Current developments in AI push systems towards more autonomous decision making and consequent action taking (think e.g. autonomous cars). In these situations it becomes ever more important for these systems to be aware about the social context in which they operate and also to have theories of minds of the people they interact with.
In this presentation I will argue that social capabilities are not modules one could add to existing systems, but rather form the fundaments of these systems. This possibly leads to a radical change of architectures for autonomous systems (and a lot of new and exciting work for us). I will briefly show some work that is already been done in this direction in practical projects using social simulations and social robotics.
Last modified: Tuesday, 18-Sep-2018 08:17:09 NZST
This page is maintained by the seminar list administrator.