Conscience as a design benchmark for social robots

For humanlike robots to behave like human beings, they would have to be designed to perform as the latter would in ethical relationships. This, however, is more than an engineering question. According to the phenomenological approach explored in this paper, every relationship is an existential relationship in which each person confirms the existence of the other as just that type of entity that can be in an ethical relationship in the first place - quite unlike, say, the relation one may hold with a rock. This paper investigates the merits of conscience as a design primitive for social robots and proposes that for social robots to function in real social relationships with human beings, it is necessary to understand the nature of human beings themselves from beyond an engineering perspective of objects of a certain kind. For an android to have a conscience, it will have been necessary for its human creators to appreciate the ontological requirements of the concept in the first place