This paper investigates the claim that artificial Intelligence Systems cannot be held morally responsible because they do not have an ability for agential self-awareness e.g. they cannot be aware that they are the agents of an action. The main suggestion is that if agential self-awareness and related first person representations presuppose an awareness of a self, the possibility of responsible artificial intelligence systems cannot be evaluated independently of research conducted on the nature of the self. Focusing on a specific account (...) of the self from the phenomenological tradition, this paper suggests that a minimal necessary condition that artificial intelligence systems must satisfy so that they have a capability for self-awareness, is having a minimal self defined as ‘a sense of ownership’. As this sense of ownership is usually associated with having a living body, one suggestion is that artificial intelligence systems must have similar living bodies so they can have a sense of self. Discussing cases of robotic animals as examples of the possibility of artificial intelligence systems having a sense of self, the paper concludes that the possibility of artificial intelligence systems having a ‘sense of ownership’ or a sense of self may be a necessary condition for having responsibility. (shrink)

According to dominant views in affective computing, artificial systems e.g. robots and algorithms cannot experience emotion because they lack the phenomenological aspect associated with emotional experience. In this paper I suggest that if we wish to design artificial systems such that they are able to experience emotion states with phenomenal properties we should approach artificial phenomenology by borrowing insights from the concept of ‘attunement to the world’ introduced by early phenomenologists. This concept refers to an openness to the world, a (...) connection with the world which rejects the distinction between an internal mind and the external world. Early phenomenologists such as Heidegger, consider this ‘attunement’ necessary for the experience of affective states. I argue that, if one accepts that the phenomenological aspect is part of the emotion state and that ‘attunement to the world’ is necessary for experiencing emotion, affective computing should focus on designing artificial systems which are ‘attuned to the world’ in the phenomenological sense to enable them to experience emotion. Current accounts of the phenomenal properties of affective states, analyse them in terms of specific types of representations. As artificial systems lack a capability for such representation mainly because of an inability to determine relevance in changing contexts (‘the frame problem’), artificial phenomenology is impossible. I argue that some affective states, such as ‘attunement’ are not necessarily representational and as such a lack of capacity for representation does not imply that artificial phenomenology is impossible. At the same time ‘attunement’ helps restrict some aspects of the ‘frame problem’ and as such, goes some way of enabling representational states such as emotion. (shrink)

How should human beings and robots interact with one another? Nyholm’s answer to this question is given below in the form of a conditional: If a robot looks or behaves like an animal or a human being then we should treat them with a degree of moral consideration (p. 201). Although this is not a novel claim in the literature on ai ethics, what is new is the reason Nyholm gives to support this claim; we should treat robots that look (...) like human or non-human animals with a certain degree of moral restraint out of respect for human beings or other beings with moral status. Although Danaher or Coeckelbergh also claim that we should treat robots with a degree of moral consideration, the reasons they give for making this claim focus on duties or rights attaching to the robot themselves (see J. Danaher, “Welcoming Robots into the Moral Circle: A Defence of Ethical Behaviourism,” Science and Engineering Ethics, (2019): 1–27 or M. Coeckelbergh, “Moral Appearances: Emotions, Robots and Human Morality,” Ethics and Information Technology, 12(3) (2010): 235–241.). Nyholm disagrees with this type of reasoning and claims that until robots develop a human or animal like inner life, we have no direct duties to the robots themselves. Rather, it is out of respect for human beings or other beings with moral status that we should treat some robots with moral restraint. Gerdes, similarly inspired by Kant, focuses on the human agent to argue that we should avoid treating robots in cruel ways because this may corrupt the human agent’s character (see A. Gerdes, “The Issue of Moral Consideration in Robot Ethics,” siggas Computers and Society, 45(3) (2015): 274–279.). Nyholm’s contribution here is to extend this view such that the corruption or harm being done is against the humanity in all of us. (shrink)