There are three distinct communities doing economics of AI research, and they don't always agree. Seth Benzell sits at the intersection of all of them and has views about where each one goes wrong.

Topics covered:

• The three factions of AI economics: digitization, macro labor, and growth
• Why "directing technical change" toward pro-worker outcomes is probably less controllable than the optimists think
• Diminishing returns to intelligence
• Capital accumulation as an underrated story
• Why human task data is a factor of production
• Why priors in economics shift more from being weakly held than from unusually strong evidence

Check out the Justified Posteriors podcast here: https://www.justifiedposterior ...