Chris Dorland
Latent Stack at Lyles & King by Adam Simon, April 20, 2026
Chris Dorland, Untitled (tensor core), 2026, Metallic polymer, pigment, gesso, UV coating on linen, 36 x 46 inches Image Courtesy of the Artist and Lyles & King
I’ve often heard it said that the advent of photography in the 1800s paved the way for abstract painting by releasing painters from the burden of representation. This sounds logical but says little about the impact of photography. It doesn’t account for the seismic shift in human self-awareness and the reification of experience that came with photography. It took further technical developments and cinema, television, handheld cameras, etc. for that shift to become apparent. We are experiencing a similar shift now in the early stages of artificial intelligence and data processing. Our experience is being drastically reshaped by these technologies. There have always been artists that embrace new technologies and plenty of artists now are using AI and computer software, but there aren’t many that are engaging these technologies on a level that addresses this cognitive shift. Chris Dorland’s exhibition, Latent Stack at Lyles & King does so by making the underlying structure of digital imaging its subject and counterintuitively employing the very analog medium of paint on canvas to do so.
To me, the past work of Dorland’s evoked dystopian science fiction. These paintings were populated with what looked like digital code and its manifestations. I was interested in the way his paintings managed to address our digital world while remaining connected to gestural abstract painting. It felt like an unlikely union, but it worked. Viewers were able to navigate unfamiliar terrain because of formal strategies that had precedents going back to the 1950s. Latent Stack takes as its subject layers of data compression and the successive operations these layers undergo before they are realized as images on our screens. As I understand it, Dorland begins these paintings with imagery generated and then degraded using artificial intelligence and custom software. At some point the work shifts to paint on canvas using metallic polymer, pigment and gesso, with a palette keyed to the CMYK color space of digital printing. Dorland builds up layers that allow earlier layers to still be visible, a kind of sequencing that directly parallels the stacking of layers in computerized image generation, hence the title. In addition, the floor of the gallery has been covered with reflective vinyl extending the paintings into an environment the viewer inhabits. Compared to his past exhibitions, these paintings have less text, patterns or distinct images that trigger recognition. What imagery exists is buried beneath the multiple layers of striated, abraded and otherwise messed with paint. Initially, these paintings may seem harder to parse as compositions, but they are convincing representations of the hidden layers undergirding our virtual lives. Each painting can be understood as portraying a segment of endless data flow, subjected to compression, diffusion and other forms of transformation.
Dorland’s paintings begin with ‘latent representations’, the hidden layers of computer-generated images. These involve processes of compression including those that are integral to AI diffusion models. Diffusion models are one way that artificial intelligence learns. Visual noise is added to images, rendering them indiscernible, and the model learns to reverse the noise giving it a structural understanding of how to generate future related images. Diffusion models involve a lot of failure modes, forms emerging from and collapsing back into noise. Dorland seems to be translating this irresolution into paint.
Chris Dorland, Untitled (latent stack), 2026, Metallic polymer, pigment, gesso, UV coating on linen, 75 x 60 inches and detail Images Courtesy of the Artist and Lyles & King
I don’t think it matters if the paintings function as actual translations of computational latent representations. I’m ok thinking they do so metaphorically, like how Barnett Newman’s Stations of the Cross both represent and don’t represent Christ’s passion. At the very least it is remarkable that Dorland has found equivalencies between data processing and the processes that make an abstract painting, both subjecting an initial impulse to multiple forms of disruption, diffusion and obfuscation before arriving at an image. The paintings in Latent Stack with their palette that combines crimson and blue/black can feel forbidding. The imagery is more subdural than submerged. The radial striations register as scan lines. Multiple pockets in the surface where the paint pulled away register as glitches in the system. And yet, against this weightiness there is Dorland’s obvious pleasure in the making of these paintings and in the limitless possibilities of paint as something that embodies both thought and experience. In fact, these paintings are exuberant. It also occurred to me that abstraction is the only mode of painting able to carry this subject. More significantly, they don’t look like most abstract paintings. They look like something sensed but never seen.
Painting has always mirrored changes in culture. The subject matter in late 19th century European painting reflected the breakdown of class hierarchies. Abstract Expressionism channeled Freudian interest in the subconscious. In our current paradigm, the approach favored by most painters is to try to ignore the world of digital code, social media and the algorithms that mine our personal data. Painting is seen by them as a counterforce, offering contemplation, slowness and sustained focus. Others, more responsive to a changing world, have adopted AI and other forms of digital software as tools for image generation. Dorland doesn’t fit either category. He is as much interested in painting as a language model as he is in digital coding or artificial intelligence. That he maintains this dual focus is remarkable and his decision to align such disparate languages amounts to a utopian project. Rather than using new technology as a tool to better do what was already being done, Dorland is treating it as a subject, describing something pervasive we are unable to see, an invisibility that is reshaping our lives.
Installation view. Image Courtesy of the Artist and Lyles & King