Fear of artificial intelligence or fear of looking in the mirror? Revisiting the Western machine-takeover imaginary

Abstract What do we fear when we fear AI? This paper presents the claim that robophobia is autophobia, the fear of AI is the fear of ourselves, in at least two fashions. First, I frame the question in relation to what I call the Western machine-takeover imaginary, and identify two historical tracks: (1) the fear of non-human autonomy and (2) the logic of the genie. The first track is rooted in the idea of the domination of creations and their possible revolt, a trope that reflects the anthropogenetic narratives that guide Western self-understanding. The second track establishes a link between cultural understanding of non-human, wish-fulfilling others, and modern AI drones and chatbots in terms of potential asymmetries between inputs and outputs, instructions and execution. When biased inputs inform system outputs that inform human inputs in a hermeneutical circle, the outputs we fear cannot be separated from the inputs we generate.