Code Dependent: Living in the Shadow of AI

In this mordant debut exposé, Financial Times editor Murgia goes into the global trenches where artificial intelligence is being rolled out and finds a proliferation of lousy jobs, impenetrable red tape, grotesque misogyny, and tyrannical surveillance. Among those she visits are poorly paid workers in Nairobi and Bulgaria tasked with labeling pictures to train AI systems; an African American engineer who learns that the facial-recognition systems he works on are prone to misidentifying Black people; an Argentinian government official who wrote an AI program to help prevent teen pregnancies that proved a useless failure; UberEats delivery contractors fed up with Uber’s opaque AI system, which routinely cheats and misdirects them (like when it dispatches them to long-shuttered restaurants); an English writer who discovered deep-fake porn of herself all over the internet; and activists battling China’s ubiquitous surveillance of Uyghurs, whose every step is analyzed by AI. Murgia’s vivid, sympathetic reportage looks beneath the grandiose promise of AI to get at the mundane reality of systems that merely automate the inept, callous, and unaccountable mismanagement and dispossession that ordinary workers and citizens already endure. She also intriguingly spotlights an interpretation of all this as a kind of “data colonialism” that extracts data from poor communities just like any other resource. The result is a biting and skeptical look at the brave new world of AI. (June)