Once again demonstrating that this puritanical push to purge all “bias” from machine learning models is a fool’s errand, and a transparently ideological one at that.