LLM is still too verbose, a real person Socratic conversation can interact a couple sentences at a time, not spew 1-3 windowfuls of low density bullet points.

I even wonder if this behavior is due to next-token prediction architectures, somehow.

There are ways to ground an LLM to be concise, and Socratic (method of inquiry, back and forth dialogue)