Also, Penrose doesn't conver if I recall correctly about modelling the quantum part too. It's just statistics after all.