emarks by Bill Gates, Chairman and Chief Software Architect, Microsoft Corporation
Maria Klawe, Dean of Engineering and Applied Science, Princeton University
Microsoft Research Faculty Summit 2005
Redmond, Washington
July 18, 2005

RICK RASHID: Now I have the distinct pleasure of being the person who gets to introduce Bill Gates and Maria Klawe for their session this morning. We're doing something a little different this time. We're going to actually have Bill and Maria on stage having a conversation about some of the subjects that they think are the most important right now to be talked about.

And for those of you who don't know Bill -- I don't think there are probably very many of you that are out there -- he's our chairman and chief software architect, and has really been a participant in this conference for the last five years or so, and I think people have really enjoyed the opportunity to interact with him.

Maria Klawe is dean of engineering at Princeton. And if you don't know Maria, she's an incredibly vibrant and exciting person. She has a tremendous amount of passion and energy. I happen to share duties with her on the board of trustees of the Anita Borg Institute for Women in Technology. And I could just tell you from my experiences with Maria, the passion she has around education and students and issues such as gender equity and diversity is just tremendous.

So with that, let me bring Maria and Bill on stage, and I'll get off, and we'll have them have their conversation. Thanks, Maria. (Applause.)

MARIA KLAWE: Thanks, Rick.

RICK RASHID: Hey, Bill.

MARIA KLAWE: So this morning, Bill, I'd like to start off with a really important question: How many of the Harry Potter books have you read so far? (Laughter.)

BILL GATES: Well, it's a phenomenon that I'm going to have to get myself immersed in. I read two, and then my daughter decided they were a little too scary. But now I understand even adults are supposed to read them. In fact, I'm curious what do you think is going to happen, what would you like to see happen in the Harry Potter series?

MARIA KLAWE: Well, I have one special dream, and I'd really love to see Hermione choose to become a computer scientist. (Laughter.) I think that would really turn around our problem, especially for young girls.

BILL GATES: Good. Well, even Harry can do it, too.

MARIA KLAWE: That would be great, too. After all, we all know that the magic in the future comes out of computer technology, and especially computer software.

Let me move to a slightly more serious question. As I'm sure you know, one of the biggest concerns of computer scientists in the United States is the decline in federal funding for academic research, and graduate education. We've taken the liberty of inserting a few statements from a recent article by Ed Lazowska and Dave Patterson, two of my favorite computer scientists. Their article is entitled: "An Endless Frontier Postponed".

And from the article they say, "U.S. IT research grew largely under support from DARPA and the National Science Foundation. NSF relied on peer review, whereas DARPA relied on vision and reputation, complementary approaches that served the nation well.

"Over the past four decades, the resulting research has laid the foundation for the modern microprocessor, the Internet, the graphical user interface, and single user workstations; and has also launched new fields such as computational science. Virtually every aspect of IT that we rely on today bears the stamp of federally sponsored research.

"However, in the past three years, DARPA IT research funding at universities has dropped by nearly half. Last year, NSF supported 86 percent of the federal funding for fundamental research in IT at academic institutions. The funding rate for competitive awards in the IT directorate fell to 16 percent, the lowest of any directorate. The average funding rate is close to 30 percent."

So why do you think the government should be spending money on computer science research in tough economic times? What does the public get out of federal funding for research?

BILL GATES: Well, I think the payoff, if there's any place you can say there's been a dramatic payoff, it's in computer science. The United States in the 1980s was viewed as falling behind, Japan had a better industrial model, the U.S. just was going to lose industry after industry; and yet what really happened in the 1990s was that our economy created more jobs, new companies, lots of amazing leadership things happened. And I think you can really point to the DOD and NSF money that went into computer science work as being one of the key elements that allowed us to turn what was a period where people thought we were falling behind into preparation for one of the greatest success periods the country has ever had.

The amount of money we're talking about here is not gigantic, I mean, compared to, say, the government budget as a whole or the defense budget or even research as a whole, the portion that computer science really should get is not that gigantic, but to have a decline is really bad.

And part of the reason it's hard for people to see the decline I think is that there was all that DOD related money, DARPA primarily, but I remember when I was a computer science student, all the reports would have things like, "thanks to the Office of Naval Research" and I always wondered what the heck was the Office of Naval Research doing on speech recognition or linguistics, but, what the heck, it was good money.

Ando we've been able to draw on all these different sources, so as some of those have become shorter term or more focused, the numeric impact on are we really tackling these frontiers has been more dramatic than I think people realize.

And it's kind of a crime that at the time when computer science is about to solve the most interesting problems, and when computer science is not only an interesting field of its own with some exciting problems, but it's also becoming the toolkit for all the sciences where biologists are turning to us and saying, OK, how do you find the pattern in this information or astronomers or physicists or basically all the sciences are becoming very data driven, you'd think, wow, there would be a shift of NIH money into computer science techniques and standards and things like that. That's also not happening to any significant degree.

MARIA KLAWE: So why is this? I mean, what are we doing wrong? I mean, I agree completely with you that computer science -- I find it hard to imagine any field more important to invest funding for research in, because of its importance both at the core of computer science in terms of sort of what we're going to be creating that's going to drive the industry, the economy for the next ten years, but also its applications in other areas. So what is the computer science community doing wrong?

BILL GATES: Well, I think there are two boundaries that we have to be very clever about. One is the boundary between academia and commercial innovation, and the other is within companies, between their research oriented groups and their product oriented groups. And if you have famous stories like Xerox, where they funded a lot of research and it was very helpful to Apple and Microsoft, that doesn't make business managers think, wow, I'm just rushing out there to increase this big research budget.

That's why Microsoft is trying to be a fantastic example of how you can manage both of these boundaries; first, being able to say that the best investments we've ever made as a company is having our Microsoft Research group, and that the impact that has on our products over the long term is amazing, but even within, say, a two-year period, that a lot of good ideas flow, and not just from the Redmond lab, from all the labs that we have, which are sort of four significant labs now.

So we're saying to companies, hey, you ought to invest more in R&D, this is defining the future of the company, it is our competitive edge that we're out there on the frontiers, and we're always surprised, at least in our field, our competitors, if you put aside IBM that's kind of a special case, the amount invested in research is very small. I'd like to see that change, I think it's a huge mistake on their part.

The boundary between academia and us as a company is another one that we've tried to make that an extremely fruitful dialogue, events like this being part of it, and I think we've had spectacular success. Our ability to fund certain faculty activities and fellowships and not have Microsoft Research think of university work as being a competition of what they do, but rather perfectly complementary so even if we're pursuing one way of doing AI internally, if there is another way that's making better progress outside, I want to know about it, I don't want my research people to sort of suppress this other approach.

And so I didn't know when we started out would this give us more visibility into the world of academia or in the extreme case it might actually give us less visibility. I think that's worked out very, very well.

So we'll certainly be as strong an advocate as we can be that the government is making a mistake here, and throughout the world I think governments should fund computer science research. I think in terms of creating great jobs, great companies in their area, what other area would people be funding. This is the change agent of the time, this is the thing that will drive forward. Even just say you're only going to do it just for education to build the tools for the future of education, you'd want to fund it just for that one little piece alone, not to mention --

MARIA KLAWE: Healthcare.

BILL GATES: -- yeah, e-government --

MARIA KLAWE: Environment, energy.

BILL GATES: Yeah, modeling the world, it's about time we understood things like the CO2 cycle and stuff like that, and definitely computer science will be at the heart of that.

So it is an incredible paradox, and you need examples of cases where it made a big difference, and certainly our company wouldn't exist without that funding that took place on those basic advances.

MARIA KLAWE: So let me go to the other half of the crisis in computer science. I mean, I think at least within the U.S., but I believe it's also true in other areas of the world, not only are we seeing a decline in research funding, but perhaps even more worrisome we're seeing a huge decline in interest in studying computer science.

So just give some data, the popularity in the U.S. of computer science as a major for incoming college students has plummeted. It's fallen more than 60 percent between 2000 and 2004, according to the Higher Education Research Institute at UCLA.

On the other hand, according to the U.S. Labor Department, the fastest growing jobs throughout 2012 include data communication analysts, health information technicians and computer software engineers.

The most recent numbers for U.S. employment in IT are the highest ever, up more than 5 percent since the peak of the bubble in 2000. In addition, salaries in the IT area have continued to grow by a compound growth rate of 4 percent.

So it seems that there's a huge mismatch between the demand for graduates with these skills and what we're willing to pay graduates with these skills and the interest among our youth.

Do you have -- I mean, do you have a sense of what's going on here and should we be really worried about this?

BILL GATES: Well, I'm certainly very worried about it. Microsoft is trying to hire every great college graduate who has basic computer science skills and we think is highly talented. When I sit down and review projects here inside the company, the topic that always comes up is how is the hiring going, we've got open headcount, these are super well-paying jobs, you can get your own office --

MARIA KLAWE: No cubicles?

BILL GATES: We're not big on cubicles. Other people like it that way, but we've always believed that developers should just have the quiet of their own office.

And so it really is gating the speed at which we do things.

I think there are two sources of talent coming to work here. One, of course, is native-born talent, which will always be the dominant part, and there we have to have the interest level and the excitement behind. And then the other is foreign talent coming in at various levels of the education system or coming in as part of a career move.

And in both cases what's going on is very troubling. I think the fact that so many smart people were willing to come to the computer science departments in this country and many of them stay and take jobs, that was a huge win-win; great for the country, great for those people. I know when we were going to start up in India, it was because we'd had really good people here who came from India and wanted to go back that that became possible, and so that worked out very well for us; likewise, the same thing in China.

Some people say, well, doesn't this mean there will just be a shift to Asia of a lot of this activity, and there will be some, but you take a company like ours, we're going to always have the vast majority of our development taking place in this location. We like to do things in a unified way, and so even though India and China are going to grow quite a bit, it's a big problem for us that we can't get these great students.

And I say to myself, what are these other fields doing, I mean, what's going on? Apparently the fastest growing major is physical education.

MARIA KLAWE: No! (Laughter.)

BILL GATES: And so I think wow --

MARIA KLAWE: I thought it was economics.

BILL GATES: You know, what is going on in that field? I mean, are they making breakthroughs like speech recognition or artificial intelligence? I'm dying to see these new games they're inventing, new rules. And I think, you know, the poor Chinese, they don't realize this is the coming field, and 10 years from now they're going to wake up and say, oh no, physical education, we completely missed that activity. (Laughter.)

MARIA KLAWE: So let me ask you, when Microsoft -- I mean, are you finding enough people to hire in the U.S.?

BILL GATES: No, absolutely the answer is no. We have this interesting paradox where in China and India we can get lots of engineers but getting people who have sort of what we call program management type skills or general management type skills, it's very hard to find enough of those, whereas here in the United States we do pretty well at getting people with those skill sets, but then it's just the engineering we're very short of what we'd like to get. And so the competition for somebody who's got the right background is just phenomenal.

MARIA KLAWE: So describe for me the perfect engineer that you'd like to hire. What kind of skills, capabilities, talents, so on would they have?

BILL GATES: Well, when we first started the hiring, doing university hiring, which was all the way back in the '80s when we really started putting a program together, we knew that we weren't looking for a specific skill, it wasn't like, oh, do you know how to write C programs or something like that. What we wanted was somebody who understood the field, the idea of algorithms and programming, and that they had done some projects that we could ask them about to get a sense of their depth of understanding, did they really model out the whole problem and try something.

For us it's ideal if they can learn, they don't just program in these garbage collected languages, but they actually get some sense of, hey, underneath it all there really is a notion of resources, memory resources, computing resources, things like that.

So it's a tiny bit of a concern for us as the curriculum doesn't teach you what are the basic mechanisms, so we'd like to see more of that.

We'd like people to have written reasonable programs, the main problem being that if they haven't done some of these things, it's harder for us to get a sense of would they really enjoy this for the long run, and do they have the type of discipline it takes to be very good at doing that.

And so it doesn't matter what the specifics are, but some degree of thinking through a tough programming problem that then as part of the discussion process we can find out how they like that.

The one thing that has worked very well for us is the intern program. Today, tomorrow, in two days I'm having interns to my house, we actually have four of these things, because I try to have about 1,200 interns come over.

MARIA KLAWE: That's a lot of people at your house. It must be a big house.

BILL GATES: Yeah. It's a lawn. Actually it's more on the lawn than in the house. So it's 300 a night, but it's a tradition, and it's fun for me to hear what kind of questions they ask, which companies they give me a hard time about, or which activities.

MARIA KLAWE: OK, which companies do they give you a hard time about?

BILL GATES: Well, there's the one that's faddishly hot right now. (Laughter.)

MARIA KLAWE: Faddishly hot; get that one.

BILL GATES: And there's always one like that. Google sort of has that pole position right at the moment.

But they don't ask so much about that as what are the interesting things, what should they do in their career, a lot of good questions.

MARIA KLAWE: One of the things that we all know is that as fields go, computer science probably has the most rapidly changing content, and technology evolves so quickly. And both as head of a computer science department and then now my second job as a dean, having computer science departments, I would always get into these discussions with people in the math department saying, it makes sense that your people teach more courses per semester than the computer scientists do because you're still teaching the same courses that you taught 50 years ago, whereas in the computer science you have to continually redevelop materials so that you really are covering the most up to date things.

And I wondered if you had just ideas that would help us or whether Microsoft has ideas that might help computer science departments stay on top of the most recent technological developments.

BILL GATES: Well, certainly it's the goal of our University Relations Group to make sure that we're talking about what we think the state of the art problems are, finding out from the universities and a lot of dialogue back and forth about that. In a certain sense, yeah, the curriculum has changed, but say somebody came for an interview and they said, "Hey, I read the 'Art of Computer Programming', that's all I ever read, I did all the problems, I would hire them right then."

MARIA KLAWE: You'd hire them right then.

BILL GATES: Yeah, that's right.

MARIA KLAWE: So would I.

BILL GATES: Even if they didn't do the double-star problems, I mean, just the fact that they'd read the whole book, you know, those are the kinds of things you need to know to be a good programmer. Actually, there's some of that you don't even need to know, but the kind of algorithmic thinking that's promoted there.

So in a sense, we want to teach the same thing, but we'd like to teach it in a forum that's most interesting. So, for example, the idea of, OK, let's take some of those basic ideas and program a robot to go somewhere or figure something out, you want to inject that into the field. But what you're really teaching the person about design is pretty much the same as you wanted to teach them 30 years ago. I mean, we still haven't gotten past that. And so there may be rich runtimes that we can give to universities that make sure that the person learning those things feels like they're doing something very cool and very interesting.

MARIA KLAWE: Well, let me ask you a little bit, and this one comes back to the research again, because throughout my life as a computer scientist there's been the sense about what is core computer science and then what is sort of cross-disciplinary or bridging into other areas, and, of course, my experience has been that 20 years ago graphics wasn't core computer science and neither was databases and now they are, so what's core computer science keeps on going like this and this and this, which I think is actually great.

But one of the things that we've been thinking a lot, both at the education level and at the research level, is the extent to which education or research should be focused on the computer science discipline or the extent to which it should be bridging an application oriented; so, for instance bioinformatics or computational science.

And one of the issues coming back to the research funding is that there's a lot of current concern that even at NSF much of the funding for research has been focused on cross-disciplinary areas, and that we worry that if you focus so much on that, that the funding for the core research areas, the traditional areas, whether it's graphics or theory or systems, is really being hurt.

And so I'm interested in what Microsoft's position is on this, both in terms of where should we focus our research priorities within the field of computing, and also when you think about it at the educational level, is it better to graduate somebody who has a double major in biology and computer science or better to graduate somebody who is really fixated on computer science.

BILL GATES: Well, there's a lot of super-important problems that are pure computer science problems. I get a chance to talk about how the world is going digital, and we're going to have digital lifestyle, digital work style and everything, video, audio, how you're going to store those things, deal with your memories, but in order for that dream to come true, there are some very tough problems that just haven't been solved, problems in security and privacy, which that's computer science, problems in data organization, how you navigate that data, how you present the information.

We have an interesting dilemma coming up in that clock speeds of microprocessors aren't going to be going up much from about 3 to 5 gigahertz, and so what the chip guys are giving us is lots of processors, and they're saying, hey, you computer science guys --

MARIA KLAWE: Figure out how to use them together.

BILL GATES: -- it's about time you guys figured out how to take n processors and get n capability out of these things, and now just go do that.

Well, that turns out to be one of the great unsolved problems, and yet there's no way around it. I mean, literally Intel is going to have 32-processor systems running at 5 gigahertz, and that's what a modern PC will be, say, within four or five years. So we certainly need brilliant people thinking about that problem, what's going to go on with it.

All of the issues around natural interface, you know, vision, speech, ink, modeling, those are areas where the economic value of getting those things right, whether it's Microsoft or anybody else, it's the next ten years that it's interesting. I think if I'd written down in the late '60s, early '70s the interesting problems in computer science, if anybody had, that we could say that a small percentage have been solved right now, but that a very high percentage will be solved in these next 10 to 15 years. And so that should make it --

MARIA KLAWE: But only if we work on them.

BILL GATES: Oh, that's true.

MARIA KLAWE: I mean, only if we have funding for the graduate students to actually work on these things and so on.

BILL GATES: No, but people should want to work on these problems. I mean --

MARIA KLAWE: Yeah, of course they should.

BILL GATES: Well, this is the most biased crowd in the world perhaps.

MARIA KLAWE: Well, let me talk about this issue about wanting to work on these problems, because I have managed to succeed in convincing one of my two children to study computer science. I'm glad I at least got one of them. The other one happens to be female, and I sometimes tell myself that the reason she doesn't want to work in computer science is because she sees her mother as a pioneering rebel, and so she wants to be a rebel, and her mother wants her to work on computer science and therefore she's going to work on anything else.

But I know from having spent a lot of time both talking to my daughter and her friends, but doing broader research as well, including surveying thousands of high school students, that one of the issues that really stops a lot of particularly young women but also minorities and others from wanting to study computer science is the image of the computing career and what a computing professional is like.

And so just to quote some of the young women I've spoken to over the last 10 years or so, they say things like, "Well, if you're going to work in a computing career, all you're ever going to do is program, it's going to be 24x7 programming, you're going to be pale because you never get outside, you never talk to anyone, and then you have to work with all these horrible other people who don't have a life, they don't know how to talk, they don't even know how to do these things; and besides, it's boring." (Laughter.)

I don't know, anyone out there had this conversation with someone?

So I know what my field is like, I know it is the most exciting area in terms of affecting the world, I know that people work in teams, I know that, yes, you have to be creative, you have to be smart, you have to be disciplined, all of those kinds of things, but you also need to know how to communicate.

What can we do about creating a more positive image for our profession and what is Microsoft doing?

BILL GATES: Well, here is another example where I think the best thing Microsoft can do is just share the examples of what kind of jobs these are, and how interesting they are. I spent a lot of time last week over in the Microsoft Office group and they're working on this new version, and it's so exciting to walk the halls and have people show me the feature they've worked on and talk about what customer meetings they were in that led them to believe they ought to do this thing, and what other group they talked to.

I mean, the nature of these jobs is not just closing your door and doing coding, and it's easy to get that fact out. And, in fact, the greatest missing skill is somebody who's both good at understanding the engineering and has good relationships with the hard-core engineers, and bridges that to working with the customers and the marketing and things like that.

And so that sort of engineering management career track, even amongst all the people we have, we still fall short of finding people who want to do that, and so we often have to push people into it.

And so I'd love to have people who come to these jobs wanting to think of it as a lot and exercise in people management and people dynamics, as well as the basic engineering skills. That would be absolutely amazing.

And we can promise those people within two years of starting that career most of what they're doing won't be coding, because there are many career paths, say, within that Microsoft Office group where you're part of creating this amazing product, you get to see how people use it, you get to then spend two years, build another version, and really change the productivity in this very deep way, take some big bets on what you're doing and do some things that are just responsive to what that customer wants.

So I think if people understood they wouldn't feel this way at all. In fact, they ought to go learn about these other jobs like --

MARIA KLAWE: Like being a lawyer, for example.

BILL GATES: Well, or going into a sales job where you're always just measured by a number basically, OK, give me a bigger number, and you're not doing anything that's really all that new and different.

I just don't get it myself, and I don't know why it's not evident to people. We do, when we do university recruiting, we have the people who are doing these jobs go out and do the interviews. We make sure they allocate time to that.

Now, by the time we're going out and interviewing, of course, we've already lost a lot of the --

MARIA KLAWE: Well, they haven't studied computer science in the first place.

BILL GATES: Right.

MARIA KLAWE: Well, so let me -- I mean, one of the things that within my lifetime has been a huge change is the number of women who have gone into medicine and law.

BILL GATES: It's phenomenal, yeah.

MARIA KLAWE: I mean, it's not 50 percent.

BILL GATES: It's greater, yeah.

MARIA KLAWE: And I have this feeling that there is some correlation between that and the fact that there were tons of TV shows around the time that I was a teenager, but they're still there, about really exciting careers in medicine and law that show women and men having lives and status and doing really important things. And I know lots of lawyers and doctors and I know that I think being a computer scientist is a lot more interesting and a lot better job, a lot more creative than the jobs that many doctors and lawyers have.

Why have we been so unsuccessful in getting the media and particularly Hollywood to create sitcoms or adventure shows or reality shows even, whatever -- (laughter) -- around our field?

BILL GATES: Well, I think it is a tough field, because it involves -- say somebody went over with a bunch of movie cameras into that Microsoft Office building, there would be some terminology that would not make sense --

MARIA KLAWE: You don't think this is true in medicine? I mean, just imagine if you were really, really looking at what's going on in surgery or in a dermatologist's office or lots of these things. I mean, it's actually -- or go film a lawyer's office and watch what they're doing in a lawyer's office.

BILL GATES: If they really did, yeah, if they told the truth. (Laughter.)

Well, that's another approach is to get the truth about these other professions out there. (Laughter.)

MARIA KLAWE: Well, do you have any thoughts about what are more effective ways to get more women into computing careers? I mean, one of the things that's really depressing from my perspective is that computer science is the only field in science and engineering where participation of women has gone down over the last 25 years. So, for instance, if you look at mathematics, when I got my PhD in mathematics in 1977, I think it was about 11 percent of the PhDs went to women, and now it's over 30 percent. If you look at undergraduate degrees in mathematics, it's about 45 percent, and it was down around 10, 15 percent.

So in computer science, our figures are now about 15 percent of the PhDs go to women, about 15 percent of the bachelor degree recipients in research universities are women. I mean, it's just unbelievable how bad it is. We're down there, we're below physics in some cases.

So what could we do to bring more -- what would be effective in getting more women into these fields?

BILL GATES: Well, I don't know the magic answer. I think everybody who thinks about the problem says you've got to get the women who are in the field to be more visible and get them --

MARIA KLAWE: No, no, no, no, that can't be the answer. OK, raise your hand if you're a female here. All right. Are we being visible? Are we serving on every committee, going to all the schools?

BILL GATES: Well, it's good, you should keep doing that.

MARIA KLAWE: Yeah, we are going to keep doing it, but I hate to say it --

BILL GATES: I applaud that.

MARIA KLAWE: -- we're not getting anywhere with it. The numbers --

BILL GATES: I think if you weren't doing that, we'd be even worse off, to be frank.

MARIA KLAWE: Yes, I think that's absolutely true. But I'm sort of coming down to Hermione, and Katy Holmes is my dream for computer science role models because it's clear that the sort of things that we're doing on the grassroots level ourselves, I mean, I've been going out to the schools since I was 17-years-old. I mean, I have never stopped doing presentations in elementary schools, high schools, middle schools. Every woman computer scientist I know is doing the same thing. So somehow there's got to be another way that we can succeed at this.

BILL GATES: Well, I think that's one of the things. I think you've got to go -- we've looked, as all these studies have, at what stages you lose people at, and unfortunately you lose --

MARIA KLAWE: You lose them at every stage.

BILL GATES: -- if you lose at about five stages, the cumulative effect is pretty incredible, and you get the fact that the rich get richer and the poor get poorer; that is, if there aren't enough women in the field, then women find it unattractive simply because of that fact. If you got everything else right, and they said, oh, there's only 15 percent there, you might still lose just because of that, whereas if you got it up to 30 percent, then it would have a very different feel to it. So I think there is a critical mass element of this.

I should know the numbers better. If you look in Asia, are they doing better on the male/female aspect?

MARIA KLAWE: No, and this is very depressing as well. So as part of a talk I was giving at ITiCSE [Innovation and Technology in Computer Science Education], which is a computer science education conference in Europe, I tried to collect statistics on what the participation rates were around the world. And I certainly don't have the numbers, this is actually work done by a couple of sociologists and anthropologists who had collected the figures and they had 22 countries, so it's not every country by any means.

So the top two countries were Ireland and Turkey. Now, for Ireland I had some conversations, fortunately there were some people from Ireland there, and they said that a part of it was that one is that this is the dominant economy in Ireland right now, but the other thing is that because there [is] still a lot of single-sex education, and so girls, there's a lot of data that shows that single-sex education is more successful in keeping girls in math and science areas.

Turkey, the reason is because you don't have a choice in Turkey. So [in] Turkey you get, based on your exams, you get assigned to a particular program, and so they were producing a lot of women in computer science, but not because the women were choosing to go there.

And even in those countries they're up around the 30 percent range, OK? And in many other countries, Germany's down to around 9 percent and so on. So it seems to be a phenomenon world wide.

And at China you can see, if you look at places like, like Jaotong or Tsinghua and so on that are the elite schools, they're again down around 10 to 15 percent female.

BILL GATES: Really?

MARIA KLAWE: So it's a worldwide problem, and certainly one we're going to work on.

I would be I think really remiss with this audience in the room if I didn't ask you some questions about your views on the future of innovation and technology. So I think everybody would want to ask you this: Of the many technologies that Microsoft is working on, which one excites you the most? Which one are you just completely passionate about?

BILL GATES: Well, it's like asking me about which children I like best. (Laughter.)

MARIA KLAWE: OK, you're like two, yes. (Laughter.)

BILL GATES: Well, I, I can say without a doubt, of all the things going on in Microsoft, the work in Microsoft Research is not really the answer, but because they have a longer time horizon on the things we're doing, that is the most exciting thing. I have time where Rick will bring in people to talk about their progress and what they're doing. That's about the funnest part of my job is to sit and listen to that and think about how we take advantage of that.

And people here probably know we have this thing called Tech Fest where all the researchers essentially create booths and all the product development people come, and it's over 10,000 go through and look at the ideas. And the energy and excitement around that makes that one of the funnest things on the yearly calendar.

Yeah, we have such a range of products. We have this Tablet PC is a very cool thing that eventually every student would just have a Tablet.

MARIA KLAWE: Yeah, I've got one, and love it, I have to say.

BILL GATES: And the hardware is going to get dramatically better, the software is going to get dramatically better. So that's something that I've been a personal advocate for internally even, against quite a bit of skepticism and so it'll, it will take a number of years, but it's neat to see things coming into place for that.

You know, we have things like Xbox that is just sexy because it's a videogame, but even there we had to invest billions of dollars. And the first generation of that we knew we didn't have the credibility or even the skill set to be the leader, and so it took a lot of money and a lot of internal skepticism to say should we be in this, and then say, no, if we do it well enough we, what do we get; we get to play again. And so now with this second round we're coming out with a product that's pretty good, and it sort of indicates that level of investment that we made.

And it's amazing how it's drawing on the deep technology that gets used in even something like that is pretty amazing, whether it's the communications or the graphics, or the security, or things of that nature.

I've always had a small bias towards the things we do for people at work because the impact Microsoft can have by making office work more effective, the way that people collaborate or know what customers are up to or plan projects, we can actually improve world productivity far better with what we do there than any of our consumer products.

And actually it's a field that there's this incredible vacuum. If you sit and say, OK, what company comes into work and thinks about, you know, making meetings effective or getting rid of meetings in large part or empowering people with information about what's going on, we're kind of in a unique position there, certainly in terms of really thinking about that and planning ahead.

And so the dreams we have there about digital work style reach down all of the way into the system, so how can a database let you find things, how can there be security that spans multiple organizations, what are the new user interface technique for these things?

And so I've always been a champion of OK, the consumer stuff is neat, we need to do that, but let's keep business productivity in mind as well.

And so we're an interesting mix where some areas like parts of graphics, parts of database, Tablet computing, we're kind of off on our own in terms of doing commercial products, and then we have a few areas where we have a competitor who's doing well, and we have to not only catch them but get ahead of them.

So take, Web search would an example of that, and that's fun, too, because you have the target has been painted for you, you know what it is you have to match and exceed and get ahead of, and there's a certain dynamic in the team where you get to go in and say OK, go for it, how are we going to beat these guys, let's go do that.

And so the way you manage the morale and excitement is slightly different between a team that's just trying to meet and exceed where you know what you're doing, versus where you're out in the frontiers by yourself just defining new areas. And you've got to make sure you've got a good balance of those, particularly the frontier things. With Xbox we're not just matching what Sony's doing, we are in the period where we're going out and really extending what can be done, and likewise with things like Office.

MARIA KLAWE: I think we probably have time for one last question, and so I'm going to save it for actually my favorite one. If you'd ask me what I thought would really make a difference in terms of getting more women in computing --

BILL GATES: OK.

MARIA KLAWE: -- one piece of it, I mean I think there's a whole bunch of things. I think Rick Rashid actually has one of the best answers, which is make the first computer science class you teach the best computer science class you teach in each of our departments and in schools and so on.

But the real question I want to ask, I think the other thing that will make a huge difference is I personally believe that the most important thing we have to do today is use technology to address societal problems, especially in developing regions. And I know that the Gates Foundation has made this really a priority. And I would like to close by hearing some of your ideas about how our technology, computing technology, algorithms software, hardware, all these kinds of things can address things like healthcare, education, water quality, all of those kinds of things. Do we have a role to play?

BILL GATES: Yes, absolutely.

MARIA KLAWE: (Laughter.) No, no, that's not enough.

BILL GATES: No, no, I know. The lawyers always tell you first you say yes or no, and then --

MARIA KLAWE: Oh, I see, OK.

BILL GATES: -- and then you explain.

MARIA KLAWE: Yeah, I haven't had as much experience there as you have. (Laughter.)

BILL GATES: No, I, I have. (Laughter.)

It seems if you don't first say yes or no it's, they say it seems evasive, like you're trying to evade the question. So it's -- (Laughter.)

MARIA KLAWE: Oh, I see. Good, thank you.

BILL GATES: One of the, the coolest things I saw in the last years was I was in Mozambique in an outpost called Manica where they were doing some work on a malaria vaccine. And the key reason that this site could do this, well, first there was an incredible doctor from Spain who was there organizing it, but he had actually tracked the people in the village and their health history using a little PC Access database.

And so the ability to actually tell who was having fever, what was that caused by, and really prove out was this drug working, it was really an IT thing was the key thing that was missing there, and that's why that was one of few sites in the world that we could do this. And so sitting down and talking with him about what are your problems, how do you get support, how do you get connectivity, how can we make these things simpler, what's expensive?

You know, it's fascinating, this technology, because it's now very low cost, high volume stuff, can reach out and be used basically anywhere in the world. The toughest problem we still have is the connectivity piece, and that's where work that I'm sure many people here are doing. Microsoft is very keen on so-called mesh networking where you can take the wireless spectrum and use that to get, particularly in areas with low density, very good connectivity at very, very low cost. That's kind of a fascinating thing. Then there's other things about how you can make the computer better for that early education thing, make it more approachable, that type of stuff.

If all we were doing, if the only thing we were doing in the field was providing computing technology to developing countries, it'd be worth it just for that.

Likewise, if all we were doing was developing computing technologies so that biologists would be able to collaborate and cure diseases and share ideas and model data, find patterns, things like that, if that's all we were doing, it would still all be extremely worthwhile. You know, that's a great field of endeavor, and that is one case where people who work on the boundaries, we need some of those at the same time as we need people who work on our fundamental problems in order to build the tools that they need.

So this is a field where - I'll just give you one more example - where if you go to any of these countries and see where they have blind people or people who have physical disabilities, and you see how computers have come in to be the tool that let's, say, the visually impaired reach out and get any information on the Internet, as opposed to in the past where their access to things that have been rendered into brail were extremely limited, you can just go in and see that's it night and day, their ability to communicate, collaborate, participate is phenomenal.

And that's all because this is a field that's taken a very high-volume, low-price approach. It's driven breakthroughs, driven the prices down. And so it's very gratifying, and it's not just for the richest 2 billion people on the planet, although there are some neat things happening there, it's for the world at large.

MARIA KLAWE: Thank you. (Applause.)