Not even the right answer much of the time, just information that sorta matches the right patterns as requested by the user. But LLMs say things in such an auth

Not even the right answer much of the time, just information that sorta matches the right patterns as requested by the user. But LLMs say things in such an authorative, intelligent seeming way that it's indeed easy to forget they have no idea what they're actually saying or how accurate it is at all...