Book Reviews

Yarden Katz. Artificial Whiteness: Politics and Ideology in Artificial Intelligence. New York, NY: Columbia University Press, 2020. 352p. Paperback, $28.00 (ISBN: 9780231194907).

In Artificial Whiteness: Politics and Ideology in Artificial Intelligence, Yarden Katz encourages the reader to step away from discourse that frames Artificial Intelligence (AI) as a purely technological development. Instead, the reader is guided through AI’s epistemological roots, the espoused values and priorities of its progenitors, its sources of research funding, and the marriage among academia, industry, and the American military that birthed it. By providing this context, Katz is able to more thoroughly examine and interrogate AI’s principal service to structural white supremacy and imperialism, as well as how that service has been masked by a “progressive veneer” in recent years.

Katz opens the book with AI’s formal birth in 1950s academe and its struggle to disambiguate itself from other forms of computing and automation like neural networks, machine learning, and cybernetics. Readers familiar with these various forms of computing will find this portion of the book valuable as it illustrates that, in some cases, the core differences among them are less epistemological than political. This is explicitly illustrated when Katz describes the financial relationship between early AI practitioners (academics in higher education) and the Pentagon and how this cemented AI’s role as a tool for American military and industry. The book follows the ebbs and flows of AI’s relevance throughout the decades, as well as its rebranding and evolution as it began to disambiguate itself from the various forms of computing that practitioners had worked so hard to distinguish it from in its early days. In this context, Katz successfully situates AI’s reputation and perceived value within American society’s changing views of industry and the military. White supremacy, imperialism, and neo-liberal ethics are all discussed with practical examples provided, like the anti-Japanese sentiment present in United States industry during the 1980s (specific publications, businesses, and a testimony to Congress are discussed). There is no room for confusion or vagueness in the use of these terms, making it easy for the reader to engage with Katz’s assertions.

The transition into neo-liberal attitudes toward corporations and the military in the United States is an important one in the world of tech and computing. Katz does a good job noting the purposeful shift in AI’s image to better suit this transition. He illustrates how AI practitioners have moved away from overt ties to the military and, instead, moved toward supporting a “progressive veneer” that frames AI as a tool for social progress. It is at this point that AI’s links to the interests of large corporations like Microsoft and Google are laid out. Katz also addresses racialized problems such as the role algorithm-based policing plays in harming Black communities. Katz also describes the ways colonialism and capitalism act as the foundation of white supremacy. A key takeaway from this book is how fundamental the academic community was—and continues to be—to AI’s ethical shortcomings. A consistent theme is how scholars looking to make a name for themselves, coupled with university research centers looking to secure funding, display only performative concerns around ethics to fuel an environment in which AI is pushed forward with naive enthusiasm around its potential good. Higher education actors pay little attention to well-established precedent around the ways technological innovations have previously exacerbated and upheld structural inequality. Katz also addresses the overwhelming whiteness of AI experts, the attempts to address this with calls for increasing diverse representation among AI practitioners, and the ways that existing ethical critiques of AI often end up replicating the very systems they claim to want to fix.

As a Black woman working in digital scholarship, I see scholars replicating many of the uninterrogated problems this book highlights. Katz offers a strong introduction to the structural inequalities academia contributes to and simultaneously denies. Those familiar with ethical problems in tech will recognize that many of the issues raised in this book are representative of shortcomings that extend well beyond AI. Katz’s writing is straightforward and manages to cover a very broad set of complex topics in a concise, organized fashion. Readers with little familiarity with AI will be able to take away plenty from this book. AI is pervasive across disciplines, with applications to criminology, economics, computer science, the social sciences, and digital humanities among others. Any librarian who will support research and projects around AI, or with an interest in technology and ethics, would find value in this publication. It’s a fairly short, easy read at 234 pages (excluding the acknowledgments, notes, bibliography, and index), and those looking for more in-depth reading on the subject will find the notes and bibliography to be very extensive. The dialog this book introduces is one worth having; I recommend the read.— Jasmine Clark, Temple University Libraries