March 2007 Archives

This week was a two conference presentation week for me. The second was the keynote for Etech, but the first was for the Outdoor Historical Museum Forum. The Forum, which our friends at The Henry Ford asked me to participate in, is a gathering of leaders in the US outdoor history museum world. I think that outdoor museums are fascinating examples of the long history of experience design. They've been designing experiences for 100 years or more, in the interest of creating environments that allow people to see things that no longer exist in their world, in environments people no longer live in. It was an honor and pleasure to spend some time with these folks.

Unfortunately, the audience of many of these museums is shrinking. My talk was about the role that technology can play in helping history museums communicate their core competitive advantage, which I defined to be authenticity, and provided some examples of projects that I think used technology particularly well to do that:

The history museum's advantage relative to other activities is direct exposure to real artifacts and experiences. You provide the examples on which explanations of contemporary life, politics, industry, etc, are based. People’s understanding of “here and now” starts with “there and then.” You’re the there.

[...]

I believe that new digital technologies can greatly lower the costs of communicating the value of authenticity. In other words, they can tell you what makes the real thing REAL.

The full text of my presentation is available as a 600K PDF.

It appears that Movable Type has been marking all comments as "junk," regardless of what their spam filtering plugin rates them at, so I appear to have not gotten any non-junk comments in weeks, if not months. I apologize. I've gone through the last couple of entries and published those comments, but it may take me a couple of days to get to the others, and there may be some that rolled off the comment list and I won't be able to find.

I sincerely apologize. I've been wondering why it's been quiet, and now I know.

If you know of a good MT solution for this, email me at blog c/o this site. Thanks.

This morning I gave a keynote at O'Reilly's Etech. It was an elaboration on the theme of magic in the design of ubiquitous computing user experience that I've been developing for a while now.

The core of the piece were three linked arguments about emergence:

• The emergence of ubiquitous computing from market forces acting on, and in concert with, CPU prices
  
• The emergence of animist reactions to devices that have behaviors that go beyond action-reaction physics
  
• The emergence of magic as a metaphor for the design of ubicomp devices
  

I've made the presentation (710K PDF) with all of my speaking notes available.

I'm reading the entry about sympathetic magic in the MIT Encyclopedia of the Cognitive Sciences in preparation for my Etech talk. In it, they talk about the three "laws of sympathetic magic" set down by late 19th and early 20th century researchers. The first law is called the law of similarity, which they define as:

The law of similarity has been summarized as “like produces like,” “like goes with like,” or “the image equals the object.” Likeness is elevated to a basic, often causal principle; the simplest example confounds likeness with identity, hence “appearance equals reality.” The adaptive value of this law is clear: generally speaking, if it looks like a tiger, it is a tiger.

[...]

Examples of similarity include burning effigies of persons in order to cause harm to them, or reliance on appearance in judging objects when the appearance is known to be deceiving (e.g., avoidance by educated adults of chocolate shaped to look like feces; or difficulty experienced in throwing darts at a picture of a respected person). In the domain of words, Piaget (1929) described as “nominal realism” the child’s difficulty in understanding the arbitrary relation of word and referent. Similarly, educated people have difficulty disregarding a label on a bottle (e.g., “poison”) that they know does not apply.

Reading this, I thought of Virginia Postrel's Substance of Style. In that book, Postrel summarizes the user experience of the value of design as I like that. I'm like that. That sounds a lot like sympathetic magic to me. We choose the things we choose to decorate ourselves and our lives because those things resemble (in some way) the people we want to be (whether or not we are). In the two universes of Carhartt clothing: the European urban youth version of the brand, invokes the American rural laborer brand, because the laborer brand brings with it some of the values its youth brand buyers want, an aura of authenticity that's impossible to acquire otherwise. That's pure sympathetic magic, and it's interesting how hidden in plain sight it is.

The nice folks at Ambidextrous Magazine asked me to contribute an essay on magic as a metaphor for ubiquitous computing user experience design to Issue 6, which launches this week. This essay fleshes out my October dorkbot presentation. Here's the premise

What’s missing in this technological vision, however, is a consistent design language that explains how these devices work to the people who will use them. No common verbal, visual, or interaction techniques have emerged to help users navigate a world filled with augmented devices.

[...]

This is where magic can help us. The desktop metaphor is largely inadequate to describe the wide range of form factors and functionality possible with devices that do not have screens or pointers. Mobile phone screens hardly resemble 1970s offices (the inspiration for the desktop metaphor). A shoe that dynamically changes its functionality using sensors and a small CPU looks even less like an office. And yet nothing currently is replacing the desktop metaphor.

I then go on to define what a magic metaphor for ubicomp user experience could contain and how it structures the design process. For me, the point of metaphor is to communicate a set of concepts not just to users but also to developers, designers, marketers, and CEOs. In other words, metaphors ease the creation of consistent experiences by providing a useful set of constraints and a shared visual, behavioral, and verbal vocabulary.

The full text of my piece is here (224K PDF), but you should go out and get the magazine. Ambidextrous is always full of great stuff.

I've been playing with Swivel lately. It's fun, but their "Compare" icon, an apple and an orange, is particularly apt. Here, I compare a data set I entered (the price of first class postage in the US, 1910 until today) to the Consumer Price Index. Kind of nonsensical--one is an emergent trend, the other is a set of conscious decisions by a small group--yet somewhat interesting in terms of how the our postage rate continues to lag behind all the other products (thus shutting up my complains of how it's so expensive)

(photos (cc) by eecue and decade_null, found on Flickr)

A couple of years ago I wrote about an idea I had for visualizing the implicit heat maps in Wifi signal strength using actual heat.

I never made the device, but I thought about what the ideas it was pointing at and generalized this as an observation I called "new data for old senses" and wrote some notes about it that I never shared here. Today PT at Makezine posted a link to a project along the lines that I was thinking about. It's a Wifi sensor that uses vibration to give you a sense of the Wifi landscape around you without having to look at anything, which was the crux of my idea in 2005. So, since the idea is now out there, here are my notes:

-----

I'm interested in the idea of using senses that don't normally get used for device communication as secondary display channels. This is a way to allow access to what John Udell calls the vast middle ground between devices that either demand our full attention or none.

We have more senses than sight and sound, which are channels already full with important information, so how do we use our "secondary" senses to communicate "secondary" information?

What are other kinds of senses and other kinds of data we can use?

Here are the somatic senses (thanks, Google!):

• touch
  
• pressure
  
• vibration
  
• heat
  
• cold
  
• pain
  
• proprioception (the feeling of joint movement)

What to visualize? Liz has been doing a bunch of stuff about visualizing people's relationship with the RF spectrum and geography, but I've been thinking that there are several granularities that would change in perceptible and interesting ways. At human scale in a city there's Wifi strength; at car scale there are things like crime maps, and at airplane scale there are political boundaries (voting records, natural phenomena).

-----

The bottom line is:

How we can introduce secondary information into people's awareness in a secondary way, using their less-used senses and without adding additional cognitive noise to the primary channels of sight and sound?

How do you define a sense of place? I've lived in the same San Francisco apartment for 11 years. It's the longest I've lived anywhere, ever. I have a great deal of affection for this neighborhood, for what was once an Ohlone settlement, and then the Spanish Mission grounds, and then a succession of immigrant neighborhoods throughout the 20th century, ending today as a gentrifying bohemia on the edge of the Castro. Walking around the spaces immediately adjacent to Ramona Ave, the street I live on, I regularly look at the buildings and think about the layers life that have happened here. A single spot may have had a shell mound, a Spanish goat grazing field, the house of a ship captain who ferried people from the east coast for the Gold Rush and settled, an Irish sheet metal worker's family, a Mexican grocer, a house of junkies, a struggling painter, a Web developer and a second-year associate at a downtown law firm. That layering is fascinating to me.

Yesterday, I decided to do a little experiment on my way to the cafe where I sometimes work and back. I grabbed a bunch of images from the San Francisco Public Library's historic photo collection of places within a two-block radius of my house, printed them, and took pictures that were as close as I could manage in a couple of minutes.

14th and Guerrero, 2007/1928

14th and Dolores, 2007/1929

15th and Dolores, 2007/1956

16th and Dolores, 2007/1929

16th and Dolores, 2007/1856

I tried to match the camera angle and the lens, but it's not a meticulous recreation. Mostly, I'm interested in documenting how the place has and hasn't changed. Really, the physical space hasn't changed much at all. There are new buildings, certainly, and in one case a particularly nice one was was burned by an arsonist in 1993, but the overall space hasn't changed much, with one exception. The trees. Holy cow is there more foliage in San Francisco now than there was even 40 years ago. You go, Friends of the Urban Forest

Here's a Platial map documenting the locations:

After all of the observation- and analysis-based discussions of terminology on this blog, I decided to do a little experiment to see if there was any data that could be collected. To get an idea of how much people used which term, I first thought that I should just tally the Google references to each of the terms that refer to ubicomp and related concepts. Then I realized that that technique suffered from the unpredictable nature of the Google estimation algorithm and it didn't recognize that some of the terms have been around a lot longer than others, so there are probably more documents that refer to them, even if they're not as popular in the field today.

I decided that a better way to do this would be to buy some search engine keywords (specifically "ubiquitous computing," "ubicomp," "pervasive computing" and "ambient intelligence") and watch how many times the keyword was searched-for and how many times people clicked on it.

Here are the results from 11 days of keyword placement on Google:

OK, what do we see? I'm going to treat clicks as active interest and impressions as a kind of semi-active interest, though I acknowledge this is projecting a lot on the audience and may be conflating several factors in terms of audience composition and intention. In this analysis, "ambient intelligence" is a nonstarter in terms of active interest, although as many people searched for it as "ubicomp." Maybe it's just an academic term and my ad for ThingM (there needs to be something people can click on ;-) wasn't interesting to academics. Still, it's interesting to note that no one clicked on an ad that mentioned "ubicomp" when they had searched for "ambient intelligence."

The next finding, though, I think is the most interesting: "pervasive computing" was searched-on as much (actually a little more) as "ubiquitous computing," but clicked significantly less. First of all, I was surprised that it was searched for as much, but it gets clicked on less than "ubiquitous computing," and that puzzles me. It shows different levels of interest, or different audiences, between the terms. Moreover, the cost-per-click on it is significantly higher, which means that other people are trying to buy the term (I don't think I had any competition for any of the other terms).

Anyway, out of this small thicket of numbers come more questions than answers. Some things we can extract reliably: "ubiquitous computing" and "pervasive computing" are both roughly equally as popular in terms of interest. "Ambient intelligence" is not so popular. "Ubiquitous computing" creates more active interest than any of the other terms, and "ubicomp" is not in nearly as active use (despite its greater popularity as a tag on del.icio.us; for comparison: ubiquitous computing).

Excuse the shovelware. I just found a position paper I wrote for a CHI2004 workshop called "Lost in Ambient Intelligence." This is a very short paper that' a recap of the argument I made in my animism essay of 2003, but formatted for CHI. I wanted to link to it mostly so that I'd have it handy. Here's the abstract:

I posit that the more widespread ubiquitous computing becomes, the more people's explanations of technology will resemble animism. As common technology becomes more interconnected and smarter, people's understanding of its functionality ceases to be based on a mechanistic model and becomes more anthropomorphic. This model in turn changes the way that the creators of ubicomp devices have to approach their designs to consider systems of objects and users' acceptance of a level of unpredictability.

It's a 150K PDF.