An observation of mine with regards to ardent Chomsky critics is that it's usually the case that the critic has a lot to gain from their attack or position while Chomsky has little to nothing to gain.

I think the whole stable diffusion and ChatGPT thing will come and go, almost as self-driving cars have. If you remember, just eight or so years ago, everybody was absolutely convinced that they were right around the corner. There was an insane amount of hype. Now, that hype has largely come and gone, with several companies abandoning their self-driving approaches or just flat out failing at them. There is still some hype, but I feel it has faded as people have come to grips with the premise, much less the difficulty of the implementations. The focus seems to be shifting back to driver assistance features.

Another thing is VR. So these stable diffusion and GPT methods are having their time in the sun. Yet again, just like VR and self-driving, proponents are trying to will these things into existence. Yes, they are impressive in some very specific cases, but they are not general. Just like stable diffusion, these GPT things will just flood the Internet with junk, either filling people with incorrect knowledge or driving people away, leaving the only remaining application to be advertising and trying to sell stuff.

Humans are seemingly absolutely obsessed with technology trying to replace humans and reality. That is the common thread between all of VR, self-driving, stable diffusion, and now ChatGPT. But this is both hard and destined for failure. People keep discovering that the role of technology is to augment humans and not to replace them. Engelbart knew this 60 years ago.