⤴ Productivity paradox
The curious case of generative AI's missing efficiency miracle — and why that's precisely the point
Item: “The U.S. experienced its first decline in productivity in nearly three years during the first quarter, turning what was once a growth engine into another potential drag on U.S. economic growth.” - Barron’s, May 8, 2025
My fellow pro-growth/progress/abundance Up Wingers,
Stick with me here: AI chatbots’ (so far) negligible economic impact — particularly in top-line productivity numbers — should encourage rather than alarm technology optimists. Why such an Up Wing take? Well, new research tracking 25,000 workers across Denmark in late 2023 and early 2024 reveals a classic pattern of technological disruption of the sort that historically precedes major productivity booms.
At first glance, the findings from “Large Language Models, Small Labor Market Effects,” a National Bureau of Economic Research working paper authored by Anders Humlum (University of Chicago) and Emilie Vestergaard (University of Copenhagen), might seem a Down Wing bummer. Two years into the era of ChatGPT — with firms rolling out in-house models and workers enthusiastically embracing AI tools — the data show virtually no impact on wages, working hours, or employment levels.
From the paper:
In this sense, our results echo Robert Solow’s famous observation about the IT revolution: “You can see the computer age everywhere but in the productivity statistics.” … Despite substantial investments, economic impacts remain minimal. … Our findings challenge narratives of imminent labor market transformation due to Generative AI.”
But here’s the thing: That apparent disappointment is exactly what economists would expect to see if generative AI is indeed a transformational general-purpose technology— one poised to reshape the economy over time, not necessarily overnight.
Keep reading with a 7-day free trial
Subscribe to Faster, Please! to keep reading this post and get 7 days of free access to the full post archives.