I would’ve said no. Then I spent last week in a hospital watching the opposite play out.
My friend was sick. Lots of unknowns, lots of tests, results that changed the picture every time they came back. To keep track, I fed everything into ChatGPT — every test, every result, every offhand comment from doctors and nurses. By the end it knew the arc of his case better than anyone who walked into the room.
Then, minutes before surgery, the anesthesiologist came in and asked if we were taking the biopsy to check for cancer. Cancer had never been mentioned. Not once. She’d walked in cold, assumed the wrong thing, and said it out loud in front of a patient who was already terrified. He heard “cancer” from a doctor right before going under. In that room, the human introduced the scary, wrong claim — and the thing holding the full, accurate picture was ChatGPT.
The roles felt flipped
But ChatGPT is a chatbot and she’s a doctor — that’s supposed to settle who you trust. The human’s the authority; the AI’s just a tool. Except in that room the roles felt flipped, and I’m not sure that line holds like it used to.
So I keep circling the same thing: we’ve always treated AI as the tool and the human as the authority you check it against. But when the AI holds more context, stays consistent, and never walks in cold — is that line still real, or are we just used to it?
Who do you trust more now — and are you sure you’d be right to?