I'll explain why the vector/covector axioms for matroids over hyperfields constitute in some sense the most important axiomatization, why philosophically they are appealing, and why in practice they are maddening. Many conjectures aiming to ameliorate the situation will be given.