Discriminating Data: Correlation, Neighborhoods, and the New Politics of Recognition

Public discourse over AI has been intensifying for years, ranging from the breathless anticipation of technolibertarianism—which Barbrook and Cameron's (1995) “Californian Ideology” noted combines C18-era libertarianism and science fiction—to questionable fears that an impending technological “singularity” will give rise to sentient AI capable of exterminating human society (Proudfoot 2012). Political theorist Wendy Hui Kyong Chun's Discriminating Data: Correlation, Neighborhoods, and the New Politics of Recognition offers hope that the era of such naïve extremes is coming to an end. In doing so, she sets cultural metanarratives in favor of critique of the data infrastructures and methodological assumptions at the technology's core. This materialist approach has the added benefit of highlighting the historical dependencies of contemporary data science methods. Chun has thereby done the various fields that constitute digital studies a great service: isolating the engine room of contemporary AI and opening it up for humanistic analysis. The book stands out as a watershed, clearing away uninformed commentary and focusing our minds on the statistical methods now deployed across private and public lives. Its overtly educative function, including accessible handwritten notes by mathematician Alex Barnett, adds to the sense of a transdisciplinary inquiry capable of initiating a new research domain.