Photography is approaching another turning point in its nearly 200-year history. Since its invention, it has transformed how we see and understand the world, making visible things we had never seen before. It has served as a scientific instrument, a document, a form of personal memory, and an artistic medium. It has been declared dead, reinvented, collected, and institutionalized. Today, photography finds itself confronted by another image-making technology: artificial intelligence, capable of generating images that look photographic without the need for a camera or a corresponding event in the physical world. The enormous accumulation of photographs that once helped us understand the world has now become part of the material through which machines learn to synthesize it.
In the nineteenth century, Charles Baudelaire famously relegated photography to the role of a humble servant of art and science. Almost two centuries later, photography risks becoming the secretary of AI. Its images are no longer valued solely for what they represent, but increasingly for what they can teach machines to generate. This raises a fundamental question: how can photography retain its agency as an image-making medium without becoming merely a resource for another technology?
New Alphabets emerges from this crossroads. I see two possible responses to this condition. One is to keep pre-digital photographs in their original analog state, withholding them from digitization and the systems that might convert them into training data. But this preservation comes at a cost: to protect these images from machine consumption, we must also restrict their visibility and circulation. The second possibility, which forms the basis of this project, is more radical and contradictory: to destroy the photographs themselves.
Over time, I have collected thousands of discarded photographs, images that had already lost their original purpose and were abandoned by their owners. Rather than recovering them as archives or restoring their histories, I physically destroy them, breaking them into small fragments and reassembling them into new compositions. The original images become illegible, their subjects and narratives dissolved into arrangements of color, texture, and photographic matter. What remains is no longer a recognizable record of the world, but something closer to a visual language whose meaning cannot be readily deciphered.
An unexpected connection emerged through this process. By destroying photographs, I inadvertently produced something resembling visual noise, the very condition from which diffusion-based AI models can begin generating images. While these systems transform noise into increasingly recognizable representations, my process moves in the opposite direction, transforming photographic representations into noise. This relationship was not intentional, but it has become central to how I understand the work. Photography, once the source material for machine-generated images, is pushed toward a state that visually resembles their point of departure.
Yet the resulting compositions are not entirely without meaning. Within these thousands of fragments, viewers begin to recognize faces, buildings, flowers, objects, and other familiar forms. These images are not necessarily present as coherent representations; they emerge through perception, as the mind attempts to organize visual noise into something recognizable. The viewer becomes an active participant in generating images from the remains of other images. What was once a photograph of something becomes an unstable field of possibilities.
New Alphabets does not propose a return to a pre-digital condition, nor does it suggest that photography can fully escape the technological systems that increasingly absorb it. Even these new compositions can be photographed, digitized, and interpreted by machines. Rather, the project explores what happens when photography is stripped of its original function as representation and transformed into matter, from information into noise, from a recognizable image into an uncertain alphabet. In this transformation, destruction becomes a form of production, and illegibility becomes a way of questioning the future of the medium.
What interests me is not the disappearance of photography, but the possibility that, after two centuries of teaching us how to see, it might now need to become something we no longer know how to read.
