Odd Lots · Tuesday, September 1, 2026
Adam Posen notes that while there's indicative evidence of AI improving productivity, significant labor market impacts are not yet apparent. He suggests that the integration of AI into businesses and the creation of specialized applications will take time, potentially delaying widespread job displacement, and that the 'messy jobs' theory, highlighting human interaction and nuanced tasks, may explain this lag.
“Much more indicative evidence of improvement in productivity, Tracy, than on effects on the labor market. So speaking of colleagues of mine, Martin Terzempa and Jed Kolko are both top economists who work at Peterson and work on these issues.”
“And you're just not seeing it in the data. So the number of hires, even of coders, you know, if you had to pick the two jobs where you most thought they're toast, would be long haul truckers and coders, lower level coders. And we're just not seeing it. Job growth continues in those industries.”
“And partly it's a matter of time, as you indicated. Time keeps coming up. It's how fast things happen. that there is arguably a period in which it is efficient to have the human and the AI working together, especially as the AI is in sort of learning mode, and especially as the businesses, just like with the internet, you have the technology and then you have to figure out how to reform your business to make use of it.”