Magic Design

A few weeks ago, I wrote about magic and how subconscious problem solving can sometimes seem magical:

When confronted with a particularly daunting problem, I’ll work on it very intensely for a while. However, I find that it’s best to stop after a bit and let the problem percolate in the back of my mind while I do completely unrelated things. Sometimes, the answer will just come to me, often at the strangest times. Occasionally, this entire process will happen without my intending it, but sometimes I’m deliberately trying to harness this subconscious problem solving ability. And I don’t think I’m doing anything special here; I think everyone has these sort of Eureka! moments from time to time. …

Once I noticed this, I began seeing similar patterns throughout my life and even history.

And indeed, Jason Kottke recently posted about how design works, referencing a couple of other designers, including Michael Bierut of Design Observer, who describes his process like this:

When I do a design project, I begin by listening carefully to you as you talk about your problem and read whatever background material I can find that relates to the issues you face. If you’re lucky, I have also accidentally acquired some firsthand experience with your situation. Somewhere along the way an idea for the design pops into my head from out of the blue. I can’t really explain that part; it’s like magic. Sometimes it even happens before you have a chance to tell me that much about your problem!

[emphasis mine] It is like magic, but as Bierut notes, this sort of thing is becoming more important as we move from an industrial economy to an information economy. He references a book about managing artists:

At the outset, the writers acknowledge that the nature of work is changing in the 21st century, characterizing it as “a shift from an industrial economy to an information economy, from physical work to knowledge work.” In trying to understand how this new kind of work can be managed, they propose a model based not on industrial production, but on the collaborative arts, specifically theater.

… They are careful to identify the defining characteristics of this kind of work: allowing solutions to emerge in a process of iteration, rather than trying to get everything right the first time; accepting the lack of control in the process, and letting the improvisation engendered by uncertainty help drive the process; and creating a work environment that sets clear enough limits that people can play securely within them.

This is very interesting and dovetails nicely with several topics covered on this blog. Harnessing self-organizing forces to produce emergent results seems to be rising in importance significantly as we proceed towards an information based economy. As noted, collaboration is key. Older business models seem to focus on a more brute force way of solving problems, but as we proceed we need to find better and faster ways to collaborate. The internet, with it’s hyperlinked structure and massive data stores, has been struggling with a data analysis problem since its inception. Only recently have we really begun to figure out ways to harness the collective intelligence of the internet and its users, but even now, we’re only scraping the tip of the iceberg. Collaborative projects like Wikipedia or wisdom-of-crowds aggregators like Digg or Reddit represent an interesting step in the right direction. The challenge here is that we’re not facing the problems directly anmore. If you want to create a comprehensive encyclopedia, you can hire a bunch of people to research, write, and edit entries. Wikipedia tried something different. They didn’t explicitely create an encyclopedia, they created (or, at least, they deployed) a system that made it easy for large amount of people to collaborate on a large amount of topics. The encyclopedia is an emergent result of that collaboration. They sidestepped the problem, and as a result, they have a much larger and dynamic information resource.

None of those examples are perfect, of course, but the more I think about it, the more I think that their imperfection is what makes them work. As noted above, you’re probably much better off releasing a site that is imperfect and iterating, making changes and learning from your mistakes as you go. When dealing with these complex problems, you’re not going to design the perfect system all at once. I realize that I keep saying we need better information aggregation and analysis tools, and that we have these tools, but they leave something to be desired. The point of these systems, though, is that they get better with time. Many older information analysis systems break when you increase the workload quickly. They don’t scale well. These newer systems only really work well once they have high participation rates and large amounts of data.

It remains to be seen whether or not these systems can actually handle that much data (and participation), but like I said, they’re a good start and they’re getting better with time.