External signal·MIT Technology Review·May 26, 2026·David Rotman·13 min read
A reality check on the AI jobs hysteria
“It could be disruptive, but the data is telling us right now that disruption is not yet here, and we have time to plan.”
Summary
MIT Technology Review's David Rotman pushes back on the narrative that AI is already gutting white-collar work. Drawing on US Bureau of Labor Statistics household data, Federal Reserve research, and ongoing surveys from Harvard's David Deming and Stanford's Digital Economy Lab, he shows that unemployment for AI-exposed occupations is actually lower than for less-exposed ones, and there's no sign of workers fleeing threatened roles for "safer" manual ones. The clearest real effect is narrow: a roughly 16% decline in entry-level headcount for 22-to-25-year-olds in highly automatable jobs like junior coding and customer service — even as overall coding employment keeps growing and wages in exposed sectors rise. The piece argues the economy is "flying blind" for lack of good data, and that history (radiologists, driverless trucks) shows job-loss forecasts routinely misjudge both the pace of change and the messy bundle of tasks that real jobs contain.
Predictions for the future of work
Rather than forecasting an imminent jobs apocalypse, the article predicts a slower, uneven transition whose speed is the decisive unknown. It suggests the traditional "earn-while-you-learn" career ladder may break for some occupations as AI absorbs the codified, entry-level tasks juniors once used to gain experience, while seasoned workers' tacit knowledge stays valuable. The economists quoted expect AI to be disruptive eventually but believe there is still time to plan — provided governments and businesses invest in reskilling and, above all, in better data to see the transition coming. The danger it flags is a sudden, China-shock-style upheaval that policymakers fail to anticipate.
Originally published by MIT Technology Review · May 26, 2026
Read the original at MIT Technology Review