David Woo pointed out in a podcast that the core risk of current AI trading is not short-term performance, but the underlying assumptions supporting the entire AI valuation system. He believes that Anthropic's valuation target of up to $2 trillion is built on the "winner-take-all" assumption, but AI is essentially commoditization and cannibalization, making it difficult to build a moat. Woo also questioned whether the current round of AI capital expenditure by tech giants is driven by "fear of missing out," and criticized Microsoft for "dressing up" its earnings figures through accounting treatments such as extending the depreciation period for AI data centers. He had previously been shorting AI-related assets and has temporarily exited to observe, but emphasized that this is only a tactical avoidance, not a change in judgment, and warned that once the AI narrative collapses, almost all investors holding the S&P 500 will face significant hidden exposure.