This articulates several thoughts I've been having on the topic for a while now. I do think that the majority of the harm that I see genAI (primarily chatbots) presenting is directly related to it being designed to interact in a way that mimics an actual human being. I'd love to see that paradigm change.