The ethical dilemma of self-driving cars - Patrick Lin

Prevodilac: Ivana Krivokuća Lektor: Tijana Mihajlović

Ovo je misaoni eksperiment.

This is a thought experiment.

Let's say at some point in the not so distant future,

Recimo da u nekom trenutku u ne tako dalekoj budućnosti

you're barreling down the highway in your self-driving car,

jurite niz auto-put u svom autu sa automatskim upravljanjem

and you find yourself boxed in on all sides by other cars.

i nađete se zarobljeni sa svih strana među drugim automobilima.

Suddenly, a large, heavy object falls off the truck in front of you.

Odjednom, veliki i težak predmet padne sa kamiona ispred vas.

Vaš auto ne može da se zaustavi na vreme da bi izbegao sudar,

Your car can't stop in time to avoid the collision,

so it needs to make a decision:

stoga mora da donese odluku -

go straight and hit the object,

ići pravo i udariti u predmet,

swerve left into an SUV,

naglo skrenuti ka terencu

or swerve right into a motorcycle.

ili se zaleteti desno u motocikl.

Should it prioritize your safety by hitting the motorcycle,

Da li treba da da prioritet vašoj bezbednosti i udari motocikl,

minimize danger to others by not swerving,

da umanji opasnost za druge tako što neće skrenuti,

even if it means hitting the large object and sacrificing your life,

čak i ako to znači da će udariti veliki predmet i žrtvovati vaš život,

or take the middle ground by hitting the SUV,

ili da odabere srednje rešenje i udari terenca,

which has a high passenger safety rating?

koji je visoko ocenjen u pogledu bezbednosti za putnike?

Dakle, šta bi auto bez vozača trebalo da uradi?

So what should the self-driving car do?

Ako bismo mi ručno upravljali tim zarobljenim autom,

If we were driving that boxed in car in manual mode,

whichever way we'd react would be understood as just that,

kako god da reagujemo bilo bi shvaćeno prosto tako kako jeste,

kao reakcija,

a reaction,

not a deliberate decision.

a ne kao promišljena odluka.

It would be an instinctual panicked move with no forethought or malice.

To bi bio instiktivan potez u panici bez predumišljaja ili loše namere.

But if a programmer were to instruct the car to make the same move,

Međutim, kada bi programer naložio automobilu da učini to isto,

given conditions it may sense in the future,

pod uslovima koji se mogu javiti u budućnosti,

well, that looks more like premeditated homicide.

pa, to više izgleda kao ubistvo sa predumišljajem.

Now, to be fair,

Sad, da budemo fer,

self-driving cars are predicted to dramatically reduce traffic accidents

predviđa se da će automobili bez vozača

drastično smanjiti broj saobraćajnih nesreća i smrtnih slučajeva

and fatalities

by removing human error from the driving equation.

tako što će ukloniti ljudsku grešku iz jednačine vožnje.

Plus, there may be all sorts of other benefits:

Plus, tu može biti svakojakih drugih pogodnosti:

eased road congestion,

olakšano zakrčenje puteva,

decreased harmful emissions,

smanjeno ispuštanje štetnih gasova

and minimized unproductive and stressful driving time.

i neproduktivno i stresno vreme vožnje svedeno na minimum.

Međutim, nesreće se mogu i dalje desiti i dešavaće se,

But accidents can and will still happen,

and when they do,

a kada se dogode,

their outcomes may be determined months or years in advance

njihovi ishodi mogu biti određeni mesecima ili godinama unapred

od strane programera ili donosioca zakona.

by programmers or policy makers.

And they'll have some difficult decisions to make.

A moraće da donose neke teške odluke.

It's tempting to offer up general decision-making principles,

Primamljivo je dati ponudu opštih principa donošenja odluka

like minimize harm,

poput svođenja štete na najmanju meru,

ali čak i to ubrzo dovodi do sumnjivih moralnih odluka.

but even that quickly leads to morally murky decisions.

For example,

Na primer,

let's say we have the same initial set up,

recimo da imamo istu početnu postavku,

but now there's a motorcyclist wearing a helmet to your left

ali sada se sa vaše leve strane nalazi motociklista sa kacigom

and another one without a helmet to your right.

i još jedan bez kacige sa vaše desne strane.

Which one should your robot car crash into?

U koga treba da udari vaš robotski auto?

If you say the biker with the helmet because she's more likely to survive,

Ako kažete da je to bajker sa kacigom jer je verovatnije da će preživeti,

then aren't you penalizing the responsible motorist?

zar onda ne kažnjavate odgovornog motociklistu?

If, instead, you save the biker without the helmet

Ako umesto toga spasite bajkera bez kacige

because he's acting irresponsibly,

jer se neodgovorno ponaša,

then you've gone way beyond the initial design principle about minimizing harm,

onda ste se udaljili od prvobitnog principa o smanjenju štete

and the robot car is now meting out street justice.

i sada robotski auto sprovodi uličnu pravdu.

The ethical considerations get more complicated here.

Etička razmatranja ovde postaju komplikovanija.

In both of our scenarios,

U oba naša scenarija,

the underlying design is functioning as a targeting algorithm of sorts.

dizajn koji se nalazi u osnovi funkcioniše

kao neka vrsta algoritma za nalaženje mete.

In other words,

Drugim rečima,

it's systematically favoring or discriminating

sistematski favorizuje ili diskriminiše

against a certain type of object to crash into.

određenu vrstu predmeta u koji će udariti.

And the owners of the target vehicles

A vlasnici vozila koja su postala meta

will suffer the negative consequences of this algorithm

će snositi negativne posledice ovog algoritma

through no fault of their own.

iako nisu ni za šta krivi.

Our new technologies are opening up many other novel ethical dilemmas.

Naše nove tehnologije otvaraju i mnoge druge nove etičke dileme.

For instance, if you had to choose between

Na primer, ako biste morali da odaberete

a car that would always save as many lives as possible in an accident,

između automobila koji bi uvek spasio što je više moguće života u nesreći,

or one that would save you at any cost,

ili automobila koji bi vas spasio po bilo koju cenu,

which would you buy?

koji biste kupili?

Šta bi se desilo ako bi automobil počeo da analizira i uzima u obzir faktore

What happens if the cars start analyzing and factoring in

the passengers of the cars and the particulars of their lives?

putnika u automobilu i njihove pojedinačne živote?

Could it be the case that a random decision

Da li može biti slučaj da je slučajna odluka

is still better than a predetermined one designed to minimize harm?

i dalje bolja od unapred određene koja je osmišljena da se umanji šteta?

And who should be making all of these decisions anyhow?

Ko bi uopšte trebalo da donosi sve te odluke?

Programmers? Companies? Governments?

Programeri?

Kompanije?

Vlade?

Reality may not play out exactly like our thought experiments,

Stvarnost se možda neće odvijati baš kao naši misaoni eksperimenti,

but that's not the point.

ali nije u tome poenta.

Oni su osmišljeni da izoluju našu intuiciju vezanu za etiku

They're designed to isolate and stress test our intuitions on ethics,

i podvrgnu je testu o stresu,

just like science experiments do for the physical world.

baš kao što naučni eksperimenti čine kada se radi o fizičkom svetu.

Spotting these moral hairpin turns now

Uočavanje ovakvih moralnih krivina sada

will help us maneuver the unfamiliar road of technology ethics,

pomoći će nam u upravljanju nepoznatim putem tehnološke etike

and allow us to cruise confidently and conscientiously

i omogućiti nam da uplovimo samopouzdano i savesno

into our brave new future.

u našu hrabru novu budućnost.