Parapsychology: the control group for science — LessWrong