Clinton Castro

Challenging, with several powerful arguments, some of our deepest beliefs about rationality, morality, and personal identity, Parfit claims that we have a false view about our own nature. It is often rational to act against our own best interersts, he argues, and most of us have moral views that are self-defeating. We often act wrongly, although we know there will be no one with serious grounds for complaint, and when we consider future generations it is very hard to avoid conclusions that most of us will find very disturbing.

Ethical concepts are, or purport to be, normative. They make claims on us: they command, oblige, recommend, or guide. Or at least when we invoke them, we make claims on one another; but where does their authority over us - or ours over one another - come from? Christine Korsgaard identifies four accounts of the source of normativity that have been advocated by modern moral philosophers: voluntarism, realism, reflective endorsement, and the appeal to autonomy. She traces their history, showing how each developed in response to the prior one and comparing their early versions with those on the contemporary philosophical scene. Kant's theory that normativity springs from our own autonomy emerges as a synthesis of the other three, and Korsgaard concludes with her own version of the Kantian account. Her discussion is followed by commentary from G. A. Cohen, Raymond Geuss, Thomas Nagel, and Bernard Williams, and a reply by Korsgaard.

Agency and identity -- Necessitation -- Acts and actions -- Aristotle and Kant -- Agency and practical identity -- The metaphysics of normativity -- Constitutive standards -- The constitution of life -- In defense of teleology -- The paradox of self-constitution -- Formal and substantive principles of reason -- Formal versus substantive -- Testing versus weighing -- Maximizing and prudence -- Practical reason and the unity of the will -- The empiricist account of normativity -- The rationalist account of normativity -- Kant on the hypothetical imperative -- Against particularistic willing -- Deciding and predicting -- Autonomy and efficacy -- The function of action -- The possibility of agency -- Non-rational action -- Action -- Attribution -- The psychology of action -- Expulsion from the garden : the transition to humanity -- Instinct, emotion, intelligence, and reason -- The parts of the soul -- Inside or outside -- Pull yourself together -- The constitutional model -- Models of the soul -- The city and the soul -- Platonic virtues -- Justice : substantive, procedural, and platonic -- Kant and the constitutional model -- Defective action -- The problem of bad action -- Being governed by the wrong law -- Or five bad constitutions -- Conceptions of evil -- Degrees of action -- Integrity and interaction -- Deciding to be bad -- The ordinary cases -- Dealing with the disunified -- Kant's theory of interaction -- My reasons -- Deciding to treat someone as an end in himself -- Interacting with yourself -- How to be a person -- What's left of me?

In this book, David Boonin examines the problem of punishment, and particularly the problem of explaining why it is morally permissible for the state to treat those who break the law in ways that would be wrong to treat those who do not? Boonin argues that there is no satisfactory solution to this problem and that the practice of legal punishment should therefore be abolished. Providing a detailed account of the nature of punishment and the problems that it generates, he offers a comprehensive and critical survey of the various solutions that have been offered to the problem and concludes by considering victim restitution as an alternative to punishment. Written in a clear and accessible style, The Problem of Punishment will be of interest to anyone looking for a critical introduction to the subject as well as to those already familiar with it.

This paper looks at philosophical questions that arise in the context of AI alignment. It defends three propositions. First, normative and technical aspects of the AI alignment problem are interrelated, creating space for productive engagement between people working in both domains. Second, it is important to be clear about the goal of alignment. There are significant differences between AI that aligns with instructions, intentions, revealed preferences, ideal preferences, interests and values. A principle-based approach to AI alignment, which combines these elements in a systematic way, has considerable advantages in this context. Third, the central challenge for theorists is not to identify ‘true’ moral principles for AI; rather, it is to identify fair principles for alignment that receive reflective endorsement despite widespread variation in people’s moral beliefs. The final part of the paper explores three ways in which fair principles for AI alignment could potentially be identified.