Important Dates
Philosophy, Computer Science & Artificial lntelligence Job Talks: Carlos Zednik - Explainable Al and the Problems of Philosophy
July 09, 2021, 11.00-11.45
Zoom: https://uni-bayreuth.zoom.us/j/62481044565?pwd=WTY5MFNQaEJrbFRxS1AxSzMrTnhHdz09
Please note that only members of the University of Bayreuth can attend the job talks. We appreciate your comprehension.
Abstract: The Black Box Problem in AI arises because machine learning methods yield computing systems that are opaque. This problem gives rise to difficulties in industry and research, but also poses new philosophical challenges. In ethics, the Black Box Problem appears to cause responsibility gaps for the actions of autonomous agents. In philosophy of mind, it prevents the application of procedural tests for machine intelligence. In philosophy of science, it threatens the possibility of distinguishing between causation and correlation in high-dimensional, complex systems.
Explainable Artificial Intelligence (XAI) aims to render opaque systems transparent through the development of analytic techniques with which to predict, interpret, and control these systems' behaviors. In this talk, I will consider whether, and if so how and to what extent, these techniques also address the aforementioned philosophical challenges. [Poster]