The entire project of 'formalizing' morality is futile and misguided, though certainly well-intentioned. Imagine trying to list--extensionally--the necessary and sufficient features of something as basic as a dog or cat. You can get most of them, but most is not all (see Doug Lenat's "Cyc" project...). Fundamentally these attempts at formalization rest on a simplistic understanding of the meaning of concepts and rules and a lack of understanding their inherent relation to acceptable patterns of social behavior (e.g., 'norms'). If ML people spent more time reading Wittgenstein (the later one, not the Tractatus one) and social theory (Habermas, Giddens, etc.) they might better off than trying to square the circle. But I suppose academic specialization keeps these ideas separate. And papers need to be published. Not saying these attempts are pointless, but they lack a basic self-reflexivity regarding their own limitations and basic assumptions.