We're building a dystopia just to make people click on ads | Zeynep Tufekci

Prevoditelj: Anja Kolobarić Recezent: Lucija Jelić

So when people voice fears of artificial intelligence,

Kad ljudi izraze svoje strahove o umjetnoj inteligenciji

very often, they invoke images of humanoid robots run amok.

često prizivaju slike humanoidnih robota koji divljaju.

You know? Terminator?

Znate? Terminator?

You know, that might be something to consider,

Znate, to je nešto što treba imati na umu,

but that's a distant threat.

ali to je daleka prijetnja.

Or, we fret about digital surveillance

Ili se sekiramo zbog digitalnog nadzora

with metaphors from the past.

s metaforama iz prošlosti.

"1984," George Orwell's "1984,"

"1984", Georgea Orwella, "1984"

it's hitting the bestseller lists again.

ponovno je jedna od najčitanijih knjiga.

It's a great book,

To je odlična knjiga,

but it's not the correct dystopia for the 21st century.

ali to nije točna distopija 21. stoljeća.

What we need to fear most

Ono čega se moramo najviše bojati

is not what artificial intelligence will do to us on its own,

nije ono što će nam umjetna inteligencija učiniti sama po sebi,

but how the people in power will use artificial intelligence

nego to kako će ljudi na vlasti koristiti umjetnu inteligenciju

to control us and to manipulate us

da nas kontroliraju i manipuliraju nama

in novel, sometimes hidden,

na nove, ponekad skrivene,

subtle and unexpected ways.

suptilne i neočekivane načine.

Much of the technology

Puno tehnologije

that threatens our freedom and our dignity in the near-term future

koja ugrožava našu slobodu i naše dostojanstvo u bliskoj budućnosti

is being developed by companies

razvijaju društva

in the business of capturing and selling our data and our attention

koja se bave prikupljanjem i prodajom naših podataka i naše pozornosti

to advertisers and others:

oglašivačima i ostalima:

Facebook, Google, Amazon,

Facebook, Google, Amazon,

Alibaba, Tencent.

Alibaba, Tencent.

Now, artificial intelligence has started bolstering their business as well.

Umjetna je inteligencija počela podupirati i njihovo poslovanje.

And it may seem like artificial intelligence

I može se činiti da umjetna inteligencija

is just the next thing after online ads.

slijedi nakon oglasa na internetu.

It's not.

Nije.

It's a jump in category.

To je posebna kategorija.

It's a whole different world,

To je jedan sasvim novi svijet

and it has great potential.

i ima velik potencijal.

It could accelerate our understanding of many areas of study and research.

Može ubrzati naše razumijevanje mnogih područja proučavanja i istraživanja,

But to paraphrase a famous Hollywood philosopher,

ali da parafraziramo poznatog holivudskog filozofa,

"With prodigious potential comes prodigious risk."

"S ogromnim potencijalom dolazi i ogroman rizik."

Now let's look at a basic fact of our digital lives, online ads.

Proučimo osnovnu činjenicu naših digitalnih života, oglase na internetu.

Right? We kind of dismiss them.

Nekako ih ignoriramo.

They seem crude, ineffective.

Čine se nedorađeno, neučinkovito.

We've all had the experience of being followed on the web

Svi smo se mi suočili sa situacijom da nas na internetu

by an ad based on something we searched or read.

prati oglas na temelju onoga što smo tražili ili čitali.

You know, you look up a pair of boots

Znate, tražite čizme

and for a week, those boots are following you around everywhere you go.

i tjedan dana vas te čizme slijede gdje god pošli.

Even after you succumb and buy them, they're still following you around.

Čak i nakon što podlegnete i kupite ih, i dalje vas prate.

We're kind of inured to that kind of basic, cheap manipulation.

Na neke smo načine oguglali na takvu vrstu jeftine manipulacije.

We roll our eyes and we think, "You know what? These things don't work."

Kolutamo očima i mislimo si, "Ma znate što? Takve stvari ne pale."

Except, online,

Osim što, na internetu,

the digital technologies are not just ads.

digitalne tehnologije nisu samo oglasi.

Now, to understand that, let's think of a physical world example.

Kako biste to razumjeli, proučimo primjer fizičkog svijeta.

You know how, at the checkout counters at supermarkets, near the cashier,

Znate kako u redovima u supermarketima, pored blagajni

there's candy and gum at the eye level of kids?

stoje slatkiši i žvakaće u visini očiju djece?

That's designed to make them whine at their parents

Osmišljeni su kako bi natjerali djecu

just as the parents are about to sort of check out.

da gnjave već izmorene roditelje.

Now, that's a persuasion architecture.

To je arhitektura uvjeravanja.

It's not nice, but it kind of works.

Nije lijepa, ali funkcionira.

That's why you see it in every supermarket.

Zato je možete vidjeti posvuda.

Now, in the physical world,

U fizičkom svijetu

such persuasion architectures are kind of limited,

te su arhitekture uvjeravanja ograničene

because you can only put so many things by the cashier. Right?

jer ne možete puno toga staviti pored blagajne. Zar ne?

And the candy and gum, it's the same for everyone,

Slatkiši i žvakaće isti su za sve

even though it mostly works

iako uglavnom funkcioniraju

only for people who have whiny little humans beside them.

samo na ljudima koji pored sebe imaju male gnjavatore.

In the physical world, we live with those limitations.

U fizičkom svijetu živimo s takvim ograničenjima.

In the digital world, though,

U digitalnom svijetu, doduše,

persuasion architectures can be built at the scale of billions

arhitekture uvjeravanja mogu se izgraditi na ljestvici od milijardi

and they can target, infer, understand

i mogu naciljati, zaključiti, razumjeti

and be deployed at individuals

te ih mogu koristiti pojedinci

one by one

jedan po jedan

by figuring out your weaknesses,

otkrivajući naše slabosti

and they can be sent to everyone's phone private screen,

i mogu se slati svima na privatne mobitele,

so it's not visible to us.

pa nam nije vidljivo.

And that's different.

A to je drugačije.

And that's just one of the basic things that artificial intelligence can do.

To je samo jedna od osnovnih stvari koje umjetna inteligencija može raditi.

Now, let's take an example.

Uzmimo primjer.

Let's say you want to sell plane tickets to Vegas. Right?

Recimo da želite prodati avionske karte za Vegas. Dobro?

So in the old world, you could think of some demographics to target

U starom svijetu mogli ste razmišljati o ciljanoj populaciji

based on experience and what you can guess.

ovisno o iskustvu i o tome što možete pretpostaviti.

You might try to advertise to, oh,

Mogli biste postaviti oglas za

men between the ages of 25 and 35,

muškarce između 25. i 35. godine

or people who have a high limit on their credit card,

ili za ljude koji imaju visok dozvoljeni minus na kartici

or retired couples. Right?

ili za umirovljene parove.

That's what you would do in the past.

To biste radili u prošlosti.

With big data and machine learning,

Zahvaljujući ogromnim količinama podataka i strojnom učenju,

that's not how it works anymore.

to više ne ide tako.

So to imagine that,

Kako biste to zamislili,

think of all the data that Facebook has on you:

razmišljajte o svim podacima koje Facebook ima o vama:

every status update you ever typed,

svaki status koji ste ikad napisali,

every Messenger conversation,

svaki razgovor na chatu,

every place you logged in from,

svako mjesto na kojem ste se logirali,

all your photographs that you uploaded there.

sve vaše fotografije koje ste ondje objavili.

If you start typing something and change your mind and delete it,

Ako počnete nešto tipkati, predomislite se i obrišete to,

Facebook keeps those and analyzes them, too.

Facebook će i to čuvati i analizirati.

Increasingly, it tries to match you with your offline data.

Sve vas više pokušava povezati s vašim izvanmrežnim podacima.

It also purchases a lot of data from data brokers.

Također kupuje puno podataka od posrednika podacima.

It could be everything from your financial records

Može kupiti bilo što od vaših financijskih izvještaja

to a good chunk of your browsing history.

do većeg dijela vaše povijesti pretraživanja.

Right? In the US, such data is routinely collected,

U SAD-u se takvi podaci rutinski prikupljaju,

collated and sold.

razvrstavaju i prodaju.

In Europe, they have tougher rules.

U Europi su pravila stroža.

So what happens then is,

Događa se to

by churning through all that data, these machine-learning algorithms --

da prekopavanjem svih tih podataka ti algoritmi za strojno učenje --

that's why they're called learning algorithms --

zato se zovu algoritmi učenja -

they learn to understand the characteristics of people

uče razumijevati karakteristike ljudi

who purchased tickets to Vegas before.

koji su prije kupili karte za Vegas.

When they learn this from existing data,

Kad to nauče iz postojećih podataka,

they also learn how to apply this to new people.

također nauče kako to primjenjivati na nove ljude.

So if they're presented with a new person,

Pa ako se nađu pred novom osobom,

they can classify whether that person is likely to buy a ticket to Vegas or not.

mogu odrediti hoće li ta osoba kupiti kartu za Vegas ili ne.

Fine. You're thinking, an offer to buy tickets to Vegas.

Dobro. Mislite, ponuda da kupim karte za Vegas.

I can ignore that.

Mogu to ignorirati.

But the problem isn't that.

Ali nije problem u tome.

The problem is,

Problem je u tome

we no longer really understand how these complex algorithms work.

što više ne shvaćamo kako ti složeni algoritmi funkcioniraju.

We don't understand how they're doing this categorization.

Ne razumijemo kako rade tu kategorizaciju.

It's giant matrices, thousands of rows and columns,

Njegove ogromne matrice, tisuće redova i stupaca,

maybe millions of rows and columns,

možda milijune redova i stupaca,

and not the programmers

a ni programeri,

and not anybody who looks at it,

a ni bilo tko tko to pogleda,

even if you have all the data,

čak i ako imate sve podatke,

understands anymore how exactly it's operating

više ne razumije kako to točno funkcionira

any more than you'd know what I was thinking right now

bolje nego što bi vi znali o čemu ja upravo razmišljam

if you were shown a cross section of my brain.

da vam pokažu presjek mog mozga.

It's like we're not programming anymore,

Kao da više ne programiramo,

we're growing intelligence that we don't truly understand.

već uzgajamo inteligenciju koju baš i ne razumijemo.

And these things only work if there's an enormous amount of data,

A takve stvari funkcioniraju samo kada imamo ogromne količine podataka,

so they also encourage deep surveillance on all of us

pa također potiču na značajan nadzor svih nas

so that the machine learning algorithms can work.

da algoritmi za strojno učenje mogu funkcionirati.

That's why Facebook wants to collect all the data it can about you.

Zato Facebook želi prikupiti što više podataka o vama.

The algorithms work better.

Algoritmi bolje funkcioniraju.

So let's push that Vegas example a bit.

Pozabavimo se još malo primjerom Vegasa.

What if the system that we do not understand

Što ako je sustav koji ne razumijemo

was picking up that it's easier to sell Vegas tickets

shvatio da bi bilo lakše prodati karte za Vegas

to people who are bipolar and about to enter the manic phase.

bipolarnim osobama koje upravo ulaze u fazu manije?

Such people tend to become overspenders, compulsive gamblers.

Takvi ljudi prekomjerno troše i kompulzivno kockaju.

They could do this, and you'd have no clue that's what they were picking up on.

Mogli bi to napraviti, a vi ne biste imali pojma da tako djeluju.

I gave this example to a bunch of computer scientists once

Ovaj sam primjer jednom dala većem broju kompjutorskih znanstvenika

and afterwards, one of them came up to me.

i nakon toga mi je jedan od njih prišao.

He was troubled and he said, "That's why I couldn't publish it."

Bio je uznemiren i rekao je, "Zato to nisam mogao objaviti."

I was like, "Couldn't publish what?"

Rekoh "Što to?"

He had tried to see whether you can indeed figure out the onset of mania

Pokušao je vidjeti može li se prepoznati početak faze manije

from social media posts before clinical symptoms,

iz objava na društvenim mrežama prije kliničkih znakova,

and it had worked,

i to je funkcioniralo,

and it had worked very well,

i to jako dobro,

and he had no idea how it worked or what it was picking up on.

a nije imao pojma kako je funkcioniralo ili na što je reagiralo.

Now, the problem isn't solved if he doesn't publish it,

Ali problem nije riješen ako to ne objavi

because there are already companies

jer već postoje tvrtke

that are developing this kind of technology,

koje razvijaju ovakvu vrstu tehnologije,

and a lot of the stuff is just off the shelf.

a puno je stvari raspoloživo.

This is not very difficult anymore.

To više nije teško.

Do you ever go on YouTube meaning to watch one video

Jeste li ikada otišli na YouTube s namjerom da pogledate jedan video

and an hour later you've watched 27?

i sat vremena kasnije pogledali ste ih 27?

You know how YouTube has this column on the right

Znate kako YouTube s desne strane ime stupac

that says, "Up next"

"Sljedeće"

and it autoplays something?

i automatski reproducira nešto?

It's an algorithm

To je algoritam

picking what it thinks that you might be interested in

koji bira ono što misli da biste vi htjeli pogledati,

and maybe not find on your own.

a što ne biste uspjeli naći sami.

It's not a human editor.

To nije ljudski uređivač.

It's what algorithms do.

Algoritmi tako funkcioniraju.

It picks up on what you have watched and what people like you have watched,

Prati što ste gledali i što su gledali ljudi poput vas

and infers that that must be what you're interested in,

i zaključuje da i vas to vjerojatno zanima,

what you want more of,

da želite još toga,

and just shows you more.

pa vam pokazuje više toga.

It sounds like a benign and useful feature,

Zvuči kao nevina i korisna značajka,

except when it isn't.

ali nije.

So in 2016, I attended rallies of then-candidate Donald Trump

Godine 2016. bila sam na skupu tadašnjeg kandidata Donalda Trumpa

to study as a scholar the movement supporting him.

kako bih kao naučnik proučavala pokret koji ga podržava.

I study social movements, so I was studying it, too.

Proučavam društvene pokrete, pa sam i to proučavala.

And then I wanted to write something about one of his rallies,

Tada sam htjela napisati nešto o jednom od njegovih skupova,

so I watched it a few times on YouTube.

pa sam ga pogledala nekoliko puta na YT-u.

YouTube started recommending to me

YouTube mi je počeo preporučivati

and autoplaying to me white supremacist videos

i automatski puštati videe o nadmoći bijele rase

in increasing order of extremism.

koji su išli prema ekstremizmu.

If I watched one,

Ako bih pogledala jedan,

it served up one even more extreme

ponudio bi mi jedan još ekstremniji

and autoplayed that one, too.

i automatski ga puštao.

If you watch Hillary Clinton or Bernie Sanders content,

Ako gledate sadržaje Hillary Clinton ili Bernija Sandersa,

YouTube recommends and autoplays conspiracy left,

Youtube vam preporučuje i pušta zavjeru ljevice

and it goes downhill from there.

i od toga ide samo nagore.

Well, you might be thinking, this is politics, but it's not.

Možda mislite da je to politika, ali nije.

This isn't about politics.

Tu se ne radi o politici.

This is just the algorithm figuring out human behavior.

Ovo je samo algoritam koji tumači ljudsko ponašanje.

I once watched a video about vegetarianism on YouTube

Jednom sam pogledala video o vegetarijanstvu

and YouTube recommended and autoplayed a video about being vegan.

i Youtube mi je preporučio i pustio video o veganstvu.

It's like you're never hardcore enough for YouTube.

Kad da nikad niste dovoljno žestoki zagovornici ičega.

(Laughter)

(Smijeh)

So what's going on?

Što se događa?

Now, YouTube's algorithm is proprietary,

YouTubeov algoritam je privatan,

but here's what I think is going on.

ali mislim da se ovo događa.

The algorithm has figured out

Algoritam je shvatio

that if you can entice people

da ako možete navesti ljude

into thinking that you can show them something more hardcore,

da misle da im možete pokazati nešto još žešće,

they're more likely to stay on the site

skloniji su tome da ostanu na stranici

watching video after video going down that rabbit hole

gledajući video za videom ulazeći sve dublje u rupu bez dna

while Google serves them ads.

dok im Google pokazuje oglase.

Now, with nobody minding the ethics of the store,

Dok nikome na pazi na etničku strukturu u trgovini,

these sites can profile people

ove stranice mogu isprofilirati ljude

who are Jew haters,

koji su mrzitelji židova,

who think that Jews are parasites

koji misle da su židovi paraziti

and who have such explicit anti-Semitic content,

i koji imaju takav eksplicitan antisemitski sadržaj

and let you target them with ads.

i dati vam da oglase usmjerite prema njima.

They can also mobilize algorithms

Također mogu mobilizirati algoritme

to find for you look-alike audiences,

kako bi pronašli publiku sličnu vama,

people who do not have such explicit anti-Semitic content on their profile

ljude koji nemaju takav eksplicitni antisemitski sadržaj na svom profilu,

but who the algorithm detects may be susceptible to such messages,

ali koje algoritam prepoznaje kao osobe koje su osjetljive na takve porukama

and lets you target them with ads, too.

i dopušta da na njih usmjeravate oglase.

Now, this may sound like an implausible example,

Ovo će možda zvučati kao nevjerojatan primjer,

but this is real.

ali stvaran je.

ProPublica investigated this

ProPublica je to istraživala

and found that you can indeed do this on Facebook,

i otkrila je da to stvarno možete raditi na Facebooku

and Facebook helpfully offered up suggestions

i Facebook je dobronamjerno ponudio prijedloge

on how to broaden that audience.

kako proširiti takvu publiku.

BuzzFeed tried it for Google, and very quickly they found,

BuzzFeed je to isprobao na Googleu i brzo su otkrili

yep, you can do it on Google, too.

da to možete raditi i na Googleu.

And it wasn't even expensive.

Čak nije bilo ni skupo.

The ProPublica reporter spent about 30 dollars

Reporter ProPublice potrošio je 30-ak dolara

to target this category.

kako bi se oglašavao ovoj kategoriji.

So last year, Donald Trump's social media manager disclosed

Prošle godine upravitelj društvenih medija Donalda Trumpa priopćio je

that they were using Facebook dark posts to demobilize people,

da su koristili tamne postove na Facebooku kako bi demobilizirali ljude,

not to persuade them,

ne kako bi ih nagovorili,

but to convince them not to vote at all.

već kako bi ih uvjerili da uopće ne glasaju.

And to do that, they targeted specifically,

Kako bi to učinili, posebno su se usmjerili na

for example, African-American men in key cities like Philadelphia,

npr. afro-amerikance u ključnim gradovima poput Philadelphije

and I'm going to read exactly what he said.

i pročitat ću što je točno rekao.

I'm quoting.

Citiram.

They were using "nonpublic posts

Koristili su "privatne objave

whose viewership the campaign controls

čiju publiku određuje kampanja

so that only the people we want to see it see it.

kako bi ih vidjeli samo ljudi za koje želimo da ih vide.

We modeled this.

Modelirali smo to.

It will dramatically affect her ability to turn these people out."

Dramatično će utjecati na sposobnost da te ljude isključi."

What's in those dark posts?

Što je u tim tamnim postovima?

We have no idea.

Nemamo pojma.

Facebook won't tell us.

Facebook nam neće reći.

So Facebook also algorithmically arranges the posts

Facebook također putem algoritma razvrstava objave

that your friends put on Facebook, or the pages you follow.

koje vaši prijatelji stavljaju na Facebook, ili stranice koje pratite.

It doesn't show you everything chronologically.

Ne pokazuje vam sve kronološkim slijedom.

It puts the order in the way that the algorithm thinks will entice you

Redoslijed je takav da vas algoritam želi namamiti

to stay on the site longer.

da duže ostanete na stranici.

Now, so this has a lot of consequences.

To ima jako puno posljedica.

You may be thinking somebody is snubbing you on Facebook.

Možete misliti da vas netko ignorira na Facebooku.

The algorithm may never be showing your post to them.

Algoritam možda nikada ne pokaže vašu objavu njima.

The algorithm is prioritizing some of them and burying the others.

Algoritam stavlja prioritet na neke od njih, a druge zakopava.

Experiments show

Eksperimenti pokazuju

that what the algorithm picks to show you can affect your emotions.

da ono što vam algoritam odluči pokazati može utjecati na vaše emocije.

But that's not all.

Ali to nije sve.

It also affects political behavior.

Također utječe na vaše političko ponašanje.

So in 2010, in the midterm elections,

Godine 2010. u prvom dijelu izbora,

Facebook did an experiment on 61 million people in the US

Facebook je proveo eksperiment na 61 milijunu ljudi u SAD-u

that was disclosed after the fact.

o kojem su izvijestili kad je završio.

So some people were shown, "Today is election day,"

Nekim su ljudima pokazivali "Danas su izbori",

the simpler one,

jednostavniju poruku,

and some people were shown the one with that tiny tweak

a nekima su pokazivali neznatno izmijenjenu poruku

with those little thumbnails

sa sličicama

of your friends who clicked on "I voted."

vaših prijatelja koji su kliknuli na "Glasao sam".

This simple tweak.

Ta jednostavna izmjena.

OK? So the pictures were the only change,

Dobro? Slike su bile jedina izmjena,

and that post shown just once

a to što je ta objava pokazana samo jednom

turned out an additional 340,000 voters

potaknula je dodatnih 340.000 birača

in that election,

da glasuju na tim izborima,

according to this research

prema ovom istraživanju,

as confirmed by the voter rolls.

što je potvrđeno popisom birača.

A fluke? No.

Slučajnost? Ne.

Because in 2012, they repeated the same experiment.

Godine 2012. ponovili su taj eksperiment.

And that time,

I tada

that civic message shown just once

građanska poruka koja je prikazana samo jednom

turned out an additional 270,000 voters.

potaknula je dodatnih 270.000 birača.

For reference, the 2016 US presidential election

Za usporedbu, predsjedničke izbore 2016. u SAD-u

was decided by about 100,000 votes.

razriješilo je 100.000 glasova.

Now, Facebook can also very easily infer what your politics are,

Facebook također može vrlo lako zaključiti vaše političko opredjeljenje,

even if you've never disclosed them on the site.

čak i ako ga nikad niste objavili na stranici.

Right? These algorithms can do that quite easily.

Ti algoritmi to mogu odraditi vrlo lako.

What if a platform with that kind of power

Što ako platforma s takvom moći

decides to turn out supporters of one candidate over the other?

odluči okrenuti pristalice jednog kandidata na štetu drugoga?

How would we even know about it?

Kako bismo uopće znali da se to dogodilo?

Now, we started from someplace seemingly innocuous --

Počeli smo od nečega naizgled bezopasnog --

online adds following us around --

oglasi na internetu što nas slijede okolo --

and we've landed someplace else.

i došli na jedno sasvim drugo mjesto.

As a public and as citizens,

Kao javnost i kao građani,

we no longer know if we're seeing the same information

više ne znamo gledamo li iste informacije

or what anybody else is seeing,

ili ono što svi ostali gledaju

and without a common basis of information,

i bez zajedničke baze podataka,

little by little,

malo po malo,

public debate is becoming impossible,

javne debate postaju nemoguće

and we're just at the beginning stages of this.

i opet smo na početku.

These algorithms can quite easily infer

Ovi algoritmi mogu vrlo lako zaključiti

things like your people's ethnicity,

o stvarima poput etničkog podrijetla,

religious and political views, personality traits,

religijskim, političkim nazorima, crtama ličnosti,

intelligence, happiness, use of addictive substances,

inteligenciji, sreći, korištenju tvari koje izazivaju ovisnost,

parental separation, age and genders,

rastavljenosti roditelja, dobi i spolu,

just from Facebook likes.

i to sve na osnovu lajkova na Facebooku.

These algorithms can identify protesters

Ti algoritmi mogu prepoznati prosvjednike,

even if their faces are partially concealed.

čak i ako su njihova lica djelomično skrivena.

These algorithms may be able to detect people's sexual orientation

Ti algoritmi mogu prepoznati seksualnu orijentaciju ljudi

just from their dating profile pictures.

prema njihovim slikama na stranicama za pronalaženje partnera.

Now, these are probabilistic guesses,

To su probabilističke pretpostavke,

so they're not going to be 100 percent right,

pa neće biti 100% istinite,

but I don't see the powerful resisting the temptation to use these technologies

ali ne vidim snažno opiranje iskušenjima da se koriste te tehnologije

just because there are some false positives,

jer postoje neki lažno pozitivni nalazi

which will of course create a whole other layer of problems.

koji će, naravno, stvoriti cijeli niz problema.

Imagine what a state can do

Zamislite što država može napraviti

with the immense amount of data it has on its citizens.

s ogromnim količinama podataka koje ima o svojim građanima.

China is already using face detection technology

Kina već koristi tehnologiju prepoznavanja lica

to identify and arrest people.

kako bi identificirala i uhitila ljude.

And here's the tragedy:

A u ovome je tragedija:

we're building this infrastructure of surveillance authoritarianism

gradimo ovu infrastrukturu nadgledačkog autoritarizma

merely to get people to click on ads.

samo kako bi ljudi kliknuli na oglase.

And this won't be Orwell's authoritarianism.

Ovo neće biti Orwellov autoritarizam.

This isn't "1984."

Ovo nije "1984".

Now, if authoritarianism is using overt fear to terrorize us,

Ako autoritarizam koristi strah kako bi nas terorizirao,

we'll all be scared, but we'll know it,

svi ćemo biti u strahu, ali znat ćemo to,

we'll hate it and we'll resist it.

mrzit ćemo to i opirat ćemo se tome.

But if the people in power are using these algorithms

Ali ako oni koji su na vlasti koriste ove algoritme

to quietly watch us,

kako bi nas tiho nadgledali,

to judge us and to nudge us,

kako bi nas procjenjivali i gurkali,

to predict and identify the troublemakers and the rebels,

kako bi predvidjeli i prepoznali one koji prave probleme i buntovnike,

to deploy persuasion architectures at scale

kako bi razvili arhitekturu uvjeravanja u stvarnom životu

and to manipulate individuals one by one

i kako bi manipulirali pojedincima

using their personal, individual weaknesses and vulnerabilities,

koristeći osobne, individualne slabosti i osjetljivosti,

and if they're doing it at scale

i ako to rade u velikom broju

through our private screens

kroz naše privatne ekrane

so that we don't even know

pa čak ni ne znamo

what our fellow citizens and neighbors are seeing,

što naši sugrađani i susjedi vide,

that authoritarianism will envelop us like a spider's web

taj će se autoritarizam omotati oko nas kao paukova mreža,

and we may not even know we're in it.

a mi nećemo ni znati da smo u njoj.

So Facebook's market capitalization

Facebookova tržišna kapitalizacija

is approaching half a trillion dollars.

iznosi pola trilijuna dolara

It's because it works great as a persuasion architecture.

jer odlično funkcionira kao arhitektura uvjeravanja.

But the structure of that architecture

Ali struktura te arhitekture

is the same whether you're selling shoes

ista je bez obzira na to prodajete li cipele

or whether you're selling politics.

ili politiku.

The algorithms do not know the difference.

Algoritmi tu ne vide razliku.

The same algorithms set loose upon us

Ti isti algoritmi usmjereni na nas

to make us more pliable for ads

kako bi nas napravili podložnijima oglasima

are also organizing our political, personal and social information flows,

također organiziraju naše političke, osobne i društvene tokove informacija,

and that's what's got to change.

a to se mora promijeniti.

Now, don't get me wrong,

Nemojte me krivo shvatiti,

we use digital platforms because they provide us with great value.

koristimo digitalne platforme jer nam puno toga pružaju.

I use Facebook to keep in touch with friends and family around the world.

Koristim Face da budem u kontaktu s prijateljima i obitelji diljem svijeta.

I've written about how crucial social media is for social movements.

Pisala sam o tome koliko su ključni društveni mediji za društvene pokrete.

I have studied how these technologies can be used

Proučavala sam kako se te tehnologije mogu koristiti

to circumvent censorship around the world.

da zaobiđemo cenzuru diljem svijeta.

But it's not that the people who run, you know, Facebook or Google

Ali ljudi koji vode Facebook ili Google

are maliciously and deliberately trying

ne pokušavaju zlobno ili namjerno

to make the country or the world more polarized

državu ili svijet dodatno polarizirati

and encourage extremism.

i poticati ekstremizam.

I read the many well-intentioned statements

Pročitala sam mnoge dobronamjerne izjave

that these people put out.

koje ti ljudi daju.

But it's not the intent or the statements people in technology make that matter,

No, nije važna namjera ili izjave ljudi u tehnologiji,

it's the structures and business models they're building.

već su važne strukture i poslovni modeli koje oni grade.

And that's the core of the problem.

To je srž problema.

Either Facebook is a giant con of half a trillion dollars

Ili je Facebook ogromni prevarant s pola trilijuna dolara

and ads don't work on the site,

i oglasi na toj stranici ne rade

it doesn't work as a persuasion architecture,

kao arhitektura uvjeravanja,

or its power of influence is of great concern.

ili je njegova moć utjecaja razlog za brigu.

It's either one or the other.

Ili jedno ili drugo.

It's similar for Google, too.

Slično vrijedi i za Google.

So what can we do?

Što možemo učiniti?

This needs to change.

To se mora promijeniti.

Now, I can't offer a simple recipe,

Ne mogu ponuditi jednostavan recept

because we need to restructure

jer moramo restrukturirati

the whole way our digital technology operates.

cjelokupan način na koji naša digitalna tehnologija funkcionira.

Everything from the way technology is developed

Sve od toga kako se tehnologija razvija

to the way the incentives, economic and otherwise,

pa do načina na koji su poticaji, ekonomski ili drugi,

are built into the system.

ugrađeni u sustav.

We have to face and try to deal with

Moramo se suočiti i pokušati nositi s

the lack of transparency created by the proprietary algorithms,

manjkom transparentnosti koju stvaraju privatni algoritmi,

the structural challenge of machine learning's opacity,

strukturalnim izazovom nejasnoće strojnog učenja,

all this indiscriminate data that's being collected about us.

svim tim podacima koji se o nama skupljaju.

We have a big task in front of us.

Pred nama je velik zadatak.

We have to mobilize our technology,

Moramo mobilizirati našu tehnologiju,

our creativity

našu kreativnost,

and yes, our politics

i da, i našu politiku

so that we can build artificial intelligence

kako bismo stvorili umjetnu inteligenciju

that supports us in our human goals

koja podržava naše ljudske ciljeve,

but that is also constrained by our human values.

ali koju ograničavaju naše ljudske vrijednosti.

And I understand this won't be easy.

Shvaćam da to neće biti lako.

We might not even easily agree on what those terms mean.

Možda se nećemo čak ni složiti o tome koji bi to uvjeti trebali biti.

But if we take seriously

Ali ako ozbiljno shvatimo

how these systems that we depend on for so much operate,

kako funkcioniraju ovi sustavi o kojima toliko dugo ovisimo,

I don't see how we can postpone this conversation anymore.

ne znam kako možemo odgađati ovaj razgovor.

These structures

Ove strukture

are organizing how we function

organiziraju način na koji funkcioniramo

and they're controlling

i kontroliraju

what we can and we cannot do.

što možemo i što ne možemo raditi.

And many of these ad-financed platforms,

Mnoge od tih platformi koje financiraju oglasi

they boast that they're free.

naglašavaju da su besplatne.

In this context, it means that we are the product that's being sold.

U ovom kontekstu, to znači da smo mi proizvod koji se prodaje.

We need a digital economy

Potrebna nam je digitalna ekonomija

where our data and our attention

gdje naši podaci i pozornost

is not for sale to the highest-bidding authoritarian or demagogue.

neće biti prodani diktatoru ili demagogiji koja ponudi najviše.

(Applause)

(Pljesak)

So to go back to that Hollywood paraphrase,

Vratimo se onoj holivudskoj parafrazi,

we do want the prodigious potential

želimo da ogroman potencijal

of artificial intelligence and digital technology to blossom,

umjetne inteligencije i digitalne tehnologije procvjeta,

but for that, we must face this prodigious menace,

ali zbog toga se moramo suočiti s ovom ogromnom prijetnjom,

open-eyed and now.

otvorenih očiju, i to odmah.

Thank you.

Hvala vam.

(Applause)

(Pljesak)