Every few weeks, another headline asks whether artificial intelligence is about to replace physicians. I doubt it. Patients don’t want to hand the most important conversations of their lives to a computer. Doctors didn’t spend years training to become little more than supervisors of software. Medicine has always been, and should remain, a profoundly human profession.
But I do think we’re asking the wrong question. The real question isn’t whether AI will replace doctors. It’s whether doctors will slowly begin relying on AI to make more of the decisions that have traditionally depended on independent medical judgment. Those aren’t the same thing.
As a medical malpractice attorney, I spend my days looking backward. My job isn’t to decide how medicine should evolve. It’s to understand what happened after something went terribly wrong. That process almost always begins with the same questions. Why wasn’t another test ordered? Why was the patient discharged? Why was an abnormal finding dismissed instead of investigated?
Behind every one of those questions is another one that’s just as important. Who made the decision?
Real physician voices, twice a week
Free, and one click to unsubscribe.
For generations, the answer has been straightforward. A physician reviewed the information available at the time, exercised professional judgment, and made a clinical decision. Sometimes that judgment met the standard of care. Sometimes it didn’t. Either way, there was no mystery about who was responsible for the choice.
Artificial intelligence is beginning to complicate that picture. Not because hospitals are replacing physicians with algorithms. Because algorithms are becoming part of the decision-making process itself.
An emergency physician may receive an AI-generated assessment of a patient’s risk of sepsis before walking into the room. A radiologist may review software that highlights suspicious findings on a CT scan. A cardiologist may see recommendations generated from thousands of patients with similar histories before deciding on the next step.
None of that strikes me as inherently concerning. In fact, much of it is exciting. If technology helps physicians recognize a stroke sooner, detect cancer at an earlier stage, or identify subtle abnormalities that might otherwise be missed, patients stand to benefit enormously.
Medicine has always advanced because doctors embraced better tools. The stethoscope was once new. So was the MRI. Electronic medical records fundamentally changed how physicians practice, even though many of us can still remember when paper charts lined every hallway. Artificial intelligence may prove to be another step in that evolution.
The question is not whether physicians should use it. Of course they should. The question is whether they will continue to challenge it.
That’s where medicine has always distinguished itself from mathematics. Medicine isn’t simply about reaching the statistically most likely answer. It’s about treating the person sitting in front of you.
Patients don’t always read the textbook. Symptoms don’t always follow predictable patterns. Sometimes the diagnosis everyone expects turns out to be the wrong one. The best physicians know that. They’ve always known that.
Experience teaches doctors to recognize when something doesn’t quite fit, even if they can’t immediately explain why. It teaches them when to order one more study, ask one more question, or resist the temptation to accept the obvious answer simply because it’s the easiest one.
I hope artificial intelligence strengthens that instinct. I worry if it weakens it. That’s not a criticism of AI. It’s simply an observation about human nature.
We all become comfortable with tools that consistently help us. Few of us memorize phone numbers anymore. Most of us couldn’t navigate across an unfamiliar city without GPS. Spellcheck catches mistakes we never bother looking for ourselves. Technology changes habits long before it changes professions. Medicine won’t be any different.
That’s why I believe physicians, hospitals, regulators, and patients should be having a broader conversation today, not about whether artificial intelligence belongs in health care, but about how we preserve independent clinical judgment as these systems become more capable. AI should inform decisions. It should improve decisions. It should challenge assumptions and reduce preventable mistakes. But it should never become a substitute for the thoughtful skepticism that defines outstanding physicians.
The practice of medicine has never been about following instructions. It’s been about knowing when the usual answer isn’t the right one.
If artificial intelligence helps doctors ask better questions, it may become one of the greatest advances modern medicine has ever seen. If it encourages physicians to stop asking questions altogether, we’ll have gained remarkable technology at the expense of something even more valuable.
The future of medicine will almost certainly include artificial intelligence. My hope is that it never comes at the expense of independent medical judgment.
Patients deserve both.
Sean C. Domnick is an attorney.
