A recent New York Times article, titled "When Doctors Use a Chatbot to Improve Their Bedside Manner," should raise questions about why physicians are turning to artificial intelligence for help talking with other humans. While GPTChat can generate things to say, what comes out of AI is impersonal, as it knows nothing about the individuality of the doctor asking them, or of their patient, or of the relationship between the two. Much of the joy of being a physician is forming personal, healing connections with patients. Are physicians unprepared to cultivate them? US Medical schools now teach physician-patient communication, with the help of standardized patients and various acronyms like "PEARLS" and "SPIKES," that are designed to guide clinician-patient interactions.  But are we failing to help physicians find their own voice -- specifically, to form personal, relationship centered connections that they can draw on, especially during challenging times?