> Aren’t LLM’s just predicting the next token with zero contextual knowledge of the world they inhabit Only partly true, especially with Q* learning coming ont

> Aren’t LLM’s just predicting the next token with zero contextual knowledge of the world they inhabit Only partly true, especially with Q* learning coming onto the scene as it is much more dynamic, able to self-learn.