Podmiotowość refleksyjna a autonomia sensomotoryczna w kontekście sprawczości sztucznych systemów poznawczych [PL]

Abstract

In the presented article, I invoke two concepts of agency: the first, formulated within the framework of analytical philosophy of intentional action, which assumes the reflexivity of the agent, and the second, developed within the enactivist conception of agency as sensorimotor autonomy. I place these two concepts of autonomous action in the context of the discussion on the agency of artificial systems, which are currently gaining increasing independence in terms of selecting methods for solving specific tasks. In contrast to my earlier arguments expressed in the book “The Agency of Extended Cognitive Systems”, I highlight the explanatory advantage of the enactivist conception, which, although it cannot fulfill some of the functions of the standard definition od agency, proves more useful in discussions about the moral implications of artificial intelligence development. These implications are currently being actively discussed within the framework of the Value Alignment Problem, which concerns the alignment of human values with those that guide, and in particular will guide in the future, artificial intelligence. In this paper, I also point to the applications of the enactivist notion of agency in explaining the functioning of group systems as agents of action.