In James Coupe’s "Swarm," Surveillance Equipment and Profiling Algorithms Highlight Demographic Identity and Social Paranoia
On September 4, James Coupe premiered the responsive media installation Swarm at the Museum of Contemporary Canadian Art (MOCCA) in Toronto. Commissioned by the Toronto International Film Festival (TIFF), Swarm is a 16-channel, real-time video installation that uses profiling algorithms to auto-composite footage of individual museum visitors into demographically similar groups. Four rows of monitors show these different “clans” occupying the museum, proliferating in number, assembling and dispersing.
I connected with James to learn more about the installation, influences behind the work and how the profiling algorithms work.
Jenny Gill: Swarm is the culmination of your Creative Capital-supported project, Surveillance Suite, a series of installations that use contemporary surveillance technology to highlight demographic profiling practices. How did this iteration of the project develop?
James Coupe: In these projects I am interested in the paradoxical nature of our relationship with surveillance. On the one hand, we associate being watched by cameras with anxiety and paranoia, yet with the proliferation of camera phones and social media, surveillance has become utterly routine, and an increasingly popular strategy for lending meaning to our lives. Now we volunteer to be watched, and offer up the innermost details of our everyday lives as a matter of course. I’m interested in how surveillance has become a central concept in how we engage with the world and other people, and the kinds of visual, narrative and social vocabularies we have developed to cope with that shift.
The Surveillance Suite projects all make use of demographic profiling algorithms, tagging people according to different combinations of age, race and gender. So in (re)collector, a city-wide surveillance camera network attempted to reconstruct Antonioni’s classic film, Blow Up, from people’s everyday activities; in The Lover, the profiling software paired up visitors and cast them as the title characters in a modified version of Harold Pinter’s play of the same name; in Today, too, I experienced something I hope to understand in a few days, Facebook’s profiling algorithms are used to generate films that use people’s status posts as scripts; and Sanctum, a public art project recently installed in Seattle, combines all of these. All these works are systems that operate according to the logic of surveillance: fully automated; fitting into a predetermined and limited set of narratives; less concerned with what people are actually doing, and more with what it looks like they are doing; using algorithms to autonomously organize social connections.
Swarm is the eighth, and possibly final, piece to fit into this series. Noah Cowan, the artistic director of TIFF, invited me to propose a new work after seeing my presentation at the Creative Capital Artist Retreat in 2012. He was planning an exhibition at MOCCA to accompany TIFF’s focus on David Cronenberg, and saw some parallels with my work. Using Cronenberg as a reference point was a really interesting direction to take the surveillance projects. The installation has a strong physical presence in the space—about half a kilometer of cable, 16 screens, four cameras and over eight feet high. The cameras identify visitors’ demographics and archive footage of them in a database. Each set of four monitors has a panorama of the gallery spread across it, showing a video feed of the space, but occupied exclusively by a unique demographic of people. So one set of monitors might show only people of the same age, gender and race as the person who is in the space at that time. Another might show the current dominant demographic based on all the people in the system, another the minority. Another might show only men in the 20s, or people with glasses, or only people of a certain race.
I wanted to develop a piece that pushed a sense of social tension and had a feeling of something waiting to happen – a kind of territorial standoff. I’m interested in the kind of menace you might find in a Pinter play, full of everyday norms but constantly threatening to unravel and break down. I think museums and galleries are in many ways ideal places for this kind of thing—full of exclusivity and threat.
Jenny: You mentioned conversations with Noah Cowan about David Cronenberg’s work. Can you talk more about his influence and how you see this project relating to his work?
James: After Noah and I started talking about relationships between my work and Cronenberg’s, I spent a weekend watching all of Cronenberg’s movies—I wasn’t so familiar with his earlier work, and it really resonated for me. Particularly the repercussions of new technologies upon human beings—his work seems always concerned with mutation, be that biological, psychological, social, even ontological.
So in Scanners and Stereo, humans have developed telekinetic powers; in Crimes of the Future, a dermatological disease has killed all sexually mature females; in VideoDrome, humans develop malignant tumors from television signals; in ExistenZ, immersion in video games causes a collapse of the virtual and the real, etc.
I also read some J.G. Ballard, and was especially intrigued by his novel High Rise, which has a lot in common with Cronenberg’s film Shivers. In High-Rise, Ballard envisions a new social type that results from people living in close proximity in a modern, thousand-resident apartment building. Initially, the residents experience insomnia, and gradually become territorial about shared spaces such as elevators, parking spaces, swimming pools, roof gardens, grocery stores, etc. Eventually, the pressures of isolated yet claustrophobic living causes the residents of the high-rise to form clans, which are organized around class demographics—with the wealthy residing on the top floors, and the working classes on the ground floor. The situation rapidly becomes monstrous, as residents begin attacking and killing each other. The book intrigued me on a number of different levels, but what stood out was this isolation/proximity dichotomy—Ballard is imagining a situation in which these two things are not mutually exclusive, where you can be surrounded by people yet feel alone. I see it as analogous to contemporary social media dynamics: highly individuated, yet immersed within a network of people, and masking a strong need to be seen and form connections in order to feel significant. Virtual environments like Facebook are colonizing our social existences, in part because of their efficiency and viral expansion into our daily lives. Users are willing to offer up their privacy to social media systems because they simplify the individual management of our social lives. In return, we are confronted with systems that take control over who we know, what we know about them and how those relationships are framed, remembered, and proceed. An awareness of such networks has in itself caused the kinds of mutations I referred to earlier, perhaps into the kinds of individuals that Ballard wrote about: cool, unemotional, desensitized, with minimal need for privacy and capable of thriving on the “rapid turnover of acquaintances, the lack of involvement with others, and the total self-sufficiency of lives which, needing nothing, were never disappointed.”
So in Swarm, I am similarly concerned with the way we are being reshaped by modern technology, particularly in the relationship between surveillance, demographics and identity. Technology seems to be increasingly concerned with differentiating us, according to whatever criteria it has available—visual, social, commercial, etc. Once differentiated, it then tries to join us into illusory communities of similar individuals. I see Swarm as attempting to take this logic and create a situation in which we can confront its repercussions immediately.
Jenny: How does the technology in this piece work—the hardware and the software?
James: The piece uses five computers, running Linux and some custom software. It has a ring of four stereo cameras that capture a 360-degree panorama of the gallery space. The cameras each have two lenses, slightly displaced, which they use to calculate how far away things are that they can see. So the cameras detect people and calculate their precise location within the gallery. An algorithm estimates the demographic profile of these people, by identifying landmark features in their faces—eyes, nose, mouth, etc.—and comparing these to a database of around 50,000 faces to estimate age, gender, race and other things.
Once a person is successfully profiled, the software masks them out of the background and saves their demographics and their coordinates in the space. The computers are constantly calculating what kind of crowd of people they can display on the monitors, based upon this data. Each set of monitors is assigned a different kind of demographic target that it creates its composition from. As more people are recorded, the compositions on each set of monitors evolves accordingly.
Jenny: What do you hope the MOCCA visitors will take away from their experience with Swarm?
James: I feel there is always a question when you turn the camera on your audience of what you expect them to do. If you expect them to ‘do something interesting’ then you are being unrealistic—people are not usually in a gallery for that reason. So I’ve been more concerned with involuntary participation than interaction—I’m not expecting people to do anything, I’m more interested in who they are than what they do. A work I made earlier this year even included self-help style instructions as subtitles to cope with people looking lost in front of the cameras. So in Swarm, most people will just appear to be waiting for something to happen; anyone who does ‘perform’ for the camera is simply likely to further stereotype themselves according to their demographic—it is very hard for us to be anything other than our demographic when we’re on camera.
Beyond that, I’m looking for Swarm to be a tense experience, something familiar yet highly systematized and algorithmic. In a way, it’s about seeing ourselves as these kinds of systems see us, and recognizing the extent to which we have adapted to meet their needs.
Jenny: What other projects do you have in the works?
James: I would be interested in showing Swarm elsewhere, and am discussing taking a work I made earlier this year, On the Observing of the Observer of the Observers, to the Dürrenmatt museum in Switzerland, which would be great because the piece uses one of Dürrenmatt’s novels as a template. Sanctum, a public art work for the façade of the Henry Art Gallery in Seattle, recently opened and will be up until November 2015. I’m also developing some smaller-scale works, more gallery-sized than the museum installations I’ve been mainly doing this year. So I’m actively looking for places to show.
James Coupe’s “Swarm” is on view at the Museum of Contemporary Canadian Art in Toronto through December 29, 2013.