Math used in LLMs is perfectly reducible and Gödel have nothing to do with it - inside commonly used axioms (which sufficient for LLM to exist and outside of Kurt Gödel scope) there are ZERO questions/uncertainties how it works, it's just a fact :)