Newsmaker: PC or people--who's the boss?

Even though the days when computer-human interaction revolved around the C: prompt are far behind us, one legacy from that era remains.

Despite sundry advances in operating systems over the intervening two decades, it's still not entirely clear who's the boss: the human operator or the PC.

For the folks who helped usher in the C:-prompt era, it's a matter of high priority. That's one reason why Microsoft's research division just hired Bill Buxton, a designer and expert in human-machine interfaces.

Buxton will focus on software design issues that stem from the "society of devices" taking shape now. As more people begin to use mobile phones and PDAs (personal digital assistants) or cars and appliances with on-board computers, software makers have a whole new set of challenges not seen in PC software.

CNET News.com spoke to Buxton about the importance of getting design right in the emerging world of ubiquitous computing.

Q: You're now a senior researcher at Microsoft. Can you tell us what you'll be doing?
Buxton: The simple answer is to keep doing what I was doing for the last 30 years, which has been the direction of my research. Basically, it's how do you make technologies that have a positive impact and good experience on the people who use them? I mean that not just in the sense of individuals but even at a cultural level. There is some science behind that. How do we increase understanding on how to do that--or even understanding the impact of technology to make better decisions? The other part is the design part.

Microsoft obviously has very wide penetration in terms of their impact. But they're very conscious of the impact and wanting to do it better. There's always room for improvement. So they brought me to add to their arsenal of expertise to really do things well.

What do you think you'll be affecting, tangibly? Is it how people work with their PCs?
Buxton: The nature of our relationship to computer-based technologies is in a state of flux. To be honest, to a large extent the PC as we know it is in many ways a success story. But it's a story that is through the introduction. It's at a mature state now. And a lot of the dynamics will be in what comes in the future and our changing relationship with technologies.

The nature of our relationship to computer-based technologies is in a state of flux.

So let's just look at things like ubiquitous computing and (how do you) put them in a more manageable form--the things that we carry with us, whether we wear them or keep them in our pockets, and things like electronic whiteboards or signage or kiosks and so forth, those types of things.

It's like (looking at) the social relationship between my watch versus my phone versus the shop window that I'm walking by versus my car and how things layer together.

How do you think things will shape up?
Buxton: There are two possibilities here...In the best of worlds, this "society of devices" starts to complement one another and, in their collective selves, reduce the overall complexity that confronts you and me in terms of engaging in the world and going about out lives. And at the same time, (they) improve the experience.

Or it could go where these things create "Modern Times" on steroids, if you remember Charlie Chaplin, where it gets out of control. Nothing is given. One of my heroes is a historian of technologies, Melvin Kranzberg. He has this law that technology is not good; technology is not bad; nor is it neutral.

There's an increasing awareness that every time you introduce any technology into a society or culture, it will have an impact. Once you acknowledge that, then it behooves you to make your best effort to understand what that impact will be and design it to so that it's a positive contribution. And that doesn't happen by accident. That's where I fit in...Microsoft recognized this and wanted to increase their bench strength in that area.

In that ubiquitous computing scenario you were describing, what is the best we can expect as users? Will people be able to interact with multiple computing interfaces easily? If you look at PCs, it's not something people pick up quickly.
Buxton: If it's an effective design, it's possible to make the stuff transparent. That doesn't mean you are not using highly developed skills to do what you're doing. But the trick here is that effective design, generally whenever possible, is based upon highly learned skills, but skills learned through a lifetime of everyday living.

Think about the example of talking on a cell phone in the car. If you put your handset up to your ears while trying to dial, that's like putting lipstick on or eating a hamburger while you're trying to drive. I don't want to be next to you on the freeway.

The interesting thing is if you can throw a cell phone on a passenger seat, the phone rings, and the phone says to the stereo, "There's a call coming in. Can you turn the music down and can I borrow the speakers? And by the way, can I use the microphone embedded in the steering column?" Now you can just talk hands-free.

Related reading
Microsoft hires user interface guru

Bill Buxton, whose roots are in music and computing, will work with Microsoft Research to improve product design.

In that case, at least in the first approximation, talking to you on a cell phone interferes with my driving no more or no less than a person sitting on the seat.

The important thing is if this were done properly, the phone would automatically sense this by its physical location, and its behavior would change transparently. And when I got up out of the car, its behavior would change again.

It just shows that you can design so the behaviors become transparent. So instead of banning cell phones in the car, maybe we should be banning bad cars. It's not about talking in the car, not so much about cell phones. It's about design.

Obviously, computing is becoming more pervasive with the use of PDAs and embedded computing. What's the state of affairs in terms of interface design?
Buxton: We're going through some real growing pains. A friend (of mine) used the analogy of the frontier. We're going through a certain period of development where things haven't gotten settled or codified yet. It's a bit of mess, and there is some lawlessness and bad stuff out there. I'm certainly not satisfied with where we are right now.

What do you think can be improved?
Buxton: My personal bias is against the super appliance: the Cuisinart or Swiss Army knife approach to design.

I like back-country skiing. And the PDA that is most valuable to me, the one I bet my life on, is my avalanche beacon. It's a digital device; it's multimodal, collaborative. It's wireless. If you try putting a clock or calendar in that thing, it will be over my dead body literally.

More Newsmakers

About time!

This is really great, even though it's ten or twenty years late.

I really hope, for the sake of the User, this guy makes a difference at all levels. That last paragraph says it all.

[Blito]

Tired of the cameraphone?

Also this could be a better thing for saftety as well like not even knowing if you have a camera on your device and it's connected somewhere. Also I think MS should take more of a design consulting role as well as they were always more behind-the- scenes. When they pushed IBM aside and came out with Windows they got beat up too much on the front lines. For example Windows being too all-in- one. So this is a step in the right direction I think.

[Charleston Charge]

Positive Changes

I think MS has been hiring some very good and talented people lately who will be making a positive change and impact for the company. Hopefully in the near future we will be seeing some good things as a result of these individuals.

[RicktheBrick]

What computers should know

There is no reason why computers should not know almost everything of importance. A computer in a automobile should know where it is located. It should know everything about that location. That is the speed limit and thus not allow the automobile to break that limit. It should also know if there is a reason to stop such as a stop sign or traffic light and also respond to that condition if the driver refuses to acknowledge it. Every automobile should be part of a wi-fi network as it should report it speed, location and direction of travel to all nearby automobiles. Corrective action should be started if the computers detects a collision. In the home the computer should also know everything. There should be a microphone and speakers in every room. It should be able to recognize by sound everything that causes flow(electricity, water, and gas) so it can monitor all of them. It should be able to initiate corrective action when it can not explain any flow. It should also be able to interact with humans. It should know when one is present and if that person needs help in any fashion. Most of which could be done with a simple yes or no question. Such as are you fine?, are you leaving?, will you be away for more than an hour?

[Jonthin]

Microsoft. Do something really well???

Fat chance... Microsoft products that I have known are designed,
then and now, to be over-engineered; incredibly complex and
inefficient.

They need to start from the ground-up and do things SIMPLY. Make
the product efficient, easy to use and understand. Why can't
Microsoft learn from Apple? They have been making software for
Apple for years but they seem so stupid that they miss the point all
the time. This guy will have his work cut out for sure.

[LordKalthorn]

Apple?

Haha. You realise of course that the only reason Apple got even the faintest chance to work from the ground up on OS X is because OS 9 was so completely dreadful that breaking any form of backward compatibility and starting from scratch was the only way to gain customers? Windows, Office, Internet Explorer, Media Player, Dynamics, everything really, work. They work perfectly fine; they may be complex and compared to Apple programs that get the chance to know precisely the system they're working with they are probably inefficient but they work on more PCs than there have ever been or ever will be Apples. Even if they started from scratch right now we both know they wouldn't be finished for ten, twelve years. The reason it is complex and inefficient because of that is that to code for such a ridiculously huge customer base you need that or the entire thing would fall through. You just plain cannot code for a Billion people Simply.