The media calls them glassholes, but one day, we may just call them our friends or even our own selves. I’m talking about Google Glass wearers. Whether or not this particular device or design becomes mainstream, it’s likely that we’re moving towards a stage of more cameras everywhere–ones that people wear, ones that could be and, often are, “on” all the time.
“The young kids growing up with this technology–for them, it might become completely reasonable to just wear a camera. So then the question is: What will they do with it?,” says Apu Kapadia, a privacy researcher at Indiana University.
Google Glass doesn’t automatically record all the time, but there are other devices already on the market that do, such as Narrative Clip and Autographer, which are designed specifically for “lifeloggers”–people who don’t want to miss capturing every banal moment. (The Clip and Autographer capture 120 and 360 images an hour respectively). In the future, there may be benefits other than vanity that encourage people to adopt similar devices and constantly capture their lives. For example, software might be able to search earlier footage and tell us where we left our keys, or Alzheimer’s patients might search that same footage to jog their memory.
Just as Facebook detects and identifies the faces of friends in photos, Kapadia wants to build software that’s smart enough to understand and flag potential privacy risks in images. He thinks this is needed because, while it’s easy enough for people to avoid privacy violations when taking intentional photos on their smartphone cameras, it’s much harder if we’re snapping hundreds or thousands of images continuously and indiscriminately throughout the day.
But before Kapadia builds software to detect privacy risks, he needed a better grasp of how people treat and manage privacy–both theirs and others’–when using life-logging devices, as well as how others react to the presence of these devices.
“We really need to understand these issues now. What if we fast forward 10 years from now, and everyone’s wearing these cameras?,” he says. “We’re very worried about collecting all of these photos, and what the privacy implications are for society and for the individuals who wear these cameras and for the bystanders captured in the photos.”
He and his collaborators, computer vision researcher David Crandall and sociologist Denise Anthony, recently recruited 36 undergraduates to spend a week wearing a phone around their neck that was custom-designed to capture first-person images, GPS locations, and other sensor data every five minutes from the hours of 8 a.m. to 10 p.m. (see above image). Each device also warned bystanders with a sign reading: “Photography in progress, IU research study, Photos taken every five minutes.”
Through an app on the phone, the students were given the option to pause data collection for a 15-, 30-, or 60- minute period or delete the last 15-, 30-, or 60-minutes of recorded data. At the end of each day, they answered surveys about their experience and why they chose to stop recorded when they did. They were also asked to delete more photos if they wanted, mark ones that were blurry or contained no useful information, and mark others they would be comfortable sharing with different groups of people or the public.
The result of all of this? A whopping 14,744 images that were part of the study, with an average of 490 photos taken per participant. The subjects were “default” sharers: They shared all photos with the public for which there was no good reason not to share. In the end, of the photos they took that weren’t deleted, the students on average were comfortable sharing 90% of their photos with the public.
What the students did do, however, was engage in a lot of “in-the-moment” control of images, rather than curating them at the end of the day. They would use the software to pause collection or delete right after the fact, or, more often, they would physically cover up the camera device on their neck. “We saw a lot of physical control of the device, so you don’t have to deal with private photos at the end of the day,” he says. “Some people are not so confident in the software.”
In general, Kapadia also found that the students were equally concerned about the privacy of bystanders they captured in their images as their own.
Kapadia says these kind of studies are needed to inform how companies like Google and Apple continue to design wearable devices. While it’s easy enough to imagine simple turning off these devices today, what if Google Glass ends up as a contact lens or a prescription set of glasses? Perhaps a way to physically cover the camera would help, he suggests.
There’s a lot more findings in the paper, which is available here and was presented at a wearable computing conference in September. The questions matter today. Because even if you are taking a photo and it stays “private,” if hackers can tap the iCloud and leak nude photos of Jennifer Lawrence, we have to be concerned about the photos we take–not just the ones that we share.