"The Big Short" is killing it! Michael Burry is sniping AI faith from the bottom of the technical ladder: When the parameter trap hits the capacity peak, the AI bull market welcomes the "liquidation moment"

Michael Burry, known as the "Big Short," has been posting pessimistic views on his Substack subscription platform, claiming that the end of the world is near. At a time when global funds are flocking to AI computing power infrastructure, he has been aggressively shorting popular AI technology stocks. The character Michael Burry, the prototype of the movie "The Big Short," has been focusing his recent short-selling operations on crowded trades related to AI computing power infrastructure and the capital expenditure cycle of AI semiconductors. For example, Burry has taken short positions on NVIDIA Corporation, Tesla, Inc., Micron, Applied Materials, and iShares Semiconductor ETF. This globally renowned top investor, who gained fame for accurately predicting the 2008 US real estate collapse and subprime crisis, is now making high-frequency short-selling operations on AI themes, which the market interprets as extremely crowded positions in storage chip themes, semiconductor manufacturing equipment, and AI GPU/AI ASIC chip related popular stocks. He is skeptical of the investment returns and overvalued fundamentals of tech stocks with large AI computing infrastructure capital expenditure. Previously, he disclosed his bearish positions on NVIDIA Corporation and Palantir through 13F filings. With the South Korean stock market plummeting again on Monday and popular tech stocks related to AI computing power experiencing intense volatility, some investors are beginning to agree with the narrative that the AI bubble is about to burst and are joining Burry's camp of short positions on AI themes. Burry considers SK Hynix and Samsung Electronics' unprecedented plans for massive storage chip expansion in South Korea as a critical signal that the AI investment cycle is "approaching the end." He views the Korean projects as a "peak capital expenditure signal" rather than an "immediate capacity signal," and calls it the "beginning of the end" of AI prosperity. His judgment is not that these factories will cause oversupply next year, but based on the classic semiconductor industry cobweb cycle: the most optimistic demand forecasts, the highest chip prices, and the most lenient financing conditions often lead all manufacturers to announce massive expansions at the peak of the cycle; as it takes several years from planning, land acquisition, power supply, clean room construction, semiconductor equipment import to mass production in advanced wafer fabs, the real supply peak typically comes when market demand has cooled down. Burry's recent short-selling operations on AI themes are not simply about saying "artificial intelligence has no value," but about betting on a dangerous misalignment between technology path, capital return, and market pricing. At the technological level, he believes the industry has mistaken language generation, the "intelligent output," for reasoning itself, continually improving model expression by increasing parameters, training data, and computational power, without necessarily building stable causal reasoning, world models, and long-term planning capabilities concurrently. At the economic level of capital return and market pricing, the industry operates on the assumption that "continuing to expand models can bring closer to general intelligence", investing billions of dollars in building chips, data centers, and power systems. If the marginal capacity improvement slows down and monetizable revenue cannot cover depreciation, electricity, storage, and financing costs, the entire AI infrastructure valuation system will face a rate of return downgrading. Burry characterizes this risk as the "wrong starting point" and the "parameter trap." His second logic is a typical reflexivity bubble and capital expenditure cycle: chip stock rallies reinforce the narrative of unlimited AI demand, making it easier for mega cloud companies to finance and expand their capital expenditures; the growing orders then raise profit expectations and stock prices of chip companies, further stimulating the supply chain to expand production. In Burry's view, this cycle appears to be demand validation in the upswing period, but some of the growth actually stems from the mutual reinforcement of asset prices, financing capacity, and corporate investment decisions, rather than the comprehensive cash flow realization of end AI. His recent short positions on NVIDIA Corporation, Micron, Applied Materials, iShares Semiconductor ETF, Tesla, Inc., and Carter's Incorporated actually cover multiple links in the AI capital cycle - core accelerators of AI clusters (AI GPU/AI ASIC/TPU), storage chips, semiconductor manufacturing equipment systems, the autonomous driving narrative, and capital goods related to data center construction, rather than targeting only a specific AI-related tech company.