The rapid evolution of artificial intelligence (AI) in healthcare—especially large language models capable of generating clinically plausible recommendations—has fueled speculation about whether human physicians might one day be replaced in patient consultations. This chapter examines that possibility through the lens of authority theory and moral responsibility. Drawing on Robert Nozick’s “experience machine” thought experiment, Joseph Raz’s service conception of authority, and Stephen Darwall’s account of the second-person standpoint, I argue that the value of human–physician encounter extends beyond epistemic accuracy. While AI may eventually match or surpass physicians in diagnostic reliability and treatment planning, it cannot replicate the moral and relational dimensions of decision-sharing. The “simulation problem” shows that replacing physicians with AI systems leaves patients to bear full responsibility for their choices, even when the AI offers authoritative recommendations. However, for many patients, consulting a physician is not merely a matter of receiving expert advice but of sharing the burdens of decision-making with a responsible co-agent. Consequently, I contend that most people will continue to prefer human physicians, with AI functioning best as an adjunct that enhances care without displacing the interpersonal and moral aspects of the medical encounter.