problem-solving techniques that we use here at Google.

One aspect of computational thinking is decomposition,

where we take a large problem, and we break it down into

smaller pieces.

Another aspect is pattern recognition.

We use pattern recognition to help us identify similarities

and differences.

The final part of computational thinking is

algorithmic design.

That allows us to create a step-by-step strategy for

solving a problem.

Google Earth is a perfect example of computational

thinking, because it takes the large problem of trying to

visualize the entire planet and makes it so that anybody

can explore the world around them.

DANIEL BARCAY: Google Earth is basically an attempt to

recreate the whole world in 3D.

And we want to make it as if it's the real world.

We don't want to make anything up.

We don't want to create anything or invent anything.

We want to make it the real world.

And initially, it seems like this crazy problem if you

really think about.

The world is huge.

It takes a tremendous amount of data.

If we were to try to send this to you, we would have to pull

up in front of your house with tractor trailers full of hard

drives of all the images.

So it seems kind of impossible,

but we make it possible.

In school, you're given these problems that are

very black and white.

You either have the answer or you don't have the answer.

You got it right or you got it wrong.

In the real world, there are many right answers.

JEREMY PACK: Google Maps is a collection of imagery and data

about places and roads all around the world.

And using Street View, we actually have images from cars

idle on all of the streets.

Once we've collected all this imagery, we have to somehow

put it together in a way to be able to

share it with the world.

Not long ago, I started getting annoyed with Pegman.

Pegman is the little yellow guy on Google Maps that you

can drag to get into Street View.

And so you drag Pegman from the corner in Google Maps, and

you drop him on the street you want to get into and look at.

That works great when you know exactly where you want to go,

when you want to zoom all the way in on a single address and

drop him next to house or something.

But say you just want to go to New York City.

When I drag and drop Pegman, he'd fall somewhere--

well, somewhere random.

In fact, he seems to prefer to land in back alleys.

He didn't mind landing in the middle of a field.

He'd land on big highways.

But he'd almost never land in front of a famous landmark.

And if you dropped on Paris, he'd never, ever land on the

Eiffel Tower.

I thought to myself, we can do better than that.

We kind of know what famous places are in the world, we

should fix this.

I began to think to myself, what makes a panorama, one of

these images in Street View, important?

And then I thought about it further.

I thought, when people go to places that are interesting,

when they physically travel there, what do they always do

when they're in a famous place?

They pull out their camera and start taking pictures.

People post a lot of these pictures to the internet, so

we could look for places that people take a lot of pictures.

We can find things that are in those photographs that people

took, and see if we can find the same things in the Street

View images.

And so if we can see the Eiffel Tower in the Street

View image and we can see it in a bunch of images people

took, we can automatically know that

it's probably important.

We've worked really hard to make Pegman smarter.

When he's dropping from the sky, we want him to land

somewhere interesting.

DANIEL BARKAY: Thinking computationally is a lot more

like art than it is like math class.

You go in and you know you want to create something, and

you have a blank canvas.

And you use math, and you use these tools to

paint on that canvas.

And you end up creating something beautiful.