The Many Names of The Machine: Fear, Neutrality, and The Cinema of Artificial Intelligence, 1920–2026

For a century, cinema has returned to a single figure—the thinking machine that turns on its maker—and to a single feeling: dread. This article asks where that dread is actually manufactured. It pairs a qualitative coding of twelve canonical texts, from Karel Čapek's R.U.R. (1920) to the generative-AI films of the 2020s, with a computational content analysis of their scripts, conducted with a purpose-built instrument that measures fear vocabulary, sentiment arcs, propaganda techniques, dehumanizing language, and the language of violence. The results are unexpected. Explicit fear language never reaches one and a half percent of sentences in any script; the three-act emotional arc of every film in the corpus registers as neutral or faintly positive; the most recent films carry the least fear language of all; and the two films most affectionate toward their machines objectify them most heavily in grammar. The article argues that these apparent null results are themselves the finding. The fear of the intelligent machine is cultivated not in words but in the audiovisual apparatus—in lighting inherited from expressionism and noir, in the bodiless machine voice, in the unblinking close-up and the withheld cut, in production design that photographs the machine like geography. Drawing on cultivation theory, the Frankfurt School, and postcolonial critique, the article then asks who benefits from a fear installed beneath the level of language, and what a century of spectacular, misaimed dread has trained audiences not to see.