The world is not ready for artificial intelligence; we have stated this repeatedly. However, today, we will demonstrate why the use of AI in art forms, particularly photography, erodes the beauty of the medium. Remember that what we are about to share—or compare—comes from our deep affection for the visual medium. Photography has offered us something worth holding on to, and to see the art form being desecrated with machine prompts is heartbreaking. While you have already heard of producing images entirely using text cues, we are here to introduce you to a new advent called AI-enhanced photos.
The Commencement of AI-Enhanced Photographs
So what does the term ‘AI-enhanced’ mean? Simply put, to improve the image with the help of AI. The term was first introduced to me by WePresent’s website a few days ago, which showcased an interview with a photographer that befuddled me. At first glance alone, the pictures, which are created by Ezra Miller, seemed to comment on plastic waste. The reason is that the images show a large plastic sheet warping around everyday objects. However, after reading through the interview, one realizes that Miller’s work hopes to emphasize memory, especially how often recalling a moment from the past leads to confusion in a visually saturated world.
While reading the interview, I wondered, ‘Okay, that has happened in the past, too. Sometimes, the images rely on text to bring out the creator’s perspective.’ However, I later learned that the pictures in question were not created using mixed-media photography or an experimental approach. Rather, they have been ‘enhanced’ using AI. Miller employed ComfyUI, an open-source software that requires text prompts to generate a result. So, Miller fed his images of mundane everyday life, which were shot during his travels, to the machine, which were warped, bent, and folded by AI. “It’s exciting and infinitely inspiring because I treat it as an image-processing tool,” he told the interviewee. “For me, it’s kind of like a really good version of Photoshop. I use it to modify and experiment, and that’s where I see its potential.”
Why An Alternative Process Would Have Worked
Foremost, the AI-enhanced photos barely did what they were supposed to do: evoke an understanding of memory. From the time one laid eyes on Miller’s project, the pictures, while enigmatic, barely achieved their goal. You begin to grasp the concept only after you read through the interview. It’s said that a photograph is worth a thousand words, but in Miller’s case, his images require thousands of words of explanation. Visuals, in general, help to simplify complex ideas. You see a photograph, and you know what something looks like. However, if a picture needs more visual cues to explain itself, then it appears we are fighting a losing battle. Is this really the direction we hope to take?
Another point I would like to add is that AI-enhanced photos appear like brand campaigns. Miller’s photograph of warped mylar metal may get your attention initially, but the more you look at it, the more it seems devoid of emotions and the human soul. In fact, some images could actually work better as brand campaigns for a metal company. I am not being pessimistic here, but rather highlighting how AI takes away the imperfections of alternate photography practices, which makes any visual impossible to hold on to. In recent brand or movie campaigns, the images that are being disseminated appear completely fake. Before AI, people used Photoshop or other software to create a collage using different photographs. Now, it seems everything is being generated by machines, leaving a large vacuum behind in the creative space. For instance, Pepsi, Omega, and Levi push artificial images, but the overall aesthetic appears as if one person was responsible for creative language.
Miller’s photographs are now being published as a photobook. It is a head-scratcher for me since the crux of the idea is created with machines. So, what is the photographer’s contribution here? Feeding prompts?
The notion of memory, in fact, would have created more depth and emotion had Miller utilized the abundance of photography practices available to him. For instance, if you look at Blue Mitchelle’s Evanescent Energy series, you realize how his damaged transparencies’ colors, forms, and contrast portray the fleeting nature of memory. Similarly, Diana Guerra’s photobook, which was created with anthotypes, also depicts how remembering is inked with fuzziness and unreliability. Another example is Broken Sea, a project by Irina Sadchikova and Nata Sopromadze that combines past and present in one frame using film photography, where the photographers created double exposures sharing the film roles.
With each of the photography practice examples, you see how the pictures welcome the viewer to delve deep into the world created by the photographer. Every spec of grain, shifting colors, and hazy focus worked well to create an emotion that would touch you deeply. That is what makes photography such a sacred medium. Thus, an AI-enhanced photo can never evoke such a sentimental impact.