So, you are correct. There is no way for an LLM or AI application to generate the exact image inside your head. And nobody has ever seriously claimed that it can. I mean, maybe Sam Altman has said some stupid shit like that, because he is known for doing so, but realistically, nobody is arguing that this is currently possible.

There are some phenomenal AI works out there. I’ve recommended some channels that I really like, and I even do the role-playing thing from time to time to pass the time at work and whatnot.

But ultimately, yes, that is a limitation of the technology. I think we’re arguing against two completely separate things here.