Inkwell: Authors and Artists

    

Joining us today is Adam Greenfield, author of "Everyware: The Dawning Age
of Ubiquitous Computing."
  

    

Adam Greenfield is an internationally-recognized information architect, user
experience consultant and critical futurist. He's worked for clients ranging
from global concerns like Toyota, Capgemini, and Sony to local nonprofits.
He's also been a rock critic for SPIN Magazine, a medic at the Berkeley Free
Clinic, a coffeehouse owner in West Philadelphia, and a PSYOP sergeant in
the US Army's Special Operations Command.

Adam has spoken frequently on issues of design, culture, technology and user
experience before a wide variety of audiences. He lives and works with his
wife, artist Nurri Kim, in New York City.

Leading the conversation with Adam is former Inkwell host Jon Lebkowsky. Jon
is Senior Consultant and CEO at Polycot Consulting, L.L.C., an innovative
team of Internet technology experts with broad experience creating and
managing information systems for businesses and nonprofit organizations. His
current consulting practice focuses on web usability and strategy and
effective use of online social technologies. He is also a strong proponent
of universal broadband access to computer networks.

Welcome, Adam and Jon!
  

    
Thanks for that kind introduction, Cynthia. I'm delighted to be here
on the legendary Well!

I'm sure Jon has some piquant questions for me on the subject of
everyware (and "Everyware"), and of course I can't want to see what
your responses, reactions, and insights will be. I can't imagine a
better forum for this particular topic, so I'm really looking forward
to seeing what comes of this discussion.
  

    
I should mention that Adam and I had a brief interview while he was in Austin 
for SXSW Interactive, which is available at WorldChanging.com: 
http://www.worldchanging.com/archives/004298.html

I'm sure some of our discussion here will be similar to that interview, but 
we'll have more time to talk, and will be opening the interview to others. 

Adam, your interest in ubiquitous computing comes from a user experience 
perspective. What were you seeing or hearing that led you to believe that 
user experience issues might not be addressed very well by developers of 
pervasive or ubiquitous systems?
  

    
Well, first I want to make sure that everyone knows what we're talking
about when we say "ubiquitous systems," OK? Because one of the things
I learned when I first set out to do the research for "Everyware" was
that there are a whole lot of "ubiquitous computings" floating around
out there, and some fairly wide divergence in the way people understand
these terms.

My sense of it comes from the work of a guy named Mark Weiser, who was
Chief Technologist at Xerox PARC in Palo Alto at the time of his
untimely passing, in 1999. Mark had a very clear understanding that the
personal computer as we know it was not the final stage in the
evolution of information technology - that, as a matter of fact, it was
something like a transitional state between the early mainframes and
something entirely new, which he set out to define.

He saw, among other things, that Moore's Law was pointing to a
situation in the not too terribly distant future where there would be
many processing devices serving each human user, where these devices
would largely be embedded in the objects and surfaces of everyday life,
and where information processing would be "invisible, but in the
woodwork everywhere." He called this state of affairs "ubiquitous
computing," or "ubicomp."

Well, as it turns out, there were quite a few people in the world who
had similar ideas regarding where computing was headed - and depending
on which aspect of this vision loomed foremost in their minds, they
called it "pervasive computing," or "tangible media," or "ambient
intelligence." Despite differing emphases, though, what they call had
in common was this central idea: that many of the transactions we
undertake in the course of ordinary life - making a pot of coffee in
the morning, say, or choosing an outfit for the day, or finding a
parking spot - were about to be reconceived as transactions with
information-processing systems.

And - to get back to your question, now - it was this that really
triggered my concern. Because here we had the prospect of these
moments, scattered through our days, when we'd be asked to interact
with technical systems when we weren't, after all, setting out to do
anything particularly technical. And what was worse, if our experiences
with PCs and mobile phones and so on are any guide, we would be
"improving" everyday life by overlaying it with something that isn't
particularly humane or even very reliable.

And that, I've got to tell you, struck me as being a thoroughly
unacceptable state of affairs.
  

    
Why did you feel that this technology wouldn't be humane or reliable? 
  

    
Experience, mostly. The kind of experience anyone who's more than
cursorily involved with information technology shares: crashing
computers, applications that lock up for no apparent reason at all,
mobile phones that drop conversations at the most inopportune moment,
iPods whose batteries fail five minutes into an hourlong commute...

And that's not even considering the amount of time and effort we put
into learning relatively arcane commands, or setting and re-setting
preferences, or adapting to the various discomforts imposed on us by
unwieldy user interfaces (like mobile-phone keypads). Nor does it
account for the time we spend undoing the damage done by "helpful"
systems, when they wrest control away from us and (e.g.) reformat
something we're typing in a word processor.

This sort of thing is bad enough in a clearly technical context, where
at least we're more or less inured to it. As sorry a state as this is,
we've come to expect it. In some sense, we know what we're getting
into when we engage artifacts that we clearly understand as
outcroppings of information technology.

But how about when this standard of pleasure in use is applied to all
those interactions that go to make up a normal day? Do you want to
configure your toothbrush, or reboot your sweater, or hunt around for
the command that will allow your tabletop to "discover" the water
pitcher? I sure don't...but that is what I see happening in everyware,
unless people start making the kind of noise that would indicate to
designers and manufacturers and marketers that this simply will not be
acceptable to them.
  

    
You attended some conferences where technologists were discussing the ubicomp 
future - do any of them express concern about usability or user experience 
issues? Or about network overhead, potential crashes, etc? 
  

    
Crashes, defaults, downtime...sure. As far as I can tell, the people
designing ubicomp have always had an acute sense of how fragile it is,
how larger systems need to be orchestrated such that they're "robust to
the failure of individual interactors." The same goes for scalability
issues; I don't think there are that many who'd mistake a
proof-of-concept system for something that would be able to hold its
own out in the real world.

But I'm not sure how many people in academic ubicomp have really
marinated themselves in a consideration of the experiential and
affective dimensions of system failure. I'm not sure to what degree
people have ever simply sat and imagined what it will feel like when
systems like these surround us...and break down, as technical systems
often do. They may have thought about the specific system at hand, but
as a gestalt? I haven't heard that many people raising the issue.

With regard to usability, I hope things are changing - and to be fair,
it's been a few years, and there have been some encouraging signs -
but when I asked questions about the usability of ubiquitous systems
back at Ubicomp02, I got a lot of blank stares.

I had described the usability of many Web sites as "atrocious," and
gotten some really quizzical responses. One of the more prominent
figures in ubicomp took issue with that characterization, pointing out
that one could (nominally) do things on the Web in 2002 that were
effectively impossible to achieve for any amount of love or money in
1992. Which is undoubtedly true - but is to mistake utility for
usability.

I have little doubt that expert users will be able to do thing with
ubiquitous systems that are indeed all but "indistinguishable from
magic." This is not at all the same thing as saying that ubicomp is
currently being designed so that it's acceptably usable for most people
at most times. But like I said, that was four years ago, and one
upside to the fact that these systems are much closer to appearing in
our lives is that the discourse of usability has begun to sneak in
around the edges - as a business necessity, if nothing else.
  

    
It would be useful if you could talk a bit about good vs bad usability at 
this point, to give more context to our discussion. What are some familiar 
examples of good vs bad usability? And some potential examples from the 
future world of ubiquitous computing?
  

    
There's a couple of ways I could answer that.

I could, for example, talk about some recognized, more-or-less
canonical examples of good usability in information-processing products
and services. And then we could spend some time talking about how hard
it will be to design everyware so that it meets that definition of the
usable.

But I'm not gonna do that, because to my mind that's not actually the
relevant standard. Given that everyware is, by definition, information
processing embedded in the objects and surfaces and contexts of
everyday life, I think we need to have much, much higher expectations
of it if it is not to drive all of us completely out of our minds with
frustration and rage.

The standard I have in mind for usability is something like a chair.
It's just *there*, right? Nobody ever had to tell you how to use it.
You can move it from one room, or context, to another, and it works
just the same there as it did before. It never crashes, it never needs
to be rebooted, it can be enjoyed by anyone from toddlers on up, and -
short of a catastrophic structural failure - will afford its users both
functionality and pleasure for many, many years.

To me, that's what we should be aspiring to when we dare to propose
remaking everyday life in the image of IT. Anything less than
that...well, I, at least, have a very easy time imagining how
unpleasant it's likely to be.
  

    
Are you thinking along the lines of Mark Weiser and Jon Seely Brown, when 
they talk about "calm technology"?
  

    
Yeah, in a way, I am. Their line in that paper - "The Coming Age of
Calm Technology" - was that we were going to have to design ubiquitous
systems so that they'd "encalm as well as inform," because otherwise
they'd drive us to distraction. And they offered a few insights into
how one might go about designing such a thing - strategies that had a
lot to do with moving information from the center of our attention to
what they called "the periphery."

To my way of thinking, though, the trouble with "Calm Technology" is
that it sort of vests all of its trust in the development community. It
implies that developers will be able to exert decisive control over
the final shape of the systems they bring into being, almost all of the
time, and that that final shape will be informed primarily by the
needs and preferences of the end user. That benevolent developers,
looking forward, will see the pitfalls awaiting their users, and plan
accordingly.

Well, we know from experience that this just isn't likely to be the
case. In fact, there are all kinds of pressures operating in the
business of technology development that tend to make it rather
unlikely.

So while I applaud this first tentative recognition in the literature
that, hey, this ubicomp stuff might turn out to be sort of a Faustian
bargain, I'm kind of frustrated by it, too. With the kind of experience
and credentials they could call on, Weiser and Brown could have gone
so much further in terms of suggesting what it might mean to "encalm as
well as inform," and how a development organization would go about
responding to a challenge like that. Simply pointing out that it would
be a good idea if our ubiquitous technology encalmed its users instead
of frustrating and overwhelming them doesn't strike me as being quite
enough, somehow.
  

    

(NOTE: Offsite readers who have questions or comments can email them to
<inkwell@well.com> to have them added to this conversation)
  

    
What would you suggest as an alternative - collaboration between 
developers and other professionals who are focused on user experience? Or 
should engineers have user experience incorporated into their educations? 
All of the above?
  

    
I think - speaking from my own experience of developing
enterprise-scale Web sites for clients like Sony and Toyota and
Capgemini - that the practice of IT development is about as close as
you can get to "broken" and still ship product. So when you ask me to
suggest alternatives...well, that's just opening up the proverbial can
of worms. I wouldn't even begin to know where to start.

OK, I lie. I don't want to get too deeply into the ins and outs of
business process design and so on, because it's really another
discussion (and a lengthy one, at that), but let me toss out a couple
of ideas:

I don't think it's reasonable to expect engineers to be
user-experience experts, any more than it's fair to expect marketing
people to be able to crank out code. We each have our skillsets, our
affinities, and our professional responsibilities, and I think that
(given the complexity of the artifacts we produce) this kind of
specialization is here to stay.

So the organization as an organic whole is going to have to get better
at incorporating user-experience perspectives into its process.

Too often I've seen UX included in a project only as a quick, hygienic
step at the last minute, once all the real structural and design
decisions have been set in stone. This fails on two counts: it's
effectively too late in the day to make any more than cosmetic changes
to the thing we're building, but still more frustratingly, it's just
about the least cost-effective use of the UX mindset available.

I can't tell you how many times I've seen a Web site or a mobile
service built that looks beautiful, works perfectly...and is of
absolutely no use to any real user, and is in fact never widely
adopted, and is thereafter left to die a quiet death. After, y'know,
hundreds of thousands or even millions of dollars have been spent.
Whereas, if you had done some contextual inquiry and user testing up
front, you might have figured out ahead of time that your product or
service was an answer to a question nobody's asking.

And this is why I argue that user experience considerations - and
allied pursuits, like user ethnographic studies - should be pushed much
earlier in the development process, into the requirements-gathering or
discovery phase of a project. If this is true in the relatively
circumscribed context of a Web site, it's ten times as true of
something as insinuative as everyware, where we really are going to
have to deal with the consequences in the personal and domestic sphere,
on *our* time.
  

    
In the book you talk about "ambient informatics" as "a state in which 
information is freely available at the point in space and time someone 
requires it, generally to support a specific decision." Are there current 
real-world examples of "ambient" applications?
  

    
Sure: the Ambient Orb, from a company called Ambient Devices, and the
Nabaztag - a WiFi-enabled "smart object" that looks like a cartoon
bunny - are both products that you can buy right now. You set them up,
and they'll take in complex information about the state of the world,
and present it to you in this very gentle, classically "encalming" way.

But the Orb and the Nabaztag both strike me as being kind of twee and
"futuristic," the kind of thing that only a relatively small number of
self-consciously early adopters will ever pick up on. And at any rate,
they only fulfill half of what I think of as ambient informatics, in
that they're constrained by being present only in the place where you
set them up. They're ambient, in other words, but not *circumambient*.

So I think we can get a much better idea of what's headed our way in
this regard from something I first saw on top of a cab here in New York
a little over a year ago: an LED signboard that told you where the
advertiser's nearest ATM was, no matter where the cab happened to be.

The ad would presumably pull the cab's present location from GPS,
cross-reference that location against a database of the bank's
branches, and serve it up usefully and unobtrusively - so that, if you
happened to be standing at 36th and Madison as the cab passed you, it
would be advertising an ATM at the corner of 37th and Fifth, a block
over.

That's a *much* better example of what I think of as ambient
informatics, in several regards: it's out in the world, it's providing
a specific piece of actionable information, and the technical
infrastructure behind what is, after all, a relatively complicated
transaction is almost entirely invisible to the user. And it's out
there right now, just sort of unglamorously working, not hip or notable
enough in and of itself to be on the cover of Wired...and that, my
friend, is what the everyware future *really* looks like!

But I want to open things up a little bit here, because, as you know,
my interest isn't really in technology at all, but in how people pick
it up and make use of it, and how it thereafter shapes our choices in
the world. My point in defining ambient informatics isn't to then think
about, well, how can we make this happen; it's to start to wonder
about what happens to the shape of a city, or our idea of what a city
is, or our distinctions between public and private, in a world that has
ambient informatics in it. To me, that's a much more engaging - an
endlessly engaging - question.
  

    
Anyone who's seen "Minority Report" has a sense of that sort of thing - the 
holo ads that are in your face as you move through parts of the city. But I'm 
interested (and I know you are, too) in the ubiquitous technology that you 
*don't* see, and the extent to which you don't see it. How does our 
relationship to the technology change when it's hidden from view?
  

    
Adam -- I wanted to say first off that I *love* "Everyware."  You talk    
about some fairly subtle and/or arcane concepts in very human and concrete
language, so that you never lose the reader in fogs of ubijargon.  Mazel  
tov!

I have a much more personal question:  how did your early experiences 
growing up, and then your Special Ops training etc., lay the groundwork   
for writing this book?  To put it another way, why did the subject choose
you to write about it -- which you did magnificently?
  

    
(Jon, if you'll excuse me I think I'll parry Steve's question first,
since he's a new voice in the conversation.)

Steve, thanks so much for those kind words. I'm delighted that you
enjoyed "Everyware," and that you seem to have gotten out of it just
what I was hoping.

As to your question...hmm. I guess I'd have to say that the thing
driving me, in writing the book, was an essential concern for the
individual human person trapped in a technological situation that is
not of their choosing. How can we ensure that that person's
prerogatives are respected? How can we represent such prerogatives in
language that they themselves would recognize?

An armchair Freudian would probably have a field day with the notion
of "advocacy" that's latent in all of that, and what that might have to
do with growing up the son of a reasonably successful member of the
plaintiff's bar. And I'm sure that there's some remnant of the
classically (if slightly self-congratulatory) Jewish concern for social
justice and for the underdog in there too.

What it all comes down to is a sense that, in everyware, we have a
technological wave that is about to break over the developed world, and
the people who live there, and that nobody's yet presented this to the
people who will be most affected by it. Nobody has really stepped up
and said, hey, this is what's in the offing, is this OK with you? How
would *you* design these systems so that they respect your
prerogatives?

It just didn't sit right with me. I could see this wave building on
the horizon - not that I had any particularly privileged position,
because there were and are people who are far better versed in the
practice of ubiquitous computing - but I couldn't see anyone spreading
the word about it. Not, anyway, in any form that would give people a
fighting chance of pushing back against the aura of inevitability that
was already gathering around it.

Now, "aura of inevitability" is an idea that I picked up in my time in
PSYOP. In that context, the object is to create such a feeling of
momentum around some desired change that resistance to it is
effectively pointless. As Sun Tzu said, 2,500 years ago, the acme of
skill in war is not to win 100 battles of 100 fought, it's to win
without fighting at all. (We're all familiar with the sadly denatured
21st century application of this idea in "shock and awe," where the
intention is to create such a titanic sense of one's overwhelming might
that there's just no credible ground for resistance. And we've all
seen how well that worked out...but that's a different story.)

But you can also invert that, undermine that sense. Because an aura is
just that. It's just a feeling. If, on the other hand, you can give
people a sense that their voices matter very much indeed, and will be
crucial to shaping the change that's about to break over us all, you
stand a much better chance of coming to some outcome that's acceptable
to everyone involved.

And that's what I think sometimes gets lost in the superficial debate
about things like RFID tagging, and other appearances of ubiquitous
computing. The intent of all this pushing back is not necessarily to
prevent everyware from coming into being. It's to ensure that wherever
possible, the ubiquitous computing we choose to embrace is sane, and
sustainable, and respectful of our other values.

One last note with regard to my military experience - which was a
positive one from beginning to end, by the way: I don't want anyone to
get the idea that my training was anything even remotely as strenuous
as, say, Ranger School, or what SEAL trainees go through in BUD/S. It
wasn't, and I want to be real clear about that, out of my deep respect
for all those who do have that accomplishment under their belt.
  

    
Adam, I'm greatly enjoying Everyware, too; thank you for writing it.

What do you think of Julian Bleecker's "blogject" concept, in which
digitally-enabled physical objects "blog" themselves -- that is, they tell
the story of their own existence (history, location, use, etc.) in a way
that's easily accessible? What kinds of overlap do you see with your own
concepts?
  

    
Hey Jamais! You're very welcome.

I think Julian's doing some very important work; his Manifesto for
Networked Objects (at
http://research.techkwondo.com/files/WhyThingsMatter.pdf) is brilliant,
clearly thought-out, and just an absolute pleasure to read. In fact, I
wish he had come out with it three months earlier, as it would have
had a big influence on some of the arguments I make in "Everyware" had
I seen it in time.

But I'm afraid I *also* think that the blogject project is perhaps
thornier than he's imagining.

The reason why has to do with the nature - the *political* nature, I
might say - of metadata, or information about information. (The ID3
tags that identify attributes of an MP3 such as "Artist," "Title" and
"Album," or the EXIF data that tell when, how, and by what model of
device a digital image was produced are two familiar kinds of
metadata.)

Any time you have an object that represents itself to the world
through metadata, there's a reliability issue, and because of this I
guess I have to challenge the notion that an object can usefully
contain its own history. This is something I've written about at some
length on my own site...

http://www.v-2.org/displayArticle.php?article_num=860

...and if you'll excuse me for quoting myself, I'll just repeat a
relevant passage here:

"We know from the Web and from various p2p applications that, in the
wild, metadata [on its own] is close to useless because it can be gamed
so easily; as a result, no credible search engine relies on it nor has
done so for years....Who has the authority to append metadata? Who has
the responsibility, or even the technical wherewithal, to verify it?"

The proto-blogjects that Julian's talking about haven't really been
subject to stress-testing in a live information ecology yet - they're
one-off art projects and Gedankenexperiments, for the most part. So
they haven't really had to face up to what happens when you have actors
who are incentivized to provide bad metadata, to spoof it or distort
it - I think he discusses Beatriz da Costa's Pigeons that Blog project,
where she basically wires up a flock of pigeons to provide real-time
pollution telemetry (!).

Well and good as a thought experiment, but the presence and location
of air pollution is hardly a neutral topic. Say you're the Chamber of
Commerce of Leadville, and you've got a flock of blogjectified pigeons
up there telling all and sundry that the airborne particulate count
above your fair city is above the acceptable threshold...but that
neighboring Pleasant Acres breathes free and easy. Are you going to
tell me that, in every such instance, you're going to have the
integrity to leave that information well enough alone, and live by the
results?

No way! People are going to be gaming this stuff silly, and all the
more so because (like all everyware) blogjects take information that
has always been latent in the world and represent it to us in a way
that profoundly affects our decisions and options in real space and
time.

Now, that doesn't mean that I don't think it would be useful for (say)
a car to identify its components, and when they were last serviced,
and what they're made out of, and how they can be recycled most
efficiently. But I think the credibility of blogject traces is always
going to be in play. It can't and it won't be neutral.

Jon, I haven't forgotten your question. ; . )
  

    
(No problem, happy to see others chiming in!)
  

    
Heh...you had asked how our relationship to ubiquitous technology
changes when it's hidden from view.

I think one way of answering that follows on very naturally from the
end of my response to Julian's work, and it's to say that we forget to
interrogate things that have disappeared from view. If something is out
of sight, if it's effectively invisible, its effects can seem neutral,
natural, just part of the way things are. Those effects reacquire some
of the aura of inevitability we were talking about.

This probably isn't a problem, 90% of the time. I'm concerned about
the 10% of the time that it is.
  

    
Can you give potential examples of that 10%?