Between all the images of Unicorn Frappuccinos, grain bowls, and scenic vistas, it doesn’t take an expert to tell you that clichés are running rampant in photography. To photographer and PhD student Max Pinckers and media artist Dries Depoorter, those unoriginal snapshots are especially problematic when it comes to photojournalism. We see the same sensationalized images over and over again. But do they tell us anything new and are they an accurate depiction of reality?
“Press photography appears to be becoming a self-referential medium dominated by tropes, archetypes, and pop-culture references,” Pinckers tells Co.Design via email. “What implications does this have on how we learn about the world through the images we are being shown?”
As a commentary on what they perceive to be the staid state of photojournalism, Pinckers and Depoorter designed the Trophy Camera, a device that uses artificial intelligence to ensure that every photo it takes is destined to be an award winner.
Here’s how it works. Pincker and Depoorter used machine learning to train an algorithm to discern the composition of award-winning photographs. They analyzed every World Press Photo winner from 1955 until today and deconstructed them into themes–simple categorizations like people, man, woman, soldiers, portrait, weapon, and so on. When a person takes a picture with the Trophy Camera, the algorithm gives the photo a rating based on how closely it hews with the most popular characteristics of World Press Photo winners. If there’s a 90% correlation or higher, the camera takes the photo and automatically uploads it to a website.
The camera is very basic–essentially a Raspberry Pi computer, a tiny OLED screen that reads out what the algorithm sees (inside, room, kitchen, etc.), and a red switch that acts like a shutter button. One thing that’s conspicuously missing from the camera? A viewfinder. Pinckers and Depoorter intentionally give people as little control as possible over the images the camera produces–a commentary on how little the creativity of an individual seems to matter today. They argue that photographers are already thinking like formulaic machines.
“Cameras are becoming more automatized everyday, with capabilities of producing ‘perfect pictures’ with the push of a button, which leaves less room for creative interventions within the program defined by machines and technology, and therefore only creates redundant imagery,” Pinckers says, nodding to an argument media theorist Vilém Flusser’s puts forth in Towards a Philosophy of Photography. “How will machine learning and computer vision define the way we make photographs in the future, and how will AI idealize our human experiences based on formulas designed for maximum visual impact?”
In reality, none of the images from Trophy would garner any accolades, but that’s not really the point. Taken at an exhibition about the camera, the images are mostly blurry and poorly lit. The camera’s merits are more conceptual, as the photos on the Trophy.Camera site show. Pinckers and Depoorter hope their project leads people to become more aware and skeptical of what media they consume. It’s not just a fake news problem; it’s photo retouching, it’s the editorial decisions surrounding what gets depicted, and it’s bias embedded in technology. Trophy Camera isn’t the first device to tackle the similarities in photography today, either; Camera Restricta uses GPS to limit whether or not someone can take a picture if its algorithm senses too many shots have been geotagged at that location.
“By making this camera, we try to implicitly comment on the current status of photojournalism–which seems to be becoming more questionable in today’s visual landscape–along with the incredibly fast development of computer vision and the relevance of artificial intelligence in our time,” Pinckers says. “The Trophy Camera raises many questions we still can’t answer.”