How Trevor Paglen anticipates a dystopian future

At first glance, Trevor Paglen’s photographs seem attractive and reassuring. The colours of the skies in his series Untitled (Drones) evoke both the azures of certain old masters and digital abstractions, while an impression of abstract virtuality is particularly at work in the series The Other Night Sky, for which he photo- graphed satellites and space debris in a clear starry firmament. While rather more blurry and ethereal, the wooded landscape in They Watch the Moon is not without certain pictorial qualities. All these images are reassuring in a certain sense, since they mobilize a known imagery that blends scientific space photography with the history of painting. Their scope and significance, on the other hand, are far more conceptual and political. They Watch the Moon, for example, is in fact a photograph of a secret U.S. Army listening station located in the forests of West Virginia and Maryland (where Paglen was born in 1974), a “silent” zone where radio transmissions are restricted and Wi-Fi and FM are prohibited so that the station can pick up signals from the ECHELON surveillance net- work. What reads as aesthetic pic- torialism comes from the fact that getting the shot required a very long exposure in moonlight.

Similarly, Untitled (Drones) is not a simple reinterpretation of the artistic motif that is the sky: the drones in the photographs – which are almost invisible – are deadly arms of combat. “A commonly held idea is that there’s an equivalence between an object’s image and its reality,” comments Paglen. “Is the image of the sky or of space that I make in 2020 so different from the images taken in the 19th century or painted on cave walls by prehistoric man?” And yet the reality of the sky in 2020 has nothing to do with a pan- theon of pagan gods or a place of fantasy waiting to be explored; in- stead it can be considered a space of generalized surveillance and mili- tary and strategic objectives. “As abstract as they may seem to us, the new forms of power and economics, like algorithms, surveillance or facial recognition, are always linked to very real infrastructures,” insists Paglen. “The landscape that has always been at the heart of my work is a landscape of infrastructures, the es- sential components in the new forms of power.”

Paglen recently teamed up with political scientist Kate Crawford, a Distinguished Research Professor at NYU, for a project that was shown at the Fondazione Prada in Milan as well as at a Prada runway show in Paris this January. Entitled Training Humans, its starting point is the ex- traordinary quantity of images that scientists use to train artificial intelli- gence. “They teach them how to recognize a nose, a smile, a mouth, and how to calculate the distance between them,” explains Crawford. The CIA’s first such experiments, in 1963, which used government pho- tos, have given way to research that uses images of people harvested on the internet without the slightest au- thorization. “The other problem,” continues Crawford, “is that these photos are used to categorize hu- man beings according to race, class, gender and even personality. Artificial intelligence is being trained to predict your emotions just by watching your face. There’s nothing scientific about it, and it recalls the worst periods of the 19th and 20th centuries when often racist theories such as physiognomy were thought to be able to establish an individual’s destiny through an examination of his or her facial characteristics.”

The exhibition takes the form of a materialist study of these images, which the scientists use without even looking at them. But, more than anything, Training Humans uncovers the ideology that is inherent in all technologies, which are never neu- tral in the manner in which they are constructed. “We give meaning to these images, to the image of a per- son for example, but this meaning is a construction which often has only a pseudo-scientific basis,” explains Paglen. “Nonetheless these images have a real impact on the world. A phrase I spotted recently in a maga- zine struck home: ‘We knew that images could kill. Today they have their finger on the trigger.’” An ob- servation to which Crawford adds, “Only a dozen companies today have the economic and logistical capacity to create artificial intelligence. There’s absolutely nothing abstract about that. They’re not just algo- rithms and data in a cloud. They are infrastructure and people. And very definite political decisions too.”