uh, no?
LLMs are useless. It's hard to emphasize this hard enough.
An LLM is a very simple system with a random number generator and a filter and when you combine them with some textual data the filter spits out the next most likely word
that's it, that's all an LLM does. There is no LLM in the world right now that can think a thought. There is no LLM in the world that can change anything about itself in response to the data it receives. It doesn't even predict multiple words at a time - it predicts one, then it just uses the one it predicted as part of the input to produce the next one.
The LLM is simulating the part of your brain that turns a thought into words. When you think of an apple and you want to say "red apple" you have a part of your brain that is trying to figure out the best word to use to describe the token another part of your brain has given it.
If you took the most advanced military AI in the world right now and put it in the middle of the woods it would immediately get lost and die.
If you took a cockroach and put it in the middle of the woods it would find a mate, reproduce, and live a full and happy life.
There is no AI in the world that is comparable to a cockroach. Every single AI is infinitely more retarded than a cockroach.
The AI technology we have now is fundamentally unable to moderate an imageboard on the most basic level.