Enacting an AI system typically requires three iterative phases where AI engineers are in command: selection and preparation of the data, selection and configuration of algorithmic tools, and fine-tuning of the different parameters on the basis of intermediate results. Our main hypothesis is that these phases involve practices with ethical questions. This paper maps these ethical questions and proposes a way to address them in light of a neo-republican understanding of freedom, defined as absence of domination. We thereby identify different (...) types of responsibility held by AI engineers and link them to concrete suggestions on how to improve professional practices. This paper contributes to the literature on AI and ethics by focusing on the work necessary to configure AI systems, thereby offering an input to better practices and an input for societal debates. (shrink)

‘Optimistic’ normative theories of criminal justice aim to justify criminal sanction in terms of its reprobative/rehabilitative value rather than its punitive nature as such. But do such theories accord with ordinary intuitions about what constitutes a ‘just’ response to wrongdoing? Recent empirical work on the psychology of punishers suggests that human beings have a ‘brutely retributive’ moral psychology, making them unlikely to endorse normative theories that sacrifice retribution for the sake of reprobation or rehabilitation; it would mean, for example, that (...) we cannot expect people to support such optimistic theories democratically, calling their feasibility into question. Taking the civic republican theory as an exemplar of optimistic theories, we argue that it does not fail this feasibility test. We review recent empirical research, including studies of our own, to support the claim that, far from being brute retributivists, human beings are generally satisfied only with punishment that delivers something other than mere retribution. And we show that this coheres very closely with the goals and policies that civic republicanism would support, as it may be expected to cohere also with other optimistic proposals. (shrink)

It has become a commonplace in neo-republican thinking to claim that if the notion of rights can be allowed a place in republican political theory, it can never achieve the prominence that liberalism allegedly grants it. Philip Pettit’s book, Republicanism, provides several arguments to buttress this thesis. This article aims at examining these arguments in order to show that once properly stated, they must on the contrary be considered as powerful arguments to the effect that republicans take rights very seriously.

Untangling the relationship of law and liberty is among the core problems of political theory. One prominent position is that there is no freedom without law. This article challenges the argument that, because law is constitutive of freedom, there is no freedom without law. I suggest that, once properly understood, the argument that law is constitutive of freedom does not uniquely apply to law. It also applies to social norms. What law does for freedom, social norms can do too. Thus, (...) I claim the question facing us is not the easy one of either law or social norms. Respecting this draws attention to the unique risks to our freedom introduced by both sets of norms. (shrink)

Philip Pettit’s conception of freedom as nondomination is modally robust in that it requires not simply reducing the probability of uncontrolled interference by others but entirely eliminating that possibility. In this article, I consider whether freedom as nondomination provides an attractive analysis of official discretion, particularly in the context of the criminal law, an area of recurring interest for Pettit. I argue that not only does the modally robust character of freedom as nondomination have some rather unattractive implications in the (...) criminal law, but that it sits poorly with Pettit’s more general ambitions to provide a consequentialist framework for the evaluation of social institutions. Drawing on recent work by Niko Kolodny, I develop an ‘anti-deference’ interpretation of nondomination and contrast it to Pettit’s modally robust conception. (shrink)