Why try to argue against the vocabulary on this one?

Of course it’s artificial intelligence.

We’ve used “artificial intelligence” to mean everything from bots in games to Deep Blue, to protein folding and ML predictions, and those are a whole hell of a lot more specialized than llms.

Modern llms pass the Turing Test for a LOT of people.

Are there lots of problems with it as it is now? Absolutely yes. Environmental, ethical, economic, and on and on and on. All this on top of the fact that it quite bad at many many tasks. But if it’s not artificial intelligence, you’re just moving the goal posts.