For the better part of two decades, Nicholas Bloom has been researching how Americans work, and his conclusions have recently taken on a more negative tone. He states unequivocally that the labor market is frozen. The majority of those caught in the middle are young, and employers aren’t firing employees at an unusually high rate, but they have also quietly stopped hiring them.
Bloom doesn’t use the word “frozen” carelessly, so it’s worth pausing on it. He is the economist who foresaw the surge in remote work before most CEOs did, and who subsequently demonstrated that hybrid schedules could reduce a company’s employee turnover by over one-third. Therefore, it carries weight when he advises Fortune that employees who currently have a job should hold onto it. The reason is not that the data is dramatic. No one is in a panic. Everyone is simply stuck.
What’s really causing the freeze is the more difficult question, one that dominates both shareholder calls and HR departments. Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen led a Stanford Digital Economy Lab study that examined ADP payroll data covering millions of workers and discovered something startling: employment for workers between the ages of 22 and 25 in the most AI-exposed occupations has decreased by about 13% compared to older workers in the same fields. A subsequent update increased that percentage to 16%. It’s clear from the title of the paper: “Canaries in the Coal Mine.” It contends that young workers are the first to be negatively impacted by automation.
That’s the depressing headline version. This is where things get complicated, though, and Bloom himself starts to doubt the popular narrative in his own field. In a different article for Fortune, he contended that the productivity increases that everyone attributes to AI actually predate its widespread adoption and can be traced back, in large part, to remote work. If he is correct, businesses that attribute recent increases in productivity to chatbots and copilots might be fabricating a convenient narrative rather than the truth.
Being both cited as evidence of an AI-driven crisis and one of the more well-known voices questioning whether AI deserves credit at all is an odd position to be in. Peter Cappelli of Wharton has made a similar argument, telling SHRM that many of the businesses announcing layoffs due to AI aren’t genuinely having financial difficulties; rather, they are using AI as a pretext for cuts that investor pressure was already calling for. Blaming a chatbot for choices that, upon closer examination, appear to be standard cost-cutting dressed in a futuristic outfit is almost too convenient.
Andrew McAfee of MIT, meanwhile, is raising a completely different alarm, cautioning that businesses that eliminate entry-level positions now might be subtly destroying their own pipeline of future leaders. AI hysteria is not necessary for that argument to succeed. All it takes is a long enough memory to remember the origins of today’s mid-level managers.
After all the conflicting papers and opinion pieces, what’s left is more of a genuinely unresolved debate than a single depressing verdict; even the researcher whose name is frequently associated with the “crisis” framing seems to be urging a little more caution about what, precisely, is being proven. For recent graduates, the job market is unquestionably chilly. It’s still, annoyingly, unclear whether AI, remote work, or just plain corporate nerve is the true culprit.
