The next time you comb your hair in a public restroom or aimlessly pick a stray strand off your coat while waiting for the subway, take a moment to think about the personal information you’re leaving behind. If you’ve seen enough crime shows on television, you know that hair follicles contain unique DNA sequences, from which a crafty scientist (or, increasingly, a savvy hobbyist) can glean all kinds of personal information about you. And while leaving your locks all over town isn’t necessarily as bad as, say, walking around with your Social Security number tattooed to your forehead, imagine what someone could do with that info.
In the case of artist Heather Dewey-Hagborg and her Stranger Visions project, she’s picking up stray hairs, cigarette butts, chewing gum and finger nails in public places like bus stops, restrooms, restaurants, and wherever else people might inadvertently leave traces of themselves behind, and using these samples to sequence and analyze the DNA contained within. She then uses this information to construct speculative portraits of what these anonymous shedders might look like based on their genetic profile. The project is currently on view at the Eyebeam Annual Showcase (through January 26th) and is part of an open studio presentation at The Clocktower Gallery on January 24th.
Dewey-Hagborg first conceived of the idea, which explores the dystopian future of genetic surveillance, while contemplating a stray hair caught in a crack in the wall at her therapist’s office. “I began thinking to myself, ’I wonder who would lay down on that couch when they go to therapy. The more I thought about it, the more curious I became about who this person could possibly be. As I just mulled it over, I started thinking about how I could figure out more about that person and, you know, kind of connecting that with all of the forensic shows you see on TV. The idea kind of stuck in my head.”
Though she didn’t have a background in biology or genetic research, she put together a proposal and started shopping it around to various residency and grant programs. She finally ended up as a 2012 resident at Eyebeam, where she began developing her ideas in earnest after taking several introductory crash courses in genomics that taught her the basics of analyzing sequenced DNA.
Dewey-Hagborg collects found hair samples and uses the DNA to construct her portraits.
Using facial recognition algorithms she had worked with in the past, Dewey-Hagborg started collaborating with biologists at Genspace, a community biolab in downtown Brooklyn, and began building a 3D modeling software that would reconstruct the hypothetical visages of her mystery strangers. By identifying known parts of the genetic code that are associated with specific physical traits, then using a 3D facial modeling software developed by some researchers in Basel, Dewey-Hagborg was able to construct portraits of people based on their DNA.
To date, she’s analyzed herself, the DNA of several people who have open sourced their genomes online, and five samples found in public. She’s working on sequencing additional found samples and perfecting her 3D modeling software. But what are the implications of a straight-from-the-pages-of-science-fiction project like this? We spoke to Dewey-Hagborg to find out more about her creative vision and the paranoia that drives it.
Dewey-Hagborg and her self-portrait.
The Creators Project: This type of genetic analysis was totally new for you when you began this project. Your previous work was more in the vain of…
Machine learning has been kind of my area of expertise. I’ve been working at the intersection of art and artificial intelligence for about 10 years and in lots of different media. I usually call myself an information artist because the thing that kind of connects the different things together is some interest in the abstract idea of information and information processing—kind of how humans process information, how machines process information, and what distinguishes us from each other. Most of my work in the past has dealt with that in some realm, but I think what connects all of that together is that there is a kind of criticality of the algorithm and [a criticality] about the way algorithms represent parts of nature and the way the coder gets represented in the algorithm they’re designing. That’s a big interest of mine.
So you decided to tackle the original algorithm—DNA.
Yes, exactly. It’s a cool way to do that.
What are you working towards? Are you trying to make the most accurate portrait possible? You don’t even necessarily have a way of verifying that, especially with the samples you pull from the streets. What is the goal here, if you were to identify one?
The goal definitely is just to get people to think about this. I think it’s completely fascinating that we can learn all these things about ourselves through DNA. I also think it’s really important that we remember that these are probabilities that we are dealing with. They’re not exact, it’s very probabilistic, and yet it’s starting to be used in law enforcement increasingly. It’s also becoming easier and easier for anyone to come along and grab that information, and it’s still largely unregulated and unlegislated. I think that really makes it something worth thinking about, something that’s worth having a dialogue about, as a culture.
What are the implications of it not being regulated?
For example, in law enforcement, the implication is a sample found at a crime scene: a police forensic biologist analyzes it and determines the ethnicity of this person is a certain ethnicity, and that information filters back out to the public and potentially causes harm or trouble or difficulty for someone because it may or may not be accurate. Whether or not it is, we tend to use and view science with a sense of authority. How does that compare to a witness saying they saw someone at a crime scene? We tend to give science even more authority when, in this case, it’s probably much less accurate than a witness, depending of course on the witnesses reliability.
I think what’s fascinating is that the person who designs this system, they’re like me designing my system. It’s filled with all of my biases, even though in many ways, it has a kind of objectivity. It is filled with all of these little choices that I made when I made my own characterizations of what people look like. For example, there’s a parameter of gender in the software, so you can determine on a scale of like negative five to positive five how male or female a face is. So when I categorize the faces male or female, I’m making a decision about how much of that trait to include for a person—very male, very female, somewhere in the middle, more neutral. Those are the kinds of decisions that we forget that scientists and engineers make every day.
Portraits created from open source DNA samples found on Github.
You describe this project as an investigation into “genetic surveillance.” What does that mean to you?
There is of course paranoia from media and I think a lot of it ultimately points to this question of identity and this big question mark that we all have when we try to think about who we are, what makes us, what defines us as people. DNA seems to be, scientifically, the closest thing we have to the source code for our identity. To think that it’s not as precious as we like to imagine, that we’re just discarding it at all times, throwing it around, to think that that could be cloned or that it could be used against us in some way… I think it’s a natural fear to have, and I don’t know that it is completely unknown.
I think it’s something we’re going to have to deal with in the next decade because when the $1,000 dollar genome sequencer comes out in the next year, we’ll have a USB stick and potentially I could take one of these hairs or take a couple of hairs, extract the DNA, put it on the USB stick, and within minutes have someone’s entire genome on my computer. The whole pace of this is about to just rapidly, rapidly expand and we still really see this as being science fiction.
All images courtesy of Heather Dewey-Hagborg.