In 1983 Hal Foster edited The Anti-Aesthetic, a super-slender lavender paperback that helped define postmodern critique. (Foster revisits the book and its import in an Art Forum interview, link down yonder.) Its subtitle is “Essays on Postmodern Culture,” and it collects writings from Jean Baudrillard, Frederic Jameson (who, as it turns out, was teaching at the PhD program I would eventually attend), Rosalind Krauss, Edward Said — artists, critics, and theorists questioning the dominant modes of art production at the time, so late 20th century, very nearly pre-mass internet.

I discovered the book by happy coincidence. It more-or-less configured some additional trajectories for myself professionally and in my scholarship. See, I was a grad student working for my rent and beer money at the HITLab, the University of Washington’s VR-centered Human Interface Technology Lab. I was researching and building immersive visual environments while also, coincidentally, shooting film on the side — what we used to call photography. So, I needed a darkroom. My manager Tanya at The Bulldog News — a newsstand in the University District where I also worked because, you know..rent and beer — introduced me to the photo department chair, Prof. Paul Berger. He became a mentor, and eventually pointed me towards this book during one of our many coffee chats. Thus began my education in aesthetic theory and critical theory, a framework that would augment my engineering skills with new ways of seeing.

If Foster’s title sounded like a rejection of beauty to you, you’d be understandably mistaken. What I read was the way he and his contributors were questioning the condition of looking. They wondered about how late 20th century culture had turned art into surface, spectacle, and consumption. They weren’t against form; they were against the assumption that form was neutral.

Recently, discussing DOGMA 95 in the context of generative AI, I found myself thinking back to The Anti-Aesthetic. The collection argued that postmodern art needed to resist the spectacle of smooth surfaces and slick production values. The anti-aesthetic was about making work that revealed its own construction, showed the seams, embraced imperfection and contradiction.

Generative AI has revived this problem at industrial scale. The tools promise infinite images, instant style transfer, and a frictionless feed of visual novelty. Every prompt produces something “finished.” But that finish is the problem: an aesthetic of smooth competence detached from intention, history, or risk.

Foster’s point feels newly relevant, if not urgent. He asked us to see aesthetic production as a social operation, not a decorative act. What’s being automated by AI is not just technique; it’s judgment through the slow negotiation between maker, material, model and meaning. When a system fills that space for us, it collapses the moment where thinking actually happens. Thinking upon this, I can relate to the feeling. Sometimes I’ll find myself wondering — wait..what am I actually trying to say with this image? Why this subject? Why this composition? The AI makes it easy to generate something that looks good, but that ease can obscure the deeper questions about why the image matters at all.

Perhaps what we need now is not better “AI art,” but better questions about what art becomes when generation replaces composition. The anti-aesthetic today might mean making work that leaves visible evidence of decision: why this subject, why this frame, why this image at all? It might mean treating AI as a drafting instrument rather than a vending machine—pausing before publishing, asking what story the image tells about the world that made it possible.

In Foster’s moment of The Anti-Aesthetic, postmodern artists responded by quoting, sampling, re-contextualizing. Their critique was to make the mechanisms of culture visible.

Ours could be similar: to show the scaffolding of generation, expose datasets, treat the prompt log as part of the work. Not to moralize about machines, but to restore a sense that meaning requires participation and that the interval between intention and the digital image is where imagination actually lives.

If the 1980s needed an Anti-Aesthetic to push back against spectacle, the 2020s may need an AI Anti-Aesthetic (not necessarily an anti-AI or anti-AI aesthetics) to push back against automation. The task isn’t to reject AI outright. It’s to reclaim that interval and locate the place where thinking happens.