Abstract:
As increasingly intelligent and autonomous robots continue to proliferate into every area of modern life, there is no doubt that society has to think deeply about the potential impact, whether negative or positive, that this will have on ordinary everyday contexts. One of the most urgent societal expectations for these robots is the need for them to behave in a manner that is respecting of human moral values. In response to this challenge, the field of machine ethics began with the goal of developing robots capable of making moral decisions. This work addresses the challenge by proposing that Exemplarist Virtue Ethics (or simply exemplarism), an ethical theory based on virtue ethics, is a viable, suitable and alternative framework for building ethical robots. Exemplarism is a moral theory that grounds key moral concepts (e.g. virtue, right act, etc.) by direct reference to exemplars of moral goodness. Essentially, it proposes that agents can develop their moral character by following the example of morally admirable agents in society. This work will demonstrate how an exemplar- ist machine ethics framework presents several advantages to building ethical robots over traditional approaches based on consequentialism and deontology. Specifically, exemplarism not only helps us formalise the concept of artificial moral agency more coherently, but it also lends itself to be a technically feasible approach for building ethical robots. This thesis will, therefore, also demonstrate the technical feasibility of actually building an exemplarist AMA and suggest ways in which it could be further improved. Since exemplarism has scarcely been applied to this area in prior literature, this thesis will provide an alternative perspective to the machine ethics project, which, in some small way can help to advance the field forward.
