Economists urge faster action as AI job-impact measures clash

More than 200 economists, including 16 Nobel laureates and the chief economists of OpenAI and Anthropic, called this week for faster work to understand and guide AI’s economic effects. Their warning is that AI could reshape jobs and living standards on a compressed timeline while researchers still disagree over how to measure its impact.

The statement, titled “We Must Act Now”, says AI may become far more capable within 10 years and could drive an economic shift larger than the Industrial Revolution, but over a much shorter period. It says that shift could include broad job displacement as well as gains in living standards.

Anton Korinek, a University of Virginia economist and one of the organizers, described the uncertainty in stark terms. “We are driving in the fog, and it is extraordinarily difficult to anticipate what will happen next,” Korinek said.

The statement does not lay out a detailed policy program. It calls on economists, policymakers and technology executives to build better incentives, safeguards and institutions so AI complements people and benefits society.

Michael Spence, a Nobel laureate at NYU, called for an “all hands on deck” effort to steer AI in beneficial directions, citing the scale, speed and uncertainty of the technology’s advance. The statement frames the problem as both urgent and unresolved: researchers see major possible consequences but lack shared tools for judging them.

Researchers disagree on the size and timing of the impact

Nela Richardson, chief economist at ADP, has said much of the public debate over AI and jobs amounts to “guesswork” because so many variables remain unsettled. Daron Acemoglu, the MIT Nobel laureate known for skepticism about some AI productivity claims, has previously told Fortune that much of the discussion around AI productivity is “brainless,” while also saying recent advances have raised his concern about near-term disruption.

Acemoglu signed the new statement and said he was “so happy to join other leading experts in calling for the urgent need to redirect AI so that its risks are minimized and it can work for the benefit of workers and society.” Stanford economist Erik Brynjolfsson, another organizer, said AI is moving faster than economists’ grasp of its consequences.

“AI capabilities are advancing far faster than our understanding of the economic implications,” Brynjolfsson said in the statement. “We must act now to guide AI to complement humans rather than simply imitate them—and to generate prosperity for the many, not just the few.”

Brynjolfsson has also worked with ADP Research on the Canaries Dashboard, which tracks 4.6 million workers across more than 730 occupations in near real time. According to the dashboard, employment among workers ages 22 to 25 in AI-exposed occupations is shrinking more than 4% annually, even while broader labor-market figures appear calm.

Even the measurement tools are disputed

Torsten Slok, chief economist at Apollo Global Management, argued in a recent blog post that the common phrase “AI exposure” can mean very different things depending on the method used. Slok said five measurement frameworks produce different answers, especially in jobs often described as vulnerable to AI.

According to Slok, some measures look at real usage, such as Claude chat logs or Microsoft Copilot activity. Others rely on expert judgments about which skills AI could replace, ChatGPT’s own ratings of its usefulness for tasks, or employer job postings that mention AI skills.

Slok wrote that theoretical measures tend to show higher exposure than usage-based measures because they do not account for whether workers or companies are adopting the tools, or whether adoption is worth the cost. He said the measures diverge most for jobs such as telemarketers, tax preparers and writers.

“When someone says a job is ‘highly exposed to AI,’ the honest first question is: Exposed by which measure, and measuring what?” Slok wrote. “Until that is pinned down, the label ‘AI exposure’ carries far less meaning than it appears to.”

This story draws on original reporting from Fortune.