First point is plausable, moving from bigger models to smaller models. But the nature thing is a bit of an overstatement, yes our brains are very efficient but

First point is plausable, moving from bigger models to smaller models. But the nature thing is a bit of an overstatement, yes our brains are very efficient but they are fundamentally different from LLMs so it doesn't really map.