What I see a lot of in my own industry (games) is that people that don't like AI either 1) see something that changes how they think of it and start the slow path of adoption or 2) continually dig their heels in and/or refuse to update their priors.

As technical qualms were solved and many other capability issues with AI have been solved, the continual heel digging and goal posting by group 2 has started to just feel more akin to a religious stance than something based on any specific fact. There is no version of AI that they want to see, they will oppose it on moral(-ish) grounds on sight, etc.

AI has real issues, but like you, observing this re-adjustment in realtime has been really interesting, especially as it seems to come from people I share 99% overlap in interest and politics with, but who hate AI like they do.

It also feels akin to de-radicalizing MAGA or something. The rejection or idea that they are wrong causes a whole worldview to collapse in something akin to ego death, so instead of facing that possibility they just insist they are right.