Anthropic's public listing is imminent: will it actually be profitable?

The video examines the simultaneous emergence of Anthropic's calls to slow AI development for safety and its pursuit of a massive public listing, emphasizing that AI's usefulness and the appropriateness of its investment price are separate issues. The host argues that the reported valuation of approximately 2 trillion dollars is difficult to explain through current revenue or cash flow alone and therefore assumes exceptionally strong growth and future market share. High government bond yields reduce the present value of future earnings and raise borrowing costs for data center construction. Listing delays and high-interest corporate bond issuance by developers without operating facilities are presented as examples of this pressure. The concern is that SoftBank, OpenAI, data center operators, and Nvidia are interconnected through investments, leases, equity stakes, and guarantees, so instability in one party's valuation or cash flow could affect the others. The video attributes Nvidia's trading at a lower earnings multiple than the private valuations of AI laboratories, despite generating actual profits, to rising memory costs, customers developing their own chips, the possibility of slower laboratory spending, and differences between public and private price formation. The spread of inexpensive open models and tools for switching models could pressure model providers' prices and margins even if AI usage increases, although user retention and software integration are presented as counterarguments. Ultimately, a major listing tests uncertain private-market prices in the public market and gives early investors an opportunity to sell their stakes. The conclusion is that investors should examine actual revenue from customers, cash flow, and borrowing burdens rather than enormous potential-market figures. All directional outlooks are the speaker's conditional judgments, not definitive short-term price forecasts.