> The people that aren't seeing a compelling use case simply aren't looking.

By all means, go do your startup, all the power to you.

But I personally fail to see it creating enough value for people to even bother use the AI, even if it's free and any problem could be ignored.

I can imagine people adopting it if it comes embedded on whatever software they already use. But that doesn't "completely change everything" or any of the other things people are repeating.

And yeah, I can't imagine it quickly improving so that happens either. I am personally bracing for a new AI winter, and believe that word is going to become more toxic than it has ever been.