Episode 2 of Post AGI is a conversation with Christian Catalini, founder of the MIT Cryptoeconomics Lab and previously associate professor at MIT Sloan.

Most of the AI debate swings between utopia and mass unemployment. Christian argues the real bottleneck is something else: verification. As the cost to automate falls toward zero, the cost to check that the output matches our intent stays bottlenecked by us, and for the first time the cost to verify something can be higher than the cost to generate it. Using his paper "Some Simple Economics of AGI" as a lens, hosts Sreeram Kannan and Soubhik Deb work through what that inversion means for labor, firms, and markets, why it could lead either to a "hollow economy" or to verifiable abundance, and why the primitives built in crypto over the last decade may be exactly what AI now needs.

Chapters:
02:41 The hollow economy, and why Amazon called an emergency meeting over AI slop
05:48 Why verification is the real bottleneck
07:08 The Chernobyl problem: complex systems failing in complex ways
09:01 Liability as a service
13:32 If it can be measured, it can be automated
16:55 Why AI hands you something that sounds right and is completely wrong
26:20 Why crypto built the infrastructure AI now needs
43:58 The missing junior problem, and why CS grads stopped getting jobs
59:12 The four zones that decide which jobs survive AI

Season 01 of Post AGI begins with economist Alex Imas, the Roger L. and Rachel M. Goetz Professor of Behavioral Science, Economics and Applied AI at the University of Chicago Booth School of Business.

The AGI debate gets most of the attention, but the harder questions are about what comes after. In this conversation, hosts Soubhik Deb and Sreeram Kannan talk with Alex about what AI actually does to labor, capital, and markets. We get into why "job exposure" misleads almost everyone, why automation may come for physical work before knowledge work, the economy of human-made goods that survives AI, the case against UBI and for broad capital ownership, and whether agents can ever run an economy better than markets. The throughline is one question Alex argues is the only one that matters after AGI: what stays scarce.

Recorded while Alex was full-time faculty at the University of Chicago.

Chapters:

00:00 Intro
04:26 Why "AI exposure" doesn't mean what people think
08:55 Which jobs automation actually comes for
13:18 The robotic "dark warehouses" being built in China
14:27 What happens to the labor market at the extreme
16:12 The relational economy and human-made goods
21:26 Capital ownership and the AI sovereign wealth fund
23:08 The case against UBI
26:56 Why disaster is what moves policy
30:23 Digital intelligence needs digital institutions
36:05 Why verification got harder than production
37:01 An agentic layer on everything within 12 months
41:14 Will agents become the new companies?
45:35 The only question in post-AGI economics: what is scarce?
48:02 Are AI agents better delegates than humans?
52:40 Can AI run a command economy?
55:05 Why markets are the original superintelligence
57:01 Does AI need government intervention?
1:01:39 The precedent for capital ownership: Intel, Alaska, Norway