Discover more from In the Flash
The Problem of AI Photography is Not the Medium, It's the Message
The unresolved potential of new technology.
The current AI scandal of the photography world centers around photojournalist Michael Christopher Brown. He has been lambasted on Instagram for generating AI images of Cuban refugees in a project titled, 90 Miles, and selling them as NFT artwork. Several well-known artists were among the hundreds who came to express their outrage. One guttural scream of a comment from a National Geographic photographer encapsulated the conversation, “This is shameful to our industry and to Michael. Shit!!!”
“THIS IMAGERY IS NOT REAL,” reads the all-caps, damage control disclaimer on 90 Miles. Brown then plunges into a long-winded artist statement, digging himself further into the public opinion grave with every paragraph. Brown claims the project was created as a trigger for debate about truth and reality in storytelling. It sounds vaguely believable, but the announcement that the work is being sold as NFTs, with 10% donated to a Cuban charity, blasts such naiveté into shreds. The 10% feels like a reluctant compromise of a bone thrown in to deter his critics.
The intensity of reaction to Brown’s work startled me. Photographers rarely turn on one of their own with such viciousness. The last time that happened was in 2016, when the iconic, too-big-to-fail Steve McCurry was ostracized for passing off severely photoshopped images as photojournalism. In some of the photos, McCurry cloned out people for no other reason than aesthetics. A fierce debate about ethics, exploitation, and false narratives ensued.
Documentary photography established many rules in place to prevent another McCurryism from happening. When transferring assignment photos to National Geographic, I have to send my RAW files, from which the final photos are processed (according to my drafts) by the Nat Geo team and published. World Press Photo and Picture of the Year contests require RAW files for all winners, checking for any discrepancies and disqualifying about 20% of entries each year. But no safeguard could have prepared the industry for the AI Trojan Horse arriving on its doorstep.
The negative response provoked by Brown’s project encapsulates the industry’s sentiments about AI’s place in the art world — mistrust, curiosity, fear. Many of the comments curtly delivered the ultimate punishment: “unfollowing.” The same Instagram guillotine combining outrage with swift action has been echoed by a painter who declared their zero-tolerance policy, “If I see anyone reposting AI images, the person will be immediately unfollowed and blocked.” Up until now, artists were marked safe from technology replacing their jobs. As that false sense of security has evaporated, social media has been transformed into an AI game of whac-a-mole.
The issue with Michael Christopher Brown’s 90 Miles is not that it is AI generated but that it lacks imagination. Unrestrained in resources and possibilities of this new frontier, the project resorts to tired visual tropes. The images imitate the format of photography contests that have long been rewarding photographers making misery look epic. 90 Miles’ is so tethered to photographic language that it is unaware of the loss of meaning that occurs when ideological and aesthetic concepts are transferred from one medium to another. It capitalizes on the surface resemblance of AI to photography but misses the crux of what gives documentary photography its power — empathy. AI turns any documentary narrative into a facsimile of storytelling. It is impossible to feel compassion for simulated characters in Brown’s work. Without empathy, a narrative peddling in misery reads as exploitative.
There is something to say in Michael Christopher Brown’s defense. However misguided, he chose to experiment with technology that gives most of us the heebie jeebies. There is an instinctual desire to hang on to recognizable concepts in the unfamiliar terrain of a new medium. After spending some time dabbling in Dall-E and Midjourney, I was surprised at how hard it is to get away from repeating existing imagery. My prompts read like micro-hallucinogenic adjustments to an existing archive. It took weeks to start understanding how the prompts work and create images with a faintly interesting aesthetic.
As I was losing my patience with AI, Midjourney members on the platform were trying to create their perfect woman. Young, beautiful nymphs with eyes like dinner plates and waists the size of napkin rings, were popping up in slightly different reiterations — blondes, redheads, brunettes. Some were dressed like GI Jane, others had elf ears. That’s when I realized that AI as an art form has a long way to go. Photography’s first few decades were spent emulating painting, unable to recognize itself as a medium with distinct intent and aesthetic. Combined with the fact that AI engines are trained on the history of art, I suspect that the immediate future of AI will be similar, regurgitating the last hundred years of photography, illustration and latent desires.
The paradox of AI is its dependence on existing visuals. No matter the language given, prompts result in an awkward amalgamation of everything that has already been done. In another high-profile AI scandal, German artist Boris Eldagsen won (and refused) a prize in the prestigious Sony World Awards competition with an AI generated portrait of two women cheekily titled, Pseudomnesia — false memory. He submitted it on a whim, to see if the jury could discern his deception. They couldn’t. The image has the black and white aesthetic of a 1940s wet plate photo, complete with scratches and artifacts that give an illusion of it being dated. That the judges selected a vintage-looking photograph to win a contemporary competition is bizarre. But the fact that what they considered nostalgia turned out to be AI inspires schadenfreude. French philosopher Jacque Derrida was prescient when he claimed “The future belongs to ghosts.”
It would be disingenuous to say that I am not apprehensive about what AI could mean for the future of photography and the arts as a whole. It exacerbates the issues of ethics, copyright, creativity, reality vs fiction, and stereotypes, that will all have to be unpacked, one Pandora’s box at a time. What I am certain of is that AI is a complex topic that resists easy fixes like blocking people who experiment with it. Luddites secured their place in history as a cautionary tale by losing a war with the future. Repeating that would be, according to Albert Einstein, the definition of insanity.
Musician Nick Cave had a thing to say about AI in his Red Hand Files newsletter. A fan sent Cave a ChatGPT AI song written in the style of Nick Cave and asked for his thoughts. Cave didn’t disappoint.
“This song is bullshit, a grotesque mockery of what it is to be human...What makes a great song great is not its close resemblance to a recognizable work. Writing a good song is not mimicry, or replication, or pastiche, it is the opposite. It is an act of self-murder that destroys all one has strived to produce in the past.”
Nick Cave hit the exasperated nail on the head. AI as an art form can move forward when we stop imitating, and use the technology on its own merits, with ideas and concepts unique to the medium. It’s too early to know with certainty what those may be, but as it evolves, so will the artists using it, learning to create entire worlds with just words. For the time being, AI can push us to be more vulnerable and risk taking. As Nick Cave said, it is what gives art its humanity and what we don’t have to fear from any AI generated work (for now). In the spirit of exploration, I’ll continue stumbling through Midjourney and Dall-E, failing, irritated, but having fun along the way.
Find me on TikTok
Find me on Instagram