The jobs we'll lose to machines -- and the ones we won't | Anthony Goldbloom

Prevodilac: Milenka Okuka Lektor: Mile Živković

So this is my niece.

Dakle, ovo je moja nećaka.

Her name is Yahli.

Zove se Jali.

She is nine months old.

Stara je devet meseci.

Her mum is a doctor, and her dad is a lawyer.

Njena majka je doktor, a njen otac je advokat.

By the time Yahli goes to college,

Kad Jali pođe na fakultet,

the jobs her parents do are going to look dramatically different.

poslovi koje njeni roditelji obavljaju izgledaće drastično drugačije.

In 2013, researchers at Oxford University did a study on the future of work.

Istraživači sa Oksforda su 2013. uradili istraživanje o budućnosti poslova.

They concluded that almost one in every two jobs have a high risk

Zaključili su da je gotovo jedan od svaka dva posla pod velikim rizikom

of being automated by machines.

da bude mašinski automatizovan.

Machine learning is the technology

Mašinsko učenje je tehnologija

that's responsible for most of this disruption.

koja je najodgovornija za ovaj raskol.

It's the most powerful branch of artificial intelligence.

To je najmoćnija grana veštačke inteligencije.

It allows machines to learn from data

Omogućuje mašinama da uče iz podataka

and mimic some of the things that humans can do.

i da oponašaju neke stvari koje ljudi mogu da rade.

My company, Kaggle, operates on the cutting edge of machine learning.

Moja firma, Kagle, se bavi najnaprednijim vidom mašinskog učenja.

We bring together hundreds of thousands of experts

Spajamo na stotine hiljada eksperata

to solve important problems for industry and academia.

kako bismo rešili važne probleme u industriji i akademiji.

This gives us a unique perspective on what machines can do,

To nam pruža jedinstvenu perspektivu na to šta mašine mogu,

what they can't do

šta ne mogu

and what jobs they might automate or threaten.

i koje poslove mogu da automatizuju ili ugroze.

Machine learning started making its way into industry in the early '90s.

Mašinsko učenje se počelo probijati u industriji tokom ranih '90-ih.

It started with relatively simple tasks.

Počelo je relativno jednostavnim zadacima.

It started with things like assessing credit risk from loan applications,

Počelo je stvarima poput bavljenja kreditnim rizikom kod molbi za zajam,

sorting the mail by reading handwritten characters from zip codes.

sortiranjem pošte čitanjem ručno pisanih slova zip kodova.

Over the past few years, we have made dramatic breakthroughs.

Tokom proteklih nekoliko godina imali smo drastična dostignuća.

Machine learning is now capable of far, far more complex tasks.

Mašinsko učenje je sada sposobno za daleko, daleko složenije zadatke.

In 2012, Kaggle challenged its community

Godine 2012. Kagle je izazvao njegovu zajednicu

to build an algorithm that could grade high-school essays.

da naprave algoritam koji bi ocenjivao srednjoškolske eseje.

The winning algorithms were able to match the grades

Pobednički algoritmi su mogli da daju podudarne ocene

given by human teachers.

kao i ljudski profesori.

Last year, we issued an even more difficult challenge.

Prošle godine smo napravili čak i komplikovaniji izazov.

Can you take images of the eye and diagnose an eye disease

Možete li da uzmete snimak oka i da dijagnostikujete očnu bolest

called diabetic retinopathy?

pod nazivom dijabetička retinopatija?

Again, the winning algorithms were able to match the diagnoses

Opet su pobednički algoritmi mogli da daju podudarnu dijagnozu

given by human ophthalmologists.

kao i ljudski oftalmolozi.

Now, given the right data, machines are going to outperform humans

Sad, uz odgovarajuće podatke mašine će da nadmaše ljude

at tasks like this.

u sličnim zadacima.

A teacher might read 10,000 essays over a 40-year career.

Nastavnik može da pročita 10.000 eseja tokom 40-ogodišnje karijere.

An ophthalmologist might see 50,000 eyes.

Oftalmolog može da pregleda 50.000 očiju.

A machine can read millions of essays or see millions of eyes

Mašina može da pročita na milione eseja ili da pregleda na milione očiju

within minutes.

za nekoliko minuta.

We have no chance of competing against machines

Nemamo nikakve šanse u takmičenju s mašinama

on frequent, high-volume tasks.

na učestalim zadacima velikog obima.

But there are things we can do that machines can't do.

Ali ima nešto što mi možemo, a mašine ne mogu.

Where machines have made very little progress

Mašine su postigle veoma mali napredak

is in tackling novel situations.

kod bavljenja novim situacijama.

They can't handle things they haven't seen many times before.

Ne mogu da savladaju nešto što nisu videle mnogo puta ranije.

The fundamental limitations of machine learning

Temeljno ograničenje mašinskog učenja

is that it needs to learn from large volumes of past data.

je što mašine moraju da uče iz obilja prethodnih podataka.

Now, humans don't.

A ljudi ne moraju.

We have the ability to connect seemingly disparate threads

Sposobni smo da povežemo naoko nepovezane niti

to solve problems we've never seen before.

kako bismo rešili za nas nov problem.

Percy Spencer was a physicist working on radar during World War II,

Persi Spenser je bio fizičar koji je radio na radaru tokom II svetskog rata,

when he noticed the magnetron was melting his chocolate bar.

kad je primetio kako magnetron topi njegovu tablu čokolade.

He was able to connect his understanding of electromagnetic radiation

Mogao je da poveže sopstveno razumevanje elektromagnetne radijacije

with his knowledge of cooking

sa poznavanjem kuvanja

in order to invent -- any guesses? -- the microwave oven.

kako bi izumeo - pretpostavljate li šta? - mikrotalasnu pećnicu.

Now, this is a particularly remarkable example of creativity.

Sad, ovo je izrazito upečatljiv primer kreativnosti.

But this sort of cross-pollination happens for each of us in small ways

Ali ovakva plodna ukrštanja nam se dešavaju na mikroplanu

thousands of times per day.

hiljadama puta tokom dana.

Machines cannot compete with us

Mašine ne mogu da se takmiče s nama

when it comes to tackling novel situations,

kad je u pitanju bavljenje novim situacijama,

and this puts a fundamental limit on the human tasks

a ovo postavlja temeljno ograničenje na ljudske zadatke

that machines will automate.

koje mašine mogu da automatizuju.

So what does this mean for the future of work?

Pa, šta ovo znači za budućnost rada?

The future state of any single job lies in the answer to a single question:

Budućnost svakog posla počiva u odgovoru na samo jedno pitanje:

To what extent is that job reducible to frequent, high-volume tasks,

do koje mere je taj posao svodiv na učestale zadatke velikog obima

and to what extent does it involve tackling novel situations?

i u kojoj meri uključuje bavljenje novim situacijama?

On frequent, high-volume tasks, machines are getting smarter and smarter.

Kod učestalih zadataka velikog obima mašine postaju sve pametnije i pametnije.

Today they grade essays. They diagnose certain diseases.

Danas one ocenjuju eseje. Dijagnostikuju određene bolesti.

Over coming years, they're going to conduct our audits,

U narednim godinama radiće revizije poreza

and they're going to read boilerplate from legal contracts.

i čitaće opšta mesta u pravnim ugovorima.

Accountants and lawyers are still needed.

I dalje ćemo trebati računovođe i advokate.

They're going to be needed for complex tax structuring,

Trebaće nam za složeno struktuiranje poreza,

for pathbreaking litigation.

za pionirske parnice.

But machines will shrink their ranks

No, mašine će suziti njihovo zvanje

and make these jobs harder to come by.

i učiniće ove poslove težim za nalaženje.

Now, as mentioned,

Sad, kao što sam pomenuo

machines are not making progress on novel situations.

mašine ne postižu napredak kod novih situacija.

The copy behind a marketing campaign needs to grab consumers' attention.

Poruka marketinške kampanje mora da zgrabi pažnju potrošača.

It has to stand out from the crowd.

Mora da se ističe u gomili.

Poslovna strategija znači nalaženje rupa u tržištu,

Business strategy means finding gaps in the market,

things that nobody else is doing.

stvari koje niko drugi ne radi.

It will be humans that are creating the copy behind our marketing campaigns,

Ljudi su ti koji će da stvaraju poruke marketinških kampanja,

and it will be humans that are developing our business strategy.

i ljudi su ti koji će razvijati naše poslovne strategije.

So Yahli, whatever you decide to do,

Pa, Jali, čime god odlučiš da se baviš,

let every day bring you a new challenge.

neka ti svaki dan donese novi izazov.

If it does, then you will stay ahead of the machines.

Ako bude tako, bićeš ispred mašina.

Thank you.

Hvala vam.

(Applause)

(Aplauz)