So, Larry sent me an email, and he basically said we've got to make sure that we don't seem like we're a couple of middle-aged boring men. I said I'm flattered by that because I'm a bit older, and he has a bit more net worth than I do.
Well, thank you.
So we'll have a conversation about the Internet, and we'll have a conversation about Google, and we'll have a conversation about search and privacy, and also about your philosophy and a sense of how you've connected the dots and how this journey that began some time ago has such interesting prospects. Mainly we want to talk about the future. So my first question: Where is Google and where is it going?
Well, this is something we think about a lot, and our mission we defined a long time ago is to organize the world's information and make it universally accessible and useful. And people always say, "Is that really what you guys are still doing?" And I always kind of think about that myself, and I'm not quite sure. But actually, when I think about search, it's such a deep thing for all of us, you know, to really understand what you want, to understand the world's information, and we're still very much in the early stages of that, which is totally crazy. We've been at it for 15 years already, but it's not at all done.
When it's done, how will it be?
Well, I guess, you know, in thinking about where we're going—you know, why is it not done?—a lot of it is just computing's kind of a mess. You know, your computer doesn't know where you are, it doesn't know what you're doing, it doesn't know what you know, and a lot we've been trying to do recently is just make your devices work, make them understand your context. Google Now, you know, knows where you are, knows what you may need. So really having computing work and understand you and understand that information, we really haven't done that yet. It's still very, very clunky.
Tell me, when you look at what Google is doing, where does Deep Mind fit?
Yeah, so Deep Mind is a company we just acquired recently. It's in the U.K. First, let me tell you the way we got there, which was looking at search and really understanding, try and understand everything, and also make the computers not clunky and really understand you—like, voice was really important. So what's the state of the art on speech recognition? It's not very good. You know, it doesn't really understand you. So we started doing machine learning research to improve that. That helped a lot. And we started just looking at things like YouTube. Can we understand YouTube? But we actually ran machine learning on YouTube and it discovered cats, just by itself. Now, that's an important concept. And we realized there's really something here. If we can learn what cats are, that must be really important. So I think Deep Mind, what's really amazing about Deep Mind is that it can actually—they're learning things in this unsupervised way. They started with video games, and really just—maybe I can show the video, just playing video games, and learning how to do that automatically.
Take a look at the video games and how machines are coming to be able to do some remarkable things.
The amazing thing about this is this is, I mean, obviously, these are old games, but the system just sees what you see, the pixels, and it has the controls and it has the score, and it's learned to play all of these games, same program. It's learned to play all of these games with superhuman performance. We've not been able to do things like this with computers before. And maybe I'll just narrate this one quickly. This is boxing, and it figures out it can sort of pin the opponent down. The computer's on the left, and it's just racking up points. So imagine if this kind of intelligence were thrown at your schedule, or your information needs, or things like that. We're really just at the beginning of that, and that's what I'm really excited about.
When you look at all that's taken place with Deep Mind and the boxing, also a part of where we're going is artificial intelligence, where are we when you look at that?
Well, I think, for me, this is kind of one of the most exciting things I've seen in a long time. The guy who started this company, Demis, has a neuroscience and a computer science background. He went back to school to get his Ph.D. to study the brain. And so I think we're seeing a lot of exciting work going on that sort of crosses computer science and neuroscience in terms of really understanding what it takes to make something smart and do really interesting things.
But where's the level of it now? And how fast do you think we are moving?
Well, this is the state of the art right now, understanding cats on YouTube and things like that, improving voice recognition. We used a lot of machine learning to improve things incrementally, but I think for me, this example's really exciting, because it's one program that can do a lot of different things.
I don't know if we can do this, but we've got the image of the cat. It would be wonderful to see this. This is how machines looked at cats and what they came up with. Can we see that image?
There it is. Can you see the cat? Designed by machines; seen by machines.
That's right. So this is learned from just watching YouTube. And there's no training, no notion of a cat, but this kind of concept of a cat is something important that you would understand, and now that the machines can kind of understand. Maybe just finishing also on the search part, it started with search, really understanding people's context and their information. I did have a video—I want to show quickly on that—from what we actually found.
Not long ago, I planted a crop of potatoes. Then suddenly they started dying one after the other. I checked out the books and they didn't tell me much. So, I went and I did a search. Potato diseases. One of the websites told me that ants could be the problem. It said sprinkle wood ash over the plants. Then after a few days, the ants disappeared. I got excited about the Internet. I have this friend who really would like to expand his business, so I went with him to the cyber cafe and we checked out several sites. When I met him next, he was going to put a windmill at the local school. I felt proud because something that wasn't there before was suddenly there. I realized that not everybody can be able to access what I was able to access. I thought that I need to have an Internet that my grandmother can use. So I thought about a notice board—a simple wooden notice board. When I get information on my phone, I'm able to post the information on the notice board. So it's basically like a computer. I use the Internet to help people. I think I am searching for a better life for me and my neighbors. So many people have access to information, but there's no follow-up to that. I think the follow-up to that is our knowledge. When people have the knowledge, they can find solutions without having to be helped out. Information is powerful, but it is how we use it that will define us.
Now, the amazing thing about that video, actually, was we just read about it in the news, and we found this gentleman and made that little clip.
When I talk to people about you, they say to me, people who know you well, say, "Larry wants to change the world, and he believes technology can show the way." And that means access to the Internet. It has to do with languages. It also means how people can get access and do things that will affect their community, and this is an example.
Yeah, that's right, and I think, for me, I have been focusing on access more, if we're talking about the future. You know, we recently released this Loon Project which is using balloons to do it. It sounds totally crazy. We can show the video here. Actually, two out of three people in the world don't have good Internet access now. We actually think this can really help people sort of cost-efficiently.
It's a balloon.
Yeah, get access to the Internet.
And why does this balloon give you access to the Internet? Because there was some interesting things you had to do to figure out how to make balloons possible. They didn't have to be tethered.
Yeah, and this is a good example of innovation. Like, we've been thinking about this idea for five years or more before we started working on it, but it was just really, how do we get access points up high, cheaply? You normally have to use satellites and it takes a long time to launch them. But you saw there how easy it is to launch a balloon and get it up, and actually again, it's the power of the Internet. I did a search on it, and I found, 30, 40 years ago, someone had put up a balloon and it had gone around the Earth multiple times. And I thought, why can't we do that today? And that's kind of how this project got going.
But are you at the mercy of the wind?
Yeah, but it turns out we did some weather simulations which probably hadn't really been done before, and if you control the altitude of the balloons, which you can do by pumping air into them and other ways, you can actually control roughly where they go, and so I think we can build a worldwide mesh of these balloons that can kind of cover the whole planet.
Before I talk about the future and transportation, where you've been a nerd for a while, and this fascination you have with transportation and automated cars and bicycles, let me talk a bit about what's been the subject here earlier with Edward Snowden. It is security and privacy. You have to have been thinking about that.
Yeah, absolutely. I saw the picture of Sergey with Edward Snowden yesterday. Some of you may have seen it. But I think, for me, I guess, privacy and security are really important things. We think about it in terms of both things, and I think that you can't have privacy without security, so let me just talk about security first. Because you asked about Snowden and all of that, and then I'll say a little bit about privacy. I think, for me, it's tremendously disappointing that the government sort of secretly did all this stuff and didn't tell us. I don't think we can have a democracy if we're having to protect you and our users from the government for stuff that we've never had a conversation about. And I don't mean we have to know what the particular terrorist attack is they're worried about protecting us from, but we do need to know what the parameters of it is, what kind of surveillance the government's going to do and how and why, and I think we haven't had that conversation. So I think the government's actually done itself a tremendous disservice by doing all that in secret.
Never coming to Google to ask for anything?
Not Google, but the public. I think we need to have a debate about that, or we can't have a functioning democracy. It's just not possible. So I think I'm sad that Google's in the position of protecting you and our users from the government doing secret thing that nobody knows about. It doesn't make any sense.
Yeah. And then there's a privacy side of it.
Yes. The privacy side, I think it's—the world is changing. You carry a phone; it knows where you are. There's so much more information about you, and that's an important thing, and it makes sense why people are asking difficult questions. We spend a lot of time thinking about this and what the issues are. I'm a little bit—I think the main thing that we need to do is just provide people choice, show them what data's being collected, you know, search history, location data. We're excited about incognito mode in Chrome, and doing that in more ways, just giving people more choice and more awareness of what's going on. I also think it's very easy. What I'm worried is that we throw out the baby with the bathwater. And I look at, you know, on your show, actually, I kind of lost my voice, and I haven't gotten it back. I'm hoping that by talking to you I'm going to get it back.
If I could do anything, I would do that.
All right, so get out your voodoo doll and whatever you need to do. But I think, you know what, I look at that, I made that public, and I got all this information. We got a survey done on medical conditions with people who have similar issues, and I look at medical records, and I say, "Wouldn't it be amazing if everyone's medical records were available anonymously to research doctors?" And when someone accesses your medical record, a research doctor, they could see, you could see which doctor accessed it and why, and you could maybe learn about what conditions you have. I think if we just did that, we'd save 100,000 lives this year.
Absolutely. Let me go...
So I guess I'm just very worried that with Internet privacy, we're doing the same thing we're doing with medical records is we're throwing out the baby with the bathwater; we're not really thinking about the tremendous good that can come from people sharing information with the right people in the right ways.
And the necessary condition that people have to have confidence that their information will not be abused.
Yeah, and I had this problem with my voice stuff. I was scared to share it. Sergey encouraged me to do that, and it was a great thing to do.
And the response has been overwhelming.
Yeah, and people are super positive. We got thousands and thousands of people with similar conditions, which there's no data on today. So it was a really good thing.
So, talking about the future, what is it about you and transportation systems?
Yeah, I mean, I guess I was just frustrated with this when I was at college in Michigan. I had to get on the bus and take it and wait for it. And it was cold and snowing. I did some research on how much it cost, and I just became a bit obsessed with transportation systems.
And that began the idea of an automated car.
Yeah, about 18 years ago, I learned about people working on automated cars, and I became fascinated by that, and it takes a while to get these projects going, but I'm super excited about the possibilities of that improving the world. There's 20 million people or more injured per year. It's the leading cause of death for people under 34 in the U.S.
So you're talking about saving lives.
Yeah, and also saving space and making life better. You know, Los Angeles is half parking lots and roads, half of the area, and most cities are not far behind, actually. And I mean it's just crazy that that's what we use our space for.
And how soon will we be there?
I think we can be there very, very soon. We've driven well over 100,000 miles now, totally automated. I'm super excited about getting that out quickly.
But it's not only you're talking about automated cars, you also have this idea for bicycles.
Well, at Google, we got this idea that we should just provide free bikes to everyone, and that's been amazing, most to the trips. You see bikes going everywhere, and the bikes wear out. They're getting used 24 hours a day.
But you want to put them above the street, too.
Well, I said, how do we get people using bikes more?
We may have a video here.
Yeah, let's show the video. I just got excited about this. So this is actually how you might separate bikes from cars with minimal cost. Anyway, it looks totally crazy, but I was actually thinking about our campus, working with the Zippies and stuff, and just trying to get a lot more bike usage, and I was thinking about, how do you cost-effectively separate the bikes from traffic? And I went and searched, and this is what I found. And we're not actually working on this, that particular thing, but it gets your imagination going.
Let me close with this. Give me a sense of the philosophy of your own mind. You have this idea of Google X. I mean, you don't simply want to go in some small, measurable arena of progress.
Yeah, I think many of the things we just talked about are like that, where they're really—I almost use the economic concept of additionality, which means that you're doing something that wouldn't happen unless you were actually doing it. And I think the more you can do things like that, the bigger impact you have, and that's about doing things that people might not think are possible. And I've been amazed. The more I learn about technology, the more I realize I don't know, and that's because this technological horizon, the thing that you can see to do next—the more you learn about technology, the more you learn what's possible. You learn that the balloons are possible because there's some material that will work for them.
What's interesting about you too, though, for me, is that we have lots of people who are thinking about the future, and they are going and looking and they're coming back, but we never see the implementation. I think of somebody you knew and read about, Tesla. The principle of that for you is what?
Well, I think invention is not enough. If you invent something, Tesla invented electric power that we use, but he struggled to get it out to people. That had to be done by other people. It took a long time. And I think if we can actually combine both things, where we have an innovation and invention focus, plus the ability to really—a company that can really commercialize things and get them to people in a way that's positive for the world and to give people hope. You know, I'm amazed with the Loon Project, just how excited people were about that, because it gave them hope for the two thirds of the world that doesn't have Internet right now that's any good.
Which is a second thing about corporations. You are one of those people who believe that corporations are an agent of change if they are run well.
Yeah. I'm really dismayed most people think companies are basically evil. They get a bad rap. And I think that's somewhat correct. I mean, companies are doing the same incremental thing that they did 50 years ago, or 20 years ago. That's not really what we need. We need—especially in technology—we need revolutionary change, not incremental change.
You once said, actually, as I think I've got this about right, that you might consider, rather than giving your money, if you were leaving it to some cause, just simply giving it to Elon Musk, because you had confidence that he would change the future, and that you would therefore—
Yeah, if you want to go Mars—he wants to go to Mars, to back up humanity. That's a worthy goal, but it's a company, and it's philanthropical. So I think we aim to do kind of similar things. And I think, you ask we have a lot of employees at Google who have become pretty wealthy. You know, people make a lot of money in technology. A lot of people in the room are pretty wealthy. You're working because you want to change the world. You want to make it better. Why isn't the company that you work for worthy not just of your time but your money as well? I mean, but we don't have a concept of that. That's not how we think about companies. And I think it's sad because companies are most of our effort. They're where most of people's time is, where a lot of the money is, and so I think I'd like for us to help out more than we are.
When I close conversations with lots of people, I always ask this question: What state of mind, what quality of mind is it that has served you best? People like Rupert Murdoch have said curiosity, and other people in the media have said that. Bill Gates and Warren Buffett have said focus. What quality of mind, as I leave this audience, has enabled you to think about the future and at the same time, change the present?
You know, I think the most important thing—I looked at lots of companies and why I thought they don't succeed over time. We've had a more rapid turnover of companies. And I said, what did they fundamentally do wrong? What did those companies all do wrong? And usually, it's just that they missed the future. And so I think, for me, I just try to focus on that and say, what is that future really going to be and how do we create it, and how do we cause our organization to really focus on that and drive that at a really high rate? And so that's been curiosity. It's been looking at things people might not think about, working on things that no one else is working on, because that's where the additionality really is, and be willing to do that, to take that risk. Look at Android. I felt guilty about working on Android. When it was starting, it was a little startup we bought. It wasn't really what we were really working on. And I felt guilty about spending time on that. That was stupid. That was the future, right? That was a good thing to be working on.
It is great to see you here. It's great to hear from you, and a pleasure to sit at this table with you. Thanks, Larry.