What we learned from 5 million books

Translator: Bjarne Poulsen Reviewer: Jonas Tholstrup Christensen

Erez Lieberman Aiden: Everyone knows

Erez Lieberman Aiden: Alle ved

that a picture is worth a thousand words.

at et billede siger mere end tusind ord

But we at Harvard

Men på Harvard

were wondering if this was really true.

spurgte vi os selv, om det egentlig er sandt.

(Laughter)

(Latter)

So we assembled a team of experts,

Så vi samlede et hold eksperter,

spanning Harvard, MIT,

både fra Harvard, MIT,

The American Heritage Dictionary, The Encyclopedia Britannica

The American Heritage Dictionary, The Encyclopedia Britannica

and even our proud sponsors,

og sågar vores stolte sponsor...

the Google.

The Google.

And we cogitated about this

Og vi har funderet over dette

for about four years.

i cirka fire år.

And we came to a startling conclusion.

Og vores konklusion er overraskende.

Ladies and gentlemen, a picture is not worth a thousand words.

Mine damer og herrer, et billede siger ikke mere end tusind ord.

In fact, we found some pictures

Det viste sig faktisk at nogle billeder

that are worth 500 billion words.

siger mere end 500 milliarder ord.

Jean-Baptiste Michel: So how did we get to this conclusion?

Jean-Baptiste Michel: Hvordan når vi denne konklusion?

So Erez and I were thinking about ways

Erez og jeg tænkte på, hvordan man

to get a big picture of human culture

kunne få overblik over menneskets kultur og historie -

and human history: change over time.

- og ændringen over tid.

So many books actually have been written over the years.

Der skrevet så mange bøger gennem tiderne.

So we were thinking, well the best way to learn from them

Så vi tænkte at man kan lære mest af alle disse bøger

is to read all of these millions of books.

ved at læse dem alle sammen.

Now of course, if there's a scale for how awesome that is,

Hvis der er en skala for, hvor fantastisk det er

that has to rank extremely, extremely high.

må det selvfølgelig ligge meget, meget højt (Awesome).

Now the problem is there's an X-axis for that,

Problemet er, at der også er en X-akse,

which is the practical axis.

og det aksen for, om det også er praktisk.

This is very, very low.

Den er meget, meget lav.

(Applause)

(Bifald)

Now people tend to use an alternative approach,

Folk bruger som regel en anden tilgang,

which is to take a few sources and read them very carefully.

Man tager nogle få kilder og læser dem meget omhyggeligt.

This is extremely practical, but not so awesome.

Dette er meget praktisk, men ikke særlig fantastisk.

What you really want to do

Det bedste må være

is to get to the awesome yet practical part of this space.

at nå til dette fantastiske men alligevel praktiske område.

So it turns out there was a company across the river called Google

Et firma på den anden side af floden - Google -

who had started a digitization project a few years back

startede et digitaliseringsprojekt for nogle år siden

that might just enable this approach.

og det kan måske gøre denne tilgang mulig.

They have digitized millions of books.

De har digitaliseret millioner af bøger.

So what that means is, one could use computational methods

Man kan således bruge computerbaserede metoder

to read all of the books in a click of a button.

til at læse alle bøgerne med et enkelt klik.

That's very practical and extremely awesome.

Det er meget praktisk og ekstremt fantastisk.

ELA: Let me tell you a little bit about where books come from.

ELA: Nu skal I høre, hvor bøger stammer fra.

Since time immemorial, there have been authors.

Der har altid eksisteret forfattere.

These authors have been striving to write books.

Disse forfattere har bestræbt sig på at skrive bøger.

And this became considerably easier

Og det blev væsentligt nemmere

with the development of the printing press some centuries ago.

da trykpressen blev opfundet for nogle hundrede år siden.

Since then, the authors have won

Siden da, er det lykkedes forfattere

on 129 million distinct occasions,

at udgive bøger

publishing books.

129 millioner gange.

Now if those books are not lost to history,

Hvis disse bøger ikke er gået tabt for historien,

then they are somewhere in a library,

findes de på et bibliotek et sted,

and many of those books have been getting retrieved from the libraries

og mange bøgerne er blevet taget fra hylderne

and digitized by Google,

og er blevet digitaliseret af Google,

which has scanned 15 million books to date.

som til dato har scannet 15 millioner bøger.

Now when Google digitizes a book, they put it into a really nice format.

Når Google digitaliserer en bog, får den et rigtig fint format.

Now we've got the data, plus we have metadata.

Nu har vi både data og metada.

We have information about things like where was it published,

Vi har f.eks. oplysninger om, hvor den blev udgivet,

who was the author, when was it published.

hvem forfatteren var, og hvornår den blev udgivet.

And what we do is go through all of those records

Og vi går gennem alle disse arkiver

and exclude everything that's not the highest quality data.

og udelukker alle data, der ikke er af højeste kvalitet.

What we're left with

Det, der er tilbage, er en samling

is a collection of five million books,

på fem millioner bøger,

500 billion words,

500 milliarder ord,

a string of characters a thousand times longer

en tegnstreng, der er tusind gange længere

than the human genome --

end menneskets arvemasse.

a text which, when written out,

Hvis teksten blev skrevet ud,

would stretch from here to the Moon and back

ville den nå herfra til månen og tilbage igen

10 times over --

10 gange!

a veritable shard of our cultural genome.

- Et sandt brudstykke af vores kulturelle arvemasse.

Of course what we did

Det vi gjorde,

when faced with such outrageous hyperbole ...

da vi stod over for så vanvittige sammenligninger...

(Laughter)

(Latter)

was what any self-respecting researchers

var, hvad enhver forskere med respekt for sig selv

would have done.

ville have gjort.

We took a page out of XKCD,

Vi gjorde som i tegneserien XKCD,

and we said, "Stand back.

og sagde "Gør plads!

We're going to try science."

Vi prøver med videnskab".

(Laughter)

(Latter)

JM: Now of course, we were thinking,

JM: Først tænkte vi selvfølgelig,

well let's just first put the data out there

"Vi gør bare data tilgængelige,

for people to do science to it.

så andre kan bruge videnskab på dem."

Now we're thinking, what data can we release?

Nu tænker vi "Hvilke data kan vi lægge ud?"

Well of course, you want to take the books

Egentlig vil vi gerne tage bøgerne

and release the full text of these five million books.

og lægge teksten fra alle fem millioner bøger ud.

Now Google, and Jon Orwant in particular,

Men Google - og særligt Jon Orwant -

told us a little equation that we should learn.

fortalte om en ligning, vi skulle lære.

So you have five million, that is, five million authors

Vi har altså fem millioner forfattere

and five million plaintiffs is a massive lawsuit.

altså fem millioner, der gerne vil sagsøge os.

So, although that would be really, really awesome,

Så selvom det ville være virkelig, virkelig fantastisk,

again, that's extremely, extremely impractical.

ville det også være helt ekstremt upraktisk.

(Laughter)

(Latter)

Now again, we kind of caved in,

Igen lod vi os overtale

and we did the very practical approach, which was a bit less awesome.

og fulgte den praktiske tilgang, der var lidt mindre fantastisk.

We said, well instead of releasing the full text,

I stedet for at lægge den fulde tekst ud ville vi

we're going to release statistics about the books.

gøre statistikker om bøgerne tilgængelige.

So take for instance "A gleam of happiness."

Et eksempel er "A gleam of happiness" - Et glimpt af lykke

It's four words; we call that a four-gram.

Det er fire ord - det vi kalder et fire-gram

We're going to tell you how many times a particular four-gram

Vi vil nu fortælle jer, hvor mange gange et bestemt fire-gram

appeared in books in 1801, 1802, 1803,

optrådte i bøger i 1801, 1802, 1803,

all the way up to 2008.

og helt op til 2008

That gives us a time series

Det giver os en tidsserie, der viser hvor hyppigt

of how frequently this particular sentence was used over time.

denne ene sætning er blevet brugt over tid.

We do that for all the words and phrases that appear in those books,

Det gør vi for alle ord og udtryk i disse bøger.

and that gives us a big table of two billion lines

Det giver os en stor tabel med to milliarder linjer

that tell us about the way culture has been changing.

som viser hvordan kulturen har ændret sig.

ELA: So those two billion lines,

ELA: Disse to milliarder linjer

we call them two billion n-grams.

som vi kalder to milliarder n-grammer...

What do they tell us?

Hvad fortæller de os?

Well the individual n-grams measure cultural trends.

De enkelte n-grammer måler kulturelle tendenser.

Let me give you an example.

Lad mig give et eksempel.

Let's suppose that I am thriving,

Jeg vil sige, at jeg trives,

then tomorrow I want to tell you about how well I did.

i morgen siger jeg så, hvor godt jeg havde det.

And so I might say, "Yesterday, I throve."

Jeg ville sige "I går trivedes (throve) jeg".

Alternatively, I could say, "Yesterday, I thrived."

Man kan også bruge "thrived" i stedet for "throve".

Well which one should I use?

Hvilket af de to ord skal jeg bruge?

How to know?

Hvor skulle jeg vide det fra?

As of about six months ago,

Indtil for seks måneder siden

the state of the art in this field

var den anerkendte metode på dette område

is that you would, for instance,

at du f.eks. kunne få fat i

go up to the following psychologist with fabulous hair,

denne psykolog med lækkert hår

and you'd say,

og spørge ham:

"Steve, you're an expert on the irregular verbs.

"Steve, du er ekspert i uregelmæssige verber.

What should I do?"

Hvad skal jeg gøre?"

And he'd tell you, "Well most people say thrived,

Og han ville sige: "De fleste mennesker bruger "thrived"

but some people say throve."

men nogle siger "throve".

And you also knew, more or less,

Og du vidste også - mere eller mindre -

that if you were to go back in time 200 years

at hvis du gik 200 år tilbage i tiden

and ask the following statesman with equally fabulous hair,

og spurgte denne statsmand med ligeså lækkert hår:

(Laughter)

(Latter)

"Tom, what should I say?"

"Tom, hvad ville du sige?"

He'd say, "Well, in my day, most people throve,

Han ville sige: "På min tid brugte de fleste "throve,

but some thrived."

mens andre brugte "thrived".

So now what I'm just going to show you is raw data.

Så nu vil jeg bare vise jer rå data.

Two rows from this table of two billion entries.

To rækker i denne tabel ud af to millarder poster.

What you're seeing is year by year frequency

Den viser hyppigheden pr. år

of "thrived" and "throve" over time.

af "thrived" og "throve" over tid.

Now this is just two

Det her er kun to

out of two billion rows.

ud af to milliarder rækker.

So the entire data set

Så hele datasættet

is a billion times more awesome than this slide.

er en milliard gange mere fantastisk end dette slide.

(Laughter)

(Latter)

(Applause)

(Bifald)

JM: Now there are many other pictures that are worth 500 billion words.

JM: Der er jo mange andre billeder, der siger mere end 500 milliarder ord.

For instance, this one.

For eksempel dette.

If you just take influenza,

Hvis vi bare ser på influenza,

you will see peaks at the time where you knew

vil I se høje udslag på de tidspunkter, hvor I vidste

big flu epidemics were killing people around the globe.

at der var store globale influenzaepidemier.

ELA: If you were not yet convinced,

ELA: Hvis du ikke er overbevist,

sea levels are rising,

stiger vandstanden i havene -

so is atmospheric CO2 and global temperature.

det gør CO2-indholdet i atmosfæren og den globale temperatur også.

JM: You might also want to have a look at this particular n-gram,

JM: Prøv også at kaste et blik på dette n-gram,

and that's to tell Nietzsche that God is not dead,

og det fortæller Nietzsche, at Gud ikke er død,

although you might agree that he might need a better publicist.

selvom du måske også synes, han har brug for en bedre ///presseagent.

(Laughter)

(Latter)

ELA: You can get at some pretty abstract concepts with this sort of thing.

ELA: Man kan få nogle ret abstrakte begreber med disse ting.

For instance, let me tell you the history

Lad mig f.eks. fortælle jer historien

of the year 1950.

om året 1950.

Pretty much for the vast majority of history,

I den største del af vores historie

no one gave a damn about 1950.

har ingen interesseret sig en pind for 1950.

In 1700, in 1800, in 1900,

I 1700 og 1800 og 1900

no one cared.

var ingen interesseret.

Through the 30s and 40s,

Op gennem 30'erne og 40'erne

no one cared.

var ingen interesseret.

Suddenly, in the mid-40s,

Pludselig, midt i 40'erne,

there started to be a buzz.

blev der hvisket i krogene.

People realized that 1950 was going to happen,

Folk indså at 1950 var noget, der ville ske,

and it could be big.

og det kunne være noget stort.

(Laughter)

(Latter)

But nothing got people interested in 1950

Men det der gjorde folk allermest interesseret i 1950

like the year 1950.

var året 1950.

(Laughter)

(Latter)

People were walking around obsessed.

Folk var som besat.

They couldn't stop talking

De kunne ikke lade være med at tale

about all the things they did in 1950,

om alt det, de lavede i 1950,

all the things they were planning to do in 1950,

alt det de planlagde at skulle gøre i 1950,

all the dreams of what they wanted to accomplish in 1950.

og alle drømmene om, hvad de ville opnå i 1950.

In fact, 1950 was so fascinating

Faktisk var 1950 så fascinerende

that for years thereafter,

at folk i flere år efter

people just kept talking about all the amazing things that happened,

bare blev ved med at tale om alle de utrolige ting, der skete -

in '51, '52, '53.

i 1951, 1952 og 1953.

Finally in 1954,

Omsider i 1954

someone woke up and realized

var der en der vågnede op og indså

that 1950 had gotten somewhat passé.

at 1950 var blevet noget passé.

(Laughter)

(Latter)

And just like that, the bubble burst.

Og uden videre sprang boblen.

(Laughter)

(Latter)

And the story of 1950

Og historien om 1950

is the story of every year that we have on record,

er historien om alle de år, vi har registreret,

with a little twist, because now we've got these nice charts.

med et lille tvist, fordi vi nu har disse fine grafer.

And because we have these nice charts, we can measure things.

Og fordi vi har disse fine grafer, kan vi nu måle ting.

We can say, "Well how fast does the bubble burst?"

Vi kan sige "Hvor hurtigt springer boblen?"

And it turns out that we can measure that very precisely.

Og de viser sig, at vi kan måle dette meget præcist.

Equations were derived, graphs were produced,

Der blev udledt ligninger, og der opstillet grafer,

and the net result

og nettoresultatet er

is that we find that the bubble bursts faster and faster

at det viser sig, at boblen springer hurtigere og hurtigere

with each passing year.

for hvert år der går.

We are losing interest in the past more rapidly.

Vi mister interessen for fortiden hurtigere.

JM: Now a little piece of career advice.

JM: Og nu et godt karrieretip:

So for those of you who seek to be famous,

For de af jer, der vil være berømte,

we can learn from the 25 most famous political figures,

kan vi lære af de 25 mest berømte politiske personligheder,

authors, actors and so on.

forfattere, skuespillere osv.

So if you want to become famous early on, you should be an actor,

Så hvis du vil være berømt tidligt, skal du være skuespiller,

because then fame starts rising by the end of your 20s --

fordi berømmelsen så begynder at stige, nrå du er sidst i 20'erne –

you're still young, it's really great.

Du er stadig ung, og det er virkelig skønt.

Now if you can wait a little bit, you should be an author,

Men hvis du kan vente lidt, skal du blive forfatter,

because then you rise to very great heights,

fordi så opnår meget stor berømmelse,

like Mark Twain, for instance: extremely famous.

som f.eks. Mark Twain: Ekstremt berømt.

But if you want to reach the very top,

Men hvis du vil helt til toppen,

you should delay gratification

skal du udskyde den tilfredsstillelse, det er

and, of course, become a politician.

at blive berømt - og selvfølgelig blive politiker.

So here you will become famous by the end of your 50s,

Her vil du blive berømt, når du er i slutningen af 50'erne,

and become very, very famous afterward.

og blive meget, meget berømt derefter.

So scientists also tend to get famous when they're much older.

Videnskabsfolk plejer også at blive berømte, når de er meget ældre.

Like for instance, biologists and physics

For eksempel biologer og fysikere

tend to be almost as famous as actors.

bliver næsten ligeså berømte som skuespillere.

One mistake you should not do is become a mathematician.

En fejl, du ikke skal begå, er at blive matematiker.

(Laughter)

(Latter)

If you do that,

Hvis du gør det,

you might think, "Oh great. I'm going to do my best work when I'm in my 20s."

tænker du måske "Herligt! Jeg leverer mit bedste arbejde, når jeg er i 20'erne"

But guess what, nobody will really care.

Men tænk engang... stort set ingen lægger mærke til det.

(Laughter)

(Latter)

ELA: There are more sobering notes

ELA: Der er mere nøgterne observationer

among the n-grams.

blandt n-grammerne.

For instance, here's the trajectory of Marc Chagall,

Her er f.eks. Marc Chagalls livsforløb,

an artist born in 1887.

som kunster født i 1887.

And this looks like the normal trajectory of a famous person.

Og dette ligner det normale forløb for en berømt person.

He gets more and more and more famous,

Han bliver mere og mere berømt,

except if you look in German.

bare ikke hvis vi ser på tysk.

If you look in German, you see something completely bizarre,

På tysk ser vi noget ganske bizart,

something you pretty much never see,

noget man stort set aldrig ser,

which is he becomes extremely famous

og det er, at han bliver ekstremt berømt

and then all of a sudden plummets,

hvorefter berømmelsen falder brat

going through a nadir between 1933 and 1945,

og er på nulpunktet mellem 1933 og 1945,

before rebounding afterward.

hvorefter berømmelsen vender tilbage.

And of course, what we're seeing

Og de vi selvfølgelig kan se

is the fact Marc Chagall was a Jewish artist

er at Marc Chagall var jødisk kunstner

in Nazi Germany.

i nazi-Tyskland

Now these signals

Disse signaler

are actually so strong

er faktisk så stærk,

that we don't need to know that someone was censored.

at vi ikke behøver at vide, at en person er blevet censureret.

We can actually figure it out

Vi kan faktisk regne det ud

using really basic signal processing.

ved hjælp af meget grundlæggende behandling af signalerne.

Here's a simple way to do it.

Her er en simpel måde at gøre det på.

Well, a reasonable expectation

Det er rimeligt at forvente

is that somebody's fame in a given period of time

at en persons berømmelse i en given periode

should be roughly the average of their fame before

vil være nogenlunde gennemsnittet af berømmelsen før

and their fame after.

og berømmelsen efter perioden.

So that's sort of what we expect.

Så det er nogenlunde, det vi forventer.

And we compare that to the fame that we observe.

Og vi sammenligner med den berømmelse, vi kan aflæse.

And we just divide one by the other

Og så dividerer vi bare den ene med den anden

to produce something we call a suppression index.

så vi får noget, vi kalder et undertrykkelsesindeks.

If the suppression index is very, very, very small,

Hvis undertrykkelsesindekset er meget, meget, meget lavt,

then you very well might be being suppressed.

er der stor sandsynlighed for at du er undertrykt.

If it's very large, maybe you're benefiting from propaganda.

Hvis det er meget højt, får du måske hjælp af propaganda.

JM: Now you can actually look at

JM: Nu kan man faktisk se på

the distribution of suppression indexes over whole populations.

fordelingen af undertrykkelsesindekser over hele populationer.

So for instance, here --

For eksempel her:

this suppression index is for 5,000 people

Dette undertrykkelsesindeks er for 5.000 personer

picked in English books where there's no known suppression --

taget fra engelske bøger uden nogen kendt undertrykkelse.

it would be like this, basically tightly centered on one.

Det ville være på denne måde, tæt centreret om ét.

What you expect is basically what you observe.

Det man kan aflæse, er grundlæggende som forventet.

This is distribution as seen in Germany --

Dette er fordelingen, som den ses i Tyskland.

very different, it's shifted to the left.

Meget anderledes... den er forskudt til venstre.

People talked about it twice less as it should have been.

Folk talte dobbelt så lidt om det, som de burde.

But much more importantly, the distribution is much wider.

Men vigtigere er, at fordelingen er meget bredere.

There are many people who end up on the far left on this distribution

Der er mange personer, der ender ude til venstre i fordelingen,

who are talked about 10 times fewer than they should have been.

som der bliver talt 10 gange så lidt om, som der burde.

But then also many people on the far right

Men der er også personer ude til højre,

who seem to benefit from propaganda.

som synes at være hjulpet af propaganda.

This picture is the hallmark of censorship in the book record.

Dette er kendetegnende for censur i bogregisteret.

ELA: So culturomics

ELA: Denne metode

is what we call this method.

kalder vi "culturomics".

It's kind of like genomics.

Det er lidt ligesom genforskning

Except genomics is a lens on biology

Genomics - genforskning - er et nærbillede af biologi

through the window of the sequence of bases in the human genome.

hvor man ser på sekvenser af baser i arvemassen.

Culturomics is similar.

Culturomics minder om dette.

It's the application of massive-scale data collection analysis

Det er en analyse af en kæmpe samling data

to the study of human culture.

anvendt på studiet af menneskets kultur.

Here, instead of through the lens of a genome,

I stedet for at bruge arvemassen som perspektiv,

through the lens of digitized pieces of the historical record.

bruges digitaliserede stykker af historisk materiale.

The great thing about culturomics

Det gode ved culturomics er

is that everyone can do it.

at alle kan gøre det.

Why can everyone do it?

Hvorfor kan alle gøre det?

Everyone can do it because three guys,

Alle kan gøre det, fordi disse tre herrer,

Jon Orwant, Matt Gray and Will Brockman over at Google,

Jon Orwant, Matt Gray og Will Brockman hos Google,

saw the prototype of the Ngram Viewer,

så prototypen af Ngram Viewer,

and they said, "This is so fun.

og sagde, "Det er så sjovt,

We have to make this available for people."

at vi må gøre det tilgængeligt for alle."

So in two weeks flat -- the two weeks before our paper came out --

På nøjagtig de to uger inden offentliggørelsen af vores rapport

they coded up a version of the Ngram Viewer for the general public.

kodede de en version af Ngram Viewer til almen brug.

And so you too can type in any word or phrase that you're interested in

Du kan så skrive et vilkårligt ord, du er interesseret i

and see its n-gram immediately --

og straks se det tilhørende n-gram,

also browse examples of all the various books

og du kan gennemse eksempler på alle bøger

in which your n-gram appears.

som dit n-gram optræder i.

JM: Now this was used over a million times on the first day,

Dette blev brugt over en million gang første dag,

and this is really the best of all the queries.

og dette er den bedste af alle søgninger.

So people want to be their best, put their best foot forward.

Så folk ønsker at yde deres bedste.

But it turns out in the 18th century, people didn't really care about that at all.

Men i det 18. årh. var folk ligeglade med alt det.

They didn't want to be their best, they wanted to be their beft.

De ville ikke gøre bedste, de ville være "beft".

So what happened is, of course, this is just a mistake.

Dette var selvfølgelig bare en fejl.

It's not that strove for mediocrity,

Man stræbte ikke efter middelmådighed,

it's just that the S used to be written differently, kind of like an F.

men tidligere skrev man S anderledes, nærmest som et f.

Now of course, Google didn't pick this up at the time,

Det opdagede Google selvfølgelig ikke dengang,

so we reported this in the science article that we wrote.

så vi skrev det i den videnskabelige artikel.

But it turns out this is just a reminder

Dette minder os om, at

that, although this is a lot of fun,

selvom det er rigtig sjovt,

when you interpret these graphs, you have to be very careful,

at fortolke disse grafer, skal man være forsigtig

and you have to adopt the base standards in the sciences.

og overholde de videnskabelige standarder.

ELA: People have been using this for all kinds of fun purposes.

Folk har brugt dette til mange sjove formål.

(Laughter)

(Latter)

Actually, we're not going to have to talk,

Vi behøver faktisk ikke tale,

we're just going to show you all the slides and remain silent.

vi viser bare alle slides og tier stille.

This person was interested in the history of frustration.

Denne person var interesseret i frustrationens historie.

There's various types of frustration.

Der er forskellige typer frustration.

If you stub your toe, that's a one A "argh."

Hvis slår tåen, er der ét A i "argh".

If the planet Earth is annihilated by the Vogons

Hvis Jorden udslettes af Vogonerne

to make room for an interstellar bypass,

for at gøre plads til en intergalaktisk ekspresrute,

that's an eight A "aaaaaaaargh."

er det et "aaaaaaaargh" med otte A'er.

This person studies all the "arghs,"

Personen undersøger alle udgaver af "argh"

from one through eight A's.

fra ét til otte A'er.

And it turns out

Og det viser sig

that the less-frequent "arghs"

at de mindst hyppige "argh" vedrører

are, of course, the ones that correspond to things that are more frustrating --

vedrører ting, der er mere frustrerende

except, oddly, in the early 80s.

men sjovt nok ikke i de tidlige 80'ere.

We think that might have something to do with Reagan.

Vi tror det kan være noget med Reagan.

(Laughter)

(Latter)

JM: There are many usages of this data,

Disse data kan bruges til mange ting,

but the bottom line is that the historical record is being digitized.

men grundlaget er, at historien bliver digitaliseret.

Google has started to digitize 15 million books.

Google er begyndt at digitalisere 15 millioner bøger.

That's 12 percent of all the books that have ever been published.

Det er 12 % af alle bøger, der er udgivet.

It's a sizable chunk of human culture.

Det er en god klump af menneskets kultur.

There's much more in culture: there's manuscripts, there newspapers,

Kultur er meget mere: manuskripter, aviser

there's things that are not text, like art and paintings.

noget er ikke tekst, f.eks. kunst og malerier.

These all happen to be on our computers,

Disse vil alle findes på vores computere,

on computers across the world.

på computere i hele verden.

And when that happens, that will transform the way we have

Og når det sker, ændrer det den måde

to understand our past, our present and human culture.

vi forstår vores fortid, vores nutid og menneskets kultur.

Thank you very much.

Mange tak.

(Applause)

(Bifald)