This essay resonated with me, speaking about the "mediocrity" of our future relationship with computer assistants.

I loved the paragraph about feeling "scammed" though I would've called it being "faked". The AI doctor can never use the stethoscope around her neck. She is hijacking totems of professionalism to appear more comforting, without the capabilities to back them up. The fake veterinarian can suggest a diagnosis but can't actually treat anyone. The real-estate chatbot cannot try to help a domestic violence victim.

Maybe that's why I interact with AI assistants the same way I interact with psycopaths. I'm comfortable interacting with them in jobs where the law will incentivize them to behave well. But for things like teaching, or medicine, or personal matters, I prefer someone with empathy.