Ghost in the Machine Examines AI’s History and Social Costs

Summary

PBS documentary Ghost in the Machine examines artificial intelligence as a collection of technologies shaped by human values, institutions, and power, rather than as an autonomous force with a single agreed definition. It traces a line from 19th- and early-20th-century eugenics and statistical tools such as correlation, regression, clustering, and IQ measurement to later ideas about machine intelligence and artificial general intelligence. The film argues that these histories can influence how people are ranked, modeled, and treated, while also emphasizing that race is not a biological hierarchy. It uses Microsoft’s Tay chatbot, which was removed after 16 hours when user interactions produced racist and pro-Nazi outputs, as an example of how systems can reflect their data and deployment environment. The documentary then follows Silicon Valley’s mythology of exceptional male entrepreneurs, technological optimism, and the pursuit of AGI and artificial superintelligence. Interviewees disagree about whether superintelligence or a singularity is plausible and when it might arrive; the film presents these claims as contested forecasts rather than established outcomes. It also focuses on the hidden labor behind AI. Kenyan workers who helped train ChatGPT through outsourced data annotation describe earning less than one dollar per hour and being exposed to graphic material without adequate counseling, while contributors characterize this structure as exploitation or digital colonialism. The film connects AI expansion to large data centers, noting a claim that training Meta’s Llama 3 used 22 million liters of water over 97 days, and it shows conflicts over water, emissions, and infrastructure in Nairobi and Memphis. Later sections examine synthetic media and the risk that high-volume generated narratives could weaken trust and informed public debate. They also discuss growing links between civilian AI companies and military systems, including surveillance, targeting, cloud infrastructure, and government procurement. The documentary concludes with a call to question whether AI is necessary for a given task, preserve human agency and accountability, and treat technological direction as a matter of collective choice rather than inevitability.