NYTimes readers debate how doctors should use AI in training

In short: New York Times readers are split on how AI should fit into medical training, with many stressing that doctors must double-check AI suggestions.

What's going on

The New York Times published reader letters responding to an essay by Dr. Rachael Bedard, a geriatrician and palliative-care doctor. The print headline was “A.I. Has Made Me a Better Physician,” and the online title was “Being a Doctor Will Never Be the Same After A.I.” In the essay, Bedard said AI can help experienced doctors widen their thinking, but she worried that medical trainees could lean on it before they build strong judgment.

Several letters supported using AI as a helper, as long as doctors verify what it says. One physician wrote that AI can quickly surface useful information, but clinicians still need to question it and check it. The idea is similar to using a GPS, it can suggest a route, but you still have to watch the road and notice when it is wrong.

Other readers pushed back on the idea that a training gap should be a reason to hold back AI. One letter argued that differences between experienced attending physicians and residents are normal in medicine, and that residents close the gap as they gain experience.

A key point across the responses was the same tension Bedard raised. If trainees “outsource” their thinking to a tool too soon, they may not develop the habits needed to judge whether an answer makes sense. Bedard mentioned using a tool called OpenEvidence for clinical questions, while emphasizing the need for independent reasoning.

What to watch

These letters do not settle whether AI improves patient care overall. The practical question is how hospitals and medical schools teach doctors to use AI like a second opinion, not a final answer, and how they test that trainees can still think through a case on their own.

Source: NYTimes