DeepMind and Stanford economists debate what stays scarce after AGI

Alex Imas of Google DeepMind and Phil Trammell of Epoch join Dwarkesh Patel to explore the economic landscape after AGI. They examine what remains scarce when intelligence and labor become abundant, arguing that 'relational' goods and services—where human involvement itself carries value, like ballerinas or baristas—may be the enduring scarce assets. The conversation covers optimal taxation and redistribution of AI-generated wealth, whether demand will collapse, how developing nations can capture gains, and why human employees may struggle to integrate into a machine-only economy. The economists caution against relying on individual forecasts, advocating instead for prediction markets to navigate unprecedented uncertainty.