I wouldnt mind if the conversations actually were smart, not sycophantic, or were otherwise useful. I find more often than not it creates more work for me than it saves me, and even if i were to break even on time invested i lose massively on comprehension/understanding.
I'd rather see this technology being used as a method of input and control of the machine, just like keyboard, mouse and monitor is. Without humanized bits of conversation or suggestions, exaggerated remarks like "Got it!" - you would ask OS to preform a task and it would do what was told. And if I'd want to have a specific question or task I'd use some dedicated application that would stay dormant up until used.
Basically sums up why i don't use any kind of voice assistant still. Until the computer can DO exactly and precisely what i asked -- not what it's faulty recognition model thinks i asked -- there's zero point to trying to talk to it
I have one device in my house on Alexa called "under cabinet lights", I asked Alexa to "turn on the cabinet lights" she said "several things share the name cabinet lights, which one do you mean?" (As if there's enough information there to answer the question) I told her "all of them" thinking maybe the lights got duplicated or something.
She turned on every smart home light in the house.
