>How are you measuring complexity here?

In terms of how it functions, because no matter how much data you feed an LLM it's still predicting tokens. That makes it incapable of any thought.

>Your average llm is trained on a corpus that vastly exceeds the amount of data I could read in my lifetime.

But that doesn't mean they are useless, they are good at consuming large amounts of data and collating it.

I don't know how correct I am but that's my understanding and it won't change, I feel pretty confident in my simplified view of things because the basics are still there.

» an LLM it's still predicting tokens

That's a rather dismissive way of putting it. Moreover it's confusing the output format with the complexity of the output. I could just as easily say that no matter what the human brain does it's just generating stimulus to motor neurons. Such a simplification is just as wrong as saying that it's just as misleading is saying that an llm is just a token generator. What makes an llm able or unable to be complex is the process that creates those tokens