Online research in philosophy

The ultimate goal of research into computational intelligence is the construction of a fully embodied and fully autonomous artificial agent. This ultimate artificial agent must not only be able to act, but it must be able to act morally. In order to realize this goal, a number of challenges must be met, and a number of questions must be answered, the upshot being that, in doing so, the form of agency to which we must aim in developing artificial agents comes into focus. This chapter explores these issues, and from its results details a novel approach to meeting the given conditions in a simple architecture of information processing.

What would it take for an artificial agent to be treated as having moral value? As a first step toward answering this question, we ask what it would take for an artificial agent to be capable of the sort of autonomous, adaptive social behavior that is characteristic of the animals that humans interact with. We propose that this sort of capacity is best measured by what we call the Embodied Turing Test. The Embodied Turing test is a test in which intelligence is operationally defined in terms of autonomous, adaptive interaction with the environment and with other animals. Three versions of the Embodied Turing test were performed with a SONY AIBO robot. Human participants were asked to differentiate between AIBO in a human-controlled mode and AIBO in a software-controlled mode. Our results indicate that the human participants were guessing at how AIBO was controlled. Our data reveals that people do not have enough experience with robots to accurately evaluate its behavior. This indicates that today’s humans do not have enough experience with artificial agents to treat them as morally valuable.

Kant’s duty of self-knowledge demands that one know one’s heart—the quality of one’s will in relation to duty. Self-knowledge requires that an agent subvert feelings which fuel self-aggrandizing narratives and increase self-conceit; she must adopt the standpoint of the rational agent constrained by the requirements of reason in order to gain information about her moral constitution. This is not I argue, contra Nancy Sherman, in order to assess the moral goodness of her conduct. Insofar as sound moral practice requires moral self-knowledge and moral self-knowledge requires a theoretical commitment to a conception of the moral self, sound moral agency is for Kant crucially tied to theory. Kant plausibly holds that self-knowledge is a protection against moral confusion and self-deception. I conclude that although his account relies too heavily on the awareness of moral law to explain its connection to moral development, it is insightful and important in Kantian ethics.

For the purpose of this paper, I assume that if a person is morally responsible for an action, this is a necessary and sufhcient condition for moral appraisal of that person for that action. For instance, if the action is morally wrong, moral blame is in order. Other morally relevant responses that are sometimes connected with moral responsibility are praise, pardon, shame, pride, reward, punishment, remorse. I now introduce two quite different concepts of moral responsibility: one grounded on the causal responsibility ofthe agent for an action, the other on the ability of the agent to do otherwise. The one based on the agent’s causal responsibility considers it a necessary condition for praising or blaming an agent for an action, that it was the agent and not something else that brought about the action. The question of moral responsibility becomes one of whether the agent was the or a cause of the action, or whether the agent was forced to act by something else. On this view, actions or choices can be attributed to agents because it is in their actions and choices that the agents, quo moral beings, manifest themselves. The second idea of moral responsibility considers it a prerequisite for blaming or praising an agent for an action that the agent could have done otherwise. This idea is often connected with the agents, sentiments or beliefs that they could have done otherwise, as well as the agents’ feelings of guilt or regret, or pride, for what they have done. Some philosophers consider the causal indeterminedness of the agent’s decision to act as necessary to warrant that the agent could have done otherwise.

Some philosophers have argued that moral agency is characteristic of humans alone and that its absence from other animals justifies granting higher moral status to humans. However, human beings do not have a monopoly on moral agency, which admits of varying degrees and does not require mastery of moral principles. The view that all and only humans possess moral agency indicates our underestimation of the mental lives of other animals. Since many other animals are moral agents (to varying degrees), they are also subject to (limited) moral obligations, examples of which are provided in this paper. But, while moral agency is sufficient for significant moral status, it is by no means necessary.

Kant’s duty of self-knowledge demands that one know one’s heart - the quality of one’s will in relation to duty. Self-knowledge requires that an agent subvert feelings which fuel self-aggrandizing narratives and increase self-conceit; she must adopt the standpoint of the rational agent constrained by the requirements of reason in order to gain information about her moral constitution. This is not I argue, contra Nancy Sherman, in order to assess the moral goodness of her conduct. Insofar as sound moral practice requires moral self-knowledge and moral self-knowledge requires a theoretical commitment to a conception of the moral self, sound moral agency is for Kant crucially tied to theory. Kant plausibly holds that self-knowledge is a protection against moral confusion and self-deception. I conclude that although his account relies too heavily on the awareness of moral law to explain its connection to moral development, it is insightful and important in Kantian ethics.

It has been argued that, on Kantian grounds, pedophiles, rapists and murderers are morally obligated to take their own lives prior to committing a violent action that will end their moral agency. That is, to avoid destroying the agent's moral life by performing a morally suicidal action, the agent, while he still is a moral agent, should end his body's life. Although the cases of dementia and the morally reprehensible are vastly different, this Kantian interpretation might be useful in the debate on the permissibility of suicide for those facing dementia's effects. If moral agents have a duty to act as moral agents, then those who will lose their moral identity as moral agents have an obligation to themselves to end their physical lives prior to losing their dignity as persons.

Utilizing the film I, Robot as a springboard, I here consider the feasibility of robot utilitarians, the moral responsibilities that come with the creation of ethical robots, and the possibility of distinct ethics for robot-robot interaction as opposed to robot-human interaction. (This is a revised and expanded version of an essay that originally appeared in IEEE: Intelligent Systems.).

We can discern three types of amoral beings in Kant’s ethical philosophy (B1 - B3 below), one kind of moral being (B4), the true moral agent, and one kind of immoral being (B5), for five kinds in all: B1) beings that are driven solely by inclination, such as animals. B2) beings that act solely out of reason and, therefore, duty, such as divine intellects.

In this paper, I examine a variety of agents that appear in Kantian ethics in order to determine which would be necessary to make a robot a genuine moral agent. However, building such an agent would require that we structure into a robot’s behavioral repertoire the possibility for immoral behavior, for only then can the moral law, according to Kant, manifest itself as an ought, a prerequisite for being able to hold an agent morally accountable for its actions. Since building a moral robot requires the possibility of immoral behavior, I go on to argue that we cannot morally want robots to be genuine moral agents, but only beings that simulate moral behavior. But then, if that is what we want for robots, why should we want something different for human beings? Robot ethics, it seems, presents something of a reductio of Kant’s ethics that points to hidden assumptions that hide in the very fabric of the Kantian moral enterprise, not the least of which is that Kant presumes humans to be fallen creatures. Religious doctrine, in other words, infects Kant’s attempt to derive morality from reason. This paper will demonstrate that this is so.