Remarks by Steve Ballmer, Chief Executive Officer
Royal Institute of Technology (Kungliga Tekniska högskolan, KTH)
October 4, 2010
STEVE BALLMER: Well, thanks! I thought I was going to have to do a little Dance Central number there to get the crowd going, but I think we're all glad I don't have to at this stage.
It's a real pleasure for me to have a chance to be here with you today and to have a chance to make some remarks and take some questions. I want to thank President Gudmundson for the opportunity to be here at KTH and to be in this auditorium. It's one of the more vertical auditoriums I've ever been in, so if you're in the back and it doesn't feel like I'm making good eye contact, I apologize in advance.
I'm going to talk about a few things and hopefully from the Kinect demo and some of the other things that we had a chance to share with you, you can get a little bit of a sense of why we're so enthusiastic about the future of technology. Everybody who sits here in this room essentially has made kind of a key life decision, which is you chose somehow to focus on some aspect of information technology as really important to your lives. You're going to make a business, you're going to make a career, you're going to become academics. And in a sense, I'll, you just at the start, you made the world's best decision. The chance to do fun things, to do exciting things, to make a difference in the world, there will be no better opportunity than for people who have studied, broadly, information technology.
And whether you look out five years, 10 years, 15 years, things are going to continue to evolve. We've done a lot of great things at Microsoft. There are a lot of great things we wish we had done. There are a lot of great things we are doing, but the thing that we can rest assured on is that the opportunity to do something new is there tomorrow. It's only, you know, sort of up to our imaginations.
And so we sit here today and we talk about technologies like natural user interface and the ability to speak, to touch, to interact visually with systems, it's only going to get better. You saw a little bit of that hopefully in the Kinect demonstration, and it just starts to scratch the surface of what we can do.
Today, we've got search engines and we've got applications and we've got websites, but what you really want to be able to do over time isn't to have these be three different things, you just want to be able to type something into the computer like, or speak something into the computer like "get me ready for my trip to Stockholm." Get me ready for my trip to Stockholm means everything to my secretary. Ah, Stockholm, what's the weather in Stockholm? That would be nice to know. Ah, Stockholm, what customers is he seeing Stockholm, what speeches is he giving -- Stockholm, Stockholm, Stockholm -- she knows exactly what to do.
My poor, dumb computer doesn't know anything to do. And yet, everything my secretary does, she does on the computer. That's a trainable, repeatable, that's just natural user interface to learn to understand our user intent and to learn to understand information well enough to bring them together.
And I just pick one technology. There are so many more that are coming. I'm going to talk today about the move to the cloud, which I think is very important. There's the ability to model the physical world in the virtual world and I know we'll have some people here in the audience who will care about biological science and environmental science. The key to the speed up of scientific discovery in a lot of fields is actually going to depend on being able to run experiments in the virtual world. It's done wonders in the world of genomics and pharmaceuticals, et cetera.
And as we sit here and we think about the environment and the desire we all have to improve the environment, it's a key question of how do we speed up the science, the science of energy, the science of the environment? And with new computing techniques, most of which I won't have a chance to talk about today, things will move faster. New programming techniques, new user interface techniques. It's all, to me, very, very exciting.
Perhaps one of the most foundational advances that will help all of these other things move forward is the move to what we, in our industry, have chosen to refer to as the cloud. I will tell you, I have given speeches of half an hour and more where I described the cloud and at the end, I've had people raise their hands and say, "But what is the cloud?" (Laughter.)
And whether I answer the question in a way that you feel like you can go home and tell somebody precisely or not, I hope you'll get a sense of the transformation, because it's a buzzword that's not a precise thing, it's a buzzword that refers to a fundamental transformation that's going on in the computing infrastructure that should enable and speed up a lot of different things.
When we use the word "cloud" we're in a sense talking about bringing together the best of the Internet with the best of smart devices -- phones, PCs, TVs, other devices -- with the best of traditional enterprise and data center computing. We're trying to bring the best of all of those together. The best of all of those.
Today, if you stop and think about it, the time and cost and complexity in creating any kind of computing product is a lot bigger and the work is a lot harder than it will be in the future.
I just spent an hour -- had a great lunch with representatives on the IT side from four regions of Sweden, and these were the folks who were responsible for healthcare and healthcare IT specifically, and their No. 1 problem is how do you really bring together information for the doctor? How do you really ensure that instead of having essentially hundreds of different systems across Sweden to help process lab results in doctors' offices and hospitals, how do you write one of those, make it available, write it once and make it available instantly across the country somehow to speed up the rate at which IT benefits healthcare?
Well, the answer is in the cloud, but the cloud doesn't just create kind of new opportunities, it also creates new responsibilities. It allows us to improve the reusability of technology, the speed and agility with which technology comes to market, but it doesn't relieve us of any of the responsibilities of doing things that are secure and private. And we spend a lot of time talking about that.
I'm sure all of us have had the experience in a doctor's office of finding out that you’re entering the same information multiple times. Certainly, sharing systems is a key aspect of that. You move to the cloud and all of a sudden, the ability for the small creator, whether you're trying to create a new application-slash-website, a new game -- hopefully for our Xbox Kinect system -- the opportunity to do that as a smaller -- as an individual or smaller company really climbs quite dramatically because all of that infrastructure, the clients, the servers, it's all there for you in the cloud. You don't have to sit there and say, oh, I've got to buy servers or I've got to reserve capacity at the hoster. You just want to write your application and have it propagated.
If you're a scientist, we're going through this right now, one of the big problems for scientists is actually sharing experimental data and all being able to operate and collaborate on the same data sets. The cloud, again, serves as a point of answer.
There's a lot of work to do. There's a lot of work to do both in terms of creating commercial infrastructure that enables opportunity, but also working on privacy and security. If you look at the kinds of things we've done in Internet Explorer, just on private browsing, just as a small example, there is a lot of work to be done to make sure that as people commit more and more of their lives to kind of shared technology infrastructure, they can do that in a way in which they have great confidence and success.
I talk about the cloud as a place that learns and helps you learn. I'm going to avoid the -- how many people here use a word processor? Everybody uses a word processor. In the old days of a word processor, we built in these things called spell checkers. People like spell checkers, they're quite handy, they're quite nice. The fact of the matter is spell checkers are better today than they ever were before because now instead of just looking at your own mistypings, we can look at your own mistypings and compare them to the mistypings of everybody else on the planet in any language on the planet. And the spell checker can get better and smarter every day.
I take that as a small example, but it leads into a whole host of examples, again, with appropriate respect for people's privacy being important. The ability to have systems learn and get smarter by seeing what people around the planet are doing is very, very helpful.
Search engines work this way, particularly one of the things we're trying to do as we break in against the Goliath in the search business with our Bing technologies, we're trying to be smarter in recognizing user intent. When you type the word "flowers" into a search engine, we're trying to understand what did you mean statistically? Not just what are the 10 most relevant links, but it turns out, nine out of 10 times, when people type the word "flowers" in, they want to buy some. Or maybe it's six out of 10, and three out of 10 they want horticultural references. How do you help understand and learn statistically what people want and anticipate the world for them?
So I like to say the cloud learns, but the cloud also is a place that helps you learn. It doesn't just learn about you and learn about others, it should help you learn and explore and navigate and take action. And I think those aspects are quite important.
During the height of the kind of economic meltdown of whatever -- it feels like a long time ago, two years ago, I guess. During the height of the economic meltdown, I was sitting there in my office over Christmas saying, "Oh my gosh, it doesn't mean much for the technology, it's going to be fine, but we might have a little less buying here in the near term. What should we do? What's happened in the past in economic downturns? This is cataclysmic." And what I really wanted to do was just -- what I did was say to myself: I want to see what happened with GDP growth in economies that had huge debt in times in the past. So, I said for 10 countries, show me what GDP -- what debt was as a percentage of GDP, and what the performance was in GDP growth.
I'm not an economist, I'm not an expert, but I kind of got a sense that that was important. So, what do you sit there and do? You go website to website to website and then you copy things down and you paste them back in. Why couldn't I just, voom, draw out the spreadsheet that I knew I wanted and have the computer, the cloud, take care of the rest of it for me?
I eventually got my answer. We had to cut back just a little bit at Microsoft and now, hopefully, things are good, as all of you enter the workforce. But we have to be better at helping people learn, decide, and take action and the cloud can do that. The cloud has access to all of the information that you would need to get there.
The cloud enhances your personal and professional interactions. I've got to say, the world is now finally different than it was when I started at Microsoft. We've gone through -- let me give you a few generations. When I got to Microsoft, this sounds just bizarre to me, but there was no e-mail. You wouldn't even really make a phone call across the ocean, it seemed very expensive, and there were these little things called telexes that we used to send that were short cables, just bizarre.
Then there was kind of the generation like when you could call people. Then there was a generation of PowerPoint and a generation of Facebook. I used to give speeches like this, and I knew nobody would ever watch them, despite the fact they were also videoed. OK? I used to be able to say that until about four years ago, probability 1.00 nobody would ever watch the video. Now the probability is somewhat reduced from 1.00, but it's still not -- we still haven't accomplished all that we want to have.
If I was to have offered to give this speech virtually, nobody would have thought it's the same. We are not yet in a place where we've done all we can, despite all the great innovations and the way we communicate, we network with one another, both professional and personally, the job is still not even remotely done. I mean, even as people are sitting here jotting notes or typing, you know that your notes are not correlated with what's on the slide or what's in this video. You can't just dash it off and send it to a colleague or a friend who's not here and say, "Boy, Ballmer wasn't making any sense right there." And have it, voom, drill exactly into that point in the video. But that's the kind of things that people really want to do.
You know, it's funny, I'm here so that you can see me and I can see you, but with this bright light, I can't even see most of you. I want a virtual interaction where I literally -- the camera can be looking through who's paying attention -- all that kind of stuff. We're not there yet. This could help my personal and my social interaction. I'm SteveB@Microsoft.com. That's my e-mail address. I give it out liberally. Turns out, most computers will spam you, human beings will only send you what they think are good ideas. So, I'm happy to hear from you. And yet, there are better ways for me to invite your questions, your feedback, than counting on individual e-mails and individual responses, and there's more to be invented.
So we're investing certainly as a company, if you look at what we're trying to do with Hotmail and Messenger, you look at what we're trying to do with Office, and a lot of new things, the way we're interacting and collaborating with Twitter and Facebook, there's a lot to be done to reinvent the way we collaborate and communicate socially and personally.
The cloud drives advances that drive the cloud. When we, and others, first started building our high-scale Internet services, one of the things that we learned was all of the technology, software, computers, storage, networking gear, all of the stuff that we were selling to our own customers for their data centers needed to be improved to actually work well for high-scale, cloud-based solutions. With the amount of traffic, the places in the world it gets used, with the redundancy required, whether it's Bing or what we're doing with Windows Live or the back-end services for our Windows Phone or what we're trying to do with Office, there's no way to run those services using just the traditional datacenter technologies. You have to write programs differently and you have to deploy them differently.
We talk about next-generation Windows Azure replacing Windows Server as a way to program and deploy applications. Most of you coming out of school will come out thinking I've got to learn to write for the cloud more than you think I have to learn to write for yesterday's systems, just like people my age never really learned how to write mainframe programs. We learned to write for the client-server paradigm. We're going through that shift right now. We're not there. It may not be as cataclysmic as some, but it's super, super important.
Even the way -- I have a son who's a freshman in university, and they're studying the history of computing. The history of computing is, in large measure, kind of what I've done most of my life, but it's now the history of computing. And yet, when we talk about the history of computing, what you will -- what students will study in this university in 10 years, what is kind of today's practice in terms of the way to build and deploy datacenters. And for many of you, I think it's important to recognize just how rapidly that shift will go on.
We see that in so many ways now that are kind of interesting. And as we think about the new applications, I mean, literally, let's talk about datacenters. Datacenters on the planet today consume about 3 percent of the world's power. The datacenter annual budget for the world is hundreds of billions, probably about $700 billion. And the applications that get deployed to those datacenters are almost always out of date very quickly.
One of the things we're doing just for our own use is how do you build a datacenter that literally comes in a box, a box that requires almost no electricity and almost no power to cool? What does that kind of thing look like? And we're designing next-generation datacenters that literally come in shipping containers. You plug in an electrical cord that hopefully doesn't have to have much power to it, you plug in a garden hose because we hope you can cool it with a garden hose, and you plug in an Internet connection, and you get a cloud. And instead of building these big facilities, I don't know what the datacenters look like here at the university, but it probably still has a fancy air conditioner and some raised roof. You really just want to pour down a slab of concrete, maybe put a roof on top of the shipping container, and call it done.
In next-generation datacenters, very few jobs should be created, in fact. The data centers we're building today, there are more jobs that go into constructing them than go into operating them because if you build the hardware and software platform correctly, that's the evolution, and now we need to keep taking cost and complexity in ways that I think benefit the technology industry, the environment, everything else under the sun.
Fifth and certainly not least is the cloud wants smarter devices. This is now more of a legacy discussion, but one of the things that people like to talk about is would the world change back to a world in which all computing was centralized in the cloud and we would all just have sort of dumb devices that would receive HTML streams? Well whether it's sort of the advent of smart phones and some of the good work that we've seen there, whether it's the kind of work we're doing in Internet Explorer 9 where we show that there are multiple ways even just of running HTML, a smart device can run standard HTML better than a dumb device can with the right browser.
If it's the kind of things you saw in Kinect, the cloud doesn't want to do everything. If it's going to recognize you and your voice and your gestures and your dance moves, you know, it's going to have to have intelligence built into it. So, we need to reconceptualize what an intelligent device looks like in the move to the cloud and the way it gets built and the way it gets managed.
You know, we're on the verge of launching the next generation of our smart phone technology, so-called Windows Phones. You know, I've gone to a world where in the old days, I actually used paper and pen when I came on a trip like this. Now, I move to a world in which a smart device actually -- and then I worry all trip about losing the paper. All trip, my No. 1 concern was I'd lose all my notes.
I still occasionally use paper. Why? Because it turns out that it's more sociable and respectful if when I make a note I'm looking at you and scribbling indeterminately than if I sit here, for example, and say, you know, maybe I'll make a note here and just say quickly -- I'll do it as an audio recording. Swedish students, hopefully, love Windows Phones. Now I'm done with that.
I probably wouldn't do that in a one-on-one meeting. I'd probably sit there and type, but these smart devices -- that note is already available to me in the cloud. I don't have to think about it anymore, it's done, it's been backed up, it's relevant -- it's unfortunately not yet synched to the video, so I won't remember what I was doing, but the cloud wants smarter devices.
In an attempt to kind of give you a little bit of a feel for how all of this comes together, we thought we'd have one of our partner companies come on stage and share with you a little bit of how they're thinking about cloud services and smart devices and maybe one you know pretty well. Important and popular kind of Swedish startup. We're going to invite Daniel Ek, CEO and founder of Spotify to come on stage and share some thoughts with you about some of these topics. Daniel. (Applause.)
DANIEL EK: (Comments off mike.) (Inaudible) As you may know, Spotify is a company that was founded in 2006. And the general idea beyond Spotify was that we see that music has changed based on the Internet. (Inaudible) -- and the reason why you do so is because they no longer just (inaudible) music from a radio station. What they do is that they (inaudible) friends.
So when we set out to build Spotify, one of the most important things that we could do was to actually make it easier for you to discover music from your friends and share music with your friends. The other thing that we really sort of wanted to do and that we cared about was the fact that we wanted to make it easy for you to just consume any song in the world.
So we started working on the Spotify app for the desktop. And that was sort of the first (inaudible) allows you to easily just consume as much music as you want -- whether it's Lady Gaga or whether it's something else. So, we really made that simple.
We also wanted you to more easily share music with your friends. So, within Spotify, you can now drag and drop any song, (inaudible) friends. We started working with Microsoft a couple of months back for the Windows 7 platform, so now we support task bars, snap, and jump list as well.
The real sort of game-shifter for us has been connecting not only with your PC, but connecting with your device. So, about a year ago, we started to introduce Spotify for the mobile platform. So, if you've got a smart phone today, you can basically consume all of your library, available from the cloud, and you can also instantly share that with your friends. So, a few months back, we started working together with Microsoft on bringing Spotify to your Windows Phones as well.
So what I'm here to announce today is that you're now going to be able to consume your Spotify music on Windows Phone. So, hopefully we can show you a little bit of a video here and showcase that technology as well, of course all the latest features that are available.
(Spotify video segment.)
DANIEL EK: So you can download the Windows Phone app today from Spotify.com, or the Microsoft Marketplace. Of course we're going to work together as well to develop for Windows Phone 7. Thank you again, Steve, for having me today, and I'm going to leave it to Steve. Keep sharing music. Thanks. (Applause.)
STEVE BALLMER: I've heard a lot about Daniel over the years, first time we met personally, so it's good to have a chance to do it, and thanks to KTH for that opportunity as well, we have a lot to talk about between us.
I hope what you get a little bit is a sense of the possibilities, the ongoing possibilities and innovation that we see happening in the marketplace. You know, we have a great relationship here at KTH, we sponsor the largest international programming competition, something we call Imagine Cup. KTH has been actively involved, and students from here have done super, super well on that. Exciting for me to have a chance to be here in that context.
We have a number of graduates of this university working for us, thank you very much, and certainly if you look around the technology industry, a lot of the great innovations and energy, particularly in mobile, but more broadly, have come from Swedish innovators and Swedish entrepreneurs. And whether it's the move to the cloud or all of the other next-generation technologies that come with it, I'm very clear in my mind that many of you are going to be out there defining the future, and many others are going to be applying it to important challenges in society and certainly it's been a pleasure to address you in both of those contexts, and hopefully our paths get a chance to cross again.
I'll look forward to some of your questions, but it's been my pleasure. Thanks again to KTH and thanks to all of you for your time. (Applause.)
The truth of the matter is it's the fundamental technology underpinning on which we all will have a chance to do more than we could before. If you look at it today, and I said this, but I'll say it in a little different way -- the No. 1 impediment to using -- the top two impediments to having information technology support a lot more of what we do is the speed with which we can build applications and the speed with which you can get them deployed in ways that people can use them. And whether it's Spotify or the favorite healthcare application of tomorrow, speed and agility of using information is going to be an important thing for everybody in their daily lives.
STEVE BALLMER: Well, I would say the challenge of the next five years, in some senses, I'll give meta-level, then maybe a little more -- you know, concreteness to it. The challenge is just moving forward fast enough on a broad enough set of fronts. I mean, the truth of the matter is new things, they don't happen every day, but you've got to be shooting far in enough in front to make a difference.
You take a look at our Xbox Kinect system, wow, I feel like we're really onto something -- and it's not going to be all about video games, it's going to be about anything you want to do from your TV set.
So the challenge is staying ahead. And then in the few cases, heaven forbid, that you don't stay -- how do you get back in the game? How do you shoot ahead of the other guy? And I would say it's probably pretty fair for me to say we've got our hands full in phones, and we've got our hands full in search. And so really inventing -- it's almost even harder -- how do you invent a future that's sort of different but around things that people are getting used to? And those two, every bit of cleverness, creativity, hard work, energy -- it's going to take all of it and then some.
Meanwhile, everything else has to be -- whether it's Xbox, it's Windows, it's what we're trying to do with Windows Azure and cloud computing -- blah blah blah -- so just -- innovating and getting ahead or, yeah, getting back ahead or getting ahead, that would be I'd say the No. 1 challenge.
What's my favorite game? Well, I'm a pretty sort of of-the-time kind of guy. I love -- my wife has not liked the Xbox very much. She thinks the Xbox isn't a great system. She thinks it's a thing that gets in the way of our kids' homework. And now that we've launched Kinect, my wife loves the new Xbox. It gets the kids up and moving, it's not all passive. My wife hasn't been able to figure out a game controller to save her life, but she does know how to speak and use her hands. They are very social. They are inclusive. I'd say a lot of games have been more for men 18 to 35.
You know, if you take a look at Dance Central or any of these things. So, I love the new system. My favorite, actually, is this new beach volleyball game that we have coming out for Xbox Kinect. You play two-on-two beach volleyball. And I'll admit, I don’t get a lot of air between my feet and the floor, I'm just not that agile, but I get a pretty good workout in a short period of time. So, that's No. 1 for me. Coming soon to a store near you. (Applause.)
QUESTION: Hello, Mr. Ballmer. My question is this: How is Microsoft going to retake the operating system market from Apple and the open source community like Linux?
STEVE BALLMER: Well, let me start by being a little cocky. Retake? (Laughter.) I mean, honestly, for all the "blah blah blah" -- I'm the first one to admit when we -- and I'm not saying there's not going to be a lot of competition, but 95 percent, this room excluded, 95 percent of the world's computers run Windows. They don't run Mac, they don't run Linux. I'm not saying that there's not good innovations happening in those areas, I'm not saying that. And there is a lot of competition whether it's the move to new form factors, phones, slates, tablets, et cetera. There's a lot of competition out there.
The key is actually to really think through broadly a lot of challenges at once. Where is hardware going? And how do you change the operating system to take advantage of next-generation hardware? That's been area where we've done a good job, but where also there have been seams. And you see that in some of the good work that our competition has done has been innovating in hardware and software at the same time.
You have to think about the software itself and how do you innovate in the experiences? How do you innovate in a way that makes it easier or better or kind of more sensible in writing applications? How do you integrate services in? How do you give an opportunity to think about the cloud?
We're trying to move on all of those fronts at once. We took some good steps with Windows 7. We have some big steps with Internet Explorer. We've got a lot of challenges with Windows Phone. We're kind of trying to move it all along at once.
STEVE BALLMER: Yeah. Small reminder -- because students are not typical -- small reminder: No. 1 player in smart phones in the world is? Nokia. No. 2 in the world in smart phones is BlackBerry, No. 3 is Apple, number four a long way behind that is Google, and then you get us as No. 5. And we used to be a lot higher up. But anyway -- (laughter.) So, we have our work cut out for us. And it is a combination of helping -- we don't make phones, but we have to help think through the design of new phones, the hardware.
We did that in generating Windows Phone 7. We actually did a core hardware design on top of which you'll see innovation from HTC, from Samsung, from others, from LG, from a number of players.
We decided that we wanted to do things differently from a user interface perspective. Another phone that just had a sea of applications, instead, what we have here are active hubs, everything that's going on, for example, with my son Sam is on one tile. I have all of his contact information, I can see what's new with him on Facebook, I can see what's new with him on Twitter -- I'm not going to show you that stuff to save the innocent, but instead of being about applications, it's about the things and tasks you want to do in your life. So, some of them are apps, some of them are people, some of them are experiences. We bring together all of your music and video, not just any one of those. Yeah, you can still get to a sea of application UI if you want it, but we tried to do things very different in the user experience.
We're trying to be very different than Android. Android is a least-common-denominator platform. The good news, some people will tell you, is you can get a lot of diversity. We've done a lot of standardization so that if you write an application, it can run across multiple environments.
So there are some things we're trying to do, but it starts with the user experience being different and to enable new classes of hardware that I hope people think are exciting and different, but with a wealth of applications that can integrate because of the basic user experience that we've built into the phone.
STEVE BALLMER: Yeah. Certainly, China is a large population country that's relatively affluent. It's a rich country. There are poor people, but it's actually a relatively rich country that has embraced information technology quite dramatically. It's the largest cell phone market in the world. There will be more PCs purchased in China this year than there are in the United States so it's a big, important market.
I think that one of the big question marks in China will be how well protected or not is intellectual property, because the opportunities for innovators, for entrepreneurs will, in part, depend on that question. If you build your intellectual property down into chips, you do very well in China. If you distribute software, quite difficult. If you build a product for which the trademark and brands are not well protected in China because there's not great protection for brands and trademarks in addition to patents and copyrights, problematic.
I think the Chinese government gets this and will want to continue to see growth in the market, but they're going to have to work really hard because the level of intellectual property protection in China is about the worst in the world, and it's gotten worse over the last several years.
STEVE BALLMER: Surf boards?
STEVE BALLMER: OK, good. No. (Laughter.) Replace phones is that what you said? Yeah, no. I can't imagine holding a surf board up to my ear, or even trying to carry it in my pocket. I mean, the truth of the matter is I think part of the lesson we're learning is society is affluent enough and needs are diverse enough that people want multiple form factors.
I happen to think people will expect more general purpose behavior. We expect this to be more than just a telephone. We expect a computer to be more than a word processor, so those expectations will live. But when you want something you can put in your pocket, you want something you can put in your pocket. You cannot put a surf board in your pocket.
Now, with that said, will the way input works on these things continue to evolve? Might we have fold-out screens that are thin and light? That's probably not next year, but it will happen in the future. Can you imagine this also being a projector so you set it down, shine it somehow on a piece of paper, maybe something like this, you shine it on a piece of paper, and that becomes a digital interactive surface onto which you can type. You can imagine a lot of things like that happening, but literally something that looks like today's slate-cum-surface surf board becoming something that you would actually lug around? I'm not going to make that bet. I think it's more likely your phone will take on attributes of your slate than the other way around.
QUESTION: The question is: As a CEO, what's most challenging for you? Is it managing the people? Is it the technical issues that you look at every day?
STEVE BALLMER: I'd say two things: People are always up on the list. No, because attracting, motivating, enabling sort of funneling talented people is tough. You can give people too much rope, and then everything fragments into a bunch of different pieces. You can try to over-centralize, and then it's tough to get the best, the brightest, the most creative. Really, finding the mix. And it's different in different kinds of businesses. You know, Daniel's got a little tighter organization. He can see, touch, feel probably most of your people most days.
You've got to remember, I've been at Microsoft since we were 30 people, so I kind of know about all sizes of companies at this stage. And what you have to do as the leader or the CEO, it does change. It starts from being the vision, making the sales. Over time, it's about enabling people and at the same time channeling them just enough so that the whole is bigger than the sum of the parts.
The other big thing is prioritizing. Prioritizing. You can't -- it doesn't matter how big you are, how small you are, you can't do everything. Startups that try to do everything fail. Actually, most startups that don't try to do everything fail, but in aggregate, startups succeed colossally well, and that's an important thing.
We are a little different. We're not just a funder of startups. We have to pick the right big bets, but we've got to pick them, we've got to make them, we've got to bet on them, we've got to decide whether it's sales or marketing or product innovations and R&D. What are we really going to bet on and then channel the talent around that? And so the combination of making the right bets and enabling and channeling people the right way I would say -- I don't know if it's my biggest challenge, but it's certainly the thing that will most determine whether I do a good job in the way I lead Microsoft.
(Break for direction.)
STEVE BALLMER: I'll take the last question from the audience just to show I can take questions from real people in addition to you. Yes, ma'am.
QUESTION: My question is: From your personal experience, what do you think are the most important qualifications to become a leader in today's high-tech industry?
STEVE BALLMER: Yeah. I think the most important thing to be a leader in our industry is to kind of have two sides to your personality, a product side -- I'm not a deep -- I've never written a commercial piece of software, so I've never been a programmer, but I have a side of me that loves product. It doesn't mean that I'm going to conceive every new idea, but I have to be willing to make bets.
At the same time, all great leaders in technology companies have a commercial side, a way to envision kind of how the business will unfold and make money. If you look at the great companies in our business, all of the leaders have had some kind of a blend. Bill Gates, who I know the best, definitely a product person, definitely a commercial person. Larry Ellison, who we compete with, definitely a product person, and also definitely a commercial person.
If you watch the movie “Social Network,” I think you'll -- I like Mark Zuckerberg quite a bit, but he definitely has a commercial side in addition to his product side at Facebook. So, I think for leaders in our business -- and to some degree, that's about the individual, and to some degree, it's about the teams that come together to lead efforts have to be able to combine commercial and product-slash-technical characteristics. But if you ask Daniel, he himself and the people around him at Spotify, it's getting those things to come together that's part of the magic.
Thank you all very, very much. Appreciate it. Been my pleasure. (Applause.)