Remarks by Bill Gates, Chairman and Chief Software Architect, Microsoft Corporation
College of Engineering, University of California at Berkeley
Berkeley, California
October 1, 2004


A. RICHARD NEWTON (Dean, College of Engineering, University of California at Berkeley): Well, Bill, let me join the chancellor in welcoming you here to Berkeley. As you know, Berkeley is the greatest public university in the United States. (Cheers, applause.) At least as important is that our College of Engineering is the most distinguished on the entire planet. (Cheers, applause.)

So let me start this dialogue with an easy one: From your seat as Chairman and Chief Software Architect at Microsoft, obviously you get a chance to see all sorts of things that most of us don't get a chance to see and you have a lot of visibility into the future of technology and how it's going to affect both business as well as consumers.

If you were one of our young engineers sitting out there in the audience today and you hadn't yet made up your mind entirely what path you were going to follow, what is it that would excite you the most today, do you think?

BILL GATES: Well, I think the interesting problems that will be solved in the next 10 or 20 years are probably even more fascinating than what we've solved in the last 10 or 20. In the area of computer science, the Holy Grail is writing perfect programs, having software that can understand speech, software that finally takes the challenges of artificial intelligence and delivers aid in that area. I'm very optimistic that we will solve those problems.

If we look at the personal computer today, it's certainly a glass half-full in terms of ease of use and things that it does for you. Wow, it's amazing compared to not having that tool in the past, the way it connects up, gives you information. But it falls so far short of what it should be, just the reliability. Take, say, helping you record your memories and letting you go back and find things that are of interest, today we just don't deliver on that and yet you can glimpse with some of the things going on that that will be possible.

So I'd say to be someone working in computer science, to be somebody working in biology, there are real advances that will improve the quality of life that the people here will get a chance to make those breakthroughs.

A. RICHARD NEWTON: So would you sort of say information technology is the path to take or more biology or nano today? I know biology is a strong personal hobby of yours.

BILL GATES: Well, certainly information technology is critical to everything that's going to happen. It's interesting by itself because we really will get natural interface solved, we'll get vision solved, we'll get reasoning solved, we'll create these little robots that will do nice things. That's pure information technology. But even as the sister science, to say biology, computer science is very important. I think a lot of the breakthroughs will be made by people who are trained in biology and in computer science because these are complex systems and the idea of state and messaging and all those things, those are concepts that we've developed in the world of computer science and seen how nature uses those things. There's so much to be discovered so you can't go wrong being in computer science.

I'd also say biology is a very hot area. Changing health, improving those outcomes, that's a very basic thing, not just in the rich world but even more importantly in the world at large where there are still a lot of diseases, some that are epidemic today like AIDS, and fundamental understandings about molecule shape space and protein expression, we will be able to model these things in a deep way and it will be the tools of computer science that will be helping there.

So any of the sciences I think you have pretty neat frontiers, but I'd highlight those two as the ones that I'd probably choose if I was a student today.

A. RICHARD NEWTON: Well, you know, in Australia we have an expression called "poor poppy syndrome." I don't know if you've heard of that but it goes like this. There's a field of poppies and one poppy sticks out above the rest. There's a tendency for everyone to want to cut its head off. (Laughter.) Microsoft obviously is the 500-pound gorilla in the software industry and as a result you've had your share of critics and your share of lawsuits over the last 10 years. What have you learned from that experience?

BILL GATES: Well, I wouldn't give up being the tall poppy. (Laughter.) It comes with its price, but then again that's what's created the opportunity for us to create Microsoft Research and take on some very tough and risky problems.

The whole way the computer industry was structured before we came along was very different. IBM did hardware, software, applications; everything was in a horizontal stack. And our proposition was that things would be much more -- I'm sorry, they had the vertical stack; we said we'll just do the software piece and we'll have specialization in chips and distribution and applications, so it was a different industry model and it succeeded beyond our wildest dreams that we created a big software industry, we created the idea of technology being created not just in one country but every country around the world can provide services for PC users, they can create components for the PC.

We sort of were late to recognize our own incredible success, in fact, because we were always looking over our shoulder thinking, OK, what is it that we're missing, what are we not doing, so stepping up in government circles to be part of the dialogue about laws of technology and really having the right kind of outreach to various groups, you know, we were perhaps late to recognize that.

I used to brag about how it was a testament to the United States that Microsoft didn't have to have anybody in Washington, D.C., and that was a great thing. Eventually I learned that I wasn't supposed to say that and that at a certain level of success you needed to become part of the dialogue and helping out there.

So I think we were kind of naïve. I think we went through a lot of things that have got us thinking about our unique role as a company. And yet everything we believe in in terms of the role of software, the degree that it's going to get better, hiring smart people, our basic principles are the same today as they were when the company was founded.

A. RICHARD NEWTON: So following up on that a little bit, I mean, there are those who would say that, in fact, obviously you were personally one of the first people to see the real value in software moving forward and what its potential would be. As a constant to that, Microsoft developed a lot of productivity tools and other sorts of things and constant to that perhaps is that we have now two classes of software engineers or scientists. We have designers and we have the technicians, and productivity tools to some extent have enabled the technician. Many of those technicians today, as we know, exist overseas or are moving overseas.

And so what's your view on the whole globalization issue, and then more importantly from obviously our perspective here is what do you think the future is for sort of the Bay Area and Silicon Valley in particular in this new world?

BILL GATES: Well, I think it's been a fantastic thing that a very large part, dramatically larger than before, of the world's population gets to participate in capitalistic opportunity; that is, as their societies are better off, more people go to college, they get to do more of the interesting work and contribute to the world economy.

And it's a little scary to me to see people thinking of this as a zero-sum game. It's not like war where you have one winner and one loser. Is it bad for the United States that China would be rich? Take as an extreme that India and China were as rich as the United States is. Now, on a relative basis in terms of political power and percentage of economic wealth, on a relative basis the U.S. would be worse off, but in any other value frame we'd be way better off just in terms of the living conditions in those countries, their ability to come up with better products, to come up with better services; that's a good thing.

And so we're starting to see we're at the start of a process where the whole world is getting into this virtuous cycle of the higher income, better living standards, all of those things.

And so it's a little scary to me to see some people in the U.S. saying, no, we need to put up walls and make sure that, Oh, they're getting good jobs; no, no, no, there's a finite number of those, let's hold that back.

And I wouldn't say that there is some divide where only the technician jobs necessarily can be done overseas and the design jobs will necessarily stay here. The U.S. will have to compete on the merits and a little bit that means saying, hey, how come we've been so successful; a little bit because we are so successful and it's never been challenged, we've never really had to ask ourselves. Certainly I believe, and it's part of why when I'm outside Microsoft I spend a lot of time coming to universities, I believe that the university system is -- I'd put it as the No. 1 thing that's allowed the U.S. to be at the center of innovation, whether it's computer science or biology. The quality of a few dozen institutions, including in a very strong way this one, create the people that will allow us to lead the way.

And as China has seen that and they're doing a great job making Tsinghua and Beijing Universities better and India is taking the jewel in their crown, the India Institute of Technology and expanding that.

It says, hey, we have to step up the excellence, we have to stay ahead, we have to go into the new areas, the risky new areas, we've got to keep the level of funding, the openness that's allowed students from all over the world to come here and made this an attractive place. That's what's going to allow the U.S. to stay at the forefront.

But there will be broad participation. I know you got a chance to go over and see our research lab in Beijing and that's been just incredible for us. I mean, that is world class research and it's one of those things where if you hire somebody super-smart, then other super-smart people want to work with that person and you get this positive momentum cycle where you get this group of people that are just phenomenal.

But I did just yesterday a review of our graphics advances and we have equal contributions from Cambridge, from our U.S. lab work and from our Beijing lab in terms of video, photos, graphics, all of those things.

And so China and India are the big change agents for the years ahead and as we embrace that and understand our unique role in that, that's the path forward.

A. RICHARD NEWTON: Obviously you have to deal with a lot of controversial matters that come up at Microsoft. Today, what do you think is the most controversial thing that you have to deal with at Microsoft and how are you dealing with it?

BILL GATES: Well, I'd have to think what is the most controversial. (Laughter.) There's controversy everywhere. I remember when we put out Windows 95, because we let you set the time zone we had these maps of the world and we thought, well, this is trivial, you just stick a map in there and show these time zones. Well, we showed Kashmir the wrong way and so we had to recall the product, and we showed the boundary between North Korea and South Korea and some islands off of Japan; people are very sensitive about these maps. (Laughter.)

I remember there was a controversy where we have these funny symbols that I came up with when we first did the PC called dingbat that are smiley faces and frown faces and actually a few of those are religious symbols, Star of David.

Anyway, there was some sequence -- if you want to enter those keys you hold down alt and type in funny things -- there was some sequence like if you typed N, Y, C, it was like thumbs down Star of David or something and people were saying that we had put that secret incantation in. (Laughter.) This was a great scoop that the New York Post came up with, so that was very controversial.

And sometimes we inadvertently get into something. We recently had an e-mail thing that classified mail of certain types where somebody called something a "white list" and, OK, that was wrong but they probably didn't realize, I don't think at the time they did that and, of course, we changed that.

Probably the biggest thing we get into is intellectual property and the whole question of, you know, take China. In China today Microsoft software to the 99 percent level is free in China, and I'm not just talking about the consumers, I'm talking about to companies and government agencies. The compliance with licensing the software is very low. Well, we can't just charge in there and say, "Oh, overnight you've got to change this." We want to make sure that they year by year get better at that. And so striking that right balance about intellectual property, in many venues that's a controversial thing and we have to do it right.

Intellectual property we think is very important. We'd like to be able to do more in China, and so over time we'd like at least the businesses there to buy their software and so we'll push ahead, but sometimes we'll do too much, sometimes we'll do too little.

A. RICHARD NEWTON: Well, that leads me to another one that, in fact, as the chancellor mentioned, this is the 40th anniversary of the Free Speech Movement. And one of the things that Berkeley Computer Science, of course, is very famous for is the Berkeley UNIX distribution BSD. And, of course, that was an open source development, but with a much different licensing model than the open source movement of today, such as the GPL license, for example.

With all of the discussion of open source and Linux, what's your view of the role of open source moving forward and how will Microsoft's role evolve in this open-source world?

BILL GATES: Well, certainly when I was first using computers the idea of getting source code was very fascinating to me, and so I when I was 13 I remember being lowered into this big garbage can where there were all these coffee grounds, going in there to find a printout of the PDP-10 operating system, and so Paul Allen and I got it out of there and tried to clean it off and really tried to understand the operating system we were using and finding bugs so that we could crash the system. (Laughter.)

We don't have this today but back then the only way we got free time was that the company that had this PDP-10 had a deal with Digital Equipment that as long as they could find bugs they didn't have to pay rent on the system, and so they decided having these 13-year olds around finding bugs was a good economic proposition for them. And so for a year this company didn't pay any rent. Eventually they decided we were good enough that even though they had to pay rent they kept us around.

So there are many cases where having the source code out there is a fantastic thing. Microsoft for a lot of its stuff puts the source code out. In some cases we do that just for everybody, in some cases we require a license. Like the source code for Windows, every government in the world can come in and get that, and we have about three dozen who have actually taken it and looked it over. They want to make sure in terms of how the cryptography is done nothing that's hidden or biased in any way in that.

Clearly Berkeley UNIX, the BSD distribution was a fantastic thing. It let a lot of computer science students understand operating systems, tinker around. It was an element that allowed Sun to get going and build its products, that had been a huge contribution.

So a lot of software will have the source code available. There will be these different licensing models and I think this is one thing where researchers, universities, people have to think carefully. We tend to favor the distribution license that was used for BSD, which is a very non-coercive open license that allows you to modify it and make your modifications available, or you can actually modify it and create a version that you build a company around, hire people, pay taxes and there's this virtuous cycle that there's lots of free software that often comes out of the universities, sometimes that just generates more free software, sometimes it generates companies and jobs that then pay taxes and that money goes back to the university to keep this ecosystem going, and that ecosystem that the U.S. has is the envy of the world.

The GPL in our view should be used, which is the license that says you can't enhance it and create a commercial product. Our view is that it should be used very narrowly, and we think people should think twice. So if you have government funded research, it's ironic that then if it goes into that GPL you can't create a company that creates jobs that pays taxes. And so most of the countries outside the U.S. have stayed away from that because they want to get the ecosystem that we have.

So over time in software we'll have free software and commercial software and the equilibrium between them will always shift as people see the support, the indemnification, the certain types of innovation. There is some innovation in terms of taking risk, like building the system that will do machine translation. That will be done because of the scale of the problem and the nature, it will be done, it will come from research but it will be done in the commercial world and then eventually there may be free versions of that.

So it really is an interplay that's working very well. It keeps us on our toes and except for a little bit of overuse of the restrictive license sometimes, I think the direction it's going in is quite good.

A. RICHARD NEWTON: So you've opened up source code for a number of your products over the last few years. I mean, how do you see Microsoft playing in this moving forward? Is Microsoft going to continue to do that or are each of these going to be sort of a decision that you make in some sense?

BILL GATES: Well, the one thing that we will do is we'll have some of our software that we charge money for because we want to pay people's salaries and be able to fund the research and those things. So we come from the commercial side of this equation.

The portion of the code that we actually give out source code, today it's about half of the stuff we do, the source code is available and it's always a tradeoff in terms of what kind of advance do we have in there, what competitors would use that information to close the gap on the things that we've invested in doing. The upper layers of the software tend to be where you give out the source code more.

A lot of customers in a sense don't want -- the notion that they would go in and tinker with the source code, that's the opposite of what's supposed to go on. We're supposed to give that to them and it's our problem to make sure that it works perfectly and they rely on us.

And so I'd say a very high percentage of the times we give out the source code, it's more of a security blanket or even a source of information than it is somebody who's editing it and recompiling it. But just being able to have it, there's a lot of interest in that and we do more and more of it.

A. RICHARD NEWTON: So if you think about all of the innovations and developments that Microsoft has been involved with other the last five years, which innovation are you the most proud of?

BILL GATES: Well, at the top of the list has got to be the whole model of the PC, the idea that computing was moving from being a tool of organizations where it was very daunting and people would say things like, take that punch card that comes in the mail and put a staple in it and see if you can bring down the big computer that has all this information. Today we think of computing as something that lets us browse the world of information, lets us be creative with the software tools that are there, so that bringing about the personal computer revolution and the business model of a very large software industry, many hardware manufacturers, not just one company but any number of companies able to make the hardware, and because we add this level of abstraction, the Microsoft operating system, any of those applications could run on any of the hardware; that's our biggest contribution. I mean, there basically was no software industry before that came along. Today there's this worldwide thriving software industry, highly competitive, very oriented towards low price, high volume.

It will be hard for us to equal that contribution. I'll feel very good when we get, say, the Tablet computer to critical mass. That's a pet project of mine with the ink and the ink recognition and those things. I love that particular project.

It will be a milestone for us when our Xbox videogame is the best-selling videogame, and it may not be good for productivity but it's fun. (Laughter.) And we're putting such neat technology into that, we're working on the next generation now of the chips, and this is sort of the high definition round of entertainment. And we want to bring in people in a much more social way so this idea of being online talking to people, having contests and spectators, it's not just one person sitting there shooting at artifacts, it's a variety of new game genres there.

So I feel super good about the general role that Microsoft research has played. Our willingness to take risks and certainly by taking risks we've hit a lot of dead ends. We got involved in TV-related software way too early and we spent $500 or $600 million before its time. Now actually this idea of delivering video streams over the IT network, so that you can individualize every video stream, have complete control over it, its time has finally come, so we're going to get some return, but we could have waited a lot longer to get involved in that.

So that just the general notion of creating a lot of partner companies, spreading the idea of developing software around the globe, I think those are our biggest contributions.

A. RICHARD NEWTON: One of the obviously important ones for us is four years ago, as the chancellor mentioned, we started this institute, the CITRIS Institute, the Center for Information Technology Research in the Interest of Society, which was pretty revolutionary at the time, and Microsoft was one of the handful of companies that supported us in that, and I think the collaboration that we've had between Microsoft researchers and our Berkeley faculty over the last few years we've been solving some of the really toughest problems.

And it's not just Berkeley, we have UC Santa Cruz, we have UC Davis, we have UC Merced involved, so it's a real critical mass activity.

But that said, I mean, you alluded earlier to the fact that universities are playing a critical role in terms of the ecosystem that you see. How do you see that role evolving moving forward? What's going to happen in terms of the relationship between big companies, small companies, industrial research labs and universities? I mean, clearly in the Bay Area we have a special example of that, but what's your view of the future?

BILL GATES: Well, I think the strength of the research university system we have here in the United States, we really have to rededicate ourselves to that, whether it's the funding, governments or alumni giving back to that, and as we look at this challenge of does the U.S. maintain its leadership role, I'd put that absolutely at the top of the list.

I remember in the 1980s, where Japan took a more monolithic approach to advances in artificial intelligence, they picked for the so-called fifth generation a very centralized approach and they picked prologue and certain things and they were just duplicating that throughout their research environment, whereas the U.S., we had no centralized notion of, OK, which approach would make sense, every university was off doing whatever it wanted.

There was a lot of fear that the Japanese model was better. Well, in fact, the U.S. model was better and it was during the '80s when we had all the humility and concern about Japan that things like the Internet were being invented, coming out of the university world and then being refined in the commercial world and now this wonderful foundation that we've got to build on.

And so I don't think we should tinker with it much. There's always this question of how you get companies and universities to work together at some level, but then again you don't want to eliminate the unique thing that universities can try out wild things and companies obviously have a somewhat more focused agenda on taking things into the market. We're lucky that with our level of success with Microsoft Research there's a real continuum between the kinds of things we do and what goes on at the university.

It's very disappointing for me that spending on corporate research in this country is down. That's a scary thing. If you think of the paragons of corporate research of 20 years ago, it was Bell Labs, it was Xerox. The fact that they weren't able to translate their research into product success set a very bad precedent that kind of has people thinking, geeze, how much should you really put into that.

We're trying to be a strong counter example, where we can say the best investments we've ever made are in our research group and that there really are methodologies for making sure that we track neat things in the universities, we partner with the universities; we then get things into Microsoft Research and then we get them into our products and that there's no walls in those different compartments and different activities.

And I think we do have the proof in hand that our success certainly over the last five or six years comes from things where our collaboration with the universities and Microsoft Research made all the difference. Even if you take a case where we're in a sense coming from behind and catching up, take the search area that's very hot right now, it's because of the work in Microsoft Research that we're able to say, OK, let's not just come up to the level of what's out there, what Google and others have done, let's move to a new level of document understanding and personalization. And so it's really fun to take those things and say, OK, let's go get them.

A. RICHARD NEWTON: Well, talking about fun, I think it's time for us to open the mikes up for some questions from the floor here. Please keep your questions short and to the point and no speeches, please. I'd like to get as many questions and as many opportunities for people as I possibly can in the time, so we'll start here.

QUESTION: Mr. Gates, what I'm wondering is, you know, we've seen you do wonderful philanthropy throughout the world, foundations in Third World countries, et cetera, and what I'm wondering is how come you've never left the software industry, and how come we haven't seen you do business ventures in other fields such as energy or healthcare that may benefit people in a way which you would often make a profit?

BILL GATES: Well, when you do philanthropy, it's a very interesting question of what do you pick, because the needs in the world are so broad and even with the scale of dollars that I'm lucky enough to have where my wealth is going back to society, it's still very necessary to pick because the scale of the problems and in comparison to government budgets it's not that much.

So the key thing for me was saying, where is there a market failure? For example, in medicine I don't do philanthropy for rich-world diseases. That's not to say I don't think that heart disease or cancer or Parkinson's, whatever diseases that are a real problem in this country, there shouldn't be research on that. But I feel like between the government dollars here that go for that and the pharmaceutical research dollars, which is about 60 billion a year, they are addressing that problem and so the market is working and that my contribution to that wouldn't be that significant.

Whereas if you take developing country diseases, take malaria, when I gave my first $50 million to malaria, people said, wow, that doubled the amount of private money going to malaria. Well, this is a disease that kills a million people a year. This is a disease that right this minute 200 million people are suffering from. It's way worse than any rich-world disease and yet .1 percent of medical R&D goes against malaria. And, in fact, the biological techniques we have, making vaccines, making drugs, this is a very probable thing. In fact, there's a collaboration here at this university to use some amazing synthetic chemistry approaches to actually make malaria medicines.

So I pick things that I thought weren't being done, and I was stunned to find that most of the research money in medicine, 95 percent goes on 5 percent of the disease burden, and 5 percent of the money goes on 95 percent of the burden, so I tried to pick something that just wasn't being addressed, so that's gone very well in terms of getting a critical mass of smart people to work on those things.

So the Foundation picked really two things, what I think is the biggest challenge in this country, which is education, scholarships, those things, and what's the biggest problem globally, which is the inequity in terms of living conditions, particularly health.

QUESTION: I've been living in Hong Kong for the last 10 years and one of our great problems is that we are running out of landfill space. And so my question is that with each new generation of computers and, of course, even software systems, that a lot of waste is generated because people tend to discard these. Even though there are some recycling programs, a large number of these computers are simply discarded.

Now, I was curious if you see any solutions to the problem, because, of course, we're still sort of at the infancy of the computer revolution and, in fact, when we look into mankind and you projected another hundred years or a couple hundred years, the problem will simply get worse, and I wondered if you see if there is any solution to the problem and is there any way that Microsoft might be willing to tackle it?

BILL GATES: Well, the right adoption of computer technology will change the way resources are used. I was at a CEO event a few months ago and I was saying that even though the idea of the paperless office has been kind of ridiculed, I still believe in this idea of the paperless office, that is I believe that eventually you'll be reading off of the Tablet PC, whether it's your daily newspaper or your textbooks, that we won't be using paper for those things. And I'd forgotten there was a CEO from the paper industry in the audience, and so he spoke up and didn't like that, and then somebody from HP said, "Hey, but then we won't need printers, so that can't be right." (Laughter.) And I thought, hey, you just can't win.

So there are elements of this information technology society that are beneficial; that is, use of paper should go down quite a bit. We promote a new piece of software we have called Live Meeting literally for the fact that you don't need to travel as much. We're not saying you don't have to go face-to-face at all, but a lot of those trips are unnecessary as we bring in rich collaboration right there to the screen and there's a lot more software can do in terms of the different video viewing and how we record those things and make that easy to do.

So I think we can cut down on the amount of travel that's necessary. I also think computer technology, the intelligent computer technology really is pro-environmental. I mean, people are talking about now with gas engines that the way you can do injection in a different way, if you have enough computer control you can get almost double the efficiency out of even a gas engine, which is a great thing. So overall if you have more information the way you do irrigation is just more targeted and smarter, the way you use fertilizer is targeted and smarter.

In terms of your specific point about recycling computers, fortunately computers are getting smaller and smaller; that is, if you look at the disk drives what they were like when I grew up and what a printer was like, those things are getting smaller and they just use less physical resources and I think that trend is very much in our favor. We're moving from, say, a CRT that uses a lot of glass and is very hard to dispose of to the LCD and a lot of the newer screen technologies, even screens eventually that you'll just fold up use very, very few materials.

We do a lot on recycling computers to help people get those into nonprofits and make sure the software -- we provide free software for the nonprofit uses in those things.

So I think there are best practices. There's the notion of the green PC in terms of how it turns on and off.

One nice thing is that as we're trying to preserve battery life in portable machines, we've been able to take those energy-saving technologies and even use them on desktop machines, because we thought, hey, the power is there, don't worry about it, but, in fact, computers are measurable part of the power load and so by taking those things we've done in portables and putting them into the desktop we can get like a factor of three improvement there.

So overall I'd say we're a positive factor and the awareness is higher now than it's been of specific things we need to do.

A. RICHARD NEWTON: And by the way, I should add that I think that all of the things you mentioned, particularly the live notes activity that Professor John (?) here has been undertaking on the Tablet and our XP use, the Microsoft conferencing technology and then the organic semiconductor work we're doing, collaborating with Microsoft, is all in that direction, so that's been fantastic.

We'll take a question here.

QUESTION: You mentioned earlier, Mr. Gates, that universities are a really important source of great innovators, and so I expect that you consider being able to recruit from that source of great innovators pretty important to you. (Laughter.) So I'd like to find out if I could have a show of hands in the audience, I'd like to find out how many people in this audience might have concerns about working for a company that's been found guilty of illegal business practices, that limits the choice that its customers have to choose a product they want to use and the type of media they want to watch, and that has also been found guilty by the Federal Trade Commission of misleading the public?

A. RICHARD NEWTON: I think we get your point.

QUESTION: Could I have a show of hands from the audience, please?

A. RICHARD NEWTON: I think we get the point. Thank you. Thank you.

BILL GATES: But what's the question?

QUESTION: Do you think you might do better at recruiting students from universities if you improve the business practices of your company?

BILL GATES: Sure. (Laughter, applause.)

QUESTION: I've had a lot of experience with doing nonprofit work and I was wondering how your work in the nonprofit world has given you any perspective or insight into your corporate life?

A. RICHARD NEWTON: Good question.

BILL GATES: Well, I hope and I think that my philanthropic work has made me smarter in my business work. I can say it's probably even stronger the other way around; that is, I spent most of my career doing my job at Microsoft and I even used to say that I would wait until I was in my sixties and basically retired from my Microsoft job before I did any philanthropy because I wanted to take time to figure out what things to do.

And it was about -- well, about 12 years ago now when I was reading about population growth and diseases and I started to get into putting computers in libraries that I said, no, I've got to break -- I'm not going to be able to have this partitioned life where I do make money for these years and then I do this giving away money thing later. I thought it would be kind of schizophrenic that you have one day where you go into work in the morning and you're saying let's make money and then in the afternoon you're saying no, no, no, let's give it away. (Laughter.) I thought my discipline might get messed up or something, but I don't think that's been a problem.

The one thing that's great on the nonprofit side is you feel you're part of something bigger than yourself, you've got this sense of excitement there that everybody is involved with that you're part of. And I think even in the commercial world, that's very, very important when people think about their job, it's not just an economic bargain. If you're going to draw the best out of those people, the notion of -- really, take at Microsoft, we go and show the employees how the hundreds of millions of dollars worth of software we've given to libraries are being used; we take, for example our accessibility work, where handicapped people, because we do special things in the software, can use the computers in a better way, we get then to meet some of those people and understand, OK, what are the unmet needs where we can do that better.

And so I think having a sense of cause, a sense of purpose about why work is important, that's something where you see it best in the nonprofit world because they couldn't survive without it, but you learn how to do that even better in the business world.

A. RICHARD NEWTON: George Soros was here a few months ago and he made a comment that I thought was interesting, that he found it harder, in fact, the work that he had to do to give money away, the thinking he had to do was actually harder than in many ways what he had to do to make money. Do you sort of see it that way?

BILL GATES: Yeah, I think that's fair because at least when you're making money, you try something like Microsoft will create a product and people either buy it or they don't buy it; you get very quick feedback and even if it sells you can call up those people and say, hey, what more do you want and because people are very involved with software, they're using it all day long, they're glad to tell us what's missing, what should be simpler, that kind of thing. And so we have this constant feedback loop, we really know are we doing something that resonates and where do we need to go. A few times we need to take a leap like the Tablet PC or vision or something where they don't know to ask for it, but a lot of this is just listening to the customer.

In philanthropy, of course, all your grantees are going to say to you, good job, you gave money to the right person. (Laughter.) So the thing where you survey your grantees, I mean, is that really going to give you any meaningful data? What about the people you didn't give money to?

And knowing what the impact is, some of these things like I have a minority scholarship thing, knowing what is the benefit of that, is it just the thousands of kids who get it directly or is it there some indirect thing in terms of they're role model, they feel an obligation to help other kids come along and so you get this multiplier effect that makes a big difference, I'm not sure I'll ever know that. I mean, we'll do some studies, we'll interview people but it's not as hard core as saying people bought millions of copies of Windows 95. It's just tougher to know. And certainly in the frontiers that Soros has gotten involved in, in philanthropy, it's very, very hard. Did his Open Society get some various countries help in that fund? I think so, but he's not sure.

QUESTION: I have a question about Windows and future Windows. I've been doing some IT consulting on the side and I've seen numerous problems that Windows has. I've also seen that some advanced users are migrating to like open source systems. And I'm just wondering what does Microsoft have to offer for the future of Windows and for advanced users who like right now are migrating out to open source systems, and Windows and Microsoft in general?

And I have a small question from an online community of users. In 2000 you released a product called Microsoft Train Simulator and there has been a second version in the works but it died. Are there any plans to release that as an open source?

A. RICHARD NEWTON: What was the program again?

QUESTION: Microsoft Train Simulator. (Laughter.)

BILL GATES: OK.

QUESTION: And just one last thing.

A. RICHARD NEWTON: I think we've got two, that will do.

QUESTION: I just want to thank you for being here today.

BILL GATES: Super. (Applause.)

Well, I always try to be prepared for whatever questions might come up but I have to admit I have not met with the Train Simulator group, but if you can get me your card I'll let you know what new versions of that we're going to do. I know it was a cool product, I remember when it was released, but I have to say I don't even know how well it sold.

In terms of the competition in operating systems, what there has been, actually, Windows' share of the market has gone up every year. There has been a huge shift within the UNIX space, that is where the Solaris UNIX or HP-UX, which is the HP UNIX or AIX, which is the IBM, they've been declining very rapidly and Linux has been gaining. So its share of UNIX today is over 50 percent and that's going to continue. And then Windows' share of servers and desktops has continued to go up as well. And so what we're going to see is over time very intense competition between the variants of Linux and Windows.

We have a lot of R&D we've put into Windows, several billion a year. We're working on a version called "Longhorn" right now. And we've got some ambitious things we're doing. Today in an operating system you have to learn a lot of commands. You learn different commands for e-mail than files, the way you navigate photos is different than the way that you navigate music and what we'd like to do is get back to this thing where you have one store where all the information is there, your address book, your photos and it's a very rich store and so you can learn a small set of commands and so the way you navigate the address book, photos, music, mail, files, those are all the same and we think we can bring a simple conceptual understanding of where's my stuff, how do I move it between machines, how do I secure it by doing this. So that's one example of something that I think if we can solve that it will be a huge contribution and it will get a lot of Windows users to want to move up and do something new.

And every day we're hearing how we need to make Windows better. So far so good, and we'll keep doing our best.

QUESTION: When you spoke at MIT a student challenged you by saying that there would never be another entrepreneur as successful as you, and you told the audience that if they're able to make some mechanism through which machines can learn artificial intelligence that would be, quote, "worth 10 Microsofts."

So as you said, artificial intelligence seems to be a major research thrust in computer science. Can you to through the other areas of engineering and discuss what you think will be the major research areas that have both academic and industrial importance?

BILL GATES: Yeah, I can touch on a few, and maybe for a comprehensive view Microsoft Research has a very good Web site going through the different things they're doing.

The range of activities is quite broad. I think in graphics this is a very exciting time, the idea of automatically analyzing videos, doing things with photos. Some of it is very mundane, like we had a paper at SIGGRAPH where if you take a picture of your family and say you take like three or four and you get none of them where everybody is smiling, now we have a piece of software, you just point to who looks good in each photo and we'll automatically make a photo where everybody looks good. (Laughter.)

Or they do this incredible thing where if you use a flash it messes up all the colors because it floods everything, and so what we do is we put into the camera something that takes a photo without the flash and takes a photo with the flash and so by combining those two together you can actually get a photo that's way better than either of them would have been.

So anyway the whole area of graphics, recognizing people in photos, motion analysis in videos, the graphics field is very, very hot today. SIGGRAPH is as neat as it's ever been.

The area of document understanding, what's the future of search? The future is that you really understand what the site is about. I know today, although people say search is really good, it's really not very good. People are just comparing to what it was before. When I type in something like "malaria," I certainly don't get back the most interesting new things; I get a bunch of random old stuff, even things that aren't even in English.

And so we need to parse documents and this has been a Holy Grail, it's been expressed, can you translate the document, can you answer a question about the document. That's an area where we're going to have some big breakthroughs and we need linguistic people, we need Bayesian inference systems, lots of rich techniques brought to that. So I'd say that's a big area.

I'll just mention a third one, which is reliable software. Today's software is written at a very low level using pointers and the amount of problems we get into by expressing software at those low levels includes things like buffer overruns or exploits that people can do. We really need to be expressing this software in a much higher-level form, richer type system and even contracts that guarantee the behavior between different parts of the system so as you change one part of the system you can understand is that breaking the contract and would that have any effect on software that's further away.

The level we express software at really hasn't changed over the last 30 years. I mean, people wrote FORTRAN programs in 1960. The things we do today, yeah, they're a little better, C#, Java, object-oriented stuff, but not a lot. When somebody wants to change a business process they can go to the board and draw out the change and describe it in 30 minutes but then it might take 10,000 lines of code to describe it; that's a mismatch. Why can it be expressed in some weird human form in 30 minutes and yet be so complex in its algorithmic expression? So we have to move up to have modeling-based constructs and provable software. And this is an area where the last few years, there's really started to be some progress.

So I'd pick those three as areas that anybody can make a breakthrough on those; that would be a huge, huge contribution.

A. RICHARD NEWTON: And in credit to Microsoft, the work that's been done at Microsoft Research in this correct software has been extraordinary over the last few years, the work on drivers and so on has been real positive.

This is I'm afraid going to have to be the last question, so you're the lucky last questioner here.

QUESTION: I just wanted to ask you what do you think is the best way to approach and hopefully solve the problem of getting technology to permeate into the poorer areas of the world, because putting a computer in front of everybody is not going to solve that problem?

BILL GATES: Yeah, that's one of the ironies of working in a field where you feel so good about the tools you're creating, and going and seeing young kids using these tools and navigating the Internet and thinking about, hey, when I was growing up I had to read the World Book in alphabetical order to try and get at information. Now it's so much easier, so much better the wealth of information you can go out and get. So the tool is so good you want everybody to have it.

And even here in the United States we fall very far short of that ideal. There are some things that have been done in this country that are very good. Putting computers into every library in the country, that was a project Microsoft and my Foundation did that's been super successful, so any kid who can reach a library can get in, use the computer and we know they're being used heavily and people are worried that they wouldn't use the books as much but actually they go in and use them more. And people were worried they go in and just browse bad things but that hasn't happened either, so that's a good thing.

Now, it's even tougher, which is really your question, as you get outside the United States. Why is it tougher? Well, the hardware is getting less and less expensive; still, it's not super cheap so that everybody can go and buy one. The software for these types of projects we're able to give grants and make the software available, because our marginal cost of making the software is basically zero, and so the software is not a limiting factor either.

The limiting factor tends to be No. 1 by far, the cost of communications, because you don't just want to give somebody a computer, you want to give them a computer connected to the Internet. Even though you want rich software there, it's that software interacting with that information.

So how do you solve this? You solve it by having community computers, and we have a lot of partners in India that have gone out and are doing that. You solve it by using wireless approaches so that you don't actually have to build the expensive infrastructure that goes out there. And one of our big projects at Microsoft Research is a mesh-networking project, and I think the Web site talks about that quite a bit.

We ought to be able to over the next five years or so connect all the villages say in India using variants of Wi-Fi and some very rich software and fairly inexpensive base stations; it will take some money, but between what people like Reliance are going to do running fiber in the country and this ability to go from that fiber through wireless out, we ought to be able to provide that connectivity.

At some point there will be mid-Earth orbit satellites that may provide that connectivity. Unfortunately, the dream of that didn't materialize in the timeframe that some people hoped, including I was an investor in a thing called Teledesics that didn't work. But hopefully that dream will be revived in the 10 to 20-year timeframe.

So the connectivity piece is the No. 1 hard piece and then the No. 2 hard piece is you've got to get the teachers and the curriculum and get enough there so that the young kids can get going and feel adept at that.

And so this cause is very important to Microsoft. I would say that it fits into an overall picture where you also want to get clean water and healthcare out to those villages as well. It would be ironic to get the computer there when people are still dying of diseases that they should die from, so the two things can go hand in hand, it doesn't have to be one or the other. But I do see software for wireless networks and things, I do see a huge contribution that Microsoft and the software industry can make to this.

A. RICHARD NEWTON: Well, I'm afraid we're out of time. Bill, I would really like to thank you very much on behalf of UC Berkeley and the College of Engineering but also on behalf of the whole UC system for everything that Microsoft and you personally have done for education, higher education and especially what you and Melinda have been doing for education in the high schools, that pipeline issue is a very important one for us. So thank you very much for being here and why don't we all thank Bill?

BILL GATES: Thank you. That was fun. That was great. (Applause.)