We should probably just agree to disagree. I think the strides made in AI are at the very least impressive and have made some things (text-to-speech, for example) better - if not enormously then at least noticeably.

But there isn’t a true analog to be had between calculated probabilities and conscious thought. The former is a mimic of varied competence, but has no logic inherent to it. It requires human maintenance, it’s only path to “growth” if we want to call it that, is a black-box of infinite probabilities it calculates at incredible speed.

It’s a super-magic-8-ball that we choose to pretend has agency of some sort. But it does not.