The World Unpacked AI’s Biggest Skeptic Sees a Bubble
Nov 21, 2025
Ed Zitron, a tech commentator known for his critical insights on AI finances, shares his views on the current AI investment frenzy. He argues that the explosive spending could lead to a market crash, warning of potential risks to retail investors and the broader economy. Ed discusses the limitations of large language models and their unsustainable business economics. He critiques major tech firms' opaque AI revenue reporting and challenges the notion that advertising could save AI profitability. The conversation raises critical questions about the future of AI amidst this financial bubble.
AI Snips
Chapters
Transcript
Episode notes
Inference Costs Vastly Outpace Revenues
- Zitron estimates OpenAI spent roughly $12.4 billion on inference while generating far less revenue, implying severe operating losses.
- He highlights inference alone can cost two to three times reported revenues for leading labs.
Price Cuts Often Mask Investor Subsidies
- Many AI companies subsidize user costs, making apparent price drops reflect investor subsidies rather than true efficiency gains.
- Zitron warns a 'subprime AI' correction may occur when companies must price services at true marginal cost.
Ads Haven't Cracked AI Monetization
- Advertising has not proven lucrative for AI-first interfaces; Perplexity made only about $20,000 after enabling ads.
- Zitron argues ad models require replicability and brand safety that LLMs struggle to guarantee.