The films "Her" and "A.I. Artificial Intelligence" both explore the longing to be understood and loved.

Emotional reality, not objective reality, often determines whether a bond feels meaningful.

These dynamics now appear in real clinical conversations about AI companions and chatbots.

By Steven E. Hyler, M.D., a member of the Committee on Technology and Psychiatry.

When Robert Spitzer assembled the team that would produce the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III), artificial intelligence was not part of the conversation. The focus was on improving the reliability of psychiatric diagnosis. Personal computers were still a novelty. The DSM-III drafts were typed on IBM Selectric typewriters. Cable television was just being commercially established. The internet, as we know it, did not exist. Chatbots belonged in the realm of fantasy.

Fast forward nearly 50 years. Today, patients talk with artificial intelligence systems about their fears, relationships, and emotional struggles. Researchers are studying AI-assisted psychotherapy. Physicians are using AI scribes to document clinical encounters. Large language models are being evaluated as diagnostic assistants — tasks that would have seemed fantastical a generation ago.

Questions that once seemed absurd have become serious clinical discussions: Should a chatbot be permitted to provide psychotherapy? Can an AI detect suicidal ideation from speech patterns? What obligations does a clinician have when a patient prefers talking to a machine?

What is striking, in retrospect, is that Hollywood may have anticipated some of these conversations long before psychiatry did. Not because Hollywood accurately predicted artificial intelligence — most of the time it did not — but because many........

© Psychology Today