What History Tells Us About the AI Revolution | Tyler Cowen & Alex Imas

What History Tells Us About the AI Revolution | Tyler Cowen & Alex Imas

More

Summary

Economists Tyler Cowen and Alex Imas join the a16z podcast to discuss what economic history can tell us about the current AI revolution. They weigh how fast AI progress is accelerating against how slowly institutions, bottlenecks, and real-world adoption tend to move.

Imas points to the shift from the chatbot era to agentic systems and to the METR report’s task-length trends, while questioning how well benchmark gains translate into economically valuable work, especially for tasks that are poorly specified. Both guests argue that benefits will exceed costs, but that the path will look more ordinary than many forecasts suggest. They contrast a world with more normal GDP growth against one with double-digit growth, and note that agents are not yet close to 10% of GDP.

A major thread is agent coordination: despite talk of agent swarms, today’s agents mostly perform parallel processing rather than true coordination, which the guests see as a key requirement for broad economic transformation. The conversation also covers alignment, including a mechanism-design view in which incentives and checks, rather than agent preferences, keep systems working for their principals. Along the way they touch on AI lab IPO expectations, Nvidia, and the gap between how labs talk about safety and how they actually approach it.


📺 Source: a16z · Published November 16, 2026
🏷️ Format: Podcast

1 Item

Channels

4 Items

Companies