How does your theory account for the Eliezer Yudkowsky type person, who clearly shows no love for any of the labs or the current progress, and yet is very much pro-"AI safety"?