to expand, since I think you misunderstood.
"doesn't contain something that's not part of the question asked"
I said it provided lots of redundant information. It, actually (apologies to Dr. Bender for the 'splainerism - truly), contained lots of extraneous, unnecessary information that wasn't needed to answer the question. If it was using theory of mind to formulate its response, it wouldn't have done that as it would have assumed enough common ground of knowledge to find the relevant information at the root of your question and provide that without the extraneous details.
No person using the process of meaning making this thread highlights would respond like that to a question. The LLM is doing something fundamentally different.
