Between the Human and the Machine: In Defense of Uncanny AI Art

As AI systems increasingly produce humanlike output, the boundary between human and machine becomes harder to locate. This paper argues that this boundary, and the uncanniness that emerges from it, is a site of revelation rather than a problem to be solved. Tracing the evolution of computer vision from early image classification through generative models, this paper shows how each technological development brings AI closer to simulating theoretical frameworks of human cognition. In building machines that process information in the ways we theorize ourselves to do so, we have made AI a mirror that reflects our self-perception back to us so convincingly that what lies beneath becomes invisible: that the machine is built from us, and that in its seamless function, we disappear into it. Drawing on philosophical ideas from Theodor Adorno, Sigmund Freud, Masahiro Mori, and Martin Heidegger, I argue that as AI grows more humanlike, we become standing-reserves within it. We prefer AI to remain invisible—working as intended—because to see it clearly is to confront that we have become resources. AI art that refuses this invisibility and stages a tension between the human and the machinic is uniquely capable of revealing this truth. Through case studies of several AI artworks, I argue that this kind of uncanny art is increasingly urgent as technology advances.