The $5.7 Trillion Bubble Had a Single Point of Failure—And It Was Wearing a Leather Jacket


(SeaPRwire) – Three weeks. That is all it took for Jensen Huang to go from preaching AI’s gospel in a converted West London church to resigning as Nvidia’s CEO. In October, he was shouting at founders in the crowd—”I’m gonna invest in you!”—the high priest of a $5.7 trillion faith. Then came the reports: U.S. intelligence agencies believed Huang had been involved with Alice Huang (no relation), a former executive at Singapore-based Megaspeed, a data center company that had bought billions of dollars in Nvidia chips, and who was allegedly working as a Chinese intelligence asset. Within a week, Huang had emailed Nvidia’s 42,000 staffers admitting he had been “unfaithful” to his wife. The board opened an independent investigation. Congress summoned him. Federal agencies circled. When the New York Times reported a second alleged affair—this one with an Nvidia executive who reported directly to him—the CEO of 33 years was gone by the next morning. CFO Colette Kress stepped in as interim CEO. No reason given.
Here is what makes the timing dangerous. Huang was not just a chip salesman. He was, as some characterized it, the central bank of AI. Nvidia had poured nearly $50 billion into leading AI model companies as of the end of July, with total equity stakes across the sector roughly double that. It pledged $170 billion in financing commitments, including $105 billion in guarantees to OpenAI for a single Ohio facility. It was even in talks with Wall Street lenders to backstop a $500 billion plan to turn AI chips into tradable securities. Critics had warned for years this was a circular machine—Nvidia funds the labs, the labs buy Nvidia chips—masking weak demand and inflating a bubble. Everyone assumed the crash would come from the balance sheet. Nobody priced in the founder. Meanwhile the core business holds, for now. Chips are sold out and contracted through next year. But customers were already building rival silicon, and governments were chasing sovereign AI before any of this broke.
So what is the end-game? Governance expert Francine McKenna argues Nvidia survives “independent of his cult of personality.” Seaport’s Jay Goldberg counters that the company risks losing Huang’s appetite for big, sometimes failing bets—without which it becomes “just another chip company.” Both can be true. The machine runs without its priest, but the faith does not. The AI sector has never proven it can thrive without Nvidia, and Nvidia has never proven it can thrive without Huang. Watch the $170 billion in pledged commitments. If Kress starts quietly renegotiating those guarantees, the circular web unwinds—and every neocloud and model lab that leaned on Nvidia’s credit becomes the next headline. That is where the real contagion starts, not in the resignation letter.