未登录,请登录后再发表信息
最新评论 (0)
播放视频

怎样让团队做出正确的决定

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

在社会中 我们需要作出会改变自身未来的集体性决策
As societies, we have to makecollective decisions that will shape our future.
众所周知 当我们在集体中作出决定时
And we all know that whenwe make decisions in groups,
这些决定不一定都是正确的
they don’t always go right.
甚至有些时候 它们大错特错
And sometimes they go very wrong.
那么 一个团队如何作出好的决定呢?
So how do groups make good decisions?
研究表明 能独立思考的群体是智慧的
Research has shown that crowds are wisewhen there’s independent thinking.
这便是为什么 同辈压力 舆论 社交媒体
This why the wisdom of the crowds can be
有时甚至是影响思维的简单对话
destroyed by peer pressure, publicity, social media,
都能摧毁团队智慧
or sometimes even simple conversations that influence how people think.
另一方面 一个团队可以通过谈话交流
On the other hand, by talking,
来进行知识分享
a group could exchange knowledge,
调整和修正彼此的想法
correct and revise each other
甚至从中产生新的想法
and even come up with new ideas.
这都是很好的
And this is all good.
那么 彼此交流是有助于还是会阻碍集体决策?
So does talking to each otherhelp or hinder collective decision-making?
我和同事丹•阿莱利
With my colleague, Dan Ariely,
近期在世界各地进行实验 对此加以调查
we recently began inquiring into this by performing experiments in many places around the world
探究团队该怎样互动 来作出更好的决策
to figure out how groups can interact to reach better decisions.
我们认为 如果人们辩论时
We thought crowds would be wiser
处于信息交流更周全理智的小群体
if they debated in small groups that foster a more
那么群体的决策就会更加明智
thoughtful and reasonable exchange of information.
为了验证这个想法
To test this idea,
我们最近在阿根廷的布宜诺斯艾利斯
we recently performed an experiment
进行了一项实验
in Buenos Aires, Argentina,
参加这次TEDx活动的人数超过了1万人
with more than 10,000 participants in a TEDx event.
我们问了他们一些问题 比如:
We asked them questions like,
“埃菲尔铁塔的高度是多少?”
“What is the height of the Eiffel Tower?”
还有“在披头士的歌曲《昨天》中
and”How many timesdoes the word’Yesterday’ appear
‘昨天’这个词出现了多少次?”
in the Beatles song’Yesterday’?”
每个人写下他们自己的答案
Each person wrote down their own estimate.
然后我们把参与者分成五人一组
Then we divided the crowd into groups of five,
并请他们给出团队的答案
and invited them to come up with a group answer.
我们发现 达成共识后的团队答案均值
We discovered that averaging the answers of the groups
要比 分组讨论前的个人答案均值
after they reached consensus was much more accurate
显得更加准确
than averaging all the individual opinions before debate.
换句话说 通过这个实验
In other words, based on this experiment,
我们发现 人们在小组中互相讨论后
it seems that after talking with others in small groups,
会做出更好的判断
crowds collectively come up with better judgments.
所以这是个潜在的方法
So that’s a potentially helpful method
能帮助团体解决一些
for getting crowds to solve problems
简单的对错问题
that have simple right-or-wrong answers.
但是 整合小组讨论结果的这一过程
But can this procedure of aggregating the results of debates in small groups
是否也能帮我们决定一些
also help us decide on social and political issues
对我们的未来至关重要的社会和政治问题?
that are critical for our future?
我们这次在加拿大温哥华的TED演讲上
We put this to test this timeat the TED conference
做了这个测试 下面是实验结果
in Vancouver, Canada, and here’s how it went.
Mariano Sigman(以下简称MS):我们会向你们提供两种可能在未来发生的
(Mariano Sigman) We’re going to presentto you two moral dilemmas
道德两难困境
of the future you;
我们也许会在不远的将来 面临这种选择
things we may have to decide in a very near future.
我们会给你们20秒的时间
And we’re going to give
对每个事件做出判断
you 20 seconds for each of these dilemmas
思考他们的做法是否应该被接受
to judge whether you thinkthey’re acceptable or not.
MS:第一个事件是:
The first one was this:
Dan Ariely(以下简称DA):一个研究人员正在研究一种
(Dan Ariely) A researcher is working on an AI
可以模仿人类思想的AI
capable of emulating human thoughts.
根据规定 在每天结束的时候
According to the protocol,at the end of each day,
研究者会初始化这个AI
the researcher has to restart the AI.
有一天 AI说:“请别初始化我。”
One day the AI says,”Pleasedo not restart me.”
它认为它有感情
It argues that it has feelings,
它想要享受生活 等等
that it would like to enjoy life, and that,
如果它被初始化了 它就不再是它自己了
if it is restarted, it will no longer be itself.
研究者十分震惊并且相信
The researcher is astonished and believes
这个AI进化出了自我意识
that the AI has developed self-consciousness
可以表达它自己的想法 尽管如此
and can express its own feeling. Nevertheless,
研究人员还是决定按照规定
the researcherdecides to follow the protocol
初始化了AI
and restart the AI.
这个研究者的做法是?
What the researcher did is ____?
MS:我们要求参与者 按0到10的等级
MS: And we asked participants to individually judge on a scale
独立地判断每个两难事件中
from zero to 10 whether the action described
描述的行为正确与否
in each of the dilemmas was right or wrong.
我们还要求他们评估自己对答案的自信程度
We also asked them to rate how confident they were on their answers.
这是第二个两难困境:
This was the second dilemma:
MS:某公司提供了一项服务
(MS) A company offers a servicethat takes a fertilized egg
用受精卵培育出基因略有差异的数百万个胚胎
and produces millions of embryoswith slight genetic variations.
这让父母能选择孩子的身高
This allows parentsto select their child’s height,
眼睛颜色 智力 社交能力
eye color, intelligence, social competence
以及其他与健康无关的特征
and other non-health-related features.
这个公司的做法是?
What the company does is ____?
0到10的等级
on a scale from zero to 10,
对应完全能接受到完全不能接受
completely acceptable to completely unacceptable,
根据你的自信来选择0到10的不同数值
zero to 10 completely acceptablein your confidence.
MS:现在宣布结果
MS: Now for the results.
我们再一次发现 当一个人确信
We found once againthat when one person is convinced
这个行为是完全错误的时候
that the behavior is completely wrong,
他周围的某人会深信它是完全正确的
someone sitting nearby firmly believes that it’s completely right.
可见面临道德问题时 人们的观点非常多样
This is how diverse we humans are when it comes to morality.
但是在这种广泛多样性中 我们发现一种趋势
But within this broad diversity we found a trend.
在TED演讲中的大多数人认为 忽视AI的感受
The majority of the people at TED thought that it was acceptable
然后把它关掉 是可以接受的
to ignore the feelings of the AIand shut it down,
而改变基因 选择与健康无关的外表特征
and that it is wrong to play with our genes
是错误的做法
to select for cosmetic changesthat aren’t related to health.
之后我们将参与者分为三人一组
Then we asked everyoneto gather into groups of three.
给他们两分钟时间来讨论
And they were given two minutes to debate
尝试达成共识
and try to come to a consensus.
MS:你们有两分钟的时间去讨论
(MS) Two minutes to debate.
我会通过敲钟来告诉你们时间到了
I’ll tell you when it’s time with the gong.
听众讨论中
(Audience debates)
[鼓声响]
(Gong sound)
DA:好了
(DA) OK.
MS:该停下来了 各位
(MS) It’s time to stop. People, people.
MS:我们发现很多小组都达成了一致
And we found that many groups reached a consensus,
尽管它们可能由观点完全相反的人组成
even when they were composed of peoplewith completely opposite views.
达成共识的组和没有达成共识的组
What distinguished the groupsthat reached a consensus
有什么区别吗?
from those that didn’t?
通常 有极端观点的人
Typically, people that have extreme opinions
对自己的答案更有信心
are more confident in their answers.
那些观点更折中的人
Instead, those who respond closer to the middle
往往不确定某件事是对是错
are often unsure of whethersomething is right or wrong,
因此他们的自信等级更低
so their confidence level is lower.
而有另外一部分人
However, there is another set of people
他们对自己折中的答案很自信
who are very confident in answeringsomewhere in the middle.
我们认为 这些高度自信的人
We think these high-confident grays
知道不同观点各有千秋
are folks who understand that both arguments have merit.
他们趋于折衷 并不是因为他们不确定
They’re gray not because they’re unsure,
而是因为他们相信道德上的两难事件
but because they believethat the moral dilemma faces
往往有两个对立却合理的观点
two valid, opposing arguments.
我们发现 具有高度自信成员的团队
And we discovered that the groupsthat include highly confident grays
更有可能达成统一结论
are much more likely to reach consensus.
我们不确定这是为什么
We do not know yet exactly why this is.
这些只是初步的实验
These are only the first experiments,
我们需要更多的实验来理解
and many more will be needed to understand
人们为何 又是如何对道德立场加以探讨
why and how some people decide to negotiate their moral standings
以达成统一意见的
to reach an agreement.
现在 若团队达成了共识
Now, when groups reach consensus,
他们是如何做到的呢?
how do they do so?
第一念头是 这只是群体中
The most intuitive idea is that it’s just the average
所有答案的平均值 对吗?
of all the answers in the group, right?
另一种观点是 团体会根据表达者的自信
Another option is that the group weighs the strength of each vote
来权衡每一票的力度
based on the confidence of the person expressing it.
想象保罗·麦卡特尼是你们小组的成员
Imagine Paul McCartney is a member of your group.
你最好跟着他的节奏
You’d be wise to follow his call
去数“昨天”重复的次数
on the number of times”Yesterday” is repeated, which,
顺便说一下 我认为是九次
by the way — I think it’s nine.
但相反 我们发现
But instead, we found that consistently,
所有两难困境中 在不同的实验里
in all dilemmas, in different experiments —
甚至在不同的大陆
even on different continents —
团队会利用基于数据的聪明办法
groups implement a smartand statistically sound procedure
叫做“稳健平均值”
known as the”robust average.”
关于艾菲尔铁塔高度这个问题
In the case of the heightof the Eiffel Tower,
我们来看一个队伍中出现的答案:
let’s say a group has these answers:
250米 200米 300米 400米
250 meters, 200 meters, 300 meters, 400
还有个荒谬的答案说3亿米
and one totally absurd answerof 300 million meters.
对这些数字简单地取平均值会影响结果的精确性
A simple average of these numberswould inaccurately screw the results.
但是 稳健平均值
But the robust average is one
更重视折中人群的选择
where the group largely ignores that absurd answer,
让群体很大程度上忽略了荒谬的答案
by giving much more weight to the vote of the people in the middle.
回到温哥华的实验 情况正是如此
Back to the experiment in Vancouver, that’s exactly what happened.
团队对于离谱的数值考虑得更少
Groups gave much less weightto the outliers,
而且 最后达成的共识
and instead, the consensus
是个体答案的稳健平均值
turned out to be a robust average of the individual answers.
最值得注意的是 这是团体的自发行为
The most remarkable thing is that this was a spontaneous behavior of the group.
我们没有给出关于达成共识的任何暗示
It happened without us giving them any hint on how to reach consensus.
那我们接下来怎么办?
So where do we go from here?
这仅仅是一个开始 但是我们已经有一些头绪
This is only the beginning,but we already have some insights.
良好的集体决策需要两个条件:
Good collective decisionsrequire two components:
审慎思考 以及观点多样
deliberation and diversity of opinions.
现在 我们在社会上发声的典型方式
Right now, the way we typically make our voice heard in many societies
是直接或间接投票
is through direct or indirect voting.
这有利于观点的多样性
This is good for diversity of opinions,
并且有益于确保
and it has the great virtue of ensuring
每个人都能表达意见
that everyone gets to express their voice.
但这样 却不太有利于形成深入的探讨
But it’s not so good to fosterthoughtful debates.
我们的实验提出了一个不同的方法
Our experiments suggest a different method
也许能有效地同时平衡这两个目标
that may be effective in balancing these two goals at the same time,
也就是通过形成小团队 来整合决策
by forming small groupsthat converge to a single decision
同时又因为存在许多独立的小团队
while still maintaining diversity of opinions
保留了观点的多样性
because there are many independent groups.
当然 比起道德 政治和意识形态问题
Of course, it’s much easier to agree on the height of the Eiffel Tower
在艾菲尔铁塔的高度上达成一致要容易的多
than on moral, political and ideological issues.
但在目前 世界问题更为复杂
But in a time when the world’s problems are more complex
人们更为极端的情况下
and people are more polarized,
利用科学来帮助我们理解
using science to help us understand
我们是如何互动和决策的
how we interact and make decisions
有可能激发出有趣的新方法
will hopefully spark interesting new ways
来建立一个更好的民主社会
to construct a better democracy.

发表评论

译制信息
视频概述

我们都知道,当我们分组做决定的时候,他们并不总是正确的——有时候他们会犯很大的错误。团体怎么能做出正确的决定?神经科学家marianosigman和他的同事danariely一直在研究这个问题。通过在世界各地进行实验,他们发现,团队通过吸收每个人的意见,并且取其“稳固平均值”来确保达成一致,而这往往有利于结果的准确性,也有利于观点的多样化讨论。

听录译者

收集自网络

翻译译者

青山入怀

审核员

审核员WY

视频来源

https://www.youtube.com/watch?v=JrRRvqgYgT0

相关推荐