Making sense of artificial agents’ behaviour

This colloquium talk is planned as an in-person event. Registration by email is only required for non-CEU members. Please check the COVID safety rules here to enter the building.

ABSTRACT
This talk addresses the need for and possibility of understanding the behaviour of artificial agents. It is argued that
anthropomorphic and anthropocentric assumptions, both on the part of philosophers and the general public, have obscured alternative and diverse courses of development of artificial systems. Human psychological vulnerabilities and practical necessities call for heterogeneous ‘semi-folk’ psychologies of artificial agents, which would substantially differ from human folk psychology and would have to be informed by expert knowledge.