adaptation, personalization. possibility, limits.
Fabio Sergio | December 2002
It all started when Dan Hill posted the presentation on Designing for Adaptation he gave at London's AIGA Experience Design forum.
A compelling introduction to the concept, with a thought provoking "manifesto" strategically placed at the very end.
I wish I could have attended the talk in person, but the presentation itself and Dan's comments were more than enough to stimulate neuron growth.
A few days earlier Anne Galloway had made (seemingly) unrelated considerations I had always meant to respond to.
In typical emergent fashion, they started to make not more, but different sense.
Anne had referred to open-mindedness as a mental disposition that allows people to let go of their own beliefs to be able to embrace those of others.
She had expressed doubts about the ability to sustain this position over time and had instead pointed at using possibility as a better term to refer to the same idea.
I agreed and sent her a quote from French poet Stéphane Mallarmé:
"To define is to kill. To suggest is to create."
I realized that, interpreted in very practical terms, both Anne's thoughts and the quote above can easily be applied to adaptation as it relates to design, architecture and its on-line information-focused step-sister, IA.
Adaptation has to do with the level (and granularity) of definition that designers want to impose over their work.
In other words: how much control over your creation are you willing to relinquish?
How much do you want to define, how much to suggest?
If you believe you have the right answers already, are you willing to let others disagree with you, or even try to formulate different questions?
As Dan says in his presentation:
"You don’t control; you humbly enable."
But thinking about yourself as an enabler can be very tricky, as any designer (in, as usual, the broadest sense of the term) will attest.
Even fascinating memes like Naoto Fukasawa's "design dissolves in behavior" can be x-rayed as pointing to the designer as a catalyzer of the habits, needs and desires of the user.
That's to say, a design might dissolve, not necessarily its designer.
Mass-media have been pushing businesses, sports, fashion, architecture and design to constantly breed new rock-stars fit for the spotlight.
How hard or idealistic would it be to start teaching young designers the value of "disappearing in the background" and letting people and time take control over the artifacts they've conceived?
Strangely enough a different conversation on Jonathan Jaynes' pages helped spark more thoughts along these lines.
Could it be that digital technologies and the increasing levels of personalization they have enabled over formerly mass-produced, industrialized products have been silently pushing history back to the early days of friction between art(s & crafts) and (industrial) design?
Think about the Nike iD concept for example, and how already such a basic level of personalization has led to heated discussions around control over the final product.
Once personalization is pushed to a certain level who can actually claim ownership over the final outcome’s appearance and qualities?
Nike has decided to let its customers have a say in the way Nike's own products end up looking, but it has opened the door only half way.
The iD concept in the end fails (from this perspective) because an Orwellian Deus ex Machina can ultimately negate the very principle behind the concept.
But, I find myself asking, what if customers had been really enabled to change the very principles that help define the product's intimate characteristics and not only its shallowest shell?
What if they had been collectively given control over the underlying rules and not just their manifestations?
In other words: what if adaptation was just as an emergent quality of personalization?
A couple of examples come to mind.
The first is BBCi’s (in)famous darkening of often-clicked link boxes feature.
Arguably while this can be perceived as a basic form of adaptation, it's actually based on an (implicit) choice of the user to indicate his/her preferences.
The second example comes from speech-recognition software, such as Dragon Naturally Speaking.
They usually require new users to dictate certain keywords that are then matched to the software’s own sound schemas.
In this case people have to actively contribute to jump-start the system into gear to then allow it to further adapt with frequent use.
To relate back to Nike iD’s example could you imagine if all of sudden in both of these cases an "entity" would suddenly advise users to "stop clicking the same links over and over" or that "their pronunciation is bad"?
While personalization is usually triggered by voluntary human action, adaptation tends to be based on established principles at the micro-level, which over time inform behavior at the macro-level.
Adaptation in its purest form is somewhat of an automated, evolutionary process.
In this sense designing an adaptive artifact is not unlike growing a child.
After a certain point a parent can only hope that the values that have been implanted in the early years of the child’s life will be strong enough to inform its entirety.
It is also important to notice that the parent-child system usually doesn’t follow a master-and-slave model. As any parent knows, values passed to a child can be strongly influenced by the child’s own inclinations (genes) and interaction with the environment (memes).
Even more importantly during this process the parent’s beliefs can be changed by the experience as well.
To make things short: a virtuous feedback loop between the parent and the child can help the parent-child system to evolve.
Let’s re-map the metaphor above.
An adaptive artifact (the child) can be given values (rules) that will influence its growth over time (adaptation) based on its characteristics and its interaction with the environment (users).
The feedback from its use and the resulting adaptation(s) can be reflected in the following generation of artifacts (evolution).
In the case of digital artifacts all of the above can be thought as on-going, continuous dialogue, and parts of its grammar have already been established (see for example Tom Coates’ comments on Dan’s talk).
For physical artifacts similar scenarios might be soon coming from nanotech labs.
In the meantime we’ll have to make do with Transformers. ;)
To design adaptation we could be thus looking at designing scalable personalization, to use a term that's dear to Information Architects.
The end result could be evolution.
Red-headed step children
If (and I am saying "if") all of the above is true another set of issues has to be considered.
It is often said that in systems showing emergent qualities "the whole is bigger than the sum of the parts".
What is not always evident though is that the whole ends up being also stronger then the individual parts.
That is to say: the principles that keep the whole together are stronger than those followed by the parts.
An ant, alone, is a dead ant. A lazy ant is also a dead ant.
Howard Rheingold has in fact called his latest book Smart Mobs and not Smart Groups realizing that the potential for collective will to exhibit negative effects can be stronger than the values professed individually by the members.
Without recalling Bill Joy-anesque scenarios, in this case I am simply referring to potentially dangerous outcomes related to allowing too much freedom either to users themselves or to the system, especially when thinking about physical artifacts.
How would sue-happy markets like the US face the possibility of products that could end up adapting and evolving in unforeseen ways?
Even more dubious: are people ready or willing to take on the increased level of responsibility these changes could bring?
As Dan points out in his presentation it's in the middle of the personalization-to-adaptation continuum that designers need to find a role to enable users, depending on their inclination, to sit back and trust the designer’s decisions, to take direct control, or to just let time have its ways.
The often-quoted control aspect finally re-fuels another old debate: how much of an "artistic" component is still inherently part of the design process?
Supposedly design should not be in itself an act in which the creator remains the only real measure of the final result.
While art can be completely unintelligible and still hold great value, design acquires relevance when put to use by people.
In purely industrial terms a designed artifact’s final value is measured according to the product’s ability to resist obsolescence on the market.
At the same time though, it's pretty evident to anyone who has ever had a desire to purchase one of Michele De Lucchi's Produzione Privata lamps that in many cases the boundary between art, craft and design can be quite blurred.
Based on cost alone in many cases design can still be often considered an industrialized form of art.
The unresolved tension between limited edition, crafted objects and mass-produced designed artifacts is reconciled in the need, in both cases, to work within limits.
To be effective in the chosen medium of expression an artist has to be a master at it, to know and embrace its limits.
Using Dan Hill's words again, to "get it".
We have thus come full circle back to "open mindedness" and "possibility".
Letting go of beliefs and limits is not and easily sustainable position for designers either.
Maybe it will be by teaching designers to embrace the limits of their own ego, of their intended users and of their chosen physical or digital medium that we'll be able to help shape a future that will adapt to people and not vice-versa.
and if you still want more...
take a look at freegorifero's weblog for daily rambling and sightings.