The AI financing landscape is being discussed primarily as a scoreboard of deals, rather than through the lens of credit risk. While large sums are being moved and lent involving companies like Amazon, Broadcom, and Nvidia, the critical question of who ultimately holds the risk is often overlooked. This mirrors a situation observed by Michael Milken in the 1960s, when he discovered investors often misjudged the risk of high-yield bonds. Milken built the high-yield market, but Drexel Burnham Lambert ultimately failed due to liquidity issues, not bad bonds. He later faced legal consequences, but his subsequent work in cancer research has been highly impactful.Current AI financing structures, like Amazon's pitch for special purpose vehicles to finance Nvidia chips, rely on borrowed ratings and may face issues if assets depreciate faster than anticipated. Broadcom's substantial loan to Anthropic, coupled with Nvidia's backstop for debt financing, creates a similar interconnectedness among suppliers and demand. These deals are being priced against a backdrop of rising interest rates and the sheer volume of AI-related debt projected. A critical concern is who will act as a market maker for these new AI-backed financial instruments if the primary sponsors withdraw. Ultimately, the author believes that while AI demand is real, the flow of capital may be overlooking the crucial human element and the companies that will actually implement and benefit from this technology.