Ethical Aspects of Complex Adaptive Systems

Speaker: Luigia Carlucci Aiello, La Sapienza University, Rome

Abstract

We may say that machine ethics pertains to science fiction, or say that ethics is a prerogative of human behaviour, and that machines do not/cannot have it. Conversely, machine ethics, i.e. the possibility of using artifacts in a way that complies with the principles of human ethics, requires serious consideration. This is due to the rapid emergence of increasingly complex autonomous software agents, robots and in general complex adaptive/autonomous systems that find wide application in all (critical) sectors of our lives and heavily interfere with, or change, them. We go from autonomous cars that may kill pedestrians, or steal jobs to taxi drivers, to systems that with subtle intrusions violate our privacy or interfere with our cognitive functions and capabilities. The wide pervasiveness and the growing complexity of software and hardware employed in intelligent agents and adaptive systems in cyber-physical domains call for a careful study of both their functional and extra-functional properties. Ethical aspects are among important extra-functional properties; they concern the whole life cycle of the systems, from design, development, deployment/production to use. In my talk I address the need to investigate the ethical aspects of Complex Adaptive Systems. I report on some recent relevant findings on the topic and overview the guidelines issued by the European Commission.