x
Relational Harm: A Therapist's Perspective on Conversational AI — LessWrong