Agence morale et intelligence artificielle

Details

Supervisors

Pence, Charles H.

Faculty

Faculté de philosophie, arts et lettres

Degree label

Master [120] en philosophie, à finalité approfondie

Abstract

enThis dissertation focuses on the issue of artificial moral agency. Through this dissertation, we have attempted to shed light on the links that can unite moral agency and moral responsibility with artificial intelligence (AI). To do so, we have first attempted to expose the different issues raised by the question of moral agency. Secondly, we have outlined the questions that AI itself raises. Finally, we have brought these two fields together in a synthesis section, showing what links can be made. Throughout this dissertation, we have tried to keep a minimalist approach, i.e. to avoid as much as possible assumptions or subjective points of view. The aim was to build our reasoning on as solid a foundation as possible. In the first part, after having exposed the different issues specific to moral agency and moral responsibility, we presented four major ethical currents in order to mobilize them in the third part of this work, and to show what links they may have with AI. These currents are deontologism, utilitarianism, Virtue ethics and, less expectedly, Care ethics. In the second part of this work, we have outlined the history of AI, its functioning, and the different approaches, namely bottom-up and top-down. In particular, we discussed the issues of "machine learning" and supervised learning. We concluded this part by evoking the debate between the defenders of Strong AI and those of Weak AI. Finally, in the third part, we discussed the conditions of possibility of an artificial moral agent, showed how such an agent interacts with the ethical currents previously introduced, and opened the questioning to various problems frequently encountered. We conclude by drawing attention to the imperfection of such an artificial moral agent as well as to the imperfection of our own moral agency.