In this paper, we take moral agency to be that context in which a particular agent can, appropriately, be held responsible for her actions and their consequences. In order to understand moral agency, we will discuss what it would take for an artifact to be a moral agent. For reasons that will become clear over the course of the paper, we take the artifactual question to be a useful way into discussion but ultimately misleading. We set out a number of (...) conceptual pre-conditions for being a moral agent and then outline how one should — and should not — go about attributing moral agency. In place of a litmus test for such an agency — such as Allen et al.'s Moral Turing Test — we suggest some tools from the conceptual spaces theory and the unified conceptual space theory for mapping out the nature and extent of that agency. (shrink)

This paper follows directly from an earlier paper where we discussed the requirements for an artifact to be a moral agent and concluded that the artifactual question is ultimately a red herring. As...

International Journal of Machine Consciousness, Volume 06, Issue 02, Page 141-161, December 2014. This paper follows directly from an earlier paper where we discussed the requirements for an artifact to be a moral agent and concluded that the artifactual question is ultimately a red herring. As before, we take moral agency to be that condition in which an agent can appropriately be held responsible for her actions and their consequences. We set a number of stringent conditions on moral agency. A (...) moral agent must be embedded in a cultural and specifically moral context and embodied in a suitable physical form. It must be, in some substantive sense, alive. It must exhibit self-conscious awareness. It must exhibit sophisticated conceptual abilities, going well beyond what the likely majority of conceptual agents possess: not least that it must possess a well-developed moral space of reasons. Finally, it must be able to communicate its moral agency through some system of signs: A "private" moral world is not enough. After reviewing these conditions and pouring cold water on recent claims for having achieved "minimal" machine consciousness, we turn our attention to a number of existing and, in some cases, commonplace artifacts that lack moral agency yet nevertheless require one to take a moral stance toward them, as if they were moral agents. Finally, we address another class of agents raising a related set of issues: autonomous military robots. (shrink)

Not all research in machine consciousness aims to instantiate phenomenal states in artefacts. For example, one can use artefacts that do not themselves have phenomenal states, merely to simulate or model organisms that do. Nevertheless, one might refer to all of these pursuits -- instantiating, simulating or modelling phenomenal states in an artefact -- as 'synthetic phenomenality'. But there is another way in which artificial agents (be they simulated or real) may play a crucial role in understanding or creating consciousness: (...) 'synthetic phenomenology'. Explanations involving specific experiential events require a means of specifying the contents of experience; not all of them can be specified linguistically. One alternative, at least for the case of visual experience, is to use depictions that either evoke or refer to the content of the experience. Practical considerations concerning the generation and integration of such depictions argue in favour of a synthetic approach: the generation of depictions through the use of an embodied, perceiving and acting agent, either virtual or real. Synthetic phenomenology, then, is the attempt to use the states, interactions and capacities of an artificial agent for the purpose of specifying the contents of conscious experience. This paper takes the first steps toward seeing how one might use a robot to specify the non- conceptual content of the visual experience of an (hypothetical) organism that the robot models. (shrink)

Understanding the relationship between concepts and experience seems necessary to specifying the content of experience, yet current theories of concepts do not seem up to the job. With Peter Gärdenfors's conceptual spaces theory as a foundation and with enactivist philosophy as inspiration, we present a proposed extension to conceptual spaces theory and use it to outline a model of the emergence of concepts and experience. We conclude that neither is ultimately primary but each gives rise to the other: i.e., that (...) they co-emerge. Such a model can then serve as the anchor to a theory of concepts more generally. Concepts are most naturally understood in symbolic and representational terms, while much of experience, in contrast, is non-symbolic and non-representational; yet the conflict between the two will, herein, be shown to be more apparent than real. The main contribution of this paper is to argue for, by means of this account of co-emergence, a continuum between "low-level" mental content that is more appropriately understood in highly context-sensitive and directly sensorimotor-based terms, and "high-level" mental content that is more appropriately understood in context-free and representational or symbolic terms. In doing so we conclude that the extreme positions of representationalism and anti-representationalism are fatally flawed. (shrink)

Concepts are the building blocks of higher-order cognition and consciousness. Building on Conceptual Spaces Theory (CST) and proceeding from the assumption that concepts are inherently dynamic, this paper provides historical context to and significantly elaborates the previously offered Iterative Subdivision Model (ISDM) with the goal of pushing it toward empirical testability. The paper describes how agents in continuous interaction with their environment adopt an intentional orientation, estimate the utility of the concept(s) applicable to action in the current context, engage in (...) practical action, and adopt any new concepts that emerge: a largely pre-intellectual cycle that repeats essentially without interruption over the conceptual agent’s lifetime. This paper elaborates utility optimization by establishing three constraints on concept formation/evaluation – non-redundancy, distinctiveness, and proportionality – embedding them in a quasi-mathematical model intended for development into a formal logic. The notion of a distinctor – a quality dimension of the conceptual space in focus at any given time, used for making what we call a difference distinction – is key. The primary contribution of the revised ISDM is the way it relates concepts to action via utility optimization/actualization and the way it describes the emergence of quality dimensions through trial-by-action (trial and error), something previous presentations of CST have failed to address. (shrink)

Standard, linguistic means of specifying the content of mental states do so by expressing the content in question. Such means fail when it comes to capturing non-conceptual aspects of visual experience, since no linguistic expression can adequately express such content. One alternative is to use depictions: images that either evoke (reproduce in the recipient) or refer to the content of the experience. Practical considerations concerning the generation and integration of such depictions argue in favour of a synthetic approach: the generation (...) of depictions through the use of an embodied, perceiving and acting agent, either virtual or real. This paper takes the first steps in an investigation as to how one might use a robot to specify the non-conceptual content of the visual experience of an (hypothetical) organism that the robot models. (shrink)

This paper serves as introduction to a significantly longer paper in progress. It argues that, within the ethics community, the wider philosophical establishment and society in general, people have been far too lax about what to accept as morally “right” behaviour – far too quick to let themselves and, all too often, each other off the hook. By drawing comparisons to artefactual behaviour and the objections people raise to calling that behaviour the morally acceptable behaviour of authentic moral agents, this (...) paper lays out a framework by which human ethics and meta-ethics can more fruitfully be approached. An earlier paper of ours argued that, for an action to be morally right, one must have a convergence of the right motivations, the right means, and the right consequences. The underlying insight is that deontological, virtue-ethics-based, and consequentialist accounts all have their necessary role to play, but each tends to get too focused on itself and its merits to the loss of the bigger picture; while utilitarian accounts, as perhaps the most prominent division within consequentialism, face the further problem of failing to allow for those occasions where the needs of the few, or the one, outweigh the needs of the many, as Ursula K. LeGuin so devastatingly addressed. Although the requirement to align motivations, means, and consequences may seem impossibly onerous, it need not be, provided one is prepared to allow that moral behaviour is far more difficult to achieve, either for artefacts or human beings, than it might seem at first glance. Mistakes will be made. Perhaps it matters more to take responsibility for those mistakes than to assure oneself, despite reasonable argument to the contrary, that one has avoided them. It is time to hold artefactual and natural agent alike to a higher standard. (shrink)