Human Rights, Science, and Technology Webinar Series
Social Bias in AI: Re-coding Innovation through Algorithmic Political Capitalism
Monday, October 7, 2024 11:00 AM to 12:15 PM EDT
Online Location
Recording of the Webinar
Panelists:
John G. Dale, Associate Professor College of Humanities and Social Sciences, Department of Sociology & Anthropology, George Mason University, Fairfax, VA USA
Samuel Carter, Ph.D. Candidate, College of Science, Department of Computational and Data Sciences, Center for Social Complexity, George Mason University, Fairfax, VA USA
Moderator:
Conor Mahoney, Principal Systems Engineer MITRE, Ph.D. student, College of Science, Department of Computational and Data Sciences, Center for Social Complexity, George Mason University, Fairfax, VA.
Karthik Ramanujam, Program Manager, Human Rights, Science, and Technology Webinar Series, Movement Engaged
Description:
Our panelists will discuss their current research paper which delves into the intricate interplay between societal and algorithmic bias, revealing how attempts to mitigate biases as a computational problem are inherently limited. Biases, deeply ingrained in social structures and conditions, were the seeds of the early generations of algorithmic development that underpin our present reality. This innovation, fostered and supported by government resources, harnessed political capitalism, shaping public policy around the development and deployment of algorithms for generations. However, this public policy is a robust, complex system, exhibiting self-organization, adaptive interactions, and emergent properties that transform individual behaviors into collective phenomena. In other words, public policy that aims to curb negative impacts must operate across multiple levels—community, local, state, and federal—and evaluate their effects through adaptive feedback mechanisms. This encompasses the use of user data in predictive analytics markets, the physical and human infrastructure affected by artificial intelligence, the optimization and feedback mechanisms that have fostered extremism and radicalization, and the underexamined user bias that has shifted an unbalanced power dynamic between user and platform owner, all of which are symptoms of this complex system. Dale and Carter discuss their analytic framework – “algorithmic political capitalism” to begin unravelling the myriad forces influencing outcomes, facilitate more effective navigation of our dynamic and evolving social landscape, and a means to democratize social power around a technology that shapes societal institutions at the speed of competition and rapid adoption.