OpenAI unveiled 'ChatGPT for Clinicians' on Wednesday, offering it for free to every verified physician, nurse practitioner, and pharmacist in the United States. They claim their GPT-5.4 model scored a 59.0 on a benchmark called HealthBench Professional, supposedly crushing the human physician baseline of 43.7. The catch? OpenAI wrote the benchmark, OpenAI did the testing, and OpenAI is the one telling you that your doctor is obsolete.

This is the first step in the 'Cognitive Enclosure' of the medical field. By offering the tool for free, OpenAI is training its models on the collective expertise of the American medical workforce. Once the model is sufficiently 'fed,' the free access will vanish, replaced by a subscription model that hospitals will use to justify cutting staff.

We have already seen this in 'vibe coding,' where companies replaced hundreds of customer service agents with AI, only to have to hire humans back when the bots began to hallucinate. But in a hospital, a hallucination isn't a bad customer review—it's a corpse. OpenAI claims they reviewed 700,000 responses to ensure safety, but they are asking patients to trust an algorithm over a person with a pulse. It is the terminal deskilling of humanity, one free app at a time.