Knowing the Facts, and Knowing from Experience

A doctor can know everything about a patient and still not really know her.

Picture a woman at a follow-up appointment, her chart open on the screen beside her. Every number is in it, every scan, every lab result, the whole history of her disease laid out in order. By one way of measuring, she is completely known. And still that record does not know that she stopped going to church a few months back, when sitting through the service began to leave her too winded to sing. It does not know that she is frightened in a way she has never said out loud to anyone. It does not know that the hardest part of her whole week is the staircase in her own house, the one she used to climb without a thought and now has to stop halfway up, holding the rail, waiting for her breath to come back. None of that lives in the chart. All of it lives in her.

There is a difference between knowing all the facts and knowing from experience, and in serious illness the difference is often everything. The facts can be written down. What it is like to live inside the illness can only be told, and only by the person living it. More than a hundred years ago the sociologist Max Weber gave that second kind of understanding a name. He called it verstehen: grasping what something means from the person's own point of view, not from the outside looking in. The woman on the stairs is verstehen, and no one needs to know the word to feel the truth of it.

We are watching the rest of the world forget this even as we write. Companies everywhere are racing to put AI agents in front of people in the name of efficiency, and the results keep telling the same story. One recent study found that nearly three out of four organizations have already been forced to pull back an AI agent they put live with customers, and the rate was actually higher, not lower, among the companies with the most mature controls. Trying harder at the machine did not close the gap. It widened it.

The cost of that gap is not abstract. A grieving man once asked an airline's chatbot about bereavement fares after his grandmother died, and the bot confidently told him he could claim the discount after he flew. The airline's real policy said no such thing. When he asked for the refund the bot had promised, the company argued in front of a tribunal that it should not be held responsible for what its own chatbot said. A real person, at one of the hardest moments of his life, was failed by a tool built to be efficient and never built to understand him.

This is the pattern, repeated until it should embarrass us. A company throws out the human in the name of speed, makes its margin for a while, and then has to walk the whole thing back once it lands on someone real. What gets discarded as inefficiency turns out to have been the part that was actually working: a person who could hear what was meant and not just what was typed.

That gap, between what can be recorded and what can only be lived, is the whole reason Qualitative Intelligence Systems exists. In rare lung disease, where so much is still uncertain and close to ninety percent of the drugs that enter clinical trials never reach approval, the lived experience of patients is not a soft addition to the real data. It is data the record was never built to hold, and we believe that missing half is part of why so much of the work falls short. The numbers can tell you what is happening inside the body. Only the patient can tell you what it is like to live in that body, and what she most needs the next decision to protect.

We are not against the machine, and we are not against the numbers. We are against deploying either one as if understanding were optional. Both halves are real, and caring for a person means holding them at the same time, with a human in the room and the patient's own voice at the center, which is exactly where it belongs.