Profesor: Greg Nalepa, AGH University of Science and Technology, Krakow, Poland

Resumen del contenido:

In this seminar we will address the issue of explainability in artificial intelligence. It regards the methodological issues related to the design of intelligent systems. The most important ones are models of knowledge. We will discuss different classes of models and their impact on explainability. Furthermore, there have been several approaches to the building of AI that include and possibly expose explanatory functionalities.

Recently the challenge from DARPA motivated by selected applications of AI brought XAI into the attention. Then we will move to the broader scope related to societal challenges, including the legal ones. Finally, we will present the perspective on the explainable AI systems and the challenges to the legal liability.