Remarks by Bill Gates, Chairman and Chief Software Architect, Microsoft Corporation
Columbia University
New York, NY
October 13, 2005

BILL GATES: Thank you. Well, I'm excited to be here at Columbia and have a chance to talk about the future of software and how software is going to change the world. Software has done a lot in the last 30 years, but that's nothing compared to what's coming. It was 30 years ago that I was a college undergraduate, and I decided to drop out. I'm not here to recommend that -- (laughter) -- but I thought that by being in early in the whole idea of the personal computer and software for that, that Microsoft could do some exciting things.

The vision at that time was not one that was widely shared. Even people at Intel who were actually making the miracle chips that allowed us to have this dream really didn't see that the nature of computing would change, that it would change from devices that were very, very small in volume, literally just thousands of computers on the planet, to something where today we have over a billion personal computers in use.

And there's something very important about that volume, because in order for a computer to be useful, you've got to have software, software is what makes it come alive. And often you have to have very specialized software. And to make that economic so that you can invest tens of millions of dollars in building that software and yet sell it for only a few hundred dollars, you've got to be able to have millions and millions of machines.

And so we created a cycle that led to the creation of the software industry. The more machines we had, the more opportunity there was for good software; the more software there was, the more attractive those machines were. As the volume went up, the price of the machines went down, and that further led to that volume cycle.

So today we have a very global software industry, it's doing lots of amazing things, but it's the hardware and platform advances of this next decade that will make it even more revolutionary, in fact, far faster than I think anyone anticipates today.

The machine that I had a chance to work on 30 years ago was called the Altair, and the feat, the first feat that Microsoft performed was creating a BASIC interpreter that ran in 3,100 bytes. Thirty-one hundred bytes is not a lot of memory. That was because in the 4k byte machine you needed to leave room for the person's programs and their data as well within that 4k allocation.

Today, of course, it's very common to have a machine with a gig of memory and in a few years we'll have 4 gigabytes of memory will be very, very typical. So what that is, that's a factor of a million change over this period of time, a million times as much memory.

And that's not the only factor that's improved by a million: the size of the disk is more than a million times bigger, the bandwidth of sending information between computers is more than a million times better; the nature of the screens we're projecting on, their resolution, their graphic performance is more than a million times better.

If we look out into the future, this kind of exponential improvement that was predicted by Gordon Moore in what's sometimes called Moore's Law, that's set to continue for all of this period. And so we're not going down, we're not at a period of diminishing returns, and once again the thing that's going to make all the difference in terms of whether we can deliver that power is what the software can do.

And the software solves the very tough problems of making these systems secure, making it so you understand as you put your information there how that will be used so your privacy is protected, make it so that the information is easily available and these systems can all talk to each other.

Those kinds of challenges require a lot of the best software people, and, in fact, a lot of the competition in our business is a competition for talent, making sure that the best minds are going into the software industry and then attracting them, having them in project groups that work together, get lots of feedback from customers, take the research ideas, pour those in and then constantly bring out improvements that drive that software forward.

Now, everybody in the company gets involved in making sure we're recruiting in all the best people, I do that some myself. In fact, I made a video that captures a little bit of one of those experiences, and I thought I'd share that with you.

(Napoleon Dynamite Video Segment.)

(Applause.)

BILL GATES: We had a lot of fun making that. It makes more sense if you've seen Napoleon Dynamite than if you haven't. (Laughter.)

Well, let me talk about what these devices will look like and how they'll be changed by software. Every computer will be connected up to a single network and every activity, whether it's TV watching or telephony, will likewise be connected to the logical Internet, which will be made up of both wired and wireless pieces.

Really the only difference between computers will be the size of their screen. You'll have a wristwatch that will be like Dick Tracy's that will just give you glanceable information. You'll have your pocket-sized device, which will be the successor to the phone but far more capable. These phones, of course, are improving very rapidly. The phone has long moved away from just being a pure voice device to being a device that you do text messages and get your mail on.

This is a Motorola Razor product that's got the nice little keyboard on it, nice and thin, of course has this camera built-in. And this is another one coming out in about the same time, this is a Treo design where we've worked together with Palm to get the latest software on it.

As you have this camera capability and the connections to the network, software can do a lot. For example, if you're in a restaurant and you get an expense receipt, you simply take and click a photo of that and the software will recognize that, understand based on your schedule what you're doing and fill out the expense form and make sure that's automatically done without you having to do any work.

Also if you're shopping for something, you take a picture of a product or a barcode, that can be recognized and your phone can tell you where you might get a better price or if you should buy a different product, all in a very automatic way. Even something like running into a sign in a foreign language when you're traveling, take that photo, the video image will get sent up to a server and then that whole power of the server can be used to do that language translation and bring that down and display that for you on the phone.

Likewise, you'll have location information, the ability to tell you about traffic and create little live 3D interfaces that show you the direction of where you should go, what's going on. That's something that you'll simply take for granted.

So that phone will replace everything you think about in your wallet -- your photos, your cash, your credentials, all put together on that device.

Moving up in size we'll have the tablet. The tablet will be a device that you can have phone-like functionality, actually a superset of everything I described there, but also something where you can read large documents, where you can annotate, where you can author them. You'll sometimes have a keyboard connection, you'll sometime just use the pen, sometimes you'll use the voice and pen together. Speech recognition is an area where we're making quite rapid progress. Twenty years ago, people were a bit naïve about how hard this would be because we underestimated human intelligence, how much we use the context to disambiguate what at the phoneme level might be the choice. You because you know the topic and you know what makes sense, you're able to resolve that. And so we're having to have computers have more of a model of what's going on in the sentences around it to do that disambiguation. But every year the error rates come down and so we can say over the next five years those get into the range where they'll be very practical, so talking to your phone, talking to your tablet will be all very standard stuff.

The desktop machines will literally be the entire desk because we'll project onto the desk surface, we'll use cameras to see whatever gestures or things you're doing there, and so you'll have many display areas that will all combine together to create one logical experience.

In fact, if we think of the home environment it's even more immersive. There will be different cameras, microphones, displaying onto walls, countertops, and so there will be a separation of the displays, the inputs, the storage, the computation, but to you you're simply saying what information you'd be interested in and it's being brought up based on whatever your location is.

So computing won't look the same as it does today. There will be devices that understand your interests and your information will move between those devices without you having to get involved. We could say today it's very device centric where you know where your files are, you know where your mail is and it's up to you to connect those things.

The future will be user centric, so, for example if somebody wants to get in touch with you, the rules about what your schedule is, how important that person is to you and what you're doing, those things will be taken into account by a software agent and then if it's appropriate to interrupt you, whatever device you're on will be used.

So we'll move away completely from the idea of multiple e-mail addresses, different phone numbers and those things to a single ID like your mail address, and that will go through the logic that makes sure things are working on your behalf.

We'll have very large sized displays. We'll certainly move away from paper-based activities almost across the board.

Today already for the encyclopedia we can say that the digital form is far superior. When I grew up I had a World Book. I started at A, then B, then C, then D. It's not a great way to read because you go back and forth in time, you go from one subject to another. And today with either a CD-based encyclopedia like Encarta or an online encyclopedia like Wikipedia, the information can be navigated in a much better way. You can get video, you can interact and so it's far, far superior.

In fact, any student, no matter where they are in the world, that has access to the Internet, has tools to pursue their curiosity dramatically better than the most privileged student had only, say, ten years ago, so that kind of empowerment is what really makes this an industry that changes things and makes it very fulfilling to be involved in these activities.

Let me quickly show you some glimpses of these directions. This is a little piece of software called (Max ?) that I've got. This just gives you a sense of photos and visualizations. Here I can start out looking at my different photos, I just want to browse all of these so I can size them any way that I want, look at these different things. I can click and put things in a list, take that and drag something else over here. And as I do this, I've got a set of photos. Then as soon as I've got that, I can think, well, how do I want this displayed. And, of course, you've got this great choice of a slideshow but you might also want something like an album type view with the multiple photos that you can move around and organize those in a nice way, or you might want something that's more like you'd put up on your mantle, a little bit of 3D look that then cycles through and anytime you see something of a particular interest you get to zoom in on that.

You're going to have over the course of your lifetime a lot of photos and so this ability to see a slice that's relevant to something, a particular friend, a particular place or memory, and then have it without much work on your part show up in a very rich way on different devices, different screen types, that's something that will be very important.

Let's move on and take a look at what's going to happen in the living room. The devices in the living room have been very complex. Even without a PC there, you've got all these different remote controls, you sort of don't get a visualization so you don't know what the state of the device is, each device has its own command set.

As soon as you get a powerful device that can put up a nice screen interface, then everything can connect through that, and so you can have simple guidance, a single universal control.

Here if I call this up and activate the controller, this is the Xbox user interface, and so I have different screens I look at to see the settings of my system, my media, the games that I'm signed up for, my social environment on Xbox Live.

Let's just go to that media and say I want to go and connect up a portable media device, a music player. So I go in here. Actually I need to plug this in, so I've got a little USB cable, just take this and plug that into the slot, and then we pretty quickly recognize it. Now, this works with all sorts of different media players, even iPods it will work very well with. (Laughter.) And so I didn't have to use one today, so that's an iRiver that connects up. We can go over and see what music is on it and it just enumerates the directory there. I can take any of these songs and here I'll just say I want to play them in some order.

Likewise, getting at your photos is very easy. So say I have a camera that I've taken some pictures with, this was a little Canon camera, but this would work with anything; again, I just take the cable and automatically connect that up and now all these photos are available to me. All I have to do is go back in here and where I had my media devices listed I have pictures as well. So I can go down to the pictures screen and then it's recognized I've got that Canon camera out there, and I can just see the photos, go ahead and play all of those.

So connecting up so that the photos you took recently are up on that living room screen or you can transfer those onto a PC, organize them in a nice way, that all becomes very straightforward because you have this visual interface and a digital device that's very, very powerful. In fact, you even get integration, I can take these songs that I like and listen to those while I'm playing videogames.

Let's switch over to this other Xbox, which actually has a new videogame on it. Xbox, this 360 ships next month, and that will be part of what you might think of as the high definition generation. Sony will have an offering sometime next year called PlayStation 3. These will be hotly competitive and dramatically better than the previous generation. You get a little bit of sense of that looking at this game here. It's a racing game called Project Gotham. I'm not too good at it, so I'll pick this easy level here and go in. And what it's going to do is it's going to load it. The disk, of course, the game ships on a DVD and so it's loading the car information here off of that DVD. I think I'm set up to race here hopefully. (Videogame sound effects.)

All right. (Laughter.)

So the level of detail is quite phenomenal, modeling the car, any dents you make on the car remembered realistically, you have different cities.

We also bring, because of this broadband connection, a sense of social interactivity. You can connect up with your friends, you can play with strangers. In fact, we put some research algorithms in to make sure that when you match up, you match up with people of the same ability. There's nothing worse than getting on and getting killed, if it's Halo, ten seconds after you get into the fight, which is what happens to me mostly. So here we have a ranking algorithm and only with just watching you play a little bit understands exactly what type of players you ought to be able to match up with.

We also bring some things that are new, the idea of contests, the idea of being a spectator, the idea of having this history of what you're good at, it will be prestigious to show your level in the different games.

We also bring in the ability to download small games that we call Arcade type games over the broadband, so you don't have to go out and buy a disk at all, and these are games that are casual, you can sit down and use them in a few minutes. It will include all of the famous old games that I grew up on or were popular in the past, and lots of new simple little games so that even different age groups will find it very attractive to get going on the Xbox in a very quick way.

So that's the near term, which shows you a pretty fast trajectory and improvement. The one last demo I want to do is a little bit further out in the future, but not much, something that's, say, two or three years away, and that's this idea of the pervasive computing. Say that I'm in an airport and I didn't bring my tablet computer with me -- that would be quite rare -- but I've just got my phone, and yet I'd like to do some work.

So I take this phone and put it down on this table. Well, this table has a camera, a little infrared light, and a projector, and so the software here recognizes that this is my phone and makes sure that I didn't steal this or something, it forces me to use my fingerprint to get on here. But now I'm going to have this whole display table in order to do whatever I want.

And when I was on the plane somebody gave me a card, I can go ahead and put that down here because I want to remember who that was, I want to follow up with them. I even scribbled a little note on the back so I'll flip it over. The camera sees that, does the recognition on that.

And now I want to show it that I'd like to put this into my contact list that's on this phone and you see it recognize that, the little blue dot means that it's being transferred there. Of course, that will be replicated up onto my PC so that's now just available to me. So I can take that off, I don't need to keep that card anymore.

And now I've got a large screen where I can do ink writing or edit things or read long memos. Here it's sending me a press release that my colleagues want me to approve. I can read it, change that as I want, and as soon as I say that that can be released, again I'll authenticate to make absolutely sure that I approve that, and that gets sent off.

And so here I am doing very rich computing using the processor and capability in this phone, but doing it on this table environment.

As soon as I pick the phone up, it will recognize that, and none of my information has been left here. I'm logged off and I'm off carrying with me the information that I care about.

And so that starts to give you a sense of computing everywhere. Because we can have very inexpensive cameras, we can have very inexpensive displays, and, of course, all those other elements getting cheap mean that software can change so many different activities.

If we think about this in terms of the home environment, what we call digital lifestyle, we're already seeing it in terms of moving from film to moving to digital photography. We're seeing it with music where there won't be a future physical format after the CD. The CD is the last popular music format, it will all be purely digital, either with a portable device with a hard disk or devices that are always connected to the Internet, so they simply stream the information down to you.

With video we're about five or six years behind that. We will have one more physical format, which is this battle between the HD format and the Blu-ray format, and then thereafter you'll have the simplicity of getting your video simply across the Internet or off a hard disk. In fact, Microsoft is working with a lot of phone companies to take TV and deliver it over the Internet. And this completely gets rid of the idea of watching it at a particular time. It's not like TiVo where you have to record it in advance, this is always there whenever you want it without any working done earlier on. The ads are targeted to you based on your profile, so they're perhaps less annoying. Even the shows are personalized, so when you sit down to watch the news, if there's certain places you like to know the weather information for that will be there, certain sports you care a lot about that will be covered in great depth, international news stories that you've shown an interest in, and skip over the things that you don't care about. And so the TV experience will become an interactive experience.

In fact, the differences between a set-top box and a PC and a videogame over time will largely go away; it's a big screen connected up to the Internet with a common computing fabric.

The digital lifestyle is moving very quickly but one thing I think people underestimate even more is digital work style. Today, people really don't have access to the information that counts in their company, knowing what customers are thinking, being able to really track numerically what's going on with quality, dive down into sales data, and to have that information at their fingertips; the ability to have software systems that they can see the business processes and see what the state of those things are, see what's taking time, what's slow, see what's profitable, and just in a very visual way change those business processes without going in and changing the software in a very complex way.

Likewise, in the IT department a lot of the activities of monitoring systems and setting them up, these are things that can be done automatically in software. And so software can make the IT environment a lot more effective than it's ever been.

So digital work style and digital lifestyle are proceeding at full pace. This is not a case where we're saying overnight, like during the Internet bubble, that these things will change, but year by year as the products get cheaper, the hardware gets cheaper, the software gets more capable, things are richer and more automatic, it just comes out and really changes the way we do things.

One way to look at this is to think of how it will affect all the sciences. Sciences have become very data driven. Think about biology with not only genomics but proteonomic data. Think of astronomy with all the observations done at different wavelengths and different times. If you're working in those fields, you need to be able to navigate through that data and visualize it and test your ideas against it. And so software is becoming this important tool so that science is as math itself was over the last several hundred years.

And so it's often people who have a software background working together with experts in the other field that really will make the advances. People in bioinformatics are in a position to really unlock some of the mysteries that will cause great advances in medicine.

As we think about these products that are being created, the fact that they become lower and lower cost means that they'll be used all over the world. I mentioned we've got a billion personal computers out there. We should get that up to 3 or 4 billion, and, in fact, getting the hardware price down is fairly straightforward, for communities that are economically disadvantaged getting the software essentially for free is very easy to do. So the only thing really standing in our way right now actually is the connectivity. That's a tough problem, but again if we take wireless breakthroughs and software that can connect together various wireless nodes, what people often talk about as Mesh, there too we can make that very approachable.

And so the equality driven by this, the ability to get information, the ability to even on the political side have tools that make it harder to suppress information is very fantastic.

It's exciting to us to see these tools being used in classrooms, to see people who have disabilities like blind people now able to reach out and get all the information on the Internet, which in the past they were restricted to just a small subset of books that had been done in Braille.

The jobs in software I think are the most interesting jobs. They're very social jobs, working with people, finding out from customers what they want, and so this is the area where you can say my work goes out there and has impact.

In order for the field to really drive forward we need lots of young, open-minded people who will see new ways of doing things, so it's really your generation that's going to come in and take this foundation that's been developed over these last 30 years and take it to a whole new level. And I'm very excited to be here and see this great work that many of you will do.

Thank you. (Applause.)

MODERATOR: We will now have a series of Q&A, so those who want to ask questions, please line up behind the two mikes, and I will point to those. Don't be shy.

QUESTION: Hi. I was wondering if you could sign my tablet. (Laughter.)

BILL GATES: Sure, I'll catch you afterwards, I'd love to. (Laughter.)

QUESTION: I think a lot of people that I know agree that the Macintosh operating systems are a lot more user friendly than Windows-based systems. Why do you think people feel this way and in the next release of Windows how do you plan to combat that without alienating hard-core Windows users like myself?

BILL GATES: You don't want us to make it easy to use? (Laughter.)

Well, I think software can be very flexible and accommodate a lot of different tastes. Apple has done good work on the Macintosh, that's a very helpful competition.

Windows has a reasonable market share, which means -- (laughter, applause.)

QUESTION: As Microsoft becomes a very large organization, how do you create an environment that still fosters innovation that startups have?

BILL GATES: Well, to be frank, the really tough innovation in our industry requires a level of skill and a long term approach that only an organization like Microsoft can do. Something like speech recognition or tablet computing, these are things that you have to work on for more than a decade before you actually get and solve those problems.

So I think the industry is great. There's a mix of startups that are very agile at doing certain types of things. Inside Microsoft we have our research group, which is our most far out group, we have literally about a hundred what we call incubation groups that are groups of about 10 to 15 people looking at new ideas, and those have often become very important products for us, and then we have our major products.

Something like Office where they're able to really go out to office workers and see how they spend time and how they'd like to do things differently. And so for a product like that you really need a large scale group, a group that is taking on a very tough problem.

Take an area like search; this is something that really brings together work we've done in research on natural language understanding over a period of a long time. The reason why search won't be the kind of treasure hunt it is today where it takes five or six minutes to find what you want is because we'll have this deep understanding, not just text indexing.

And so it's important for Microsoft that we reach out to those startups because they do neat new applications on the Windows system and that's one of the big strengths is the variety of applications that are out there, but it's also important that we take some of the deep platform things, the broad applications that require the unique characteristics we have, and we do those as well. So I think it's a very healthy ecosystem.

QUESTION: How do you think client-side software will be impacted by Web-based software?

BILL GATES: Well, the idea that you have some computing going on right where the user is and some going on at a distance in the so-called server, we've had that in computing for many, many decades. The databases have typically been up on the server, the IT managed environment has been on the server, and the really responsive, flexible computing has been down on the desktop.

Historically, people have thought about this as a dichotomy where you've had flexibility and freedom on the desktop and you've had nice controlled things overly administered centrally.

I think we could get the best of both worlds. Certainly the client is where the action is in terms of changing the environment. Speech recognition is on the client, ink recognition is on the client, the idea of being in an environment where you're visually recognized, that's a client type thing. But we can take that and combine that with advances out on the Web. We take what were servers in the past and make those available either as a service on the Web or for something that you want to say can run the software on the server.

And so typically in the Web you'll have the big data-oriented applications and down on the client you'll have the rich interactive presentation, the ability to go to multiple sites using Web Services and XML and combine those in the way that that user is interested in.
The main drawback to that client environment in terms of backing it up and administering it and things like that, the magic of software can reduce those so that we can still have that empowerment, that flexibility and yet have a cost of ownership and a control of information that makes central IT feel good about it.

QUESTION: If we're moving toward a world with more and more information, more and more personal information stored on small devices, how do we protect against identity theft and piracy issues?

BILL GATES: Well, identity theft is a very interesting problem where people attempt to get you to reveal your secret information, that is your password, and then they can go in and take money out of your bank account or charge things to you.

We have some recent breakthroughs where we can monitor what's going on in terms of how you're using passwords and where you're using them that's an almost perfect detector of this phishing, digital identity theft type activity.

And so I believe that technology, the right technology can stay out in front of these things and prevent those things from blocking this digital vision.

Obviously people are going to have to feel like they're not having their identity stolen, their information is kept secure, or else these dreams about these systems simply won't happen and, in fact, a huge part of the investment in research that the universities and companies like Microsoft need to do are on those parts of the system that will really be hidden. When you no longer get spam, you won't notice it, you just won't think about it. Underneath in the system will be this deep machine learning, Bayesian thing that's constantly evolving to make sure that new spam attacks aren't coming in and wasting your time. And so we need breakthroughs on these things that I'm confident it's getting the right focus so that those won't be blockers.

QUESTION: Hi. My name is (Jay Maziano ?). I'm a student at the school of engineering, I'm a sophomore.

First of all, I'm a Gates Scholar, and I just want to take the opportunity to thank you for your generosity because -- (applause) -- I know that for myself if it wasn't for your scholarship program, I wouldn't be here where I am today. And I speak for a lot of my fellow scholars, and I wanted to thank you for that.

BILL GATES: That's great to hear.

QUESTION: My question is I come from a school where students aren't necessarily very academically motivated, and I just wanted to know your opinion as to what you recommend to students like myself who have dreams of pursuing further education; how do you think we should approach our fellow students and how to motivate them to seek better education.

BILL GATES: That's a great question, because the goal of the whole scholarship program is that those of you who receive the scholarship will then in turn set an example and mentor other people, so that more and more kids understand that they should achieve their potential and get a college education.

In fact, the economy is shifting worldwide so that the new jobs being created are the jobs that require a lot of education. Sometimes people talk about, oh, the manufacturing jobs are moving to China. Well, in a certain sense that's true, but, in fact, the whole move of manufacturing jobs to China isn't increasing, it's just automation is constantly eliminating a number of those jobs, even as they take a higher share of the world's manufacturing activities.

And so particularly in the United States what's important is that we get everybody to be college ready. And I think if kids really understood how they're affecting their life to not take the early opportunities, to take advantage of the features that are good and to do a lot of reading and enjoy learning, the impact that that's going to have later on in their life as they see that, they would make different choices. And so hopefully you're enjoying your college time, as you're getting out into the workforce, that's a message you can take back to your community and nothing speaks as well as an example of someone like yourself.

QUESTION: Again thank you very much. (Applause.)

QUESTION: I was kind of wondering if I could get one of those Xbox 360s.

But for a serious question, I was wondering how you view the importance in your circumstances of dropping out of Harvard and then, one, getting into the business early certainly, but then also working outside of the confines of higher institution (off mike)?

BILL GATES: Well, I think it's something that worked for me. (Laughter.) And seriously, if somebody thinks that you can be at absolutely the ground floor of an entire new industry, a year more or less makes all the difference, you can take that chance. In fact, I'm not even really a dropout, I'm on leave. (Laughter.) As soon as things go bad, I'm headed back to Harvard. (Laughter.)

So, you know, for me it wasn't a high risk type thing, and I loved my college education. I sat around the dorm, I didn't do much. (Laughter.) A lot of other smart kids, they fed me every day, you know, I'm pretty good at tests so I got positive feedback; it would have been fun to stay another year, and it's possible I could have stayed another year and we still would have been okay. In fact, we were the very first company doing software, a lot of other ones came along, none of those are around today.

QUESTION: Hi there. My name is Mark Johnson, and as the student before I'm a Gates Scholar. I attend Columbia College, I'm not an engineer so I don't have any technical questions. But I just wanted to know like in your experience as being a leader and a technology leader and a world leader, humanitarian and everything, what do you find to be your greatest accomplishments?

BILL GATES: Well, I have two things that I love a great deal. I've got my work at Microsoft where I was lucky enough in terms of finding and the people who got hired in to be part of starting this personal computer revolution, the idea of empowering people, and so that's very gratifying. The daily work is a lot of fun because the people are smart and the problems are tough problems and sometimes you're successful, sometimes you're not, and you always know what you need to do next.

I'd say my foundation work, which is part time for me, is equally rewarding, but there again we're gathering in very smart people, taking on tough problems, often global health problems where nobody has really focused on those because there's no marketplace for those drugs, and yet there's these brilliant people, if given the resources, who want to go and create those new vaccines.

And so it's a lot like being in the Microsoft team where we'll have successes and I can tell them, well, don't get too optimistic or we'll have some setbacks, and I'll say, well, don't get too depressed, and try to get the team just to drive forward on things that sometimes take more than a decade.

So I've been surprised at how similar my work at the foundation is to the work that I do at Microsoft. I'm very excited about both of those things. Not many people have one job as fun as I have as in both of these activities.

QUESTION: You said the Apple iPod works with the Xbox 360. Is that just media encoded in MP3 or also Apple's DRM protected format? (Off mike).

BILL GATES: Yeah, that is a very good question. (Laughter.) In fact, the answer is that we can decode native MP3 and native AAP, but we cannot decode FairPlay and that's because Apple won't let us. So as soon as Apple lets us, because it's your music that you paid for, then the Xbox will do that well. But so far no luck on that part.

QUESTION: What role do you feel that Open Source software has in the future of personal computing and what role will it have at Microsoft as you develop your software?

BILL GATES: The one thing I'm not sure is widely recognized is that this dynamic of having free software and commercial software goes back to the very beginning of the industry. UNIX was created by AT&T, was this hard-core piece of software, but then BSD created this BSD UNIX that was out there and was free. The very first browser that was done at Illinois, the NCSA Browser was free, the original TCP/IP stack was free, the original sendmail was free. So it's very common that software at a certain level comes out of some free or academic type environment that's available for use, but then often a company will get started around that and they'll add features to it, they'll support it, they'll make it connect up to other things. And when it comes to looking at which software is the best choice, the commercial software with all those other things will often get the lion's share of what the best choice is.

So we're always going to have free software, that's a very vibrant community, we're always going to have commercial software.

For Microsoft, as long as our employees want to send their kids to school and eat, we'll probably mostly offer commercial software. (Laughter.) Now, we're going to do that under this very high volume, low price type approach, which has worked so well for us, where we can take something like Office and for like $100 a year give somebody the very best, including the ongoing support and capabilities there.

So we think commercial software is quite a bargain and it's wonderful that people have got all those choices.

QUESTION: Thank you.

MODERATOR: I'll just give you a heads up; we have time for only a few questions so not all of you will be able to ask.

QUESTION: This won't take long. I just thought this would be a good time to ask you if maybe you weren't too busy here this weekend, you'd like to help me with my computer science homework. (Laughter.)

BILL GATES: I'd love to but --

MODERATOR: I don't know, that might be cheating though. (Laughter.)

QUESTION: (Off mike/inaudible.)

BILL GATES: Well, certainly Microsoft doesn't sell censorship software to anyone. I'm not even sure what censorship software would be.

The whole point of software is that it's empowering. When you get a word processor you get to write anything you want, when you have e-mail you can send that out to anyone. And, in fact, to the degree that countries have tried to repress information, the fact that digital technology is so pervasive and that you can encrypt things, send it off to different people, there's really no way to in a broad sense repress information today, and I think that's a wonderful advance we can all feel good about.

Now, if that's used to spread secrets about making nuclear bombs, okay, we don't feel as good about that part of it. In Germany they have rules about pro-Nazi statements, that they don't want those to be up there, various countries have things about pornography, child pornography where they have restrictions.

And so when we do business in various countries, we do try to make sure that we're adhering to those rules. So, for example, if somebody put up on Microsoft Spaces something that's child pornography, we would participate in taking that down.

But we only do what is appropriate for each of those local rules. In Germany, for example, if it was a Nazi Web site, we would block that from German access because that's banned in Germany but allowed in the United States. It's actually fairly complicated for a company when you have these disparate regimes and right now the Internet doesn't make it easy to wall things off so things are available in only certain countries. But this is a medium of total openness and total freedom and that's what makes it so special.

QUESTION: As we saw from your presentation the innovation that people at Microsoft and I guess in the computer industry in general, it's going forward at a very, very fast pace, but then you also spoke about in order to make that available for I guess everyone in the world, that getting prices down is very important. But in your opinion what else needs to happen in order to I guess make that technology really accessible to the billions of people who are on the outside looking in right now?

BILL GATES: Okay, well, the first thing I'd say is that getting computers to everyone is not the top priority in terms of the inequities that exist in the world today. The biggest inequity is that we don't treat children who die in developing countries, we don't treat that as being important. It literally takes only a few hundred dollars of vaccine technology to save lives. And so if you go to those mothers or families and say what counts to you the most, would you like your children to live or would you like to have computers, they've got a hierarchy there that's pretty important. So the inequities in the world are quite stark and quite dramatic.

I believe we can work on all of these at once, we don't just need to focus on the health one. Companies in the computer industry like Microsoft, Intel and others should absolutely have their employees very involved and be making sure computers are getting out pervasively.

And there's wonderful progress on this. We have what we call Partners in Learning agreements now with other a hundred countries in the world where we've sat down and made a plan of how we're going to donate software into all of their schools, do teacher training, help them buy some new hardware and work with the hardware partners to get those things out there.

The toughest thing, as I've said earlier, is the connectivity costs because the phone companies are used to much more low volume, high priced type approach, you don't have the density of users, and so it's really this broadband peak that we have to make sure we solve. And over time I do think that can be solved through various wireless technologies.

So this is an industry that can be proud of the fact that it's about pervasive access, and within the next decade we'll drive to, oh, 75 percent of the planet having access to computers.

MODERATOR: I'm sorry for those who are still in line, this is the last question. If this is quick, we'll have another one.

QUESTION: With the conflict you mentioned between HD DVD and Blu-ray, how do you see the industry enacting your vision of uber connectivity without more conflict?

BILL GATES: Well, the beauty of once we get past physical media, then even though you have multiple formats, typically -- unless there's blocks like with FairPlay -- but typically you can translate between these various formats. And so you can translate between Windows Media and MPEG 4 and those things.

So as soon as we get rid of physical media, you won't ever face the thing that's a library that you bought becomes obsolete. In the media world you face that many times. If you bought a lot of cassettes -- maybe that dates me -- then, yes, they know I'm switching over to CDs. Now as you move from CDs, you've got to rip those and get those onto the computer. So the fact that this is the last format battle is a very good thing.

Eventually there will be a compromise. The problem right now is that one of the formats isn't very compatible with the PC. On striking the balance between protecting the content and making things really hard for the user, we feel like the Blu format, the way they're planning to protect and make things way too hard for the user and so the studios have gone overboard in their protection scheme, and that's why we're arguing the battle because we think there should be ease of use in terms of getting movies onto PCs. But this will get resolved. There will be some sort of grand compromise and then we'll have our last physical format.

MODERATOR: And now our last physical question. (Laughter.)

QUESTION: With all the progress that you expect to be made over the next generation in computer science, how do you see the role of the applied mathematician evolving over time?

BILL GATES: Well, mathematics is the mother science of all sciences, including computer science. And I'm a big believer that even if you major in computer science, you ought to take as many math courses as possible, because the kind of thinking, very concrete, proof oriented thinking that that forces you into I think is very helpful.

I also think that having pure mathematicians and applied mathematicians help us think about algorithms, lower bounds, upper bounds, different approaches to things, not just cryptography but everything around graphics, everything around data mining type things, these are really mathematical problems, and so finding the exact boundary when computer science stops and when applied math starts, I think that they are deeply entangled at this point.

So I would take my statement about anyone here taking computer science will have lots of exciting job opportunities, no limit, I'd extend that out to say that anybody who's working in applied math and wants to apply it to these kinds of problems, you'll have great opportunities as well.

MODERATOR: Let's give Bill a round of applause. (Applause.)

Finally, I would like to give Bill not one, not two, but three gifts. The first is our school official cufflinks. You can wear them without a tie and without a jacket. (Laughter.) These are the (crown here, that's our sign ?), so you'll remember the memorable visit that you had here.

The second is a (Columbia sword ?). (Laughter.)

But the third is most important, the third actually was made by our alum, who is here, Jack (Hoberman ?), stand up. (Applause.) And it's not only cute, it's also symbolic. It's the -- Jack Hoberman was a student of mechanical engineering, but he's also an artist. And the students of Fu Foundation School of Engineering Applied Science, we train them as renaissance men and women, so they have a very broad education. So I'm pleased to give you the Hoberman Ball, Hoberman is here, which might symbolize what you and Microsoft have done to the world. This is the world in 1980 and this is the world today. (Laughter, applause.)

END