Patient trust still depends on time, not just technology
Trust sits at the center of the patient-physician relationship, and in this discussion it emerged as the organizing principle behind burnout, misinformation and the growing use of AI in care. Gary Price, M.D., MBA, president of The Physicians Foundation, argued that the human side of medicine remains the part patients notice first: eye contact, a seated conversation, active listening and the sense that their clinician has time to absorb what matters to them. Those are not soft skills in the usual sense. In practice, they shape whether a patient accepts a diagnosis, follows a treatment plan and returns when symptoms worsen.
The pressure on that relationship is rising as more patients turn to AI before they ever enter an exam room. Price described a familiar problem: patients often ask AI tools highly filtered questions that reflect fear, bias or incomplete understanding, and the output can reinforce mistaken ideas rather than correct them. Clinicians, by contrast, are trained to interrogate the answer, compare it with the full clinical context and recognize when a polished response still misses the diagnosis. AI can rapidly retrieve recent publications and compress search time, but it cannot reliably replace the judgment that comes from history-taking, physical examination and a face-to-face encounter.
A single question asked in isolation can distort the whole clinical picture.
AI may be fast, but it does not know the patient in front of it
Price’s comments drew a sharp line between information retrieval and clinical reasoning. A system can surface references or summarize a topic quickly, yet the real work of medicine depends on knowing which symptom is dominant, which fear is driving the conversation, and which detail in the history changes the differential diagnosis. That is why patient skepticism about using AI alone to diagnose or recommend treatment is not misplaced. Physicians are skeptical too, not because they reject technology, but because they are trained to test outputs against reality.
The most consequential limitation of AI in this setting is not merely factual error. It is the absence of relationship. A model can generate a plausible answer, but it cannot notice hesitation, track the meaning of a pause, or see that a patient is minimizing symptoms because of embarrassment or fear. Those cues often determine whether a clinician uncovers depression, medication nonadherence, substance use, or a serious physical illness that the patient had not planned to mention. In that sense, the value of direct interaction is not sentimental. It is diagnostic.
Price also raised the possibility that some surface behaviors could eventually be mimicked by AI, including smiling language, eye contact or conversational rhythm. The deeper question is whether mimicry can substitute for actual human presence. Clinical care often depends on the patient sensing a real person who is accountable for the decision made in the room. That credibility is built through consistency over time, not through a single polished exchange.
Stigma around mental health care still alters physician behavior
The interview also addressed physician burnout and the continuing stigma attached to mental health treatment. Survey findings cited in the conversation showed that 71% of physicians agreed stigma remains around physicians seeking mental health care, while 36% said they feared, or knew a colleague who feared, seeking care because of licensing, credentialing or insurance questions. Price pointed to the work of the Dr. Lorna Breen Foundation as a major driver of change, with many states now removing licensing questions that suggested mental health treatment could threaten a physician’s license.
Progress is real, but it is incomplete. The pressure on physicians did not disappear when the pandemic eased, and the burdens that drive exhaustion, moral injury and isolation still sit inside many practices and health systems. Even where formal barriers have fallen, cultural ones remain. Physicians still work in environments where mental health care can feel compartmentalized from “regular” medical care, and some institutions effectively require clinicians to seek help inside the same system where they work, which can deter them from opening up. The result is a system in which asking for care can still feel professionally risky.
This is not a peripheral workforce issue. When physicians avoid care, the consequences can extend to decision-making, attention, empathy and retention. Burnout does not only harm the doctor; it affects the patient encounter, the safety of the plan and the quality of communication that carries a patient through uncertain treatment. The connection between clinician well-being and patient trust is direct. A clinician who is depleted has less bandwidth for listening, and a patient who feels unheard is less likely to engage fully with treatment.
Listening remains a clinical skill, not a courtesy
Price described formative experiences from training that shaped his own approach. He recalled mentors who modeled behaviors that are easy to observe but hard to institutionalize: touching each patient on rounds, sitting beside the bed, making eye contact and allowing enough time to listen. He also described an elective focused on patient interview and active listening, beyond the required curriculum. Those lessons matter because they translate compassion into a repeatable clinical method. Sitting down and listening changes what a physician hears, and what a physician hears changes the plan.
That approach has a sociological dimension as well. Price linked medicine’s interest in connection to work in sociology that treats human relationship as an object of study rather than an undefined virtue. The implication is important for modern practice: connection can be taught, reinforced and measured indirectly through patient experience, adherence and outcomes. If medicine treats listening as optional, it becomes easy to replace it with speed. If medicine treats listening as a core clinical function, then workflow, scheduling and documentation have to be built around it.
For physicians practicing in an era of AI tools, that may be the most practical takeaway. Technology can help with search, summarization and administrative work, but it cannot carry the emotional and cognitive labor of a difficult conversation. Patients still need a clinician who can hear what is said, detect what is left out and translate both into care. The unanswered question is not whether AI will continue to spread through medicine. It already has. The more immediate question is whether practices will protect enough time for the part of medicine that no machine can convincingly replace: sustained human connection at the point of care.