Paul Kedrosky: AI is the First Bubble With Every Ingredient at Once | #648
Today’s guest is Paul Kedrosky, a fellow at the MIT Institute for the Digital Economy, partner at SK Ventures and former sell-side analyst. In today’s episode, Paul Kedrosky explains why AI sits at the intersection of every force behind the biggest bubbles. He walks through why tokens are the fastest-deflating commodity ever, why more than half the data center buildout runs on debt instead of cash flow, and why an IPO wave pressures the market’s biggest winners. To close, Paul argues AI already drives most US GDP growth, and revisits how badly humans misjudge scale.
Key Points
- AI is experiencing unprecedented growth, contributing significantly to GDP and creating massive financial inflows into data centers, which are increasingly financed through external debt rather than internal cash flows.
- The rapid adoption and convergence of large language models are leading to deflationary pressures and potential oversupply issues in the semiconductor industry, posing risks to companies reliant on continuous exponential growth.
- The negative sentiment towards data centers and AI in the US is partly due to the perceived loss of personal agency and the unique tie between employment and healthcare, contrasting with more positive views in developing countries where AI is seen as an opportunity for advancement.
Chapters
| 0:00 | |
| 1:19 | |
| 3:14 | |
| 7:25 | |
| 10:45 | |
| 18:42 | |
| 22:26 | |
| 28:01 | |
| 34:04 | |
| 40:22 |
Transcript
Loading transcript...
- / -


