Jim Sachs talked about the philosophy of rapid prototyping that he picked up at Stanford; Dean Hovey talked about this was a perfect project for someone who came out of Stanford-- not too electrical, not too mechanical, not too big--
Kelley: Not too big-- you can't do 37 iterations of a Space Shuttle in the time we had--
Pang: One of the things I want to understand how the development of the mouse reflected the educations of the people who created it. It sounds like there are ways of working that you can blame on McKim or your other teachers--
Kelley: Right, right--
Pang: But were there other ways of working or approaches that were important to its success that you learned when you were in that program?
There's two. Rapid prototyping is the most important. The Stanford program has this thing about fluency and flexibility. That's what we teach in the very first class. In order to have breakthrough ideas, you have to have a lot of ideas, all different from one another. Fluency means that it's easy to come up with those ideas. So many inventors have an idea, and they lock into it and hold onto it, and it's like their baby, and they think they're never going to have another good idea. You have to get the students to break out of that. We give them a brick, and tell them, "Come up with 50 things you can do with a brick, other than use it in a building." And they say you can grind it up and make it into paint, or use it as a doorstop, and so on. You show them that in a very short period of time they can come up with a lot of ideas. So flexibility means that the ideas are different, one from the other: a doorstop to hold down papers, or using it to prop up a chair, those are all close to the same idea; but grinding it up is a different idea.
You start from a very early time in the program valuing lots of ideas and different ideas. So when to a problem like the mouse you come up with all these ideas, and you show them to all the smart people you can find, and they give you feedback, and it's easy to make the creative leap from the world of the possible. The trouble comes when you have one idea, and you show it around, and people say, "Yeah, yeah, that's okay;" then when you implement it, there's some "gotcha," some weirdo thing, that breaks it. If you have a whole bunch of ideas, someone's going to point that problem out. So prototyping is probably the easiest thing to talk about.
The other thing that's important, but is more subtle, is human values. A Stanford kid would never get in trouble with design a VCR no one can use, where the main person you're trying to go after can't program it. That's a lack of understanding of human values. We always have a central person in mind, who we're trying to design for; it might be an actual person, it might be a composite or someone we made up. But always in our minds, we Stanford designers are really clear about who it is we're designing for. And so you're taking these prototypes and thinking about that person: how big's that person's hand, what are they likely to do, what things will affect its use.
Recently we designed a recharging system for GM's electric car. They had it in the basement of a parking garage, and our central persona was a 30 year-old woman, and she's not going to go into the basement. She'd be afraid to go into the basement of a garage; she's going to be more comfortable with an "energy spot" next to the handicapped spot.
The mouse might have worked because we had that human values point of view, and we were showing it to people all the time. That's why I think it worked, and worked better than if all the engineers in the lab had just worked on it. One way to think about it is to look at the early Xerox mouse prototypes: I'm sure those were engineered in the lab, and while they did incredible work, they weren't human-oriented. In one sense they were, in that they made computers easier to use, but down at the detail level they weren't.... I don't really know, I'm just making that up.