Odd Lots Alex Imas on Why Economists Might Be Getting AI Wrong
496 snips
Apr 18, 2026 
Alex Imas, a University of Chicago economist studying decision-making and applied AI, explores why AI may not follow the usual technology script. The conversation covers labor market disruption, AI agents that can take actions, which white-collar and logistics roles look vulnerable, why demand matters for jobs, the breakneck pace of change, and even his grumpy “Marxist robots” experiment.
AI Snips
Chapters
Books
Transcript
Episode notes
Why Job Exposure Does Not Equal Job Loss
- AI exposure scores do not equal job loss because jobs bundle many tasks, and automating weaker tasks can raise a worker’s value.
- Alex Imas says economists map tasks well with O-Net, but struggle to measure complementarity between tasks, which determines whether automation helps or replaces workers.
Consumer Demand May Decide Who Keeps Jobs
- Whether AI destroys jobs depends heavily on consumer demand elasticity after productivity lowers prices.
- Alex Imas uses software as the live debate: if cheaper output sparks much more demand, firms may hire more; if not, they may downsize despite higher productivity.
Three Paths To An AI White Collar Wipeout
- Alex Imas says a white-collar wipeout could happen through full automation, weak demand absorption, or stronger firm incentives to automate one-task roles.
- If automation lets a company eliminate a worker entirely, the return on investing in automation rises sharply.
