Performing the Future : Gendered Bodies and Cinematic Scripts of “AI Professions” in AI Imagery

Sammanfattning

This study investigates whether generative AI produces genuinely novel visions of future bodies or whether it merely automates the recycling of pre-existing cultural scripts about gender, technology, and power. Using a cross-media methodology, 384 images of three emerging AI-related professions (AI Personality Designer, AI Ethics Officer, AI Content Creator) were generated via Midjourney across default, male, female, and non-binary prompts, then compared with cinematic AI representations from Her, Ex Machina, and Alita: Battle Angel. Findings reveal interconnected patterns. First, under default conditions, Midjourney overwhelmingly materializes the AI body as a human-machine hybrid, yet this hybridization remains strictly bound to humanoid templates. Crucially, anthropomorphism is inseparable from gender, with female-presenting figures dominant. Second, explicit gender prompts expose female -presenting AI professionals as more anchored in organic humanity compared to males. Most strikingly, when prompted for “non-binary AI professionals,” the model retreats to the purely organic human realm (87.5%), stripping away technological markers and reducing non-binary identity to depoliticized stylistic signifiers. Third, comparative analysis demonstrates that these patterns echo those found in the cinematic archive. In Her, Ex Machina, and Alita: Battle Angel, femininity appears in three mutually exclusive forms: warmth without body, mechanical body without warmth, and both mechanical body and warmth as pathology, a logic that recurs in Midjourney’s outputs. The study concludes that Midjourney does not produce genuinely novel visions but rather automates the recycling of gendered cultural scripts encoded in the cinematic archive. Changing this imaginative landscape requires not merely inclusive prompts but the deliberate production of new images that expand the boundaries of what is culturally imaginable.