Sergey Brin and Larry Page: The genesis of Google

Sergey Brin: I want to discuss a question

I know that's been pressing on many of your minds.

We spoke to you last several years ago.

And before I get started today,

since many of you are wondering,

I just wanted to get it out of the way.

The answer is boxers.

Now I hope all of you feel better.

Do you know what this might be? Does anyone know what that is?

Audience: Yes.

SB: What is it?

Audience: It's people logging on to Google around the world.

SB: Wow, OK. I didn't really realize what it was when I first saw it.

But this is what helped me see it.

This is what we run at the office, that actually runs real time.

Here it's slightly logged.

But here you can see around the world

how people are using Google.

And every one of those rising dots

represents probably about 20, 30 searches,

or something like that.

And they're labeled by color right now, by language.

So you can see: here we are in the U.S.,

and they're all coming up red.

There we are in Monterey -- hopefully I can get it right.

You can see that Japan is busy at night,

right there.

We have Tokyo coming in in Japanese.

There's a lot of activity in China.

There's a lot of activity in India.

There's some in the Middle East, the little pockets.

And Europe, which is right now in the middle of the day,

is going really strong with a whole wide variety of languages.

Now you can also see, if I turn this around here --

hopefully I won't shake the world too much.

But you can also see, there are places where there's not so much.

Australia, because there just aren't very many people there.

And this is something that we should really work on,

which is Africa, which is just a few trickles,

basically in South Africa and a few other urban cities.

But basically, what we've noticed is these queries,

which come in at thousands per second,

are available everywhere there is power.

And pretty much everywhere there is power, there is the Internet.

And even in Antarctica -- well, at least this time of year --

we from time to time will see a query rising up.

And if we had it plotted correctly,

I think the International Space Station would have it, too.

So this is

some of the challenge that we have here,

is you can see that it's actually kind of hard to get the --

there we go.

This is how we have to move the bits around

to actually get the people the answers to their questions.

You can see that there's a lot of data running around.

It has to go all over the world: through fibers,

through satellites, through all kinds of connections.

And it's pretty tricky for us to maintain the latencies

as low as we try to. Hopefully your experience is good.

But you can see also, once again -- so some places are much more wired

than others, and you can see all the bandwidth across the U.S.,

going up over to Asia, Europe in the other direction, and so forth.

Now what I would like to do is just to show you

what one second of this activity would look like.

And if we can switch to slides --

all right, here we go.

So this is slowed down.

This is what one second looks like.

And this is what we spend a lot of our time doing,

is just making sure that we can keep up

with this kind of traffic load.

Now, each one of those queries

has an interesting life and tale of its own.

I mean, it could be somebody's health,

it could be somebody's career, something important to them.

And it could potentially be something

as important as tomato sauce,

or in this case, ketchup.

So this is a query that we had --

I guess it's a popular band that was more popular in some parts of the world than others.

You can see that it got started right here.

In the U.S. and Spain, it was popular at the same time.

But it didn't have quite the same pickup in the U.S.

as it did in Spain.

And then from Spain, it went to Italy,

and then Germany got excited, and maybe right now the U.K. is enjoying it.

And so I guess the U.S. finally,

finally started to like it, too.

And I just wanted to play it for you.

Anyway, you can all enjoy it for yourselves --

hopefully that search will work.

As a part of --

you know, part of what we want to do to grow our company

is to have more searches.

And what that means is we want to have

more people who are healthy and educated.

More animals, if they start doing searches as well.

But partly, we want to make the world a better place,

and so one thing that we're embarking upon is the Google Foundation,

and we're in the process of setting that up.

We also have a program already called Google Grants

that now serves over 150 different charities around the world,

and these are some of the charities that are on there.

And it's something I'm very excited to be a part of.

In fact, many of the organizations that are here --

the Acumen Fund, I think ApproTEC we have running, I'm not sure if that one's up yet --

and many of the people who have presented here are running through Google Grants.

They run Google ads, and we just give them the ad credit

so they can let organizations know.

One of the earlier results that we got --

we have a Singaporean businessman who is now sponsoring a village

of 25 Vietnamese girls for their education,

and that was one of the earliest results. And as I said, now there have been

many, many stories that have come in,

because we do have hundreds of charities in there,

and the Google Foundation will be an even broader endeavor.

Now does anybody know who this is?

A-ha!

Audience: Orkut.

SB: Yes! Somebody got it.

This is Orkut. Is anybody here on Orkut?

Do we have any?

Okay, not very many people know about it.

I'll explain it in a second.

This is one of our engineers.

We find that they work better when they're submerged and covered with leaves.

That's how we churn those products out.

Orkut had a vision to create a social network.

I know all of you are thinking, "Yet another social network."

But it was a dream of his, and we, basically,

when people really want to do something, well, we generally let them.

So this is what he built.

We just released it in a test phase last month,

and it's been taking off.

This is our VP of Engineering.

You can see the red hair, and I don't know if you can see the nose ring there.

And these are all of his friends.

So this is how -- we just deployed it --

we just decided that people would send each other invitations to get into the service,

and so we just had the people in our company initially send them out.

And now we've grown to over 100,000 members.

And they spread, actually, very quickly, even outside the U.S.

You can see, even though the U.S. is still the majority here --

though, by the way, search-wise, it's only about 30 percent of our traffic --

but it's already going to Japan, and the U.K., and Europe,

and all the rest of the countries.

So it's a fun little project.

There are a variety of demographics. I won't bore you with these.

But it's just the kind of thing that we just try out for fun

and see where it goes.

And --

well, I'll leave you in suspense.

Larry, you can explain this one.

Larry Page: Thank you, Sergey.

So one of the things -- both Sergey and I

went to a Montessori school,

and I think, for some reason,

this has been incorporated in Google.

And Sergey mentioned Orkut, which is something that,

you know, Orkut wanted to do in his time,

and we call this -- at Google, we've embodied this as "the 20 percent time,"

and the idea is, for 20 percent of your time,

if you're working at Google, you can do what you think is the best thing to do.

And many, many things at Google have come out of that,

such as Orkut and also Google News.

And I think many other things in the world also have come out of this.

Mendel, who was supposed to be teaching high-school students,

actually, you know, discovered the laws of genetics --

as a hobby, basically.

So many, many useful things come out of this.

And News, which I just mentioned,

was started by a researcher.

And he just -- he -- after 9/11, he got really interested in the news.

And he said, "Why don't I look at the news better?"

And so he started clustering it by category,

and then he started using it, and then his friends started using it.

And then, besides just looking cute on a baby's bottom,

we made it a Googlette,

which is basically a small project at Google.

So it'd be like three people, or something like that,

and they would try to make a product.

And we wouldn't really be sure if it's going to work or not.

And in News' case, you know, they had a couple of people

working on it for a while, and then more and more people

started using it, and then we put it out on the Internet,

and more and more people started using it.

And now it's a real, full-blown project with more people on it.

And this is how we keep our innovation running.

I think usually, as companies get bigger,

they find it really hard to have small, innovative projects.

And we had this problem, too, for a while, and we said,

"Oh, we really need a new concept."

You know, the Googlettes -- that's a small project that we're not quite sure if it's going to work or not,

but we hope it will, and if we do enough of them,

some of them will really work and turn out, such as News.

But then we had a problem because then we had over 100 projects.

And I don't know about all of you,

but I have trouble keeping 100 things in my head at once.

And we found that if we just wrote all of them down

and ordered them -- and these are kind of made up.

Don't really pay attention to them.

For example, the "Buy Iceland" was from a media article.

We would never do such a crazy thing, but --

in any case, we found if we just basically wrote them all down and ordered them,

that most people would actually agree what the ordering should be.

And this was kind of a surprise to me, but

we found that as long as you keep the 100 things in your head,

which you did by writing them down,

that you could do a pretty good job deciding what to do

and where to put your resources.

And so that's basically what we've done

since we instituted that a few years ago, and I think it has really allowed us to be innovative

and still stay reasonably well-organized.

The other thing we discovered is that people like to work on things that are important,

and so naturally,

people sort of migrate to the things that are high priorities.

I just wanted to highlight a couple of things

that are new, or you might not know about.

And the top thing, actually, is the Deskbar.

So this is a new -- how many of you use the Google Toolbar?

Raise your hands.

How many of you use the Deskbar?

All right, see? You guys should try it out.

But if you go to our site and search

for "Deskbar," you'll get this.

And the idea is, instead of a toolbar, it's just present all the time

on your screen on the bottom,

and you can do searches really easily.

And it's sort of like a better version of the toolbar.

Thank you, Sergey.

This is another example of a project that somebody at Google

was really passionate about, and they just, they got going,

and it's really, really a great product, and really taking off.

Google Answers is something we started, which is really cool,

which lets you -- for five to 100 dollars,

you can type a question in,

and then there's a pool of researchers

that go out and research it for you, and it's guaranteed and all that,

and you can get actually very good answers to things

without spending all that time yourself.

Froogle lets you search shopping information,

and Blogger lets you publish things.

But all of these -- well, these were all sort of innovative things that we did that --

you know, we try many, many different things

in our company.

We also like to innovate in our physical space,

and we noticed in meetings, you know, you have to wait a long time

for projectors to turn on and off,

and they're noisy, so people shut them off.

And we didn't like that, so we actually,

in maybe a couple of weeks, we built these little enclosures

that enclosed the projectors, and so we can leave them on all the time

and they're completely silent.

And as a result, we were able to build some software

that also lets us manage a meeting,

so when you walk into a meeting room now,

it lists all the meetings that are happening,

you can very easily take notes, and they just get emailed automatically

to all the people that were present in the meeting.

And as we become more of a global company,

we find these things really affect us --

you know, can we work effectively with people who aren't in the room?

And things like that. And simple things like this can really make a big difference.

We also have a lot of engineers in those meetings,

and they don't always do their laundry as much as they should.

And so we found it was pretty helpful

to have laundry machines, for our younger employees especially, and ...

we also allow dogs and things like that,

and we've had, I think, a really fun culture at our company,

which helps people work and enjoy what they're doing.

This is actually our "cult picture."

I just wanted to show quickly.

We had this on our website for a while,

but we found that after we put it on our website,

we didn't get any job applications anymore.

But anyway, every year we've taken

the whole company on a ski trip.

A lot of work happens in companies from people knowing each other, and informally.

And I think we've done a good job encouraging that.

It makes it a really fun place to work.

Along with our logos, too, which I think really embody

our culture when we change things.

In the early days, we were actually advised

we should never change our logo because

we should establish our brand, you know,

because, you know, you'd never want to change your logo.

You want it to be consistent.

And we said, "Well, that doesn't sound so much fun.

Why don't we try changing it every day?"

One of the things that really excites me about what we're doing now

is we have this thing called AdSense,

and this is a little bit foreshadowing --

this is from before Dean dropped out.

But the idea is, like, on a newspaper, for example,

we show you relevant ads.

And this is hard to read, but this says "Battle for New Hampshire:

Howard Dean for President" -- articles on Howard Dean.

And these ads are generated automatically --

like in this case, on the Washington Post --

from the content on the site.

And so we use our over 150,000 advertisers

and millions of advertisements, so we pick the one

that's most relevant to what you're actually looking at,

much as we do on search.

So the idea is we can make advertising useful,

not just annoying, right?

And the nice thing about this,

we have a self-serve program,

and many thousands of websites have signed up,

and this let's them really make money. And I --

you know, there's a number of people I met --

I met this guy who runs a conservation site at a party,

and he said, "You know, I wasn't making any money.

I just put this thing on my site and I'm making 10,000 dollars a month.

And, you know, thank you.

I don't have to do my other job now."

And I think this is really important for us, because it makes the Internet work better.

It makes content get better, it makes searching work better,

when people can really make their livelihood

from producing great content.

So this session is supposed to be about the future,

so I'd thought I'd talk at least briefly about it.

And the idea behind this is to do the perfect job doing search,

you really have to be smart.

Because you can type, you know, any kind of thing into Google,

and you expect an answer back, right?

But finding things is tricky, and so you really want intelligence.

And in fact, the ultimate search engine would be smart.

It would be artificial intelligence.

And so that's something we work on,

and we even have some people who are excited enough

and crazy enough to work on it now,

and that's really their goal.

So we always hope that Google will be smart,

but we're always surprised when other people think that it is.

And so I just wanted to give a funny example of this.

This is a blog from Iraq,

and it's not really what

I'm going to talk about, but I just wanted to show you an example.

Maybe, Sergey, you can highlight this.

So we decided --

actually, the highlight's right there. Oh, thank you.

So, "related searches," right there. You can't see it that well,

but we decided we should put in this feature

into our AdSense ads, called "related searches."

And so we'd say, you know, "Did you mean 'search for'" -- what is this,

in this case, "Saddam Hussein," because this blog is about Iraq --

and you know, in addition to the ads,

and we thought this would be a great idea.

And so there is this blog

of a young person who was kind of depressed, and he said,

"You know, I'm sleeping a lot."

He was just kind of writing about his life.

And our algorithms -- not a person, of course,

but our algorithms, our computers --

read his blog and decided that

the related search was, "I am bored."

And he read this, and he thought a person had decided

that he was boring,

and it was very unfortunate,

and he said, "You know, what are these, you know, bastards at Google doing?

Why don't they like my blog?"

And so then we read his blog, which was getting -- you know,

sort of going from bad to worse,

and we said the related search was, "Retards."

And then, you know, he got even more mad,

and he wrote -- like, started swearing and so on.

And then we produced "You suck."

And finally, it ended with "Kiss my ass."

And so basically, he thought he was dealing with something smart,

and of course, you know,

we just sort of wrote this program and we tried it out,

and it didn't quite work,

and we don't have this feature anymore.

So with that, maybe I can switch back to the world.

I wanted to end just by saying that

there's a couple things that really make me excited

to be involved with Google,

and one of those is that we're able to make money

largely through advertising, and one of the benefits that I didn't expect from that

was that we're able to serve everyone in the world

without worrying about, you know, places that don't have as much money.

So we don't have to worry about our products being sold,

for example, for less money in places that are poor,

and then they get re-imported into the U.S. --

for example, with the drug industry.

And I think we're really lucky to have that kind of business model

because everyone in the world has access to our search,

and I think that's a tremendous, tremendous benefit.

The other thing I wanted to mention just briefly

is that we have a tremendous ability and responsibility

to provide people the right information,

and we view ourselves like a newspaper or a magazine --

that we should provide very objective information.

And so in our search results, we never accept payment for our search results.

We accept payment for advertising,

and we mark it as such.

And that's unlike many of our competitors.

And I think decisions we're able to make like that

have a tremendous impact on the world,

and it makes me really proud to be involved with Google.

So thank you.