To Dream the Invisible Dream

John Seely Brown When it is written, the history of computers will, I believe, be quite simple. In the beginning was the computer. Then it disappeared. Of course, it didn't go away completely. It just dissolved. Either it became part of the physical background, forming part of ordinary objects such as tables, chairs, walls, and desks. Or it became part of the social background, providing just another part of the context of work. Indeed, this second phase of the history of computing is already under way. The modern car is really a four-wheel computational platform. Yet I'm rarely made aware of this when I drive it. Furthermore, when I go to the automotive showroom, I don't have to ask what operating system or presentation manger the car uses. Here, at least, computers have finally gotten out of the way. The field of human-computer interaction is really configured around this central paradox. Designers struggle to produce simplicity out of complexity, direct connectivity out of mediation. Instead of drawing attention to itself, the best design lets us reach through computers into the world, allowing us to focus on creating value, not manipulating tools. So, for example, in panic stops and radical curves, the computational power in my car doesn't add to my problems by drawing attention to itself. Instead, it invisibly helps connect me to the road and the world outside. From this perspective, I see the new paradigms for design and use developing hand-in-hand. As they adapt to current practice, new technologies become less visible. Yet, simultaneously, by adopting these new technologies, current practice continuously evolves. Clearly, this is not a view of radical transformation. We all love to be radical and to pursue radically new ideas. But our experience shows that fundamentally new technologies seldom get adopted in a discontinuous fashion. When put to real use doing real work, new inventions almost always miss their mark, no matter how many tests ran in the lab. Adaptation and adoption require extensive fine-tuning in the real world. The passage from the Lisa to the Mac is a famous example of this. At the same time, the mistakes Apple made in over-hyping the immediate potential of the Newton show how hard it is for any of us to learn this lesson. Nevertheless, I think we all need to learn it. Instead of focusing all our attention on radical transformation, we should try to understand the dynamics of "radical incrementalism." This is what turns radical invention into innovation. © ACM, 1996. "To Dream the Invisible Dream" opened a special section of the August 1996 issue of Communications of the ACM, entitled "New Paradigms for Computing," edited by Ted Selker. It is posted here with permission.