Are We in an A.I. Bubble? We're About to Find Out.

Max Reed opens with the stakes: Anthropic going public is the first real look inside the AI boom, and it could reshape the economy, tech, or humanity. The company might raise a hundred billion dollars or trigger a market crash.

Erin Griffith explains that Anthropic is historically significant because it's the first time investors get to see real numbers on a large language model company's business model.

Erin breaks down the S-one: eighty billion dollars in net loss is actually an accounting quirk; real operating losses are six to seven hundred million and shrinking. But the company is still locked into massive non-cancellable data center contracts.

Revenue exploded to sixteen-point-three billion in the first half of twenty-twenty-six with insane growth rates, but early Q3 numbers show a sharp slowdown as competition from OpenAI catches up and customers pull back on token spending.

Erin and Max read actual quotes from the S-one, including language about catastrophic risks to humanity, model welfare, and AI safety concerns—all written in passive voice so Anthropic doesn't directly own these statements.

Matt Levin and Joe Weisenthal from Bloomberg highlighted the absurdity: Anthropic wrote that concerns about model welfare may lead them to delay deprecating old models—implying they think their language models are sentient.

Erin explains Anthropic's five classes of stock, founder voting control at 50.1 percent, and a secret list of approved successors. It's designed so founders can never lose control, reflecting their belief they're uniquely qualified to shepherd AI safely.

Jasmine Sun explains her reporting: data centers are driving local elections and recall campaigns, but in statewide races, they rank twenty-second in voter priorities. It's a municipal issue, not yet a national one.

Mike Isaac explains the massive PHP vulnerability that broke internet infrastructure, and how AI made finding it easy. But Anthropic—the company supposedly best at AI safety—got hacked by a startup crew, showing even AI-forward companies are vulnerable.

Erin notes the grim irony: Microsoft is offering voluntary severance to twenty-six percent of staff, framed as 'giving employees more control.' It's part of a pattern of tech companies bloating during hype cycles and then cutting back.

Erin reports on how one of the biggest crypto investors is quietly dropping 'crypto' from their fund name, rebranding as 'frontier' to chase AI instead. It's a sign of just how out of favor crypto has become.

Jasmine notes the weird contradiction: Anthropic was founded partly as a nonprofit question, employees go through tests asking if they'd work for zero equity, founders pledged 80% to charity—yet they're pursuing a 2 trillion valuation IPO.