surrounding AI appears to be supporting the US economy for the time being, there are concerns that this could turn into a “bubble” similar to the dot-com bubble of the late 1990s.
“The reason people are worried about an AI bubble is because seven companies are driving more than 400 companies,” Campbell Harvey, a finance professor at Duke University, told Al Jazeera.
If you look at the S&P 500 index, you’ll see that the seven tech companies that are deeply involved in AI are generating the most growth.
Harvey acknowledges that it’s difficult to tell whether these tech stocks are overvalued because the adoption and growth of AI is still in its early stages.
AI adoption rate is starting to slow down
Karl Frey, associate professor of AI and work at the University of Oxford, said: “While stocks seem to be rising a bit, there are also real revenues behind the massive push to build data centres.” “A bubble may be forming, but we are far from the realm of tulip mania,” he said, referring to the huge rise in tulip prices in the Netherlands in the 17th century, an event often cited as a characteristic of bubbles.
“The concern is that early adopters of AI are having second thoughts. Big companies that rushed in have narrowed their projects down to a small number of projects that clearly save money or make money, and leave the rest on ice,” he said.
For example, major companies like IBM and Klarna cut thousands of jobs in their customer service departments and replaced them with AI, but it wasn’t long after that decision that they began to reverse course. They found that technology could not do everything they expected compared to human workers.
That could be a serious problem for AI companies if large companies that spent a lot of money deploying AI tools decide that those tools aren’t actually of much use to their business. Eventually, the stock could start to fall as the company loses customers and expected profits decline.
A report released by MIT in August found that 95% of companies that have adopted AI fail to achieve significant revenue acceleration from AI. U.S. Census Bureau data shows that AI adoption by large companies has recently begun to slow.
People seem to be starting to question the usefulness of these AI tools, which are often used to replace people in jobs such as customer service, software engineering, and several other entry-level jobs.
“There’s a growing recognition that over the past year, the hype around AI’s capabilities and the fear of falling behind have led many companies to race to implement AI into their operations,” said Cal Newport, a computer science professor at Georgetown University. “But it turns out that integrating generative AI, especially generative AI, into existing workflows in a very useful way is more difficult than people thought.”
Newport says the models underlying these AI programs are currently “too unreliable” to successfully automate jobs. He points out that the idea that AI will rapidly replace jobs is “absolutely not happening” at this point.
A recent study from Stanford University found that the adoption of AI tools at large companies has led to a 13% decline in entry-level jobs in customer service, accounting, and software development from 2022 onwards.
It is not yet clear whether AI has reached “bubble” territory, but it is a possibility, and if the bubble were to burst, it could cause significant damage to the U.S. economy.
Frey says the dot-com bubble was extremely expensive for investors, but “ultimately left us with technology and infrastructure that improved productivity.” The question is whether this AI situation will play out the same way.
“Unless an AI failure brings down a major financial institution and triggers a full-blown financial crisis, the bigger risk today is otherwise,” Frey said. “AI has yet to deliver the clear and widespread productivity gains that our stagnant economy needs.”
