How do we know when someone knows us? Does it matter whether the knower is a human or a machine? Following the theory of interpersonal knowledge, a between-subjects experiment investigated whether a doctor’s incorporation of individualized knowledge about a patient’s social or medical history enhances doctor-patient relationships in online conversations. Patients in this study conversed with either a human doctor, an AI doctor, or an AI-assisted human doctor. Following previous research, additional factors such as perceptions of effort, relational closeness, privacy intrusiveness, and the provision of privacy control were assessed. Results showed that an AI doctor enhanced patient satisfaction when it employed social individuation messages, which triggered perceptions of increased effort, but only when patients could activate privacy control. Perception of relational closeness with a human doctor and an AI-assisted human doctor did not seem to require social individuation and privacy control. The study concludes with implications for the theory of interpersonal knowledge and AI-mediated communication research, as well as practical implications for improving chatbot medical systems.

中文翻译：

---

当人工智能医生变得个性化时：社会和医疗个性化对人类和人工智能医生的影响


我们怎么知道有人认识我们？认知者是人还是机器重要吗？根据人际知识理论，一项受试者间实验调查了医生结合患者社会或病史的个性化知识是否可以增强在线对话中的医患关系。这项研究中的患者与人类医生、人工智能医生或人工智能辅助的人类医生交谈。根据之前的研究，还评估了其他因素，例如对努力的看法、关系亲密程度、隐私侵犯程度以及隐私控制的提供。结果表明，人工智能医生在使用社交个性化消息时提高了患者满意度，这会引发人们更加努力的感觉，但前提是患者可以激活隐私控制。与人类医生和人工智能辅助人类医生的关系亲密感似乎并不需要社会个性化和隐私控制。该研究的结论对人际知识理论和人工智能介导的沟通研究产生了影响，并对改善聊天机器人医疗系统产生了实际影响。