This paper will examine the systemic risks of the Artificial Intelligence market upswing of 2022-2026 and the economical mechanism underlying it. While Generative AI is very useful technologically, its present market valuation seems to have broken away from its realized revenue, supported by a "circular economy" of infrastructure investment. This paper will analyze "round-tripping" of capital flow between cloud hyperscalers, hardware manufacturers, and AI startups. It will also study the systemic risks imposed by debt-financed data center expansion, the rapidly evolving "patience deficit" among investors, and the physical resource limitations that will serve as a ceiling for further expansion-specifically, shortages of energy and materials (such as silver and high bandwidth memory). Ultimately, we'll compare this market cycle to the 2000 dot-com bubble. We argue that while a bubble exists, its nature and origin have led to tangible assets rather than "vaporware". It seems likely that an inevitable "deflationary correction" will be needed for market value to stabilize with unit economics.