Remember that what will work in the future must not make a lot of sense now. A lot of what makes sense (in naive first order logic) has already been tried --and failed. @rorysutherland