AI’s Profitability Mirage: The Astronomical Costs and Unmet Promises

A Nobel Prize-winning economist has warned that the path to profitability for artificial intelligence is increasingly unlikely due to the staggering capital requirements needed to build and maintain its infrastructure.

Daron Acemoglu, a professor at the Massachusetts Institute of Technology, stated that companies heavily investing in AI are betting on productivity gains that have yet to materialize at scale. “You’re right. The numbers are astronomical,” he said when addressing concerns about AI data centers costing tens of billions of dollars each.

He noted that even a single data center using 1 gigawatt of power could cost an estimated $80 billion, and committing to build out 20 to 30 gigawatts would amount to $1.5 trillion in capital expenditures—an investment roughly equal to Tesla’s current market cap. Acemoglu emphasized that such a level of investment assumes a future where one or two companies dominate multiple industries and generate trillions in profits. “That’s the only way you would rationalize this,” he said, describing it as “a very, very long shot.”

Acemoglu added that even the largest technology firms have never generated sufficient profits to justify trillions in capital expenditures, particularly given how quickly AI hardware becomes obsolete every three to five years and must be replaced. While some consumers are willing to pay modest subscription fees for AI tools, he said businesses remain reluctant to spend heavily due to limited real-world productivity gains. “There aren’t that many applications that have proven to be very productive in the wild,” he explained, noting that many tools work well in controlled environments but falter when dealing with complex real-world challenges.

He also highlighted an emerging cost: companies increasingly need to hire additional employees to monitor, verify, and correct AI-generated output. “Integrating AI actually is very difficult,” Acemoglu said. “You need to understand your organization, what your employees really add, and then bring AI to help them. Rote automation is not going to work.”

Furthermore, he noted that businesses feel pressured by consultants, boards, and public narratives to adopt AI even when returns are uncertain. “Businesses aren’t spending all that much,” he said, “and when they do, they’re not getting all the returns.” Acemoglu stressed that AI’s long-term success will ultimately depend on whether its economics can deliver sustainable profits rather than technological promise alone.