What to trust in a "post-truth" world | Alex Edmans

翻译人员: jacks peng 校对人员:

Belle Gibson was a happy young Australian.

贝尔·吉布森是一个 快乐的澳大利亚年轻人。

She lived in Perth, and she loved skateboarding.

她住在珀斯， 她喜欢玩滑板。

But in 2009, Belle learned that she had brain cancer and four months to live.

但在2009年， 贝尔得知自己患有脑癌，

并且只有四个月可活。

Two months of chemo and radiotherapy had no effect.

两个月的化疗和放疗没有见效。

But Belle was determined.

但是贝尔的意志很坚强，

She'd been a fighter her whole life.

贝尔一直都是一位斗士。

From age six, she had to cook for her brother, who had autism,

从6岁起， 她就得给患自闭症的哥哥，

and her mother, who had multiple sclerosis.

还有患多发性硬化症的母亲做饭。

Her father was out of the picture.

她的父亲一直缺位。

So Belle fought, with exercise, with meditation

因此，贝尔用锻炼、冥想抗癌

同时，她也用蔬果代替肉食。

and by ditching meat for fruit and vegetables.

And she made a complete recovery.

她完全康复了。

Belle's story went viral.

贝尔的故事迅速走红。

It was tweeted, blogged about, shared and reached millions of people.

她的故事在推特和博客上广为流传。

It showed the benefits of shunning traditional medicine

它展示了传统医学以外的

for diet and exercise.

饮食和锻炼的意义。

In August 2013, Belle launched a healthy eating app,

2013年8月，贝尔发布了 一款健康饮食的应用软件，

The Whole Pantry,

“健康厨房”

downloaded 200,000 times in the first month.

首月下载量达到20万次。

But Belle's story was a lie.

但贝尔的故事是个谎言。

Belle never had cancer.

贝尔从来没有过癌症。

People shared her story without ever checking if it was true.

人们分享她的故事时 从未去检验其真实性。

This is a classic example of confirmation bias.

这是肯证偏差的典型例子。

We accept a story uncritically if it confirms what we'd like to be true.

我们会不加批判地接受一个故事，

当它证实了我们认为是真实的事时，

And we reject any story that contradicts it.

并且我们会拒绝 任何与之相悖的故事。

How often do we see this

我们看到了多少这种情况？

in the stories that we share and we ignore?

在我们分享却忽略的故事中。

In politics, in business, in health advice.

在政治上，在商业中， 在健康建议上。

The Oxford Dictionary's word of 2016 was "post-truth."

牛津词典2016年的 年度词汇是“后真相”。

And the recognition that we now live in a post-truth world

人们意识到 我们正处在后真相世界中，

has led to a much needed emphasis on checking the facts.

所以如今我们非常强调核查事实。

But the punch line of my talk

但我演讲的重点是

is that just checking the facts is not enough.

仅仅去核查真相还不够。

Even if Belle's story were true,

即便贝尔的故事是真的，

it would be just as irrelevant.

它也是一个不相关的故事。

Why?

为什么？

Well, let's look at one of the most fundamental techniques in statistics.

让我们看看统计学中的 一个基本原理。

It's called Bayesian inference.

贝叶斯推理。

And the very simple version is this:

它的核心观点就是，

We care about "does the data support the theory?"

我们关心：“数据是否支持这个理论？”

Does the data increase our belief that the theory is true?

这个数据是否能够证实 这个理论为真

But instead, we end up asking, "Is the data consistent with the theory?"

但相反，我们最终会问， “数据是否与理论一致？”

But being consistent with the theory

但与理论一致

does not mean that the data supports the theory.

不等于数据支持这个理论。

Why?

为什么？

Because of a crucial but forgotten third term --

因为有一个关键但被人遗忘的点——

the data could also be consistent with rival theories.

数据也可以和对立的理论一致。

But due to confirmation bias, we never consider the rival theories,

但由于肯证偏差， 我们从不考虑对立的理论，

because we're so protective of our own pet theory.

因为我们袒护 我们的宠物理论。

Now, let's look at this for Belle's story.

现在，让我们看看贝尔的故事。

Well, we care about: Does Belle's story support the theory

我们关心的是：贝尔的故事

支持饮食治愈癌症的理论吗？

that diet cures cancer?

But instead, we end up asking,

但相反，我们最终问的是：

"Is Belle's story consistent with diet curing cancer?"

“贝尔的故事 等同于饮食治愈癌症吗？”

And the answer is yes.

答案是肯定的。

If diet did cure cancer, we'd see stories like Belle's.

如果饮食可以治愈癌症，

我们会看到像贝尔这样的故事。

But even if diet did not cure cancer,

但即便饮食不能治疗癌症，

we'd still see stories like Belle's.

我们仍然会看到像贝尔这样的故事。

A single story in which a patient apparently self-cured

比如一个病人自我治愈

just due to being misdiagnosed in the first place.

只是因为一开始被误诊。

Just like, even if smoking was bad for your health,

或者，即使吸烟有害健康，

你仍然会看到一个烟民活到100岁。

you'd still see one smoker who lived until 100.

(Laughter)

（笑声）

Just like, even if education was good for your income,

又或者，即使接受教育 有助于增加你的收入，

you'd still see one multimillionaire who didn't go to university.

你仍会看到没上过大学的千万富翁。

(Laughter)

（笑声）

所以贝尔故事最大的问题 不在于它是虚假的。

So the biggest problem with Belle's story is not that it was false.

It's that it's only one story.

在于它只是一个故事。

There might be thousands of other stories where diet alone failed,

也许有成千上万仅靠饮食失败的故事，

but we never hear about them.

但我们从没听到这些故事。

We share the outlier cases because they are new,

我们分享异常个案 只是因为它们是新奇的，

and therefore they are news.

因此它们成了新闻。

We never share the ordinary cases.

我们从不分享普通案例。

They're too ordinary, they're what normally happens.

它们太普通， 它们就是日常发生的事情。

And that's the true 99 percent that we ignore.

这是我们忽略的99%的真相。

Just like in society, you can't just listen to the one percent,

就像在社会中，

你不能只听1%的异常个案，

the outliers,

and ignore the 99 percent, the ordinary.

去忽略99%的普通事实。

Because that's the second example of confirmation bias.

因为这是第二个肯证偏差的例子，

We accept a fact as data.

我们接受事实作为数据。

The biggest problem is not that we live in a post-truth world;

最大的问题不在于 我们生活在后真相世界；

在于我们生活在后数据世界。

it's that we live in a post-data world.

We prefer a single story to tons of data.

比起大量的数据， 我们更喜欢简单的故事。

Now, stories are powerful, they're vivid, they bring it to life.

那些强大的，生动的，鲜活的故事。

They tell you to start every talk with a story.

他们告诉你 演讲要用故事开场。

我也是这样做的。

I did.

But a single story is meaningless and misleading

但一个简单的故事 是没有意义且误导人的，

unless it's backed up by large-scale data.

除非它有大量的数据支持。

But even if we had large-scale data,

但即便我们有大量的数据，

that might still not be enough.

这可能仍然不够。

Because it could still be consistent with rival theories.

因为它可能仍然与对立结论一致。

Let me explain.

让我解释一下。

A classic study by psychologist Peter Wason

心理学家彼得·沃森的一项经典研究

gives you a set of three numbers

给你一组三个数据

and asks you to think of the rule that generated them.

并让你思考产生这些数据的规律。

So if you're given two, four, six,

如果他们给了你三个数字： 2，4，6，

what's the rule?

规律是什么？

Well, most people would think, it's successive even numbers.

很多人会认为，这是连续的偶数。

How would you test it?

你会如何检验它？

Well, you'd propose other sets of successive even numbers:

你可以提出其他连续偶数的组合：

4, 6, 8 or 12, 14, 16.

4，6，8或者12，14，16.

And Peter would say these sets also work.

彼得说这些数组也行。

But knowing that these sets also work,

但知道这些数组也行，

knowing that perhaps hundreds of sets of successive even numbers also work,

知道数百组连续的偶数也可以，

tells you nothing.

这个结论形同虚设。

Because this is still consistent with rival theories.

因为这仍然与对立理论一致。

Perhaps the rule is any three even numbers.

也许规则可能是任意三个偶数。

Or any three increasing numbers.

或者任何三个不断增加的数字。

And that's the third example of confirmation bias:

这是第三个肯证偏差的例子：

accepting data as evidence,

接受数据作为证据，

even if it's consistent with rival theories.

即便它与对立结论一致。

Data is just a collection of facts.

数据只是事实的组合。

Evidence is data that supports one theory and rules out others.

证据是支持一种理论 排除其他理论的数据。

So the best way to support your theory

所以支持你的理论最好的方法是

is actually to try to disprove it, to play devil's advocate.

试图去反驳它， 做魔鬼的代言人（唱反调）。

So test something, like 4, 12, 26.

所以检验一下， 比如4，12，26.

If you got a yes to that, that would disprove your theory

如果你的答案是肯定的，

of successive even numbers.

那么就证明 你的连续偶数理论是不成立的。

Yet this test is powerful,

这个检验很有力，

because if you got a no, it would rule out "any three even numbers"

因为如果答案不是， 就可以排除“任何三个偶数”

and "any three increasing numbers."

和“任何三个不断增长的数字”。

It would rule out the rival theories, but not rule out yours.

它会排除对立理论， 但不排除你的理论。

But most people are too afraid of testing the 4, 12, 26,

但大部分人 不敢用4，12，26检验，

because they don't want to get a yes and prove their pet theory to be wrong.

因为他们不想肯定，

也不想承认他们的宠物理论是错的。

Confirmation bias is not only about failing to search for new data,

肯证偏差不仅指 未能搜索到新的数据，

but it's also about misinterpreting data once you receive it.

它也与你误解数据有关。

And this applies outside the lab to important, real-world problems.

这个理论也适用于 实验室外的现实世界。

Indeed, Thomas Edison famously said,

的确，托马斯•爱迪生有句名言

"I have not failed,

我没有失败，

I have found 10,000 ways that won't work."

我成功地发现了1万种行不通的方法

Finding out that you're wrong

发现你的错误

is the only way to find out what's right.

是通往成功的唯一道路。

Say you're a university admissions director

假设你是大学招生办主任，

and your theory is that only students with good grades

你的理论是只有来自富裕家庭

from rich families do well.

成绩好的学生才表现好。

So you only let in such students.

所以你只招这些学生，

And they do well.

他们都表现很好。

But that's also consistent with the rival theory.

但这也跟对立理论一致。

Perhaps all students with good grades do well,

可能所有成绩好的学生都表现好，

rich or poor.

不论富裕或贫穷。

But you never test that theory because you never let in poor students

但你永远不会测试这个理论，

因为你不会招贫穷学生，

because you don't want to be proven wrong.

因为你不想被证明错误。

So, what have we learned?

那么，我们学到了什么？

A story is not fact, because it may not be true.

一个故事不是事实， 因为它可能不是真的。

A fact is not data,

一个事实不是数据，

it may not be representative if it's only one data point.

如果只是一个数据点， 它可能不具有代表性。

And data is not evidence --

数据不是证据——

如果它与对立理论一致， 就不具有支持性。

it may not be supportive if it's consistent with rival theories.

So, what do you do?

你会怎么做？

When you're at the inflection points of life,

当你处在人生的转折点，

deciding on a strategy for your business,

去选择生意的策略，

a parenting technique for your child

孩子的育儿技巧，

or a regimen for your health,

或者健康养生，

how do you ensure that you don't have a story

你如何确保你不是基于故事

but you have evidence?

而是你拥有证据？

Let me give you three tips.

让我给你们三个提示。

The first is to actively seek other viewpoints.

首先是积极寻求其他观点。

Read and listen to people you flagrantly disagree with.

阅读和倾听你公然不同意的人。

Ninety percent of what they say may be wrong, in your view.

在你看来，他们说的90%都不对。

But what if 10 percent is right?

但如果还有10%是对的呢？

As Aristotle said,

亚里士多德说过，

"The mark of an educated man

“受过教育的标志是，

is the ability to entertain a thought

你可以不接受一种观点，

without necessarily accepting it."

但你能容纳它。”

Surround yourself with people who challenge you,

和挑战你的人在一起，

and create a culture that actively encourages dissent.

创造一种积极鼓励异见的环境。

Some banks suffered from groupthink,

一些银行受到群体思维的影响，

where staff were too afraid to challenge management's lending decisions,

员工不敢挑战管理者的借贷决策，

引发了金融危机。

contributing to the financial crisis.

In a meeting, appoint someone to be devil's advocate

开会的时候， 指定某人充当魔鬼的代言人

against your pet idea.

挑战你的宠物理论。

And don't just hear another viewpoint --

不要只让这些观点从你的脑后飘过，

listen to it, as well.

请认真倾听。

As psychologist Stephen Covey said,

正如心理学家斯蒂芬·柯维所说，

"Listen with the intent to understand,

“抱着理解的态度倾听，

not the intent to reply."

别只想着怎么回答。”

A dissenting viewpoint is something to learn from

对立的观点是值得学习的，

not to argue against.

而不是不假思索地反对。

Which takes us to the other forgotten terms in Bayesian inference.

这使我们想到 贝叶斯推断中被遗忘的部分

Because data allows you to learn,

因为数据给你学习的空间，

but learning is only relative to a starting point.

但学习只是一个起点。

如果开始就完全确信 你的宠物理论必然成立，

If you started with complete certainty that your pet theory must be true,

then your view won't change --

那么你的观点不会改变——

regardless of what data you see.

不管你看到的数据是什么。

Only if you are truly open to the possibility of being wrong

只有你真正接受犯错的可能性时，

can you ever learn.

你才能学习。

As Leo Tolstoy wrote,

正如列夫·托尔斯泰所写，

"The most difficult subjects

“最难的事情

can be explained to the most slow-witted man

也可以向最迟钝的人解释清楚，

if he has not formed any idea of them already.

只要他还没有 形成任何关于此问题的见解。

But the simplest thing

但最简单的事情

cannot be made clear to the most intelligent man

却无法向最聪明的人说清楚，

if he is firmly persuaded that he knows already."

如果他确信他已经知道答案。“

Tip number two is "listen to experts."

第二个提示是：“听专家的。”

Now, that's perhaps the most unpopular advice that I could give you.

这可能是我给你的 最不流行的建议了。

(Laughter)

（笑声）

British politician Michael Gove famously said that people in this country

英国政治家迈克尔·戈夫曾说过:

这个国家的人民已经受够专家了。

have had enough of experts.

A recent poll showed that more people would trust their hairdresser --

最近的调查显示 更多人相信他们的理发师——

(Laughter)

（笑声）

or the man on the street

或者街上的路人，

than they would leaders of businesses, the health service and even charities.

而非商界领袖、医疗服务机构、 甚至慈善机构的领导人。

So we respect a teeth-whitening formula discovered by a mom,

所以我们敬仰一位母亲 发现的牙齿美白配方，

or we listen to an actress's view on vaccination.

或者我们会听 女演员对疫苗接种的看法。

We like people who tell it like it is, who go with their gut,

我们喜欢说实话、凭直觉做事的人，

and we call them authentic.

我们觉得这叫真实。

But gut feel can only get you so far.

但直觉只能让你走这么远。

Gut feel would tell you never to give water to a baby with diarrhea,

直觉会告诉你永远不要 给腹泻的婴儿喝水，

because it would just flow out the other end.

因为它会从另外一端流出。

Expertise tells you otherwise.

专家告诉你，事实并非如此。

You'd never trust your surgery to the man on the street.

你绝不会把你的手术交给街上的人。

You'd want an expert who spent years doing surgery

你想要一个拥有多年手术经验

and knows the best techniques.

并且有最佳技巧的专家。

But that should apply to every major decision.

但这点应该应用到每个重要的决定中。

Politics, business, health advice

政治，商业，健康建议都需要专家，

require expertise, just like surgery.

就跟做手术一样。

So then, why are experts so mistrusted?

那么，为什么专家如此不被信任呢？

Well, one reason is they're seen as out of touch.

一个原因是他们脱离群众。

A millionaire CEO couldn't possibly speak for the man on the street.

一个年薪百万的总裁 不可能为街头的人发声。

But true expertise is found on evidence.

但真正的专业知识来自于证据。

And evidence stands up for the man on the street

证据支持街上的人，

and against the elites.

反对精英。

Because evidence forces you to prove it.

因为证据迫使你证明它。

Evidence prevents the elites from imposing their own view

证据不允许精英们强加他们的观点

without proof.

在没有证明的情况下。

A second reason why experts are not trusted

第二个专家不被信任的理由是，

is that different experts say different things.

不同的专家观点不同。

For every expert who claimed that leaving the EU would be bad for Britain,

只要有专家说 脱欧对英国而言弊端重重，

another expert claimed it would be good.

就会有另外一个专家说 脱欧对英国而言好处良多。

Half of these so-called experts will be wrong.

这些所谓的专家， 他们一半的观点都是错的。

And I have to admit that most papers written by experts are wrong.

我不得不承认，大多数专家 写的论文也都是错的。

Or at best, make claims that the evidence doesn't actually support.

换一种好听一点的说法，

做出证据并不支持的断言。

So we can't just take an expert's word for it.

所以我们不能只相信专家。

In November 2016, a study on executive pay hit national headlines.

2016年11月，一个关于 高管薪酬的研究登上国家头条。

Even though none of the newspapers who covered the study

尽管报道这项研究的报社

had even seen the study.

没有一家看过这项研究。

It wasn't even out yet.

这项研究甚至还没有发表。

They just took the author's word for it,

他们只是把作者的话当真了，

just like with Belle.

就像贝尔的故事一样。

Nor does it mean that we can just handpick any study

这不意味着我们可以随便挑选一个

that happens to support our viewpoint --

刚好支持我们观点的研究——

that would, again, be confirmation bias.

这也是肯证偏差。

Nor does it mean that if seven studies show A

也不意味着7项研究表明A，

and three show B,

三项研究表明B，

that A must be true.

A就必然是真的。

What matters is the quality,

重点在于质量，

and not the quantity of expertise.

而不是专家的数量。

So we should do two things.

所以我们应该做两件事。

First, we should critically examine the credentials of the authors.

首先，我们应该 严格审查作者的资历，

Just like you'd critically examine the credentials of a potential surgeon.

就像你会谨慎地审视一个

外科医生的资质一样。

Are they truly experts in the matter,

他们真的是这个领域的专家吗，

or do they have a vested interest?

或者他们有没有既得利益？

Second, we should pay particular attention

第二，我们应该特别注意

to papers published in the top academic journals.

发布在顶级期刊上的论文。

Now, academics are often accused of being detached from the real world.

如今， 学术界常常被指责与现实世界脱节。

But this detachment gives you years to spend on a study.

但这种脱节 给了你充足的时间去研究。

To really nail down a result,

去真正确定一个结果，

to rule out those rival theories,

去排除那些对立的理论，

and to distinguish correlation from causation.

并且区分因果关系。

And academic journals involve peer review,

学术期刊涉及同行评议，

where a paper is rigorously scrutinized

在这个环节，论文会被严格审查

(Laughter)

（笑声）

by the world's leading minds.

被学术界的尖端代表审查。

The better the journal, the higher the standard.

越好的期刊，标准越高。

最顶级期刊的论文拒绝率高达95%

The most elite journals reject 95 percent of papers.

Now, academic evidence is not everything.

如今，学术证据并不是一切。

Real-world experience is critical, also.

现实世界的经验也很重要。

And peer review is not perfect, mistakes are made.

同行评议也不尽完美，常犯错误。

But it's better to go with something checked

但有检查

than something unchecked.

总比没有检查好。

If we latch onto a study because we like the findings,

如果我们青睐一个研究 是因为我们喜欢这个发现，

without considering who it's by or whether it's even been vetted,

而不考虑它是谁做的 或者它是否经过审查，

there is a massive chance that that study is misleading.

这个研究就很有可能误导人。

And those of us who claim to be experts

我们这些自称专家的人，

should recognize the limitations of our analysis.

需要认识到我们分析能力的局限性。

Very rarely is it possible to prove or predict something with certainty,

确切地证明或预测某事 的可能性是很小的，

yet it's so tempting to make a sweeping, unqualified statement.

然而，发表一份全面、 不够格的声明十分诱人。

It's easier to turn into a headline or to be tweeted in 140 characters.

它们往往能成为头条 或者微博热点

But even evidence may not be proof.

即便证据并不充分详实。

It may not be universal, it may not apply in every setting.

它可能不是通用的， 可能不适用于任何条件。

So don't say, "Red wine causes longer life,"

所以不要说， “红酒能延长寿命，”

when the evidence is only that red wine is correlated with longer life.

当证据只是在 红酒与长寿相关时，

And only then in people who exercise as well.

并且样本局限在运动人群中。

Tip number three is "pause before sharing anything."

提示三： “分享任何事情前先三思。”

The Hippocratic oath says, "First, do no harm."

希波克拉底誓言（医者誓言）说， “首先，不要伤害。”

What we share is potentially contagious,

我们分享的东西 有可能会快速蔓延，

so be very careful about what we spread.

所以要谨慎地对待 我们散布的东西。

Our goal should not be to get likes or retweets.

我们的目的不应该是为了 获得点赞或转发。

Otherwise, we only share the consensus; we don't challenge anyone's thinking.

否则，我们只会分享共识；

我们不会挑战任何人的思考。

Otherwise, we only share what sounds good,

否则，我们只分享听起来好的，

无视其是否是证据。

regardless of whether it's evidence.

Instead, we should ask the following:

反过来，我们应该问如下问题：

If it's a story, is it true?

如果这是个故事，这是真的吗？

If it's true, is it backed up by large-scale evidence?

如果这是真的， 有大量的证据支持吗？

If it is, who is it by, what are their credentials?

如果有，证据是谁提供的， 他们的凭证是什么？

Is it published, how rigorous is the journal?

它发表了吗？ 这个期刊是否足够权威？

And ask yourself the million-dollar question:

并且郑重地问自己，

If the same study was written by the same authors with the same credentials

如果同样的研究 是同等资质的同一作者写的，

but found the opposite results,

但发现的是对立理论，

would you still be willing to believe it and to share it?

你仍然愿意相信和分享它吗？

Treating any problem --

处理任何问题——

a nation's economic problem or an individual's health problem,

国家经济问题或者个人健康问题，

is difficult.

很难。

So we must ensure that we have the very best evidence to guide us.

所以我们必须确保 有最佳证据指引我们。

Only if it's true can it be fact.

只有它是真的，才能成为事实。

Only if it's representative can it be data.

只有具有代表性，才能成为数据。

Only if it's supportive can it be evidence.

只有被支持，才能是证据。

And only with evidence can we move from a post-truth world

只有是证据，我们才能从后真相世界

to a pro-truth world.

走向支持真相的世界。

Thank you very much.

谢谢。

(Applause)

（鼓掌）