You Lookin’ at Me? Reflections on Google Glass
There is but one remedy for the Glass wearer – a bucket of iced water in the face whenever you suspect he has taken you unawares.
With the public beta launch of Glass there has been a lot of discussion on why it will or won’t fail. The ultimate benchmark for success is high: after someone has tried Glass can they imagine life without it?
It’s the wrong question.
Glass is Google’s unintentional public service announcement on the future of privacy. Our traditional bogeyman for privacy was big brother and its physical manifestation – close circuit TV, but the reality today is closer to what I call little sister – and she is socially active, curious, sufficiently tech-savvy, growing up in the land of “free”, getting on with life and creating a digital exhaust that is there for the taking. The sustained conversation around Glass will be sufficient to lead to a societal shift in how we think about the ownership of data, and to extrapolate a bit, the kind of cities we want to live in. The argument that Glass is somehow inherently nefarious for me misses a more interesting point: it is a physical and obvious manifestation of things that already exist and are widely deployed today, whose lack of physical, obvious presence has limited a mainstream critical discourse.
As a product that is both on-your-face and in-your-face, Glass is set to become lightning rod for a wider discussion around what constitutes acceptable behaviour in public and private spaces. The Glass debate has already started but these are early days; each new iteration of hardware and functionality will trigger fresh convulsions. In the short term Glass will trigger anger, name-calling, ridicule and the occasional bucket of thrown water (whether it’s iced water I don’t know). In the medium term as societal interaction with the product broadens, signs will appear in public spaces guiding mis/use and lawsuits will fly, while over the longer term, legislation will create boundaries that reflect some form of im/balance between individual, corporate and societal wants, needs and concerns.
So Shoot Me
Of all of the companies and organisations that could bring Glass to market, I’m pleased Google is the one making a significant investment: a company with a recent record of genuine innovation that stretches/defines social and behavioural norms; with a strong revenue stream and deep enough pockets to have a fighting chance of medium to long-term success. It also helps that the project is considered of strategic importance, and has key executive sponsorship. Less obvious but no-less relevant in this equation is that the company has a lot to lose; is no-longer the media darling; has fucked-up enough times in public to know it can do so again (and again); has been humbled by more nimble competitors; has experienced talent drain and understands the impact of this on its culture and its bottom line. Of course Google can financially afford to fail again: experimentation and failure is a critical part of its DNA, but while privacy-snafu fines are low, the internal and external cultural costs of Glass failing are high.
All technology challenges the status quo, and if a technology is noticed by consumers/users/constituents at all, it presents for some an opportunity and for others a threat. The perceived and actual threat from Glass comes not from crimes against taste. (Many have commented on the perceived inelegance of the design.) Their design team appears to have done a sterling job if you assume that particular design direction and constraints. Our sense of what is tasteful succeeds or fails as part of a far broader narrative which they too are exploring. Yes you can find a hundred and one designs of “wearable computing” from the past decade that look similar, but very few are packing the same experience into the same form factor. However, as a connected, sensing object it is capable of recording and transmitting photos, video and sound directly, through content analysis or indirectly through proximate connected devices other data such as location, temperature, trajectory and so on. In other words in a worst/best case scenario it could record and measure “everything” and associate that data to a person. How will this play out?
I want you to try a little experiment. Find somewhere where you can sit and observe people interact with one another. Pick somewhere just out of the throng — the edge of a cafe looking in, a park bench, a doorway close to a market. It’s easier if you choose somewhere you don’t know so well, you’ll have less to unlearn.
Give yourself 30 minutes to view and reflect upon the scene in front of you: who visits that space and why; the differences in ritual greetings and indeed whether or not a person is greeted; how people project out who they are; things that signify status, social hierarchy; where objects are placed; the level of interaction with those objects when not in use. What can you see being documented online or off? What can you imagine being documented? Pay particular attention to things that fit your definition of “technology” and reflect upon on the things in front of you that once fit this definition but no longer do (my list of were-once-technologies includes the pencil, the wrist watch and the smartphone).
If you’re close to enough to other people you’ll overhear conversations plus bits of conversations that the speaker’s will allow you to hear, raised, projected, sotto voce and in whispers, combined with body language all serve to emphasize what is said, and the intent of what is communicated. The challenge for Glass is that the costs of ownership falls on people in proximity of the wearer, and that its benefits have yet to be proven out. How much of that conversation is directed at the “listener”, and how much of it is directed at others in proximity, including you? This rich social choreography is playing out hundreds of billions of times a day across our planet, and is as subtle and delicate as anything appearing in a BBC2 nature documentary.
Of course people and systems are already capturing (and channeling) content and data in this space in the form of photos, video, background noise on phone or video calls, who is connected to what, and what they are doing. It is likely that Google, Microsoft and Nokia’s Navteq (to name but three) have already systematically mapped this space and are serving up street views online. The difference with Glass is that it threatens surreptitious, unexpected or continuous recording from the perspective of the human-eye/ear view. At this point it doesn’t matter whether it can support sustained recording for long periods or not; what matters is that the form factor supports this, that it could at some point, and that we all know Google is in the business of selling ads against insight drawn from large volume of data. Continuous, indiscriminate recording in this space is drag-net fishing of data collection – it’s a destructive technology, a conversation and privacy killer.
And back with our experiment, take in the scene in front of you. Who owns this space both legally and figuratively? Who has the rights to do what? By what authority? Who enforces that authority? How do these rights differ for regulars or a first time visitor? What are the ways that people signal the beginning or the end of an activity? And how does that signalling make something more or less acceptable? The obvious clue to activities people have deemed socially unacceptable are often found on hand-scribbled “do not” signs as in “staff will refuse to serve customers who are on their mobile phone” “do not ask for credit”. The more sustained the infringement, the more official looking the sign.
Today we assume that our conversations and our image is not by default recorded by other people in proximity. Not having a persistent record allows us to present a nuanced identity to different people, or groups of people, provides with the space to experiment with what we could be. The risk that what we say will be broadcast, narrowcasted to people we don’t know, or may underpin someone’s future business fundamentally changes what we want to talk about. The challenge for Glass is that the costs of ownership falls on people in proximity of the wearer, and that its benefits have yet to be proven out.
A number of years ago while working at Nokia I was asked to explore use cases using an appearance model (a non-working prototype) of a form factor similar to Glass but clunkier and definitely less refined. In the first phase of this make-it-up-as-you-go-along-and-see-what-works study we hired students in Tokyo to act out various scenarios including content browsing, viewing and game-play using gestures and voice commands in a range of contexts: at home; on a commuter train; on a long distance train; in a hotel lobby; in a park; cafe and while walking along. The research team then noted interaction issues with the glasses, carefully observing social reactions from people in proximity; before finally interviewing the actor/actresses for their own experience.
Fans of Milgram’s New York subway experiment will be happy to note that our actors and actresses felt extremely self-conscious about wearing non-standard glasses, and awkward about acting out the scenarios particularly in contexts where there were others in close proximity. A number of the things we learned from this study surprised us:
- Most of what we ”see” is at any time is out of focus in the periphery where as long as the things going on in peripheral vision don’t trigger a threat response will probably pass the glance test. Glass is, in effect a blanket tax on the collective attention of society. It will be interesting to see whether Glass is perceived as a threatening object and thus may force others in proximity of a wearer to maintain a hyper-awareness of the wearer and their own actions in places — whereas today, they are currently able to relax. This would be, in effect, like a blanket tax on the collective attention of society
- Spoken interaction is awkward for almost everyone in confined spaces on systems with less than 100% accuracy. An interface built around short responses to contextually understood events will be the dominant form of interaction.
- Gesture interaction is just as awkward in close spaces and in many instances will restrict regular use and/or in a vocabulary of “quiet-gestures”. To get a sense of how this plays out the next time you are on the subway and have people sitting on either side, raise your hands in front of your face or look down and move your hands in your field of vision. Even simple gestures require upper-arm/shoulder movements, which when you are sitting shoulder-to-shoulder impacts fellow passengers. A Glass wearer that wants to maintain the social cohesion in that context (and not all will be that self aware or considered) can mitigate this by pausing interactions for the moment when they are appropriate, or more likely by avoiding interactions in that context.
- In contexts where social interaction is required e.g. sitting with friends around a table in a cafe, it will create a situation where people are not sure whether they, or the contents of the display are engaging the wearer.
- In-ear or close-to-ear (inductive) audio changes the wearers enjoyment of food and drink – a problem for an otherwise prime use-case: watching movies at home where snacks and beverages might naturally be consumed.
- Humans tend to fall asleep in contexts where they are seated, safe, and there is minimal physical movement – providing opportunities to design for disengagement.
- Humans have a vested interest in tracking changing emotional states of the people around them. This will introduce are “you looking at me” moments where others in proximity assume that a smile, tear or frown is triggered by their own presence and will spur people to send inappropriate content to their Glass wearing peers, with a weary inevitability will include this but is far less likely to include this (or was it the other way around?). In some contexts these moments will lead to confrontation. Read the footnote in this article in the Atlantic by Ta-Nehisi Coates, and imagine introducing erratic behaviour into the equation. Amplify to billions of of social interactions a day.
What starts out as a fairly broad set of use cases rapidly starts to narrow.
I’ve got a confession to make. frog has recorded thousands of conversations around the world, videotaped many more, tailed people around town, nosed around people’s homes – opening cupboards and drawers, asking personal questions where there were none. All with their permission and all in the name of research. There are a few things we’ve learned that relate to the broader discussion of what is collected by whom, how and why, and how it is used, you’ll see why these are relevant in a moment.
- Anyone can collect data. The real issue is how to collect data in such a way that that meets both moral and legal obligations and still manages to deliver some form of value.
- Ownership. People are naturally suspicious of what they don’t know. The simple act of giving them control over the process or the objects/technologies we carry defuses initial suspicion. A few simple field research techniques can rapidly build trust. These include handing someone you’ve just met on the street a five-thousand-dollar camera and then ignoring them to concentrate on a conversation with their friends. This shows we trust them. And then they trust us.
- Clear On/Off States. Most people have, at least initial concerns about being recorded. There are numerous ways that we emphasize the transition between on and off: from how a camera or other recording device is held when not in use. It is useful to think of a camera as a gun: understand the impact that bringing it out can have on any given context; only take it out if you’re prepared to use it; be careful where you point it.
- Reciprocity. Today it is easy to maintain a persistent connection between the researcher and the participant – often in the form of a social media or email address. You’ve asked something from them, they have the right and now have a channel through which to ask something of you.
- Full circle: we give participants the opportunity to review, delete or own any of the data collected on them by the research team. This is normally carried out at the end of the session, after any reward is handed over (so they are not pressured into letting us keep data) and before any data consent form is signed (so they better understand the implications of what they are signing). A team that knows the data will be reviewed by the participant changes what they collect in the first place, it becomes self policing. More than any training, this simple principle helps keep teams honest and operating within social norms.
A few simple steps lower the more obviously anti-social aspects of Glass. The evolution of body language that helps communicate Glass’s current state e.g. pushed above the head to show that not in use; a literacy around the spoken commands that communicate the current task that the user is engaged in “take panorama” or “grindr lookup”; and showing whether the camera and other recording mechanisms are in use, or disabled.
Glass has four design principles for developers that focus on the Glass-wearers user experience: “design for Glass,” “don’t get in the way,” “keep it timely,” and “avoid the unexpected”. As challenging as it is to find a compelling use-case (beyond porn) these principles are aimed at the wrong people – Glass wearers, rather than those in proximity. Two complementary principles will go some way to accommodate the concerns of people in proximity and lower social barriers to adoption.
- Proximate Transparency: allow anyone in proximity to access the same feed that the wearer is recording or seeing and e.g. view it through a device of their choosing. Make it easy to identify the Glasses themselves and to trace back to the wearer. This simple act help demystify the technology, create a broader sense ownership of its inclusion in any given space. The reality is very few people would be interested in jacking in and the act of having an open stream will change the behaviour of what is watched. For many this won’t be enough of a step, it is after all an opt-out measure for people who have the technological nous and literacy – forcing people in proximity to do something for dubious gain.
- Remote Control: allow identifiable people in proximity to control Glass’s recording functionality and have access to the output of what was recorded. Allowing others to demonstrably benefit from the utility of glass will make it part of the social landscape.
Pedestal or A Pauper’s Grave?
One could argue that the form taken by Glass offers up a lazy-futurists vision of what might be – take the trajectory of one product (displays becoming smaller/cheaper/more efficient over time) and integrate it with another (eye-glasses) sprinkle in connectivity and real-time access to content and big-data-analytics. Our expectations of what it could be are raised in part because this Glass offers up a lazy-futurists vision of what might be - a join-the-dots vision of the future fits neatly into western un/popular young-male culture from The Terminator through to Halo, Glass has a certain inevitability about it like the weight of expectation of child born to a great composer or if you prefer, to a middle-aged suicide. As any visitor toYodobashi camera over the past decade will tell you the hardware technologies that make Glass hardly feel novel (and for recent competitors see Sony, Golden-i or the Telepathy Device Prototype) but neither do they need to be, because this is all about how they are brought together into a holistic experience.
There are of course alternative, complimentary visions of this connected future that are far more discrete, taking connected, sensing things and embedding them in the world around us to inform, guide, direct, cajole, tax, enrich, … us and the things around us. It’s an area worthy of an essay in its own right but for now a few pointers to people, places and things that have helped inform my sense of this space: Dan Hill writing on the City of Sound; the MIT Senseable Cities Lab; Design Interactions at the Royal College of Art, Tisch ITP, BERG, Nicholas Nova and Julian Bleecker at The Near Future Laboratory help stretch our understanding of what could be, Curious Rituals in conjuction with students at the Arts Center College of Design is a particularly lovely piece of work; living for more than a decade in Tokyo, Shanghai and frequent trips to the cities that define this century’s urban experience: the Seoul/Nairobi/Mumbais/Rio/Chongqing’s of this world; products like Nike+, FitBit, Moves (to take one narrow category) through to less well known but arguably more impactful services that for me are at the very center of the internet of things – services like Kilimo Salama and Sarvajal; through to business units/activities in large corporations such as Cisco, IBM, Disney, and Ericsson with more of a how to make money/make a difference at scale.
Our moral character dwindles as our instruments get smaller
This essay starts with a paraphrasing a quote – here is the original in full: “There is but one remedy for the amateur photographer. Put a brick through his camera whenever you suspect he has taken you unawares”. It could be written about Glass today but is in fact taken from September 18th 1885 edition of Amateur Photographer magazine, seven years after the introduction of dry plates a technology that supported far more surreptitious photography. The full article by Bill Jay is worth reading in full.
The same article contains another quote from Amateur Photographer, twenty five years later, when cameras were becoming less noticeable: “Our moral character dwindles as our instruments get smaller”. In due course the technologies to deliver Glass’s emerging functionality will truly disappear from view — this is window of opportunity for reflection, discussion, debate and action.
I’m thankful to Google for putting so much effort into Glass at this moment in time.
That passion? Channel it.
That anger? Channel it.
My first book Hidden in Plain Sight explores the opportunities that come from seeing the world through fresh eyes.