Front Burner If AI is a bubble, what happens when it pops?
29 snips
Nov 17, 2025 
In this discussion, Paul Kedrosky, a venture capital partner at SK Ventures and a research fellow at MIT, dives into the precarious state of AI financing. He highlights the alarming shift from cash-flow funding to private credit, which skews risk assessments. Paul explains how vendor financing and circular cash flows mask real financial exposure. He also addresses the fears stemming from potential government backstops and outlines why the current AI boom echoes past financial crises. His insights suggest that the implications of a bubble burst could severely ripple through the economy.
AI Snips
Chapters
Transcript
Episode notes
Bubble Is In The Infrastructure Spend
- AI isn't a useless craze; the bubble is in the overspending to build AI infrastructure like data centers.
- That build-out's scale and cost create financial risk separate from the technology's value.
Shift From Cash To Debt Financing
- Data-center build-outs shifted from being funded by profitable tech cash flow to heavy debt financing this year.
- That debt dependence magnifies systemic risk because companies dodge balance-sheet exposure using special-purpose vehicles.
Circular Vendor Financing Masks Risk
- Vendor-style circular financing hides real exposure when suppliers lend money expecting customers to buy their gear.
- That circularity masks growth and fuels further fundraising and debt accumulation.