Algorithms at Work: The Practice of Prediction - III: Reframing Algorithmic Accountability in Practice

Abstract

This session builds on an ongoing debate about fairness, accountability, transparency, and bias in the context of algorithms, machine learning, and AI. Having demonstrated the inherent inscrutability and opacity of these sociotechnical systems, the literature is now shifting to unpacking the concepts that have thus far been taken for granted in this debate. The papers in this session interrogate our understandings of fairness, transparency, and intelligence. We then re-ground the accountability debate in an examination of the real social implications of algorithmic technologies for race and gender inequality. Finally, we attempt to connect these two perspectives by introducing questions of literacy and learning.

Chair

• Francis Lee, Uppsala University

Discussant

• Francis Lee, Uppsala University

Individual Presentations

• Defining Artificial Intelligence: Regulating a Moving Target - Peaks Krafft; Karen Y Huang, Harvard University; Michael Katell, University of Washington (Information School); Meg Young

• It’s Technically Fair, But You Might Not Like It - Christian Sandvig, University of Michigan

• The Username of the Father: Predictive Sentencing Algorithms, Race, and the Shame of Law - Aaron Neiman, Stanford University

• Where is Algorithmic Literacy in Algorithmic Accountability? - Elisabeth Sulmont, McGill University; Elizabeth Patitsas, McGill University

Session Organizer

• Sarah E. Sachs, Cornell University