Ethical Implications of Robot Inner Speech

John Sullins
Professor of Philosophy Sonoma State University

Stevenson Hall 1300
12:00 PM - 12:50 PM

In this talk we will explore work done in collaboration between SSU and The Robot Lab at the University of Palermo to explore the use of robot inner speech and developing better ethical reasoning in human robot teams. Robot inner dialog is different from human inner dialog in that it is an additional process that is added for the user's benefit, not so much for the help of the machine that is doing the reasoning.  Here we will explain the robot's inner dialog and how it has been used to build systems that display more conscious and trustworthy actions. Whereas machine morality seeks to give systems rudimentary moral reasoning capabilities, Artificial Phronesis aims to build machines that are more highly skilled in producing actions that are not only acceptable but wise and nuanced.  Such machines need to be conscious of their situation and the subtle variables that go into skillfully navigating the moral terrain they find themselves.  We provide a way of testing for the perceived presence of aspects of artificial phronesis between the human user and the robot.  We find that the robot's inner speech helps the user generate skilled practical moral reasoning during the experiment.