I don't use LLMs in my workflow frequently. When I do I have a hard time making sense out of the very specific and long answers to my questions. Especially if I

I don't use LLMs in my workflow frequently. When I do I have a hard time making sense out of the very specific and long answers to my questions. Especially if I don't know the answer it is hard to figure out if the long answer of the model indicates the right direction or misses my point completely. Maybe I'm not knowledgeable enough. But asking questions I already know the answer of has no real life use case other than testing the model. Which of course might be a valid use case for some. Having a way to let the user specify his own level of knowledge might help to receive answers that are better tailored to the user asking the question.