yeah but taking what lecun says then training an AI on that special skill set to prove him wrong is not exactly proving him wrong because you are just missing the bigger picture, just like LLMs are