April 23, 2006

toward a ubicomp reformation?

With his recently released Disputation of Adam Greenfield on the Power and Efficacy of Ubicomp, the information architect and experience designer has resoundingly nailed eighty-one theses to the church doors of the future, inviting discussion and debate on the nature and direction of ubiquitous computing.

"Out of love for the truth and the desire to bring it to light, the following propositions will be discussed at many bookstores and venues, under the presidency of Adam Greenfield, Architect of Information and of Experiences, and Lecturer in Ordinary on the same. Wherefore he requests that those who are unable to be present and debate orally with us, may do so by email or weblog."

Heh. Okay, the book is really called Everyware: The Dawning Age of Ubiquitous Computing, and it has little in common with Martin Luther's protest regarding the Indulgences of the Church. But it is structured as a sequence of 81 theses, in an apparent allusion to Luther's 95; moreover I suspect Adam has in mind to spark a kind of Reformation in the way that designers, engineers and academics approach the emerging field of ubicomp.

I first encountered Adam in the basement of a Ropponggi nightclub, the Super-Deluxe. It was 1IMC, the first international moblogging conference, and he was the organizer, host and master of ceremonies (as well as the coiner of the term moblog). In the summer of 2003 cameraphones were just coming on the scene, and there was a palpable buzz of something new in the air. The idea of taking grainy little pictures on your phone and beaming them wirelessly to your blog was just, well, cool, and so was 1IMC. The conference, like the pictures, was small, grainy, and a bit out of focus, but because of the caliber and cross-disciplinary nature of the people Adam was able to attract, it was enormously interesting and fun. Adam's closing speech at the end of the day was a heartfelt rant about the good, moral, positive things that we could all be doing with these wonderful new capabilities. We all took the happy pills.


Everyware is a bit like 1IMC was. With ubiquitous computing, there are hundreds of new ideas emerging, dozens of core technologies maturing all at the same time, and the sense that something quite different is about to happen. Enter Adam with another felicitous neologism, one that captures the ubiquity, physicality and technical nature of the phenomenon with surprising grace. Although Everyware is not (yet) a conference full of diverse, interesting people, Adam employs a similar cross-disciplinary synthesis to his analysis of ubicomp, one that draws on technology, design, user experience, societal concerns, ethics, and academic and commercial viewpoints to create a fresh and nuanced perspective. In stark contrast to 1IMC, however, here Adam is not an optimist. In fact, he seems fairly worried.

The core logic of Everyware follows a path something like this:

Ubiquitous computing will imbue the fabric and flow of everyday life with information technologies. Operating at body, room, building, and city-scale, immersive and ambient, everyware will be a very different kind of computing than what we are used to. It will require new models of interaction and design, and it will have profound implications for human society. Furthermore, the emergence of everyware seems to be largely inevitable, as it is driven by deep trends in society and technology.

Because everyware causes latent information in the world to become explicit and explicit affordances of computing to become imperceptible, and because interacting systems of distributed everyware are likely to yield unexpected emergent properties, both its design and its user experience are quite problematic. In a sense, everyware itself is an emergent phenomenon; it will be an aggregate of many distinct, separate systems created by people trying to solve particular problems in their area of specialization, with no sense of an overall cohesive architecture or even a general desired outcome.

Everyware represents a complex, long range problem; however most of the technical underpinnings are sufficiently mature, there are more than enough early existence proofs, and the potential for poor outcomes is high enough to require that designers begin to act consciously and now. Given the high social costs of getting a truly pervasive computing wrong, and our demonstrably suboptimal track record at designing so-called smart things, it would be useful to have certain principles for the design of ethical and efficacious everyware systems.

At this point, Adam offers five high level design principles:

* Everyware must default to harmlessness.

* Everyware must be self-disclosing.

* Everyware must be conservative of face.

* Everyware must be conservative of time.

* Everyware must be deniable.

I have to admit, having thoroughly enjoyed the book (including my own unexpected cameo appearance*) and agreed with much of it, my first reaction to these principles was "That's it? Five thou shalt nots? He's not much of a reformer after all, is he?" But on reflection, I believe Adam has done a wise thing here. These recommendations have the virtue of appearing straightforward, human-centered, and conceptually easy to support, yet each of them leads directly into a thicket of difficult and worthy social, technical and design issues and tradeoffs (not the least of these is, how do you specify design principles and goals for a system whose properties emerge from disparate and distributed subsystems). There are no answers articulated here, but these are good starting points for inquiry. Adam has built us a scaffolding to think with, and the ultimate success or failure of his book will be in what we as a broad community of researchers, designers, engineers, artists, social scientists, business people, policy makers, and human beings now do with it. My one disappointment with Everyware is that it is not actually made of everyware; perhaps that should be our next collective project: to design the ongoing conversation as a ubiquitous system that embodies the principles Adam has described.

Since you have come this far with me, I am going to take the liberty of assigning a little light reading for the journey ahead. In addition to Everyware , go get your brain wrapped around John Thackara's In the Bubble : Designing in a Complex World, Peter Morville's Ambient Findability, and Bruce Sterling's Shaping Things. Then come back and we'll talk more.

* Adam quotes my post on the lessons of Prada's Epicenter in a couple of places, I nearly fell off my chair when I came to those bits ;-)

Posted by Gene at April 23, 2006 10:42 AM

Wow! Gene, that's absolutely the most concise, elegant summary of the book I've yet seen. In fact, I should probably worry that having read that, folks'll barely need to actually go out and buy the thing. ; . )

Oh, also: I was thinking of Guy Debord's theses far more than Martin Luther's. I should've aimed higher, huh?

Posted by: AG at April 23, 2006 2:10 PM

Hey Adam, nice to see you! So I did worry briefly about spoiling the ending, but I figured the punchline is actually meant to be discussed and debated. Maybe I should warn about spoilers.

OK listen up folks, lest there be any question, you still need to buy the book!

And dang, Debord. I was afraid that whole Lutheran thing would be off the mark, guess I've got some reading to do. See what happens when people come at things from different contextual places...

Posted by: Gene at April 23, 2006 3:19 PM
Post a comment

Remember personal info?