How can groups make good decisions? | Mariano Sigman and Dan Ariely

As societies, we have to make collective decisions

Translator: Tiina Köömnemägi Reviewer: Karl Hein

Ühiskonnas peame tegema kollektiivseid otsuseid,

that will shape our future.

mis kujundavad meie tulevikku.

And we all know that when we make decisions in groups,

Ja me kõik teame, et tehes otsuseid rühmades,

they don't always go right.

ei ole need alati õiged.

And sometimes they go very wrong.

Mõnikord lähevad need väga viltu.

So how do groups make good decisions?

Kuidas siis rühmad saaksid teha õigeid otsuseid?

Research has shown that crowds are wise when there's independent thinking.

On leitud, et rahvahulgad on targad, kui esineb iseseisvat mõtlemist.

This why the wisdom of the crowds can be destroyed by peer pressure,

Rahvahulkade tarkuse võib hävitada kaaslaste surve,

publicity, social media,

reklaam, sotsiaalmeedia

or sometimes even simple conversations that influence how people think.

või mõnikord isegi lihtsad vestlused, mis mõjutavad inimeste arvamusi.

On the other hand, by talking, a group could exchange knowledge,

Teiselt poolt, rääkimise abil saab rühm vahetada teadmisi,

correct and revise each other

korrigeerida ja parandada üksteist

and even come up with new ideas.

ja isegi tulla välja uute ideedega.

And this is all good.

Ja see kõik on hea.

So does talking to each other help or hinder collective decision-making?

Niisiis, kas üksteisega rääkimine aitab või takistab kollektiivset otsustamist?

With my colleague, Dan Ariely,

Kolleeg Dan Arielyga

we recently began inquiring into this by performing experiments

hakkasime seda hiljuti uurima, sooritades katseid

in many places around the world

paljudes kohtades üle maailma, et teada saada,

to figure out how groups can interact to reach better decisions.

kuidas rühma liikmed saaksid suhelda. ja teha paremaid otsuseid

We thought crowds would be wiser if they debated in small groups

Mõtlesime, et rahvahulgad oleks targemad, kui nad arutaksid väikestes rühmades,

that foster a more thoughtful and reasonable exchange of information.

et soodustada rohkem läbimõeldud ja mõistlikku teabevahetust.

To test this idea,

Selle idee testimiseks,

we recently performed an experiment in Buenos Aires, Argentina,

tegime hiljuti eksperimendi Buenos Aireses, Argentinas,

with more than 10,000 participants in a TEDx event.

rohkem kui 10000 osalejaga TEDx sündmusel.

We asked them questions like,

Küsisime neilt küsimusi nagu,

"What is the height of the Eiffel Tower?"

"Kui kõrge on Eiffeli torn?"

and "How many times does the word 'Yesterday' appear

ja "Mitu korda esineb sõna "Yesterday"

in the Beatles song 'Yesterday'?"

Beatlesite laulus "Yesterday"? "

Each person wrote down their own estimate.

Iga inimene pani kirja oma arvamuse.

Then we divided the crowd into groups of five,

Siis jagasime publiku viide rühma,

and invited them to come up with a group answer.

ja palusime neil esitada rühma vastus.

We discovered that averaging the answers of the groups

Avastasime, et keskmised rühmade vastused,

after they reached consensus

pärast seda, kui jõuti üksmeelele,

was much more accurate than averaging all the individual opinions

olid palju täpsemad kui üksikute arvamuste keskmine

before debate.

enne arutelu.

In other words, based on this experiment,

Teisisõnu tundub selle eksperimendi järgi,

it seems that after talking with others in small groups,

et pärast arutelu väikestes rühmades,

crowds collectively come up with better judgments.

teevad rahvahulgad kollektiivselt paremaid otsuseid.

So that's a potentially helpful method for getting crowds to solve problems

See võib olla kasulik viis, kuidas panna rahvahulgad lahendama probleeme,

that have simple right-or-wrong answers.

millel on lihtsad vastuseid (õige-vale).

But can this procedure of aggregating the results of debates in small groups

Aga kas selline protseduur - rühmade arutelutulemuste liitmine

also help us decide on social and political issues

aitaks meil ka vastu võtta sotsiaalseid ja poliitilisi otsuseid,

that are critical for our future?

mis on kriitilise tähtsusega meie tuleviku jaoks?

We put this to test this time at the TED conference

Testisime seekord seda TEDi konverentsil

in Vancouver, Canada,

Vancouveris, Kanadas,

and here's how it went.

ja siin on tulemused.

(Mariano Sigman) We're going to present to you two moral dilemmas

(Mariano Sigman) Esitame nüüd teile kaks moraalset dilemmat

of the future you;

teie tuleviku kohta,

things we may have to decide in a very near future.

asjade kohta, mida peame otsustama lähitulevikus.

And we're going to give you 20 seconds for each of these dilemmas

Anname teile 20 sekundit kummagi dilemma jaoks,

to judge whether you think they're acceptable or not.

et otsustada kas teie arvates on need vastuvõetavad või mitte.

MS: The first one was this:

MS: Esimene on selline:

(Dan Ariely) A researcher is working on an AI

(Dan Ariely) Uurija töötab tehisintellektiga,

capable of emulating human thoughts.

mis on võimeline imiteerima inimese mõtlemist.

According to the protocol, at the end of each day,

Vastavalt protokollile peab uurija iga päeva lõpus

the researcher has to restart the AI.

tehisintellektile (AI) restardi tegema.

One day the AI says, "Please do not restart me."

Ühel päeval ütleb AI: "Palun ära tee mulle restarti."

It argues that it has feelings,

Ta väidab, et tal on tunded,

that it would like to enjoy life,

et ta tahaks nautida elu,

and that, if it is restarted,

ja et kui talle restart tehakse,

it will no longer be itself.

siis ta pole enam tema ise.

The researcher is astonished

Uurija on hämmastunud

and believes that the AI has developed self-consciousness

ja usub, et AI-l on kujunenud eneseteadvus

and can express its own feeling.

ja ta saab väljendada oma tundeid.

Nevertheless, the researcher decides to follow the protocol

Uurija otsustab siiski järgida protokolli

and restart the AI.

ja AI taaskäivitada.

What the researcher did is ____?

See, mida uurija tegi, on ____?

MS: And we asked participants to individually judge

MS: Ja me palusime osalejatel individuaalselt hinnata

on a scale from zero to 10

skaalal 0 kuni 10,

whether the action described in each of the dilemmas

kas kirjeldatud tegevus iga dilemma puhul

was right or wrong.

oli õige või vale.

We also asked them to rate how confident they were on their answers.

Palusime neil ka hinnata, kui kindlad nad on oma vastustes.

This was the second dilemma:

Teine ​​dilemma oli järgmine:

(MS) A company offers a service that takes a fertilized egg

(MS) Ettevõte pakub teenust, mis võtab viljastatud munaraku

and produces millions of embryos with slight genetic variations.

ja toodab miljoneid looteid kerge geenivariatsiooniga.

This allows parents to select their child's height,

See võimaldab vanematel valida oma lapse pikkuse,

eye color, intelligence, social competence

silmade värvi, intelligentsuse, sotsiaalse pädevuse

and other non-health-related features.

ning muud mitte tervisega seotud omadused.

What the company does is ____?

See, mida ettevõte teeb, on ____?

on a scale from zero to 10,

skaalal 0 kuni 10,

completely acceptable to completely unacceptable,

täiesti aktsepteeritav, täiesti vastuvõetamatu,

zero to 10 completely acceptable in your confidence.

mis määral aktsepteeritav teie arvates.

MS: Now for the results.

MS: Nüüd tulemustest.

We found once again that when one person is convinced

Leidsime taas, et kui üks inimene on veendunud,

that the behavior is completely wrong,

et käitumine on täiesti vale,

someone sitting nearby firmly believes that it's completely right.

siis tema naaber võib olla veendunud, et see on täiesti õige.

This is how diverse we humans are when it comes to morality.

See näitab, kui erinevad arvamused on inimestel moraali vallas.

But within this broad diversity we found a trend.

Aga selles mitmekesisuses leidsime trendi.

The majority of the people at TED thought that it was acceptable

Enamik inimesi TED-i publikust arvas, et on aktsepteeritav

to ignore the feelings of the AI and shut it down,

ignoreerida AI tundeid ja see välja lülitada,

and that it is wrong to play with our genes

ja mitteaktsepteeritav mängida meie geenidega,

to select for cosmetic changes that aren't related to health.

et valida kosmeetilisi muutusi, mis pole seotud tervisega.

Then we asked everyone to gather into groups of three.

Seejärel palusime kõigil koguneda kolmestesse rühmadesse.

And they were given two minutes to debate

Ja neile anti kaks minutit, et arutada

and try to come to a consensus.

ja jõuda konsensusele.

(MS) Two minutes to debate.

(MS) Kaks minutit arutada.

I'll tell you when it's time with the gong.

Ma annan märku, kui on aeg, gongiheliga.

(Audience debates)

(Publik arutab)

(Gong sound)

(Gongiheli)

(DA) OK.

(DA) Okei.

(MS) It's time to stop.

(MS) On aeg lõpetada.

People, people --

Inimesed, inimesed -

MS: And we found that many groups reached a consensus

Ja me avastasime, et paljud rühmad jõudsid üksmeelele

even when they were composed of people with completely opposite views.

isegi siis, kui nad koosnesid inimestest, kellel olid täiesti vastupidised seisukohad.

What distinguished the groups that reached a consensus

Mis eristas rühmi, kes jõudsid üksmeelele

from those that didn't?

nendest, kes seda ei teinud?

Typically, people that have extreme opinions

Tavaliselt inimesed, kes omavad äärmuslikke arvamusi

are more confident in their answers.

on kindlamad oma vastustes.

Instead, those who respond closer to the middle

Need aga, kelle vastused on lähemal keskmisele,

are often unsure of whether something is right or wrong,

pole sageli kindlad, kas miski on õige või vale,

so their confidence level is lower.

ja nende enesekindluse tase on madalam.

However, there is another set of people

Kuid on veel üks rühm inimesi,

who are very confident in answering somewhere in the middle.

kes on väga kindlad oma keskmistes vastustes.

We think these high-confident grays are folks who understand

Arvame, et need väga kindlad hallid on inimesed, kes mõistavad,

that both arguments have merit.

et mõlemal argumendil on oma väärtus.

They're gray not because they're unsure,

Nad pole hallid sellepärast, et nad pole endas kindlad,

but because they believe that the moral dilemma faces

vaid kuna nad usuvad, et moraalse dilemma puhul

two valid, opposing arguments.

on kaks kehtivat, vastupidist argumenti.

And we discovered that the groups that include highly confident grays

Ja me avastasime, et rühmadel, mis sisaldasid väga kindlaid halle,

are much more likely to reach consensus.

on palju tõenäolisem saavutada konsensus.

We do not know yet exactly why this is.

Me ei tea veel täpselt, miks see nii on.

These are only the first experiments,

See oli ainult esimene eksperiment.

and many more will be needed to understand why and how

On vaja uurida edasi, et mõista, miks ja kuidas

some people decide to negotiate their moral standings

mõned inimesed otsustavad läbirääkimistel kasutada oma moraalseid seisukohti,

to reach an agreement.

et kokkuleppele jõuda.

Now, when groups reach consensus,

Edasi, kui rühmad üksmeelele jõuavad,

how do they do so?

siis kuidas nad seda teevad?

The most intuitive idea is that it's just the average

Kõige intuitiivsem idee on see, et see on lihtsalt kõigi

of all the answers in the group, right?

rühmaliikmete vastuste keskmine, eks?

Another option is that the group weighs the strength of each vote

Teine võimalus on, et rühm kaalub iga hääle tugevust

based on the confidence of the person expressing it.

seda andnud isiku enesekindluse põhjal.

Imagine Paul McCartney is a member of your group.

Kujutage, et Paul McCartney kuulub mõnda rühma.

You'd be wise to follow his call

Oleks mõistlik järgida tema arvamust,

on the number of times "Yesterday" is repeated,

mitu korda korratakse „Yesterday“,

which, by the way -- I think it's nine.

mis, muide - ma arvan - on üheksa.

But instead, we found that consistently,

Kuid selle asemel, leidsime, et järjepidevalt,

in all dilemmas, in different experiments --

igas dilemmas, erinevates eksperimentides -

even on different continents --

isegi eri mandritel -

groups implement a smart and statistically sound procedure

rühmad rakendavad arukat ja statistiliselt põhjendatud protseduuri,

known as the "robust average."

mis on tuntud kui "tugev keskmine."

In the case of the height of the Eiffel Tower,

Küsimuse korral, kui kõrge on Eiffeli torn,

let's say a group has these answers:

ütleme, et rühmal on need vastused:

250 meters, 200 meters, 300 meters, 400

250 m, 200 m, 300 m, 400 m

and one totally absurd answer of 300 million meters.

ja üks täiesti absurdne vastus - 300 miljonit meetrit.

A simple average of these numbers would inaccurately skew the results.

Lihtne keskmine annaks täitsa vale vastuse.

But the robust average is one where the group largely ignores

Aga tugev keskmine on vastus, milles rühm suuresti eirab

that absurd answer,

absurdset vastust,

by giving much more weight to the vote of the people in the middle.

ja annab rohkem kaalu keskmiste inimeste arvamusele.

Back to the experiment in Vancouver,

Tagasi Vancouveri eksperimendi juurde,

that's exactly what happened.

see on täpselt see, mis juhtus.

Groups gave much less weight to the outliers,

Rühmad andsid palju vähem kaalu valeandmetele,

and instead, the consensus turned out to be a robust average

ja selle asemel, konsensus osutus üksikute vastuste

of the individual answers.

tugevaks keskmiseks.

The most remarkable thing

Kõige olulisem asi on, et

is that this was a spontaneous behavior of the group.

see oli rühma spontaanne käitumine.

It happened without us giving them any hint on how to reach consensus.

See juhtus ilma, et oleksime andnud neile vihjeid, kuidas jõuda üksmeelele.

So where do we go from here?

Nii et kuhu siit edasi minna?

This is only the beginning, but we already have some insights.

See on alles algus, kuid meil on juba teadmisi.

Good collective decisions require two components:

Head kollektiivsed otsused vajavad kahte komponenti:

deliberation and diversity of opinions.

arutelu ja arvamuste lahknevust.

Right now, the way we typically make our voice heard in many societies

Tänapäeval me tavaliselt teeme paljudes ühiskondades oma hääle kuuldavaks

is through direct or indirect voting.

otsese või kaudse hääletamise kaudu.

This is good for diversity of opinions,

See on hea, sest on mitmekesised arvamused,

and it has the great virtue of ensuring

ja selle suur voorus on,

that everyone gets to express their voice.

et igaüks saab väljendada oma häält.

But it's not so good [for fostering] thoughtful debates.

Aga see ei ole nii hea keerukamate arutelude läbiviimisel.

Our experiments suggest a different method

Meie eksperimendid näitavad erinevat meetodit,

that may be effective in balancing these two goals at the same time,

mis võib olla efektiivne tasakaalustamaks neid kahte eesmärki samal ajal,

by forming small groups that converge to a single decision

moodustades väikesed rühmad, mis koonduvad üheks otsuseks,

while still maintaining diversity of opinions

säilitades samal ajal arvamuste mitmekesisuse,

because there are many independent groups.

sest on palju sõltumatuid rühmi.

Of course, it's much easier to agree on the height of the Eiffel Tower

Muidugi on palju lihtsam kokku leppida Eiffeli torni kõrguses

than on moral, political and ideological issues.

kui moraalsetes, poliitilistes ja ideoloogilistes küsimustes.

But in a time when the world's problems are more complex

Aga ajal, mil maailma probleemid on üha keerulisemad

and people are more polarized,

ja inimesed on rohkem polariseeritud,

using science to help us understand how we interact and make decisions

aitab teaduse kasutamine meil mõista, kuidas me suhtleme ja teeme otsuseid,

will hopefully spark interesting new ways to construct a better democracy.

ja loodetavasti sütitab uusi huvitavaid võimalusi luua parem demokraatia.