Think clearly about AI's economic impact

If you've sat through an AI economics presentation and found yourself quietly noting that the speaker just assumed away the most interesting problems — you're in the right place!

The economics of AI is genuinely hard. Not because the data is sparse (though sometimes it is), and not because the technology is too new to analyze (we have decades of productivity and labor data to work with), but because the standard frameworks economists reach for were built for a different kind of good. When something is non-rival — when my use of it doesn't diminish yours — standard production theory starts to creak. When marginal cost approaches zero, competition models behave strangely. When a technology functions primarily as a prediction input that augments human tasks rather than simply replacing them, the automation-versus-labor framing most people use gets the question wrong from the start. These aren't small technical footnotes. They're foundational issues that reshape what the analysis looks like.

What I've built here is the course I kept wishing existed: one that takes the standard models seriously enough to show precisely where they break down, then works carefully through the frontier frameworks that researchers are actually using — ideas economics, general-purpose technology theory, superstar firm models, platform and data network effects, Schumpeterian creative destruction — and applies them to real empirical literature and real industries. We don't pretend the productivity paradox is resolved. We don't pretend the labor effects are settled. We do develop a rigorous terminology for thinking about both, so you can engage with the primary literature and the live debates rather than just receiving secondhand summaries of them.

The policy and strategy implications follow from the frameworks — not the other way around. Too much AI economics commentary starts with a conclusion and works backward. This course does the opposite: it builds the analytical foundation first and earns whatever conclusions follow. That means being explicit about what the evidence shows, what it doesn't show, and what we genuinely don't know yet. I think that kind of intellectual honesty is more useful to you, in the long run, than confident-sounding answers to questions that aren't yet closed.

If you're an economist, policy analyst, strategist, or intellectually serious professional who wants to think about AI's economic impact without relying on hand-waving — I built this for you. Come in with your skepticism intact. You'll need it, and we'll put it to work.