The entire AI jobs debate is a distraction. Anthropic's own Head of Economics admitted there has been 'no material increase in the unemployment rate to date.' Meanwhile, Anthropic's Economic Index and OpenAI's Economic Research Exchange exist to keep you focused on theoretical labor displacement while the actual financial story goes unexamined. AI coding tools are making developers slower. The productivity revolution is not showing up in data.
The real numbers are brutal. The entire AI industry, including OpenAI and Anthropic's cloud spend, generated roughly $110 billion in trailing twelve-month revenue. OpenAI alone raised $122 billion in March 2026. AI startups collectively raised $255 billion in Q1 2026. The industry is spending more on fundraising rounds than it earns in a year. Roger McNamee has called the investing structure a 'humongous bubble.' Venture capital's AI exposure is largely on-paper gains with no credible path to liquidity.
Ed Zitron's piece goes further than the headline math. He maps the compounding problem: as infrastructure costs rise, hyperscalers and neoclouds must charge more for compute, but the only customers who could absorb that cost cannot actually pay it. The longer the bubble inflates, the harder the unit economics become. Read this one for the structural argument, not just the numbers.
[READ ORIGINAL →]