In the poster for her lecture “Nonhuman Photography,” Professor Joanna Zylinska clutches a film camera. There is a puff of smoke covering her eyes. This picture by electronic artist Rafael Lozano-Hemmer is cryptic upon first glance, but in light of Zylinska’s research, it illustrates the modern-day relationship between computers and people. With the mechanization of factory jobs and agriculture, viral videos of the social robot developed by Hanson Robotics named Sophia, more nuanced portrayals of artificial intelligence in popular films like Alex Garland’s “Ex Machina” and increasing hysteria about the robot takeover, it is easy to consider AI an enemy of humankind, a grave mistake. Zylinska delivered the talk, sponsored by the Media Studies program and Vassar College Journal of Philosophy, on Wednesday, April 3 in Sanders Auditorium. She is a Professor of New Media and Communications at Goldsmiths College, University of London, as well as a curator and a photomedia artist. She has presented her work at the Moscow Museum of Modern Art, the Serpentine Galleries in London and the art and digital culture festival in Berlin, called “transmediale.” She is also an author of seven books, including “Minimal Ethics for Anthropocene,” “The End of Man: A Feminist Counterapocalypse” and the titular “Nonhuman Photography.”
Vassar Professor of Philosophy and Director of Media Studies Giovanna Borradori introduced Zylinska with gusto, calling her a “magical creature and a magical scholar.” At the beginning, the lecturer acknowledged the contemporary interest that likely brought people to the event: “AI is kind of a big buzzword at the moment.” Then she showed a promotional poster for a photography contest held by the technology company Huawei in July 2018. The Huawei P20 smartphone uses AI photography and, according to the company’s website, the camera recognizes more than 500 objects and categorizes pictures into 19 scenes. Opening with this introduction to AI photography, she encouraged consideration of the fact that the algorithm itself is, in fact, the artist. Zylinska emphasized machine learning in the artistic realm and the relationship between robots and humans who are equally creative. She argued that image-making, including photography, has always been artificially intelligent in that it necessitates a mechanical apparatus; one cannot take pictures without a camera, or edit and process the pictures without algorithms.
As if to help reconcile the startlingly large role of machines in image-making, Zylinska likened the parergon-ergon relationship in visual art to that between robots and the pictures they create. Jacques Derrida distinguished between the ergon, or the work of art itself, the picture, and the parergon, or its frame, an accessory that embellishes the ergon and defines its limits but necessarily cooperates with it. The frame of a painting is just as beautiful as the painting itself and is part of the painting. The outside comes into the inside. Likewise, in photography the frame or ergon includes the creative machines that take the pictures and cultural commentaries about the pictures.
Today, algorithms and AI networks have enabled computers to see and take photos autonomously. Images can be generated automatically and no longer require a human viewer. Increasingly, pictures are taken by machines, for machines. For example, CCTV, satellites and cameras on drones take pictures independently, and not for human aesthetic appreciation. Even images taken by humans involve nonhuman elements, like the camera, editing software and cultural algorithms—subject matter the photographer is “programmed” to think is valuable or beautiful. This requires humans to reevaluate their role in artmaking, which Zylinska revealed to be smaller than many think, and their relationship to AI.
“It Began as a Military Experiment” is a 2017 piece by Trevor Paglen. It is a set of ten pigment prints of American government employees curated by Facial Recognition Technology, a database of thousands of these employees photographed between 1993 and 1996. “Humans cannot sift through all these data sets,” said the lecturer. Zylinska’s own art illustrates the possibilities of machine vision. For her project “View from a Window,” she used Amazon Mechanical Turk (MTurk), a crowdsourcing marketplace which allows users to commission employees online to do anything, from survey participation to, say, writing poetry or drawing—a fluid labor force that forms part of the cloud. She got 100 MTurk workers to take a photo outside of the room they were in at the time and made a flipbook of their creations.
The parergon of nonhuman photography not only includes image-making apparatuses, but also the socioeconomic consequences of machines performing “human” activities. Zylinska said that instead of concerns about robots taking our jobs, the world should begin to take heed of crowdsourcing and gig technology in labor, because of the emergence of services like MTurk and Uber which monetize daily activities. When such legions perform labor, workers are subject to exploitation and wealth is further concentrated in the hands of a few. These are subpar conditions for making art. Consider the entertainment and fashion industries, in which one person is often hailed as the sole creator of a company’s corpus. Accordingly, the lecturer urged attendees—as university students, consumers of gig services and users of social media—to resist the sharing economy and disengage a little without being a “technophobe.”
Assistant Professor of Music Justin Patch inquired about machine hearing: “Robots have produced things people want to look at, but [they] haven’t produced things people want to listen to.” Although this is “in an age when anything can sound perfect,” he mentioned that people enjoyed the “random deviations that…animate human behavior” and distinguish sounds made by people from those made by robots. The positions of Patch and Zylinska, however, intersect: Cooperation between humans and AI in art-making is necessary to consider, yes, but equally important is the creation of conditions that allow humans to satisfy their creative needs alongside AI. Reconsidering machine vision, computational photography and technology-facilitated services, she asked, “How can we act better in the world?” How can humans function optimally in the age of AI?
Unreal Person is a AI image generator that is trained on billions of human faces to generate a brand new face that does not exist.
The faces are the result of AI generated images, thus you can make sure that this person does not exist.