Why a chatbot might seem more empathetic than a human physician
An article in JAMA Internal Medicine, “Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum,” generated a great deal of discussion — much of it horrified. In particular, people are focusing on the study’s conclusions: that “chatbot responses were longer than physician responses, and the study’s health care professional evaluators preferred chatbot-generated responses over physician responses 4 to 1. Additionally, chatbot responses were rated significantly higher for both quality and empathy, even when compared with the longest physician-authored responses.”
While this study is thought-provoking and has generated much interesting discussion, that conversation isn’t focusing on the real root of the problem in physician-patient communication: time.
When I read the article, I found myself thinking about a recent conversation with my extended family of nonphysicians. I explained why doctors have hours of work after we’re finished seeing patients. “What is it that you have left to do when all the patients have gone home?” they asked. It has to do with electronic medical records, I told them — more specifically, something called the EMR in-basket. Unfortunately, most of us have no dedicated time in the day to respond to them.
All of this provides important context to the recent uproar over the “empathetic chatbot” research. Before we rush to accept that generative AI is more empathetic than human physicians, let’s take a moment to dive into the details and methods of the study. It is important to note that this was not a comparative study of “chatbot versus human” in real-world conditions, i.e., the EMR. The real heart of the matter isn’t whether generative AI truly has the human capacity of empathy (it doesn’t), but how the shortage of physician time and corresponding high rates of burnout affect physician empathy...Read more
How can an AI chatbot truly compare to a human physician?
