Since LLMs are improving using data out in the open, do you think humanoid robots will also improve in a similar way, doesn't that put our privacy at risk ?