Could the vision system not be the AI foundational model? Car vision, specifically Tesla state-of-the-art is not mentioned. I see bit of NLP bias.