Ed Zitron: The AI Bubble is Bleeding Cash, Here Are The Receipts

Ed Zitron, author of the Where's Your Ed at newsletter and host of the Better Offline podcast, joined Monetary Matters to lay out the case that the artificial intelligence industry is not a revolution but a financial bubble sustained by opaque accounting, circular capital flows, and a quasi-religious belief in future breakthroughs. The core of his argument is not that large language models are useless, but that the economics of the companies building them are fundamentally broken and unlike anything seen in previous tech booms.

The Unprecedented Scale of the Losses

Zitron’s central claim is that the losses at the two leading AI labs, OpenAI and Anthropic, are historically anomalous. He obtained access to OpenAI’s 2025 financials and found a net loss from operations of roughly $21 billion on $13.07 billion in revenue. The company spent $34 billion total, with $5.73 billion alone going to sales and marketing—more than Coca-Cola’s annual budget. Zitron argues this category is suspiciously inflated, likely containing the cost of free compute credits given to customers or the cut Microsoft takes for selling OpenAI’s models, rather than traditional advertising. He contrasts this with the dot-com era: Uber burned roughly $32 billion total before becoming messy profitable, and Amazon Web Services cost about $55 billion in cumulative CapEx over 14 years. OpenAI has raised $1