9 Thanks for coming out

8 Appearing at Harvard GSD on 16 March

Thanks to Yasmine Abbas, I’ll be speaking at a meeting of Harvard GSD’s Critical Digital student group the day after my appearance at Emerson College.

More details as they become available.

7 Appearing at Emerson College on 15 March

On March 15th, I’ll be speaking in Boston, as a part of Emerson College’s Floating Points 3 lecture series. The theme of the panel I’m on will be questioning some of the premises that are used to promote ubiquitous information technologies and “working creatively to subvert them.”

6 Internet Actu interview: AG on ubiquitous computing

The following is the English translation of an interview recently conducted by the French magazine Internet Actu.

IA: Can you tell us something about your background and how you came to be interested in ubicomp?

Professionally, I’m an information architect and user-experience consultant. A consistent concern in my work in this field over the past half-decade has been the restoration of human users and their needs to a place of rightful centrality in the design of technical systems.

All too often, complex technical artifacts - Web sites, mobile phones and automated bank-teller machines, to name just a few of the more common we encounter - are designed without a deep understanding of how people receive, process and act on information, without regard for a natural interaction flow between user and tool.

Not only is this obviously a cause of enormous, daily frustration whenever people are exposed to such poorly-designed (but increasingly inescapable) tools, but we find empirically that users tend to blame themselves for the defaults that result. You will literally hear people berating themselves for their supposed “stupidity,” or say things like “I guess I’m just too old to learn how to use this,” when in actuality it’s the design team which is at fault.

More precisely: in my experience it is not generally the case that designers are not up to the task of providing good user experiences. It is, rather, either through time or budget pressure, or lack of a respected internal constituency for sound design practice, that users and their requirements are pushed to the periphery.

This is distressing enough at the scale we currently encounter, but, as we’ll see, as the ambit of technical intervention and interaction begins to migrate from the desktop out into broader realms of everyday life, and from theoretical to actual, the prospect of bad user experience becomes intolerable.

The bottom line for me personally is that, having seen how the development process works from the inside, in some of the world’s largest and best-known companies, and having seen the products that result, I do not believe that we can safely entrust the domestic and public realms to any technology founded on these or similar methodologies.

I’ve seen a great deal of techno-optimism and even -utopianism around ubicomp, including a fair amount from people who should know better.

But despite a deep folk understanding of some of the risks involved - Philip K. Dick was writing stories featuring recalcitrant doors and dilatory automated taxis in the late 1950s, and we’ve all heard of HAL9000 - there hasn’t really been much in the way of people pushing back against the idea of ubicomp, in a measured and knowledgeable way. And so I’ve started to make some noise about what I see coming down the road, describing the reality I see lurking behind the marketing hype that’s already beginning to build about the ostensible “conveniences” that await us.

IA: How would you define Ubicomp? Is it only about “smart objects”?

Any current definition would have to be somewhat diffuse and contentious, which is plainly evident in the diversity of terms people have been using to refer to what I think of as a broadly overlapping set of practices and techniques: some researchers think of temselves as working on “ambient informatics,” others on “pervasive computing,” and still others on “tangible media.”

Then there’s the group of researchers and developers I’m most interested in, operating on a more granular and probably more decisive scale. Their concerns will be, variously, RFID, or wireless networking, or biometrics, or using the body’s own electrical field to carry and transmit information. It is not at all certain that people working in this latter group will think of themselves as embarked upon the ubicomp project - in fact, it’s possible that they never would have heard the word “ubicomp.” But these are the enabling technologies underlying any persuasive ubicomp scenario.

Ubicomp is far more than “smart” objects, which might be best regarded as a symptom of a deeper paradigm just now unfolding. For me, it’s fundamentally about the surfacing of information that has always been latent in our lives; pattern recognition and machine inference based on large amounts of such information; and about the domain and scale of technical mediation contemplated - both wider and narrower, higher and lower than has been the case previously.

IA: What is really new about Ubicomp? Why do the stakes suddenly look higher than, say, those of “Internet regulation” or “cyberspace”?

Because, as we’ve seen, here interaction models derived from technical development begin to intervene in and affect us in all the most intimate circumstances of our “real” lives.

When I talk about “surfacing information that has always been latent in our lives,” I mean putting precise numerical values on one’s present location, on what task we might happen to be currently engaged in, and in whose company; even on things like daily caloric intake or voice stress or urine chemistry. I mean making those values broadly accessible. I mean permitting operations to be performed on such values or aggregations of same, such that algorithmic guidance and control can be installed.

The ubicomp discourse always starts with “convenience.” But at the limit - I shouldn’t even say that, as these are still reasonably tame compared to true limit cases I can imagine - we’re talking about enabling scenarios like refrigerator doors that refuse to open if you’ve exceeded your caloric limit for the day, child monitors that notify parents that their “troubled” kid is in the company of known delinquents and in proximity to tagged controlled substances, cars that automatically slow to a halt when overridden by the needs of emergency vehicles. (Some readers may be thinking that these sound like entirely reasonable things to desire, and I’ll get to the tension between that appeal and its consequences later on.)

And even far short of these scenarios, if they’re too alarmist or unrealistic for your taste, we have a wide belt of circumstances where human agency, judgment and will are progressively supplanted by compliance with external, frequently algorithmically-applied, standards and norms.

The clear net effect, over the longer term, is to deprecate human insight. This unrestricted exercise of technical criteria for competency operates (if you’ll forgive the term) discursively, providing a template against which we come to measure ourselves. I see this already in a small way watching people interact with Web sites, and I dread what this would look like if applied to the scale of household management, or getting around in the city.

IA: You say Ubicomp is, in part, the child of IPv6 - that is, in a way, of the Internet Founding Fathers’ life work. However, when talking to many Internet pioneers, we notice that many of them are not involved in Ubicomp, nor aware of its implications. Is it your feeling as well, that there is a fracture between a community raised in a vision of an autonomous “Cyberspace” and another that links (or makes no difference between) the digital and the physical?

The significance of IPv6 is simply that it’s a necessary piece of the puzzle - if it wasn’t this specification it would be something else. But true ubicomp needs a schema that provides arbitrarily for the communication of anything with anything else, and IPv6 would appear to fill that requirement admirably.

The implication of “ordinary” wireless networking on IP as a practical framework for ubicomp raises one of the those tropes in the history of technology that would be embarrassing to mention if it weren’t so usefully true, which is that once you’ve let the genie out of the bottle it’s impossible to get it back in. That people did not for the most part intentionally set out to create an effective backbone for ubiquitous ambient mediation when they devised IPv6 does not make it less suited to that role. “The street always finds its own uses…”

Regarding the second part of your question, if you’ll forgive me I’ll choose to say little about that here. I do have certain beliefs about this question - as about the role played in the structuration of user experience by the dynamics of product development in primarily market-driven institutions, and many other relevant points - but as they’d distract from the main thrust of my argument, they are perhaps better left for other venues and other times.

IA: You describe Ubicomp as “Unintentional social software”. What do you mean by that? (i.e., Why is it social? Why unintentional?) What is the difference with “Intentional” social software?

I describe ubicomp as “always already social” because all that information about you or me going into the system implies that it comes out again somewhere else, which has real consequences for how we go about constructing a social self - a presentation designed to maintain our status or credibility in the eyes of others - and therefore for polities comprised of such selves.

What if every fact about which we generally try to dissemble, in our crafting of a mask to show the world, was instead made readily and transparently available? I’m not just talking about obvious privacy issues - histories of various sorts of irresponsibility, or of unpopular political, religious or sexual predilections - but about subtler and seemingly harmless things as well: who you’ve chosen to befriend in your life, say, and what kinds of intimacy you choose to share with them but not others.

When these private and unspoken arrangements become public and explicit, embarrassment, discomfort, even resentment are frequent results, for all parties involved - we’ve seen this already in the contemporary generation of “intentional” social networking software, the Friendsters and Orkuts. Unless a good deal of thought is devoted to reproducing those mechanisms of long standing, by way of which functional human societies maintain plausible deniability and what I’ve elsewhere called “protective hypocrisy” on behalf of their members, we’ll see this still more seriously in the domain of ubicomp. (I describe ubicomp as “unintentionally social” because, while it has not explicitly been designed to forge or facilitate social connections between people, as in the case of the systems named above, it surely has that effect in any event.)

I should also mention that, as Fabio Sergio has observed, we’re increasingly being socialized (by “reality television,” among other phenomena) to regard the occurences of everyday life as content for the consumption of others. I’m just as concerned about this prospect of my own life as entertainment as I am the prospect of surveillance-as-control.

IA: Despite the risks you mention, you say that most people will demand the sort of many-to-many surveillance Ubicomp provides. How, and why?

Well, I’m willing to consider the idea that this may be an ethnocentric viewpoint, but I believe that attention is an attractive prospect in our culture: look at the cultural machinery around the production and marketing of celebrity, and especially fake celebrity.

So for those reasons, as well as for deeper psychosocial reasons having to do with contemporary families and the development of personality within them, I believe that the idea of perpetual, individualized, intimate attention and care will be greatly appealing to a great many people, and they will by comparison esteem only lightly the things they will have to trade away in order to purchase that attention.

We also live in a world where we have been taught - and not always or exclusively without reason - to fear our own environments and the other people with whom we share them. Ubicomp offering at least the latest gloss on a reassurance of security and what we might call maintaining the integrity of the perimeter, I don’t think it will be too hard of a sell.

IA: There have been several discussions and initiatives revolving around privacy in Ubicomp. How does your thinking and the “Guidelines” that you propose go beyond that?

Privacy has historically been one area where the mainstream of ubicomp researchers have been very aware of the risks involved, the tension between institutional and personal or public and private prerogatives. We find these issues addressed already in the first “Active Badge” papers, dating to 1991 or ‘92. So it’s something people have been on top of, and have evolved meaningful responses to.

My concerns, however, are broader than just privacy, as important as that is, and stem mainly from my experience of observing people in interaction with technological systems. Once you’ve seen how upset people get when - through no fault of their own - the system they’re dealing with doesn’t work as advertised, and especially once you’ve seen, repeatedly, how people then internalize the blame for this, it’s hard to accept the prospective deployment of similar interfaces across broader swathes of our lives.

That’s what the guidelines I proposed were intended to deal with: the perceived affective quality of the interaction on the part of its human participant(s), as well as the threats to privacy and civic autonomy embedded in any robust ubicomp scenario.

IA: Are there ways to embrace Ubicomp in a way that makes it a tool for human development, or do you believe the vision is inherently flawed? Or is there another scenario where “a-ethical” Ubicomp will be met by countermeasures such as citizen initiatives, hacking, private services designed to fool Ubicomp systems, etc.?

I’m inclined to believe that ubicomp as it already exists - and as its primary proponents wish it to exist - reflects a deeply unhealthy relationship between the self and itself, the self and others, the self and the world, but that’s just personal prejudice. Any social system large and unruly enough to require the kind of intervention and (micro)management we’re discussing has probably grown beyond a level that I’d personally think of as organically healthy.

But I also believe that, barring some massive collapse of technological civilization on this planet, the technical arisal and therefore the social and political reality of the milieu under contemplation is effectively inevitable, at least locally. And that therefore the responsible thing to do is help it find a point of balance with the other prerogatives we’ve discussed, to ensure that wherever possible it does actually make life more convenient and safer and all the other things it promises, without undue deformations of quality of life.

There’s no doubt that the emergence of a robust ubicomp milieu will present attractive commercial opportunities to people paying attention. My favorite example along these lines mirrors how it’s currently a selling point for a café or hotel to trumpet their wireless Internet hotspot - I’m willing to bet that in time we’ll see precisely the opposite offered as an enticement, that such-and-such café is a deadspot where customers are excluded from the purview of the various surveillant systems they would otherwise be subject to. There’s also no doubt in my mind that a ubicomp product along the lines of TV-B-Gone would do well, and merely awaits the entrepreneur to bring it to life.

Finally, your point about hacking is well-taken; one of the things I cherish about people is how reliably they’ll attempt to subvert anything imposed from without that tends to limit their autonomy. There are, though, limits in turn on this capacity for resistance, not least of which is the sheer amount of time and effort effective circumvention of really ubiquitous systems will require. Securing acquiescence by exhaustion has been a valuable tool for various sorts of oppression in the past, and I see no reason why ubicomp would present a different case in this regard, no matter how Panglossian a picture its enthusiasts paint.

IA: Are there currently projects and initiatives that sort of follow (or precede…) the “Guidelines” you proposed?

I’m happy to say that I’m engaged in some early discussions with actual practitioners and developers, as well as theoreticians, but no, there’s nothing concrete that can be pointed to as yet.

My hope is that some greatly fleshed-out version of the guidelines will eventually form the basis of something along the lines of an international standard, like ISO9001: a public-facing certification that ethical and interpersonal prerogatives have been respected in the development of the system in question. I’d further hope that there would be some pressure on developers to meet such a standard, and effort on the part of customers to avoid purchasing or participating in systems not so certified.

IA: What, if any, are the next steps for this work you have been doing on Ubicomp?

For now, I’m working on a book for general audiences, provisionally entitled Everyware. Everyware will explain to potential developers, consumers, end users, policy makers, and anyone else who has an interest just what ubicomp is, how it might affect them, and how they in turn can affect its eventual development in ways congenial to them and their aims. Hopefully I can do this effectively without either exposing people to too much jargon or dumbing things down.

I’m also working with people who share the concerns I do, notably Mike Kuniavsky, formerly of Adaptive Path, to make presentations at developer conferences, and otherwise begin to popularize the notion of making space for these ideas in any prospective ubiquitous system. We have a long, long way to go, but we also know the journey of a thousand miles begins with a single step.