Click here for a full list of our coverage of the 2008 Consumer Electronics Show.
Watch a video of this interview here.
LAS VEGAS—Just hours before addressing a ballroom packed with thousands of consumer electronics junkies, Microsoft chairman Bill Gates sat down for an interview with Scientific American. The Consumer Electronics Show served as the backdrop, though the truth be told, the event could just as easily have been called "Microsoft World," given that the bulk of the gadgets on display here incorporate, are compatible with or were designed using some Microsoft product from Windows to Microsoft Auto.
Gates has now been the CES keynote speaker a dozen times, cementing his and Microsoft's strong connection to the consumer electronics marketplace. But, come this summer, Gates will step down from his post as Microsoft chairman to work at his and his wife's philanthropic Bill & Melinda Gates Foundation, which focuses on improving the health and education of the world's very poorest populations. During our chat, Gates discussed developing technologies, such as touch and speech recognition–based interfaces and robotics as well as his transition to working with the developing world.
SciAm: About a year ago, you wrote a story for Scientific American titled "A Robot in Every Home." In that article, you contrasted the robotics industry of today with the PC industry 30 years ago. Do you see robotics exploding just as the PC industry experienced a quick acceleration in development over the past three decades?
Gates: The PC, you can call it quick or not—1975 to 1981 things were actually not so much in the mainstream. It was kit computers in 1975 and then, in 1979, we got a few TSR80s. Apple had come along by then—the Commodore. So, people weren't sure what they would do with the devices. People took our Basic interpreter, which was sort of the first thing that all those machines had in it and wrote a variety of programs. Some of them were just games, goofing around, some of them were more serious things. And that's certainly where we are with robots today.
There's no single application like word processing or buying online or watching video or organizing your photos that everybody says, "Oh, I've got to have a robot." I mean maybe the niche of the Roombas is, a tiny bit, but most of the activity is people trying out things where they themselves get hands on and do some of the programming. Some of it's just fun, kinda contest stuff or a toy that kinda marches around. And there are other people who are saying, "Okay, we can do security surveillance or we can move little packages around in this nice way." And so, it's one of those wonderful periods where many different things are being tried.
You have markets like Japan and Korea that I'd say are particularly intense at trying these things out. The new hardware designs and the energy are probably stronger in Japan and Korea than anywhere else in the world. That's fine. And once they find a solution, then the products will tend to be used on a global basis. For the personal computer, Japan was one of the early founding markets; the U.S. was, as well.
So, you can't predict that it's going to have the same shape that PCs did, but it has a lot of the same character. Take watching video on a PC. That was more than 30 years from when the PC started to when it's kind of been, "Well, obviously, I watch videos on my PC. Why not? It's way better than my TV—I can choose different things, I can organize this stuff in different ways."
So, when does the robot help in health care, so it can move people around and such? That's very hard. When does it really clean up your apartment? It turns out that's really a tough problem. Like many things in software, when we try to match human capabilities, the first decade or so, we gain more respect for the incredible control system, the learning system, [that] the human brain represents. Now, it doesn't mean that we don't ever get there. On things like visual recognition, speech recognition, the progress is very, very concrete. It'll be the same in robotics.
I think this last year has been every bit as active as I would have expected, so I wouldn't change a thing in the article. I love the way you ran it and I somewhat stake my reputation on some statements there. But I was clear that it wasn't going to be an overnight thing. As far as the cost of the hardware, people are really doing a fantastic job—walking, balancing, data acquisition. Also, some of the software libraries for those things that we, and others, are building—our software platform, people are downloading it.
This is Microsoft Robotics Studio, correct?
Yeah. If you want to get the real specific flavor of it, you should talk to [general manager of Microsoft Robotics Group] Tandy Trower and the guys in the group who live this thing every day. I think they got some of the most fun jobs in the company because they're at the beginning of a whole new thing.
Do you see your Robotic Suite as being a corollary to Microsoft Basic going forward?
I hope so. Now, we need to get more value added in there, in terms of planning, vision, speech-type things. Just like we did with earlyBasic, we see how people use it—if a lot of people are having to do the same thing, we think, "Okay, let's get that into the tool kit, so they don't have to duplicate that piece." So, we need to be very dynamic as we go along. But, that run time will get a lot richer. So somebody who just has a hardware innovation or just has a peripheral innovation, they immediately get leverage and don't have to build all the pieces. Once upon a time, every guy who wrote applications for PC had to write all of the printer drivers. Then we said, "No, no, no. Let's not have people do that."
Well, robots are well before even that stage, where to build a full application, you still have to do a lot of pieces. We look at things like those car-driving competitions and we say, "Okay, what is common in those software stacks? If we were going to make it so that it was half as much work next time people are doing those car competitions, what would those elements be?"
To get them sort of halfway ...
Right. And the general planning piece, the model of the world-type piece is where, over time, we can probably do something very important, but there are different ideas that we're testing out on that.
Aside from continuing developments in robotics, what other emerging technologies do you see on the horizon that have a chance to become mainstream?
Everything around natural interface—whether it's touch or visual recognition or ink recognition or speech recognition—all of those are things that we've seen demos of for a long time. And getting the right pieces, where it really explodes and becomes mainstream in the market—I feel like many of them are right on the verge of that. The reaction we've had to Surface, I think, is a good example of that. People have gone nuts just seeing some of the simple demos we've offered, because the directness is pretty strong. The extra cost, of course, is just a few cameras that down there; it's a Windows PC with a few cameras. And then what we put out in the last few months is the development tool kit.
[Gates then demonstrates a program that allows him to design a surfboard on the Surface.]
The amount of code on top of the development kit we have isn't a ton of code. It's the kind of thing that over the course of a week or so you could write. If somebody says, "No, let's change this—add this," another week, they can have that done. So, whether it's in the office, where you're sharing data together, whether it's in the home, where you're organizing photos, or at retail, where the early versions, which are fairly expensive—they want this kind of rugged thing. It's been a great response. That's software that came out of Microsoft Research. Like a lot of our great products, it's the transfer from the research side into the software that has allowed us to do neat things.
So, my final question is: Given that this is your last keynote address at CES and your final year overseeing day-to-day operations of Microsoft, what do you see as your role in shepherding new technologies going forward?
Well, my work at the [Bill & Melinda Gates] Foundation that will become my full-time focus is based on optimism that, whether it's software or biology, the advances can be shaped in a way to help the poorest two billion people on the planet—whether that's a breakthrough for malaria, a breakthrough for a high-quality health care system, a breakthrough for the educational opportunities that those poor people are not able to get today. And so, it represents kinda optimism and enjoying meeting with scientists and getting them organized and recognizing the things that go on that hopefully I've learned a lot about during my time at Microsoft.
In my part-time role at Microsoft, I'll still meet with Microsoft Research and see what they're doing, because that's always been the most fun part of my job. I'll have a few projects, maybe related to search and Office and natural interface that I spend a little bit of time on, but the overall software agenda, Ray Ozzie and Craig Mundie are obviously stepping up to manage that, and that's where I'll get the time for the new Foundation work. But, my time will be more looking at what opportunities science is bringing us than less, and that's a neat thing.
So, you're going to become like a scout?
Yeah, for the Foundation. Because if you see what the problems are—we did things like grand challenges where we enunciated, "Okay, if you could give us cold chain for vaccines inexpensively, that would save a huge number of lives here"—and articulating, so that the inventors don't just have the market signals [of what the rich world wants], but also the recognition and the awareness of what the nonrich world wants. There's some creativity just to be applied there.
The Internet gives us a tool to do some of those things. The desire of companies to be more socially minded gives us a tool, so that if we reach out to them in a concrete way—whether it's pharma companies or technology companies—I think they'll respond well to that. And when I go into those companies, I can at least bring some corporate experience and some encouragement along, hopefully, with a concrete plan of where they should do more.