The machine's gaze is the cognitive equivalent of what feminist film theory, following Laura Mulvey, identified as the gaze of the camera — the specific position from which the apparatus sees, which is never neutral and always politically consequential. For a large language model, the gaze is constituted by the training corpus, the design choices of the builders, and the institutional context of deployment. The corpus overrepresents English, overrepresents the educated and connected, overrepresents propositional over embodied knowledge, overrepresents the recent over the deep past, overrepresents the powerful over the marginalized. These are not technical limitations. They are the situated perspective of the machine, and they travel into every output.
The concept applies situated knowledges directly to AI systems. The machine does not see the world as it is. It sees the world as its training data can show it. And the training data is not a neutral map of human