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.
I even wonder if this behavior is due to next-token prediction architectures, somehow.
I know you probably don't consider it dense but wondering if someone can shed insight.
I find them like empty calories, like programming youtube tutorials. They maximize for feeling learnt instead of steady progress
It's dense along the wrong axis and verbose Along the wrong axis.
Like having acres of cardboard for dinner: you have too much dinner with not enough nutrition. Your dinner is plentiful and still not enough.