Aug 13, 2026 · 33 min
AI Compute Meets Wall Street’s Circularity Problem
“AI Compute Futures” — Has Wall Street Gone Too Far?
The episode connects a speculative market for AI computing power with inflation, inequality, weak hiring, and uncertainty around OpenAI.
- 1AI compute could become a tradeable financial product, but standardization and collateral risks complicate the comparison with commodities.
- 2Mark Zandi says persistent inflation, unequal purchasing power, and weak job creation leave the U.S. economy fragile.
- 3OpenAI’s senior departures add another signal of uncertainty for a company already central to AI markets.
Don't miss
Mark Zandi grades the still-growing but fragile U.S. economy C-minus, citing insufficient job creation and persistent economic discomfort.
The brief
Rohit Krishnan examines whether AI computing power can be standardized and traded like oil, electricity, or steel—and whether financial markets are ready for that leap.
The harder question is financing: Nvidia’s arrangements with financial firms and AI companies could spread risk, but they may also create circular structures that make demand look stronger than it is.
Ed Elson revisits Enron’s failed bandwidth market as a warning, while Krishnan argues that compute differs from bandwidth in ways that could make the analogy incomplete.
Mark Zandi finds inflation still above the Federal Reserve’s target, purchasing power under pressure, inequality entrenched, and job creation too weak; his grade for the economy is C-minus.
The episode closes with Brad Lightcap’s departure from OpenAI, set against an unusually large wave of executive exits that could matter for the company and its markets.
Featuring
Mark Zandi
Nvidia Corporation
CoreWeaveListen to the full episode and explore every guest, topic, and moment on PodLume.