The Epistemology of Explainable AI by Siavash Bigdeli

Siavahs Bigdeli, Assoc. Professor, Department of Applied Mathematics and Computer Science Visual Computing, DTU

Speech given at Int. Conference on Artificial Intelligence: Risks, Regulations and Roadmaps, Dec 7th 2023. The conference addresses the balancing act of regulation: seeking to ensure neither too little nor too much. While we'll present and discuss major societal threats posed by AI, we'll also consider the missed opportunities if stringent regulations hinder AI-based scientific, economic, and general human progress.

Short abstract
In this talk, I will explore the intersection of explainability, AI, and epistemology. I'll delve into recent approaches to explainable AI and examine their alignment with various epistemological schools.

Additionally, I will discuss how this interplay can guide the selection, adaptation, and construction of explanations for specific applications.

Bio
Siavash Bigdeli is an associate professor of computer vision at the Technical University of Denmark. Prior to that, he was a scientist at the Swiss Center for Electronics and Microtechnologies.

His interests lie in the Ante-/Post-Hoc explainability of ML models by employing statistical models in learning/inference processes, and by integrating philosophical methods in ML with a focus on Generative AI.