Abstract

The purpose of this essay is to determine the domain of validity of the notions developed in Machine Ethics [ME]. To this aim, I analyse the epistemological and methodological presuppositions that lie at the root of such technological project. On this basis, I then try and develop the theoretical means to identify and deconstruct improper applications of these notions to objects that do not belong to the same epistemic context, focusing in particular on the extent to which ME is supposed to feedback onto moral philosophy. By highlighting the inadequacy of many approaches to the supposed philosophical implications of ME, I wish to redirect attention to its actual scope and to stress its relevance for the social acceptance of autonomous technologies. The essay is structured as follows. After a brief introduction (§ 1), in § 2 I shed light upon the link between the current trend of robotic development toward greater degrees of autonomy and the corresponding need for artificial moral agents that fuels research in ME. I then present the epistemological profile of ME in § 3, focusing on its main component, i.e., the modelling of human moral agency in the language of robotics and computer science. In § 4 I deal with cases in which such model is brought to bear on human ethics and moral philosophy as well, whilst in § 5 I develop a criticism of this extension based on an account of the actual epistemological relations that obtain between philosophical and technological knowledge in the context of the ME project. In § 6 I discuss the role that the ordinary use of language plays in the process of extending machinerelated notions to other domains of knowledge and, finally, I clarify what the appropriate scope of machine ethics is notwithstanding its many utopian or dystopian misinterpretations.