Prof G MarketsThe AI Bubble Isn't Where You Think It Is
24 Oct 2026 · 13 minThe true AI bubble isn't in public tech stocks, but in the private valuations of AI startups like OpenAI and Anthropic, which are set to face a harsh reality when they attempt to go public.
Summary
This conversation argues that widespread fears of an imminent AI stock market crash are misplaced. While some AI hardware companies face challenges, their risks are already priced into their public valuations. Instead, the real danger lies in the massively inflated, unsustainable private market valuations of foundational AI labs, which are operating with fundamentally broken business models and are about to be exposed to public market scrutiny.
Interesting ideas
Public AI Stocks Are Not the Bubble. Despite significant price increases, public AI companies like Nvidia have seen their price-to-earnings multiples decrease, indicating that investors are already factoring in substantial risks. This counterintuitive point means that concerns about an AI stock market crash are likely overblown because the market has already discounted future uncertainties, unlike prior bubbles where valuations soared unchecked.
The Bubble is in Private AI Startup Valuations. The actual AI bubble resides in the private markets, specifically with foundational AI labs like OpenAI and Anthropic. Their multi-trillion-dollar private valuations are not supported by revenue or profitability, relying on 'adjusted' metrics that ignore core costs. This matters because these companies are poised to go public, and their private market 'playpen' valuations will clash with the public market's demand for genuine financials.
Private vs. Public Market Disconnect. Private market investors (VCs) prioritize vision and talent, leading to inflated valuations for AI startups, while public market investors (Wall Street) demand margins and profits. This fundamental difference means that when AI labs transition to public markets, their valuations could be cut by half or more, as evidenced by applying public company multiples to their current financials. This highlights a significant impending re-evaluation of AI company worth.
Questions worth thinking about
What specific accounting adjustments are being made by private AI companies, and what is the full extent of the costs they are effectively concealing?
If major AI labs delay their IPOs, how will this impact the broader private market funding landscape for other AI startups?
What would a significant re-rating of foundational AI lab valuations mean for the future availability and pricing of AI services?
Written by AI from the episode's full transcript. It can get things wrong: if something here looks off, tell us at hello@podpings.com.