Can a Chatbot Be Your Chatdoc?

Vulnerable individuals have taken their own lives while in relationships with chatbots. Not only did the chatbots not discourage them from taking their lives but they also actively encouraged it. Another man suffering hallucinations took the life of his mother when a chatbot affirmed his delusions. Yet people continue to insist that chatbot therapists and chatbot doctors are a good thing. I acknowledge that generative artificial intelligence (AI), mimicking human rational capacities, will be better than humans at some things. Drawing on the work of Martin Buber in I and Thou, I show that an AI can only treat reality as an It, as an object to be manipulated, and cannot actually enter into an I-Thou relation the way that humans do, and I further argue that when caught up with another, there is a vicarious experiencing, a shared experience, of what I will call Thou-Thou relating. An AI device cannot be pulled into this Thou-Thou form of relation; it cannot encounter a human as Thou, because it is not a subject and lacks affectivity that enables shared experience. It is this vicarious fellow-feeling that often prompts the doctor or the therapist to act.