Ryan Greenblatt says economists underestimate AI’s growth impact even in a 100 million worker scenario

RyanGreenblatt · x · 2026-07-22

Ryan Greenblatt uses a bar-stool thought experiment to argue that people are underestimating AI’s growth impact. He walks through a scenario where 100 million AI workers are effectively super-human: they match top experts across economically relevant fields, work longer and faster than humans, save almost all of what they earn, and could do substantial cognitive labor.

The punchline is that if economists would call that transformative for human immigrants, they should not suddenly become conservative when the same properties are attached to AIs. The post argues that claims of only 3% annual growth from AI ignore how much bottleneck thinking changes once labor becomes vastly more capable and scalable.

Related event: AI Economic Impact Forecasts Vary by Trillions(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →