Massive-scale online collaboration | Luis von Ahn

Prevoditelj: Silvija Florijan Recezent: Tilen Pigac - EFZG

How many of you had to fill out a web form

Koliko vas je moralo ispuniti neku vrstu mrežnog obrasca

where you've been asked to read

gdje su vas tražili da čitate iskrivljeni niz znakova poput ovog?

a distorted sequence of characters like this?

Koliko vas smatra ovo stvarno, stvarno dosadnim?

How many of you found it really annoying?

U redu, izvanredno. Ja sam to izmislio.

(Laughter)

OK, outstanding. So I invented that.

(Smijeh)

(Laughter)

Ili sam bio jedan od ljudi koji su to napravili.

Or I was one of the people who did it.

Ta se stvar zove Captcha.

That thing is called a CAPTCHA.

I razlog zbog kojeg je tu jest da ste vi, entitet koji ispunjava obrazac,

And it is there to make sure you, the entity filling out the form,

doista čovjek, a ne neka vrsta kompjuterskog programa

are a human and not a computer program

napisana da podnosi obrazac milijune i milijune puta.

that was written to submit the form millions of times.

Ona djeluje zato što ljudi,

The reason it works is because humans, at least non-visually-impaired humans,

barem oni koji nemaju oštećen vid,

nemaju problema s čitanjem tih iskrivljenih vijugavih znakova,

have no trouble reading these distorted characters,

dok to kompjutorski programi jednostavno još ne mogu tako dobro.

whereas programs can't do it as well yet.

In the case of Ticketmaster,

Npr., u slučaju Ticketmastera,

the reason you have to type these characters

razlog zbog kojeg morate tipkati te iskrivljene znakove

is to prevent scalpers from writing a program

jest da se spriječe preprodavači da napišu program

that can buy millions of tickets, two at a time.

koji može kupiti milijune ulaznica, dvije odjednom.

CAPTCHAs are used all over the Internet.

Captche se koriste na cijelom Internetu.

And since they're used so often,

I budući da se koriste tako često,

a lot of times the sequence of random characters shown to the user

puno puta precizan niz nasumičnih znakova koje vidi korisnik

nije tako sretan.

is not so fortunate.

So this is an example from the Yahoo registration page.

Ovo je primjer Yahooove stranice za registraciju.

The random characters that happened to be shown to the user

Nasumični znakovi koji se pokazuju korisniku

bili su W, A, I, T, koji, naravno, sriču riječ (čekaj).

were W, A, I, T, which, of course, spell a word.

No najbolji dio je poruka

But the best part is the message

that the Yahoo help desk got about 20 minutes later.

koju je Yahooov odjel za pomoć dobio oko 20 minuta kasnije.

[Help! I've been waiting for over 20 minutes and nothing happens.]

Tekst: "Upomoć! Čekam već 20 minuta, i ništa se ne događa."

(Laughter)

(Smijeh)

This person thought they needed to wait.

Ovaj je čovjek mislio da mora čekati.

This, of course, is not as bad as this poor person.

Ovo, naravno, nije tako loše kao ovaj jadnik.

(Laughter)

(Smijeh)

CAPTCHA Project is something that we did at Carnegie Melllon over 10 years ago,

Projekt Captcha je nešto što smo napravili ovdje na sveučilištu Carnegie Mellon prije 10 godina,

i otada se koristi posvuda.

and it's been used everywhere.

Let me now tell you about a project that we did a few years later,

Sad ću govoriti o projektu koji smo napravili prije nekoliko godina,

nešto nalik sljedećoj evoluciji Captche.

which is sort of the next evolution of CAPTCHA.

Ovo je projekt koji nazivamo reCaptcha,

This is a project that we call reCAPTCHA,

što je nešto što smo započeli ovdje na Carnegie Mellonu,

which is something that we started here at Carnegie Mellon,

zatim smo to pretvorili u malu tvrtku.

then we turned it into a start-up company.

I nakon godine i pol,

And then about a year and a half ago, Google actually acquired this company.

Google je zapravo kupio ovu tvrtku.

Reći ću vam što je pokrenuo ovaj projekt.

Let me tell you what this project started.

Dakle, ovaj projekt je krenuo od sljedeće realizacije:

This project started from the following realization:

Ispada da otprilike 200 milijuna Captcha

It turns out that approximately 200 million CAPTCHAs

tipkaju ljudi svaki dan diljem svijeta.

are typed everyday by people around the world.

When I first heard this, I was quite proud of myself.

Kad sam prvi put čuo za to, bio sam poprilično ponosan na sebe.

Mislio sam, pogledaj na utjecaj koje je imalo moje istraživanje.

I thought, look at the impact my research has had.

Ali zatim sam se počeo osjećati loše.

But then I started feeling bad.

Evo o čemu se radi, svaki put kad tipkate Captchu,

Here's the thing: each time you type a CAPTCHA,

u biti gubite 10 sekundi svog vremena.

essentially, you waste 10 seconds of your time.

And if you multiply that by 200 million,

I ako to umnožite s 200 milijuna,

you get that humanity is wasting about 500,000 hours every day

dobivate da čovječanstvo u cijelom gubi oko 500.000 sati svaki dan

typing these annoying CAPTCHAs.

tipkajući te dosadne Captche.

(Laughter)

I tad sam se počeo osjećati loše.

So then I started feeling bad.

(Laughter)

(Smijeh)

And then I started thinking, of course, we can't just get rid of CAPTCHAs,

I tad sam počeo razmišljati, pa, naravno, ne možemo se samo riješiti Captchi,

zato što sigurnost Mreže na neki način ovisi o njima.

because the security of the web depends on them.

Ali zatim sam počeo razmišljati, postoji li bilo koji način na koji možemo upotrijebiti ovaj trud

But then I started thinking, can we use this effort

for something that is good for humanity?

za nešto što je dobro za čovječanstvo?

So see, here's the thing.

Gledajte, evo o čemu se radi.

While you're typing a CAPTCHA, during those 10 seconds,

Dok tipkate Captchu, tijekom tih 10 sekundi,

vaš mozak radi nešto nevjerojatno.

your brain is doing something amazing.

Your brain is doing something that computers cannot yet do.

Vaš mozak radi nešto što kompjuteri još ne mogu.

Pa, možete li napraviti koristan posao za tih 10 sekundi?

So can we get you to do useful work for those 10 seconds?

Drugim riječima,

Is there some humongous problem that we cannot yet get computers to solve,

postoji li neki grdan problem koji kompjuteri još ne mogu riješiti,

ali možemo podijeliti na sićušne dijelove od 10 sekundi

yet we can split into tiny 10-second chunks

such that each time somebody solves a CAPTCHA,

tako da svaki put kad netko rješava Captchu

riješi i mali dio ovog problema?

they solve a little bit of this problem?

I odgovor na to je "da", i to upravo radimo.

And the answer to that is "yes," and this is what we're doing now.

Ono što možda ne znate je da danas, dok tipkate Captchu,

Nowadays, while you're typing a CAPTCHA,

ne samo da se identificirate kao čovjek,

not only are you authenticating yourself as a human,

nego nam također zapravo pomažete digitalizirati knjige.

but in addition you're helping us to digitize books.

Objasnit ću vam kako to funkcionira.

Let me explain how this works.

Postoji puno projekata koji pokušavaju digitalizirati knjige.

There's a lot of projects trying to digitize books.

Google ima jedan. Internetski arhiv ima jedan.

Google has one. The Internet Archive has one.

Amazon, sad s Kindleom, pokušava digitalizirati knjige.

Amazon, with the Kindle, is trying to digitize books.

U osnovi to funkcionira na način da

Basically, the way this works is you start with an old book.

započinjete sa starom knjigom.

You've seen those things, right?

Vidjeli ste te stvari, točno? Poput knjige?

Like a book?

(Smijeh)

(Laughter)

So you start with a book and then you scan it.

Započinjete s knjigom, zatim ju skenirate.

Sad, skeniranje knjige je

Now, scanning a book

is like taking a digital photograph of every page.

kao da snimate digitalnu fotografiju svake stranice knjige.

Daje vam sliku za svaku stranicu knjige.

It gives you an image for every page.

Ovo je slika s tekstom za svaku stranicu knjige.

This is an image with text for every page of the book.

Slijedeći je korak u postupku

The next step in the process is that the computer needs to be able

taj da kompjuter mora biti kadar dešifrirati sve riječi u toj slici.

to decipher the words in this image.

Koristeći tehnologiju OCR,

That's using a technology called OCR, for optical character recognition,

optičko prepoznavanje znakova,

which takes a picture of text

koja snima sliku teksta

and tries to figure out what text is in there.

i pokušava odgonetnuti koji je tekst unutra.

Now, the problem is that OCR is not perfect.

Sad, OCR nije savršen.

Posebno za starije knjige

Especially for older books

where the ink has faded and the pages have turned yellow,

gdje je tinta izblijedjela i stranice su požutjele,

OCR cannot recognize a lot of the words.

OCR ne može prepoznati puno riječi.

For things that were written more than 50 years ago,

Npr., stvari koje su bile napisane prije više od 50 godina,

kompjuter ne može prepoznati oko 30 posto riječi.

the computer cannot recognize about 30 percent of the words.

Ono što sad radimo jest

So now we're taking all of the words that the computer cannot recognize

uzimanje svih riječi koje kompjuter ne može prepoznati

and we're getting people to read them for us

i davanje ljudima da ih čitaju za nas

while they're typing a CAPTCHA on the Internet.

dok tipkaju Captchu na Internetu.

Sljedeći put kad tipkate Captchu, te riječi koje tipkate

So the next time you type a CAPTCHA, these words that you're typing

su zapravo riječi koje dolaze iz knjiga koje se digitaliziraju

are actually words from books that are being digitized

that the computer could not recognize.

koje kompjuter nije mogao prepoznati.

The reason we have two words nowadays instead of one

I sad, razlog zbog kojeg danas imamo dvije riječi umjesto jedne

jest taj što je, vidite, jedna od riječi

is because one of the words

is a word that the system just got out of a book,

riječ koju je sustav tek dobio iz knjige,

nije znao što je, i namjerava je prikazati vama.

it didn't know what it was and it's going to present it to you.

Ali kako ne zna odgovor na nju, ne može ju ocijeniti za vas.

But since it doesn't know the answer, it cannot grade it.

Dakle, dajemo vam drugu riječ,

So we give you another word,

for which the system does know the answer.

za koju sustav zna odgovor.

We don't tell you which one's which and we say, please type both.

Ne govorimo vam koja je koja, i kažemo da utipkate obje.

I ako utipkate pravilnu riječ

And if you type the correct word

za onu koju sustav već zna odgovor,

for the one for which the system knows the answer,

pretpostavlja da ste čovjek,

it assumes you are human

and it also gets some confidence that you typed the other word correctly.

i također dobiva neku sigurnost da ste utipkali drugu riječ pravilno.

I ako ponovimo taj proces s 10 različitih ljudi

And if we repeat this process to 10 different people

i svi se slože oko toga što je nova riječ,

and they agree on what the new word is,

tad mi dobivamo još jednu riječ točno digitaliziranu.

then we get one more word digitized accurately.

Evo kako sustav funkcionira.

So this is how the system works.

I u osnovi, otkad smo ga objavili prije tri ili četiri godine,

And since we released it about three or four years ago,

puno se mrežnih mjesta počelo prebacivati

a lot of websites have started switching from the old CAPTCHA,

sa stare Captche gdje su ljudi gubili vrijeme

where people wasted their time,

na novu Captchu gdje ljudi pomažu digitalizirati knjige.

to the new CAPTCHA where people are helping to digitize books.

Npr., Ticketmaster.

So every time you buy tickets on Ticketmaster,

Svaki put kad kupite ulaznicu na Ticketmasteru, pomažete digitalizirati knjigu.

you help to digitize a book.

Facebook: Svaki put kad dodate prijatelja ili bocnete nekog,

Facebook: Every time you add a friend or poke somebody,

pomažete digitalizirati knjigu.

you help to digitize a book.

Twitter i oko 350.000 drugih mjesta koriste reCaptchu.

Twitter and about 350,000 other sites are all using reCAPTCHA.

I u stvari, broj mjesta koji koriste reCaptchu je tako visok

And the number of sites that are using reCAPTCHA is so high

da je broj riječi koje digitaliziramo na dan, stvarno, stvarno velik.

that the number of words we're digitizing per day is really large.

To je oko 100 milijuna na dan,

It's about 100 million a day,

što je ekvivalent oko dva i pol milijuna knjiga godišnje.

which is the equivalent of about two and a half million books a year.

I sve rade s jednom po jednom riječju

And this is all being done one word at a time

samo ljudi koji tipkaju Captchu na Internetu.

by just people typing CAPTCHAs on the Internet.

(Pljesak)

(Applause)

Now, of course,

Naravno,

since we're doing so many words per day,

budući da obrađujemo toliko riječi na dan,

mogu se dogoditi smiješne stvari.

funny things can happen.

This is especially true because now we're giving people

I to je posebno istinito zato što sada dajemo ljudima

dvije nasumično izabrane engleske riječi jednu pokraj druge.

two randomly chosen English words next to each other.

Pa se mogu dogoditi smiješne stvari.

So funny things can happen.

Npr., prikazali smo ovu riječ.

For example, we presented this word.

To je riječ "Christians" (kršćani); sve je u redu s njom.

It's the word "Christians"; there's nothing wrong with it.

Ali ako ju prikažete zajedno s drugom nasumično odabranom riječi,

But if you present it along with another randomly chosen word,

loše se stvari mogu dogoditi.

bad things can happen.

Pa dobivamo ovo. (Tekst: loši kršćani)

So we get this.

[bad Christians]

Ali još je gore, zato što se određena mrežna mjesta gdje smo ovo pokazali

But it's even worse, because the website where we showed this

zapravo nazivaju Poslanstvo Kraljevstva Božjeg.

actually happened to be called The Embassy of the Kingdom of God.

(Smijeh)

(Laughter)

Joj.

Oops.

(Laughter)

(Smijeh)

Evo još jednog stvarno lošeg.

Here's another really bad one.

JohnEdwards.com

JohnEdwards.com

[Damn liberal]

(Tekst: Damn liberal - prokleti liberal)

(Laughter)

(Smijeh)

Nastavljamo vrijeđati ljude na sve strane svaki dan.

So we keep on insulting people left and right everyday.

Naravno, ne vrijeđamo samo ljude.

Of course, we're not just insulting people.

Evo o čemu se radi, budući da prikazujemo dvije nasumično izabrane riječi,

Here's the thing. Since we're presenting two randomly chosen words,

zanimljive se stvari mogu dogoditi.

interesting things can happen.

So this actually has given rise to a really big Internet meme

Ovo je zapravo dovelo do

stvarno velikog internetskog fenomena

that tens of thousands of people have participated in,

u kojem su sudjelovali deseci tisuća ljudi,

koji se naziva umjetnost Captcha.

which is called CAPTCHA art.

I'm sure some of you have heard about it.

Siguran sam da su neki od vas čuli za to.

Here's how it works.

Evo kako funkcionira.

Imagine you're using the Internet and you see a CAPTCHA

Zamislite da ste na Internetu i vidite Captchu

that you think is somewhat peculiar,

za koju mislite da je pomalo neobična,

like this CAPTCHA.

kao ova Captcha. (Tekst: nevidljivi toster)

[invisible toaster]

What you're supposed to do is you take a screenshot of it.

Ono što bi trebali učiniti jest snimiti sadržaj ekrana.

Zatim, naravno, ispunite Captchu

Then of course, you fill out the CAPTCHA because you help us digitize a book.

zato što nam pomažete digitalizirati knjige.

Ali zatim, prvo snimite sadržaj ekrana,

But first you take a screenshot

and then you draw something that is related to it.

i zatim nacrtate nešto što ima veze s tim.

(Smijeh)

(Laughter)

That's how it works.

Tako to funkcionira.

(Laughter)

There are tens of thousands of these.

Postoje deseci tisuća takvih.

Some of them are very cute.

Neke od njih su stvarno ljupke. (Tekst: stisnuti ga)

[clenched it]

(Smijeh)

(Laughter)

Neke od njih su smiješnije.

Some of them are funnier.

(Tekst: napušeni osnivači)

[stoned Founders]

(Laughter)

(Smijeh)

And some of them, like paleontological shvisle ...

A neki od njih,

kao "paleontološki shvisle",

(Laughter)

imaju Snoop Dogga.

they contain Snoop Dogg.

(Laughter)

(Smijeh)

OK, so this is my favorite number of reCAPTCHA.

U redu, ovo je moj najdraži broj reCaptche.

Ovo je najdraža stvar koju volim u cijelom ovom projektu.

So this is the favorite thing that I like about this whole project.

Ovo je broj određenih ljudi

This is the number of distinct people

koji su nam pomogli digitalizirati najmanje jednu riječ iz knjige putem reCaptche.

that have helped us digitize at least one word out of a book through reCAPTCHA:

750 milijuna,

750 million, a little over 10 percent of the world's population,

što je malo preko 10 posto svjetske populacije,

has helped us digitize human knowledge.

pomoglo nam je digitalizirati ljudsko znanje.

And it is numbers like these that motivate my research agenda.

Upravo brojevi poput ovih motiviraju moju istraživačku agendu.

So the question that motivates my research is the following:

Dakle, pitanje koje motivira moje istraživanje glasi:

ako pogledate velika ljudska dostignuća,

If you look at humanity's large-scale achievements,

te doista velike stvari

these really big things

that humanity has gotten together and done historically --

koje je čovječanstvo steklo zajedno i učinilo povijesno --

like, for example, building the pyramids of Egypt

kao npr., gradnja egipatskih piramida

or the Panama Canal

ili Panamskog kanala

or putting a man on the Moon --

ili slanje čovjeka na Mjesec --

there is a curious fact about them,

zanimljiva je činjenica

and it is that they were all done with about the same number of people.

da su sve napravljene s otprilike istim brojem ljudi.

Čudno je; sve su napravljene s oko 100.000 ljudi.

It's weird; they were all done with about 100,000 people.

I razlog je tomu taj što je, prije Interneta,

And the reason for that is because, before the Internet,

koordinacija više od 100.000 ljudi,

coordinating more than 100,000 people,

a da i ne spominjem njihovo plaćanje, bilo u osnovi nemoguće.

let alone paying them, was essentially impossible.

Ali sad s Internetom, upravo sam vam pokazao projekt

But now with the Internet, I've just shown you a project

gdje nam je 750 milijuna ljudi

where we've gotten 750 million people to help us digitize human knowledge.

pomoglo digitalizirati ljudsko znanje.

Pa pitanje koje motivira moje istraživanje glasi,

So the question that motivates my research is,

ako možemo poslati čovjeka na Mjesec sa 100.000,

if we can put a man on the Moon with 100,000,

what can we do with 100 million?

što možemo učiniti sa 100 milijuna?

So based on this question,

Na temelju ovog pitanja,

we've had a lot of different projects that we've been working on.

imamo puno različitih projekata na kojima radimo.

Reći ću vam nešto o jednom zbog kojeg sam najviše uzbuđen.

Let me tell you about one that I'm most excited about.

This is something that we've been semiquietly working on

To je nešto na čemu smo napola tiho radili

zadnju godinu i pol.

for the last year and a half or so.

Još nije pokrenut. Naziva se Duolingo.

It hasn't yet been launched. It's called Duolingo.

Budući da još nije pokrenut, pssst!

Since it hasn't been launched, shhh!

(Smijeh)

(Laughter)

Yeah, I can trust you'll do that.

Da, vjerujem vam da ćete to učiniti.

Ovo je projekt. Evo kako je počeo.

So this is the project. Here's how it started.

Počeo je s pitanjem koje sam postavio svojem diplomiranom studentu,

It started with me posing a question to my graduate student, Severin Hacker.

Severinu Hackeru.

Ok, ovo je Severin Hacker.

OK, that's Severin Hacker.

Dakle, postavio sam pitanje svojem studentu.

So I posed the question to my graduate student.

Uzgred budi rečeno, dobro ste me čuli;

By the way, you did hear me correctly; his last name is Hacker.

njegovo je prezime Hacker.

(Laughter)

Postavio sam mu pitanje:

So I posed this question to him: How can we get 100 million people

Kako pridobiti 100 milijuna ljudi

translating the web into every major language for free?

da prevode Mrežu na sve veće jezike besplatno?

Ok, ima puno stvari koje se mogu reći o ovom pitanju.

There's a lot of things to say about this question.

Prvo, prevoditi Mrežu.

First of all, translating the web.

Trenutno je Mreža podijeljena na brojne jezike.

Right now, the web is partitioned into multiple languages.

Velik dio je na engleskom.

A large fraction of it is in English.

Ako imalo ne znaš engleski, ne možeš joj pristupiti.

If you don't know English, you can't access it.

Ali postoje veliki dijelovi na drugim jezicima,

But there's large fractions in other different languages,

i ako ne znaš te jezike, ne možeš joj pristupiti.

and if you don't know them, you can't access it.

Želio bih prevesti svu Mrežu, ili barem većinu Mreže,

So I would like to translate all of the web,

or at least most of it, into every major language.

na svaki veći jezik.

Dakle, to bih želio napraviti.

That's what I would like to do.

Now, some of you may say, why can't we use computers to translate?

Sad, neki od vas bi mogli reći, zašto ne možemo upotrijebiti kompjutore da prevode?

Zašto ne možemo upotrijebiti automatski prijevod?

Machine translation is starting to translate

Automatski prijevod danas počinje prevoditi tu i tamo neke rečenice.

some sentences here and there.

Zašto ga ne možemo upotrijebiti da prevedemo cijelu Mrežu?

Why can't we use it to translate the web?

Pa, problem je u tome da još uvijek nije dovoljno dobar

The problem with that is it's not yet good enough

i najvjerojatnije neće biti 15 do 20 godina.

and it probably won't be for the next 15 to 20 years.

Radi puno grešaka.

It makes a lot of mistakes. Even when it doesn't,

Čak i kad ne griješi,

since it makes so many mistakes, you don't know whether to trust it or not.

budući da radi tako puno grešaka, ne znate da li da mu vjerujete ili ne.

Pokazat ću vam jedan primjer

So let me show you an example

of something that was translated with a machine.

nečega što je bilo prevedeno sa strojem.

Zapravo je to bio post na forumu.

Actually, it was a forum post.

It was somebody who was trying to ask a question about JavaScript.

Netko je pokušavao postaviti pitanje o JavaScriptu.

It was translated from Japanese into English.

Preveden je s japanskog na engleski.

So I'll just let you read.

Pustit ću vas da pročitate.

This person starts apologizing

Ova osoba počinje se ispričavati

for the fact that it's translated with a computer.

zbog činjenice da je prevedeno kompjuterom.

So the next sentence is going to be the preamble to the question.

Pa će sljedeća rečenica biti uvod u pitanje.

On samo objašnjava nešto.

So he's just explaining something.

Zapamtite, ovo je pitanje o JavaScriptu.

Remember, it's a question about JavaScript.

[At often, the goat-time install a error is vomit.]

(Tekst: Na često, jarac-vrijeme instalira greška je izbljuvak.)

(Laughter)

(Smijeh)

Then comes the first part of the question.

Zatim dolazi prvi dio pitanja.

[How many times like the wind, a pole, and the dragon?]

(Tekst: Koliko puta kao vjetar, štap, i zmaj?)

(Laughter)

(Smijeh)

Zatim dolazi moj omiljeni dio pitanja.

Then comes my favorite part of the question.

[This insult to father's stones?]

(Tekst: Ovo uvreda očevom kamenju?)

(Laughter)

(Smijeh)

Zatim dolazi kraj, moj najdraži dio cijele stvari.

And then comes the ending,

which is my favorite part of the whole thing.

(Tekst: Molim ispričati se za vašu glupost. Ima mnogo hvala vam.)

[Please apologize for your stupidity. There are a many thank you.]

(Laughter)

(Smijeh)

OK, so computer translation, not yet good enough.

Ok, kompjuterski prijevod, još nije dovoljno dobar.

Nazad k pitanju.

So back to the question.

Trebamo ljude da prevode cijelu Mrežu.

So we need people to translate the whole web.

Dakle, sljedeće pitanje koje bi mogli imati glasi,

So now the next question you may have is,

pa zašto ne možemo samo platiti ljudima da to učine?

well, why can't we just pay people to do this?

Možemo platiti profesionalnim prevoditeljima da prevedu cijelu Mrežu.

We could pay professional translators to translate the whole web.

Mogli bi to učiniti.

We could do that.

Unfortunately, it would be extremely expensive.

Na žalost, bilo bi ekstremno skupo.

For example, translating a tiny fraction of the whole web, Wikipedia,

Npr., prijevod malenog, malenog dijela cijele Mreže, Wikipedie,

na jedan drugi jezik, španjolski.

into one other language, Spanish.

OK? Wikipedia exists in Spanish,

Wikipedia postoji na španjolskom,

but it's very small compared to the size of English.

ali je jako mala u usporedbi s veličinom na engleskom.

Iznosi oko 20 posto veličine na engleskom.

It's about 20 percent of the size of English.

Da smo željeli prevesti ostalih 80 posto na španjolski,

If we wanted to translate the other 80 percent into Spanish,

koštalo bi nas najmanje 50 milijuna dolara --

it would cost at least 50 million dollars --

i to kod najiskorištenije, najiznajmljivačke zemlje koja postoji.

and this is even at the most exploited, outsourcing country out there.

Bilo bi vrlo skupo.

So it would be very expensive.

Dakle, ono što želimo jest pridobiti 100 milijuna ljudi

So what we want to do is, we want to get 100 million people

da prevode Mrežu na svaki veći jezik

translating the web into every major language for free.

besplatno.

Ako je ovo to što želite učiniti,

If this is what you want to do, you quickly realize

vrlo brzo ćete shvatiti da ćete naletjeti na dvije prilično velike prepreke,

you're going to run into two big hurdles, two big obstacles.

dvije velike poteškoće.

The first one is a lack of bilinguals.

Prva je nedostatak bilingvista.

So I don't even know

Pa čak ni ne znam

if there exists 100 million people out there using the web

postoji li 100 milijuna ljudi koji koriste Mrežu

who are bilingual enough to help us translate.

koji su dovoljno bilingvalni da nam pomognu prevoditi.

To je velik problem.

That's a big problem.

The other problem you're going to run into is a lack of motivation.

Drugi problem na koji ćete naići jest nedostatak motivacije.

Kako ćemo motivirati ljude

How are we going to motivate people to actually translate the web for free?

da ustvari prevode Mrežu besplatno?

Inače moraš platiti ljudima da to naprave.

Normally, you have to pay people to do this.

So how are we going to motivate them to do it for free?

Kako ćemo ih motivirati da to naprave besplatno?

Kad smo započinjali razmišljati o tome, zaustavile su nas ove dvije stvari.

When we were starting to think about this, we were blocked by these two things.

Ali tad smo shvatili da ustvari postoji način

But then we realized, there's a way

da riješimo oba problema istom solucijom.

to solve both these problems with the same solution.

Postoji način da se ubiju dvije muhe jednim udarcem.

To kill two birds with one stone.

A to je da se transformira prevođenje

And that is to transform language translation

into something that millions of people want to do

na nešto što milijuni ljudi žele učiniti,

and that also helps with the problem of lack of bilinguals,

što također pomaže kod problema s nedostatkom bilingvista,

and that is language education.

a to je učenje jezika.

So it turns out that today,

Ispada da danas,

there are over 1.2 billion people learning a foreign language.

postoji preko 1,2 milijarde ljudi koji uče strani jezik.

Ljudi doista, doista žele učiti strani jezik.

People really want to learn a foreign language.

I to nije samo zato što ih se tjera na to u školi.

And it's not just because they're being forced to do so in school.

Npr., samo u Sjedinjenim Državama,

In the US alone, there are over five million people

ima preko pet milijuna ljudi koji su platili više od 500 dolara

who have paid over $500 for software to learn a new language.

za softver za učenje novog jezika.

Pa ljudi doista, doista žele naučiti novi jezik.

So people really want to learn a new language.

Dakle, ono na čemu smo radili zadnju godinu i pol je novo mrežno mjesto --

So what we've been working on for the last year and a half

Duolingo --

is a new website -- it's called Duolingo --

gdje je osnovna ideja da ljudi uče novi jezik besplatno

where the basic idea is people learn a new language for free

while simultaneously translating the web.

dok istovremeno prevode Mrežu.

I u osnovi uče dok prevode.

And so basically, they're learning by doing.

Način na koji ovo funkcionira

So the way this works

is whenever you're a just a beginner, we give you very simple sentences.

jest da kad god ste samo početnik, dajemo vam vrlo, vrlo jednostavne rečenice.

Postoji, naravno, puno vrlo jednostavnih rečenica na Mreži.

There's a lot of very simple sentences on the web.

Dajemo vam vrlo, vrlo jednostavne rečenice

We give you very simple sentences along with what each word means.

zajedno s onim što svaka riječ znači.

And as you translate them

I kako ih prevodite, i dok gledate kako ih drugi ljudi prevode,

and as you see how other people translate them,

počinjete učiti jezik.

you start learning the language.

I kako postajete sve više napredni,

And as you get more advanced,

we give you more complex sentences to translate.

dajemo vam sve složenije rečenice na prijevod.

But at all times, you're learning by doing.

Ali cijelo vrijeme, učite dok prevodite.

Now, the crazy thing about this method is that it actually really works.

Sad, luda stvar u svezi s ovom metodom

jest da ona doista funkcionira.

People are really learning a language.

Prvo, ljudi doista, doista uče jezik.

We're mostly done building it and now we're testing it.

Uglavnom smo završili s izradom, sada ga testiramo.

Ljudi doista mogu naučiti jezik s njim.

People really can learn a language with it.

I uče ga gotovo jednako dobro kao s vodećim softverom za učenje jezika.

And they learn it about as well as the leading language learning software.

Pa ljudi doista nauče jezik.

So people really do learn a language.

I ne samo da ga nauče jednako dobro,

And not only do they learn it as well, but actually it's more interesting.

zapravo je puno zanimljivije.

Zato što s Duolingom ljudi zapravo uče sa stvarnim sadržajem.

Because with Duolingo, people are learning with real content.

U suprotnosti s učenjem s izmišljenim rečenicama,

As opposed to learning with made-up sentences,

ljudi uče sa stvarnim sadržajem koji je po sebi zanimljiv.

people are learning with real content, which is inherently interesting.

Pa ljudi stvarno uče jezik.

So people really do learn a language.

Ali možda što je još veće iznenađenje,

But perhaps more surprisingly,

prijevodi koje dobivamo od ljudi koji koriste mjesto,

the translations that we get from people using the site,

even though they're just beginners,

iako su samo početnici,

the translations that we get

prijevodi koje dobivamo su jednako točni kao oni profesionalnih prevoditelja,

are as accurate as those of professional language translators,

što vrlo iznenađuje.

which is very surprising.

Pokazat će vam jedan primjer.

So let me show you one example.

This is a sentence that was translated from German into English.

Ovo je rečenica prevedena s njemačkog na engleski.

Na vrhu je njemački.

The top is the German. The middle is an English translation

U sredini je prijevod na engleski

that was done by a professional translator

profesionalnog prevoditelja engleskog jezika

who we paid 20 cents a word for this translation.

kojeg smo platili 20 centi za riječ za ovaj prijevod.

Na dnu je prijevod korisnika Duolinga,

And the bottom is a translation by users of Duolingo,

od kojih ni jedan nije znao njemački

none of whom knew any German before they started using the site.

prije nego što su počeli koristiti mjesto.

If you can see, it's pretty much perfect.

Možete vidjeti da je prilično savršen.

Of course, we play a trick here

Naravno, izveli smo trik

to make the translations as good as professional language translators.

da napravimo prijevode jednako dobrima kao one profesionalnih prevoditelja.

Kombinirali smo prijevode više početnika

We combine the translations of multiple beginners

da dobijemo kvalitetu jednog profesionalnog prevoditelja.

to get the quality of a single professional translator.

Now, even though we're combining the translations,

Iako kombiniramo prijevode,

the site actually can translate pretty fast.

mjesto zapravo može prevoditi prilično brzo.

Pokazat ću vam

So let me show you,

this is our estimates of how fast we could translate Wikipedia

koje su naše procjene koliko brzo možemo prevesti Wikipediu

s engleskog na španjolski.

from English into Spanish.

Remember, this is 50 million dollars' worth of value.

Zapamtite, to je vrijednost od 50 milijuna dolara.

So if we wanted to translate Wikipedia into Spanish,

Da smo htjeli prevesti Wikipediu na španjolski,

napravili bismo to u pet tjedana sa 100.000 aktivnih korisnika.

we could do it in five weeks with 100,000 active users.

I mogli smo to napraviti za otprilike 80 sati s milijun aktivnih korisnika.

And we could do it in about 80 hours with a million active users.

Budući da su svi projekti na kojima je radila moja grupa dosad pridobili milijune korisnika,

Since all the projects my group has worked on so far

have gotten millions of users,

nadamo se da ćemo biti u mogućnosti prevoditi

we're hopeful that we'll be able to translate extremely fast.

ekstremno brzo s ovim projektom.

Ono što me najviše uzbuđuju kod Duolinga

Now, the thing that I'm most excited about with Duolingo

jest to da omogućuje pošteni poslovni model za učenje jezika.

is I think this provides a fair business model for language education.

Evo o čemu se radi:

So here's the thing:

Sadašnji poslovni model za učenje jezika sastoji se u tome da

The current business model for language education

učenik plaća,

is the student pays,

and in particular, the student pays Rosetta Stone 500 dollars.

i posebno, učenik plaća Rosetti Stone 500 dolara.

(Smijeh)

(Laughter)

To je sadašnji poslovni model.

That's the current business model.

Problem je s tim poslovnim modelom u tome

The problem with this business model

da 95 posto svjetske populacije nema 500 dolara.

is that 95 percent of the world's population doesn't have 500 dollars.

To je krajnje nepošteno prema siromašnima.

So it's extremely unfair towards the poor.

Ovo je u potpunosti podređeno bogatima.

This is totally biased towards the rich.

Sad, u Duolingu,

Now, see, in Duolingo,

because while you learn, you're actually creating value,

zato što, dok učite

vi zapravo stvarate vrijednost, prevodite stvari --

you're translating stuff --

which, for example, we could charge somebody for translations,

za koje bi npr.mogli platiti nekoga za prijevod.

so this is how we could monetize this.

Ovo je način na koji bi to mogli unovčiti.

Since people are creating value while they're learning,

Budući da ljudi stvaraju vrijednost dok uče,

ne moraju plaćati novcem, oni plaćaju svojim vremenom.

they don't have to pay with their money, they pay with their time.

Magična je stvar u tome da plaćaju svojim vremenom,

But the magical thing here

is that is time that would have had to have been spent anyways

ali to je vrijeme koje bi ionako potrošili

učeći jezik.

learning the language.

So the nice thing about Duolingo

Dakle, lijepa stvar u svezi s Duolingom je u tom što, mislim, omogućuje pošten poslovni model --

is, I think, it provides a fair business model --

onaj koji ne diskriminira siromašne ljude.

one that doesn't discriminate against poor people.

Ovo je stranica. Hvala vam.

So here's the site. Thank you.

(Pljesak)

(Applause)

Ovo je stranica.

We haven't yet launched,

Još nije pokrenuto,

but if you go there, you can sign up to be part of our private beta,

ali ako odete tamo, možete se upisati da budete dio naše probne verzije,

which is probably going to start in three or four weeks.

koja će najvjerojatnije početi za tri do četiri tjedna.

Još uvijek nismo pokrenuli Duolingo.

We haven't yet launched it.

By the way, I'm the one talking here,

Uzgred budi rečeno, ja sam taj koji govori,

but Duolingo is the work of a really awesome team,

ali zapravo Duolingo je rad stvarno fenomenalnog tima, od kojeg su neki ovdje.

some of whom are here. So thank you.

Hvala vam.

(Applause)

(Pljesak)