Plus these so-called “LLM”(s) have a pretty good grasp of the “shape” of language, so they appear to be quite intelligent or produce intelligible response (when they’re actually quite stupid really).