The Future of Computing

Lately I've been traveling quite a bit, so I've had time to read some books. This time around, I'll discuss three of them: The Design of Future Things (Don Norman), Leonardo's Laptop (Ben Schneiderman) and Everyware (Adam Greenfield). I've chosen these three for discussion because in many ways they are about the same thing: the future of computing. Norman's book of course touches a whole lot of other matters about future things, whereas Greenfield's book is about ubiquitous computing specifically, but all three books send the same message. It's time for users to step up.

The machines are about to take over. Not like in the movies of course, but subtly. We are about to become servants to our machines. Clearly, it should be the other way around but without enough thought given to human-centered design, we might end up with machines that think they know us and what's best for us, restricting our actions when they see fit. Machines, that demand our attention and maintenance. Networks that share everything about us without our consent. When computing will be everywhere, using it is no longer a matter of choice. And if the interfaces suck? We are screwed, and we have to learn them. These are some of the scenarios discussed in these books.

If, for some reason, this does not scare you, then think about what user-centered design for future computing could give us. A lot of possible scenarios for the future are presented in Leonardo's Laptop and they all share one central theme: computing should support, not replace or automate, human tasks. Technology should be making our lives easier, not harder. Computing should be available, not integrated. And, contrary to some beliefs, it should never be entirely seamless or invisible. Which computing do you want?

Jumping to another topic, the idea behind the concept of interactive spaces is in many ways similar to ideas described in these books - especially The Design of Future Things. Humans and machines do not talk to each other - not today, probably not in the future - they signal. When communication is limited to signaling, it is better for one participant be completely predictable. In this case of course, it should be the machine. The user knows what he wants to do - there's no need for the machine to try and guess, because no matter how much sensor information and whatnot it has, it cannot read our thoughts (hopefully).

Norman presents a good example from human-to-human communication where this also holds true. Here's another one: sometimes when walking, you find yourself walking on a collision course with another walker and when you avoid - yup, the other guy avoids to the same direction, and probably then the both of you avoid again to the same direction. This is a scenario where neither participant is predictable. Indeed, if you had just kept walking straight, the other guy would've avoided you. This scenario becomes a lot more dangerous when bicycles or cars are involved. It's happened to me with bicycles, and if I hadn't landed on my butt, I might not be writing this.

To wrap up, here are some of the principles of interactive spaces:
- Interfaces are available to the user - not seamless but gracefully seamful.
- Interaction is initiated by the human.
- Interfaces are precitable - they don't make decisions on their own.

Out of these books, I would highly recommend Everyware for anyone even remotely interested in the future of computing, and especially defending human rights in the coming* age. The other two books are good reading for designers, and especially Norman's book is once again really easy to read and even fun. All of these books made me more confident that what I and we are doing is relevant and important.

* In fact, we are partly there. The latest of these books (Everyware) was printed in 2006, and already it was said ubiquitous technology is in part already here.