J. Louise Makary

My research considers how theories of vision are complicated by generative AI, with particular attention to earlier hypotheses about vision attending the invention and use of cameras. Film and photography, before they emerged as artistic mediums, developed as technologies to aid in scientific studies of movement, and the camera uncovered questions about how we see and how we can be tricked by our eyes. Thus, we can regard photography and cinema as preliminary experiments in what it means to externalize vision through the use of machines. What we see with the aid of AI is also the externalized vision of machines—machines which, though not equipped with sensory organs or even lenses, demonstrate computational sight. However, while we humans tolerate and often find pleasure in vagueness and imprecision in visual cognition, task-based computer vision is oriented toward certitude. My work with film engages with its states of indeterminacy and their effects on cognition and subjectivity (assessing, for instance, the experience of viewing images of split-second duration), and compares our encounters with cinematic and AI-generated images to better understand what it means to see.