Microsoft's head of AI doesn't understand why people don't like AI, and I don't understand why he doesn't understand because it's pretty obvious

Is it really “mindblowing” that people are skeptical of software that consistently doesn’t do the things we’re told it can do?