Discriminating Data

Discriminating Data

Correlation, Neighborhoods, and the New Politics of Recognition

By Wendy Hui Kyong Chun

How big data and machine learning encode discrimination and create agitated clusters of comforting rage.

Overview

Author(s)

Praise

Summary

How big data and machine learning encode discrimination and create agitated clusters of comforting rage.

In Discriminating Data, Wendy Hui Kyong Chun reveals how polarization is a goal—not an error—within big data and machine learning. These methods, she argues, encode segregation, eugenics, and identity politics through their default assumptions and conditions. Correlation, which grounds big data's predictive potential, stems from twentieth-century eugenic attempts to “breed” a better future. Recommender systems foster angry clusters of sameness through homophily. Users are “trained” to become authentically predictable via a politics and technology of recognition. Machine learning and data analytics thus seek to disrupt the future by making disruption impossible.

Chun, who has a background in systems design engineering as well as media studies and cultural theory, explains that although machine learning algorithms may not officially include race as a category, they embed whiteness as a default. Facial recognition technology, for example, relies on the faces of Hollywood celebrities and university undergraduates—groups not famous for their diversity. Homophily emerged as a concept to describe white U.S. resident attitudes to living in biracial yet segregated public housing. Predictive policing technology deploys models trained on studies of predominantly underserved neighborhoods. Trained on selected and often discriminatory or dirty data, these algorithms are only validated if they mirror this data.

How can we release ourselves from the vice-like grip of discriminatory data? Chun calls for alternative algorithms, defaults, and interdisciplinary coalitions in order to desegregate networks and foster a more democratic big data.

Pre-Order Hardcover

$29.95 T ISBN: 9780262046220 344 pp. | 6 in x 9 in 69 b&w illus.

Endorsements

  • “A shattering book! Chun unveils and dispels many lazy ideas that we—data and network scientists—heedlessly adopted. Her book opens questions critical to our disciplines. We urgently need new methodological tools to tackle them.”

    Giulio Dalla Riva

    Senior Lecturer in Data Science, University of Canterbury

  • “Chun teaches readers exactly how digital networks amplify racism and discrimination, guiding the reader toward a different future by analyzing our networked past. This is a brilliant book!”

    Lisa Nakamura

    Gwendolyn Calvert Baker Collegiate Professor, University of Michigan; author of Digitizing Race: Visual Cultures of the Internet

  • “With Discriminating Data, Wendy Chun hits on a core idea of the contemporary internet: homophily, or grouping together 'like' people. In brilliant, probing essays, she gives us the analytical, ethical, and political tools to challenge the resulting prejudice and polarization. A very important book.”

    Peter Galison

    Physics and History of Science, Harvard University; coauthor of Objectivity; Director of Black Holes: The Edge of All We Know

  • Discriminating Data achieves the Herculean task of teaching us to see contemporary uses of data and algorithms, especially but not exclusively in social media, as historically contingent instead of inevitable. This book is a gift to and for our times!”

    Kara Keeling

    author of Queer Times, Black Futures