Alle Publikationen
- <<
- <
- 1
2018
-
(2018) : Moral robots In: Johnson, L. Syd M.; Rommelfanger, Karen S. (Hg.): The Routledge handbook of neuroethics: New York, NY: Routledge/Taylor & Francis Group (Routledge handbooks in applied ethics), S. 363-377
Abstract: The main reason for raising the question about the ethical behavior of robots is the rapid progress in the development of autonomous social robots that are specifically created to be deployed in sensitive human environments, from elder and health care settings to law enforcement and military contexts. Clearly, such tasks and environments are very different from traditional factory environments (for example, for welding robots). Hence, these new social robots will require higher degrees of autonomy (i.e., a capacity for independent, self-directed action) and decision-making than any previously developed machine, given that they will face a much more complex, open world. They might be required to acquire new knowledge on the fly to accomplish a never-before-encountered task. Moreover, they will likely face humans who are not specifically trained to interact with them, and robots thus need to be responsive to instructions by novices and feel ’natural’ to humans even in unstructured interactions. At these levels of autonomy and flexibility in near-future robots, there will be countless ways in which robots might make mistakes, violate a user’s expectations and moral norms, or threaten the user’s physical or psychological safety. These social robots must therefore also be moral robots. Thus, for autonomous social robots deployed in human societies, three key questions arise: (1) What moral expectations do humans have for social robots? (2) What moral competence can and should such robots realize? (3) What should be the moral standing of these machines (if any)? The first question, about moral expectations, follows from the well-established fact that autonomous social robots, especially those with natural-language abilities, are treated in many ways like humans, regardless of whether such treatment was intended or anticipated by the robot designers. In fact, there is mounting evidence that humans have very clear expectations of robot capacities based on the robot’s appearance and people’s perceptions of the robot behaviors. We will review some of this work in the third section. The second question, about moral competence, arises from the need to endow robots with sufficient capacities to operate safely in human societies. Answers to this question would ideally build on answers to the first question and provide mechanisms for robots to process social and moral norms in ways that humans expect. In addition, the design of the robots’ control systems should ensure that robots behave ethically according to the norms of the society in which they are deployed. In this chapter, we will focus on the first two questions, discussing both human expectations and computational architectural mechanisms that will allow robots to live up to those expectations while leaving a detailed discussion of the third question to legal experts and philosophers We will not focus on the philosophical debate about agency and personhood. Rather, we will assume an operational behavioral definition of a ’moral robot’. (PsycINFO Database Record (c) 2019 APA, all rights reserved)
- <<
- <
- 1