The Brooklyn Rail

Ignorance is bliss, or we can at least agree that it is what allows magic to exist. For the few lucky ones unacquainted with James Casebere’s work, the recent solo exhibition Seeds of Time at Sean Kelly was a concise but powerful opportunity to be amazed and fooled by his work for the first time. In an age where natural language description has made artificial intelligence (AI) image generation and manipulation widely accessible and has, consequently, flooded the internet with its residues, it is rather unusual and almost nostalgic to be intrigued in such a way by an image again. Especially now that the uncanniest images ever made are at the tip of our fingers and only a few taps away—Salvador Dalí, Giorgio de Chirico, and Leonora Carrington would be jealous of the ease with which anyone can produce surrealistic imagery nowadays. Nonetheless, one hundred years have passed since the Surrealists had their apogee, and culture (meaning the processes through which we produce and consume images) has since revolved 720 degrees, redefining what it takes to create something magical today.

The exhibition showcased ten new, lusciously printed works on the main gallery walls. As a collection, the images depict quasi-architectural objects in rich, saturated colors that starkly contrast against their monochromatic backgrounds. Sharp lines and directional lighting accentuate the centralized structure of most works, drawing the viewer’s gaze toward a hidden center or emphasizing the singularity of their depicted subjects. In all of them, water substitutes the ground, operating as a mirror, reflecting the scene in a distorted double. Compositionally, the works are clear and straightforward. Emotionally, on the other hand, the works negate the simplicity of their formal composition by conveying mystery and intrigue. Something is always subtly off in these architectures, creating a powerful tension with our memories of the real, raising suspicion and inviting the beholder to scrutinize the images for a little longer.

At first glance, one could mistakenly discard the works in the show as amateur AI-generated artwork because of these aesthetic qualities. Semiotically, Casebere’s works and deep-learning images rely on the ambiguous deployment of discernible signs to construct/suggest their meanings through a strategy of approximation that ties the two together. They both represent what we can refer to as distorted realism, a sur-realism akin to the fiction of dreams. This strategy is manifested more explicitly in the transfigurations of familiar subjects (architecture) via unusual settings, unnatural colors, and unexpected materials, thus producing an approximate representation of a referenced real. During my visit, I even began to formulate fictional prompts as I walked through the exhibition—“colorful architectural model of five houses in the style of Balkrishna Doshi standing on a pool of water” or “abstract architectural form made of greenery on a reflecting water surface”—simulating what the artist would have had to type on his computer to generate these images. However, for those familiar with Casebere’s work, these images continue a career-long experimentation with large-scale physical models, photography, and lighting. 

Eventually, I realized these were, in fact, not AI-generated images but photographs, a revelation that delivered a sudden sense of satisfaction similar to what one experiences with a magic act. Casebere now executes his (inter?)medium with an undeniable level of mastery like a seasoned magician pulling a rabbit out of the hat. The polished craft and skill in this collection of work are more than sufficient to land a simple trick; more importantly, the careful construction of the physical models and their translation into flat images through photography enables them to flirt with the limits of perception and positions the work in conversation with contemporary tools of digital media production. In other words, the shared medium of the digital image serves as a common ground to interpret Casebere’s images beyond a simple stylistic coincidence and further consider them as a disciplinary response to works produced by text-to-image machine learning technologies. 

In these aesthetic and material overlaps between two radically different media types, one might identify a potential new line of investigation on Casebere’s work, which might be particularly interesting to architecture. As architectural production integrates AI tools into its workflows and theory, a division between those experimenting with AI-generated images and those building and photographing large-scale models is noticeable in the trend of representational strategies favored by contemporary practices. In Seeds of Time, James Casebere proposes the possibility of producing work that operates as a mediator, establishing the grounds for potential dialogue with new methods and tools of representation. This work also inevitably raises the question of how ease of production and reproduction influences our perception and evaluation of any artistic endeavor. When something is readily available, easy to use, or commonplace, it not only loses its power to amaze but also its ability to question the status quo. 

In that sense, unlike AI-generated images, Casebere’s process is not left to the devices of automatized probability. The behind-the-scenes documentation of his practice reveals the precision necessary to produce these photographic works, adding a layer of intention to the work that remains perceptible in the finished product. This fundamental difference in the production processes of the images points to the true possibilities that working in this way enables: the opportunity to make decisions. The dreamlike qualities in Casebere’s images are thus rational, establishing a distance from the surrealistic architectures produced by the subconscious of the internet. Casebere’s work ultimately provides a model of what architecture should aim to create: critical, precise, and intentional objects of the mind that deliver surprise, awe, and delight.