Zach Stein-Perlman comments on Why not to solve alignment by making superintelligent humans?