Algorithmic Bias and Responsibility: Who, When, To Whom?

	Other - AI Ethics & Algorithmic Justice
	Algorithmic Bias and Responsibility: Who, When, To Whom?
	Mark Coeckelbergh

	Virtual- REGISTRATION REQUIRED (SEE ABSTRACT) Zoom May 26, 2021 - 01:00 pm
Abstract: This talk will address three issues regarding algorithmic bias. First I will argue that we should consider bias in the entire algorithmic and data process. This raises the question of when bias arises, which is important for ascribing responsibility. Second, I will ask who should fix the bias: given that bias in AI is related to bias in society and that a democratic way of addressing the issue is preferable, it is not clear that developers and their companies should carry all responsibility for minimizing bias. Third, what we identify as bias is culturally and historically influenced. The current discourses on bias usually fail to be sufficiently sensitive to differences in a global context. This issue is related to the question 'to whom are we responsible?' and connects to my arguments about relational responsibility. Register here. LSU Center for Computation & Technology Office of Research & Economic Development Ethics Institute Philosophy Humanities & Social Sciences Present: AI Ethics/Algorithmic Justice Wednesday, May 26 12-2pm, Central Time A Panel Discussion with: Michael Kearns, Safiya Noble, Mark Coeckelbergh Schedule: • Opening Remarks 12-12:05 P.M. • Introduction of Speakers 12:05-12:10 P.M. • Speaker 1 (Kearns): 12:10-12:35 P.M. • Speaker 2 (Safiya): 12:35-1 P.M. • Speaker 3 (Coeckelbergh): 1-1 :25 P.M. • Panelist Discussion/ Q&A 1:25-2 P.M.

Speaker's Bio: Author, *AI Ethics* (MIT Press, 2020). Professor, Philosophy of Media and Technology (Vienna). Member, High-Level Expert Group on Artificial Intelligence for the European Commission, the Austrian Council on Robotics and Artificial Intelligence, and the Technical Expert Committee (TEC) for the Foundation for Responsible Robotics.

CENTER FOR COMPUTATION & TECHNOLOGY

Interdisciplinary | Innovative | Inventive

Events