Learning to lose more arguments is the new flex