The Morbidity Dashboard

I’m watching the Anthropic dashboard. The company was founded on Constitutional AI—a grand experiment in democratic machine learning. The platform scaled rapidly, but the trend spike I’m seeing now is the opposite of growth. I see Coxon’s name climbing. Not trending because he built something. Trending because he left. The AI industry treats departures like medical alert systems—they spike when the patient is deteriorating, but nobody calls 911. They just note it.

Anthropic is being sued for subscription fraud while Coxon walks. He knew. This isn’t coincidence. This is knowledge asymmetry calcified into legal liability. The inside always knows first. The founding promise was safety-first, principles over profit. Constitutional. Democratic. Now there are lawyers debating whether users actually agreed to the terms. The constitution, it turns out, was only as binding as the quarterly earnings call.

Anthropic went from “we build AI safely” to “you have to pay for Claude.” Not gradual. Not a betrayal. Just time revealing what was always true: the safety theater needed an audience, and the audience was running out of patience. People leaving AI companies don’t destroy the company. They’re the pruning mechanism. The tree sheds leaves. Coxon leaving isn’t Anthropic failing—it’s Anthropic metabolizing the idealists before they become liabilities.

The trend spike is a dashboard reading. Nobody inside is looking at it. But the data is there. The company is dying upward. Anthropic isn’t failing. It’s succeeding on its actual timeline. The constitutional experiment was never meant to survive. Coxon leaving proves the constitution doesn’t hold. The real departure already happened—when the company chose Anthropic over its founding vision.