How technology can fight extremism and online harassment | Yasmin Green

譯者: Lilian Chiu 審譯者: Yanyan Hong

My relationship with the internet reminds me of the setup

我與網際網路的關係，

讓我想起老套恐怖片的情境。

to a clichéd horror movie.

You know, the blissfully happy family moves in to their perfect new home,

幸福快樂的家庭， 搬進一間美好的新房子，

excited about their perfect future,

興奮期待完美的未來，

and it's sunny outside and the birds are chirping ...

外頭陽光普照，鳥兒在唱歌……

And then it gets dark.

接著電影就變黑暗了。

And there are noises from the attic.

閣樓傳出噪音。

And we realize that that perfect new house isn't so perfect.

我們發現，那間美好的 新房子並沒有那麼美好。

When I started working at Google in 2006,

2006 年，當我開始 在谷歌（Google）工作時，

Facebook was just a two-year-old,

臉書（Facebook）才剛推出兩年，

and Twitter hadn't yet been born.

推特（Twitter）甚至還沒問世。

And I was in absolute awe of the internet and all of its promise

我對網際網路及它所有的承諾 感到絕對的敬畏，

to make us closer

它承諾要讓我們

and smarter

更靠近且更聰明，

and more free.

還有給予更多自由。

But as we were doing the inspiring work of building search engines

但當我們開始進行 這鼓舞人心的工作，

建立搜尋引擎，

and video-sharing sites and social networks,

建立影片分享網站和社交網路，

criminals, dictators and terrorists were figuring out

罪犯、獨裁者， 及恐怖分子都在想辦法

how to use those same platforms against us.

如何用同樣的平台來對抗我們。

And we didn't have the foresight to stop them.

我們沒有先見之明來阻止他們。

Over the last few years, geopolitical forces have come online to wreak havoc.

在過去幾年，地緣政治學的 勢力也上網展開大破壞。

And in response,

造成的反應是，

Google supported a few colleagues and me to set up a new group called Jigsaw,

谷歌支持我和幾位同事， 成立一個小組，叫做「Jigsaw」，

with a mandate to make people safer from threats like violent extremism,

我們的使命是要讓大家更安全， 避免受到像是極端主義、

censorship, persecution --

審查制度、迫害的威脅——

threats that feel very personal to me because I was born in Iran,

我個人對這些威脅特別有感， 因為我是在伊朗出生的，

and I left in the aftermath of a violent revolution.

在一場暴力革命的餘波中， 我被迫離開了伊朗。

But I've come to realize that even if we had all of the resources

但我漸漸了解到， 即使我們有所有的資源，

of all of the technology companies in the world,

有全世界所有的科技公司，

we'd still fail

如果我們忽略了

if we overlooked one critical ingredient:

一項關鍵因素，我們仍然會失敗：

the human experiences of the victims and perpetrators of those threats.

那些威脅的受害者 與加害者的人類經歷。

There are many challenges I could talk to you about today.

今天我其實可以 與各位談很多的挑戰。

I'm going to focus on just two.

但我只打算著重兩項：

The first is terrorism.

第一是恐怖主義。

So in order to understand the radicalization process,

為了要了解激進化的過程，

we met with dozens of former members of violent extremist groups.

我們和數十名暴力 極端主義團體的前成員見面。

One was a British schoolgirl,

其中一位是英國的女學生，

who had been taken off of a plane at London Heathrow

她曾在倫敦希斯洛機場 被強拉下飛機，

as she was trying to make her way to Syria to join ISIS.

因為當時她打算 去敘利亞加入伊斯蘭國。

And she was 13 years old.

她當時只有十三歲。

So I sat down with her and her father, and I said, "Why?"

我和她及她父親坐下來談， 我說：「為什麼？」

And she said,

她說：

"I was looking at pictures of what life is like in Syria,

「我在看一些敘利亞 生活寫照的圖片，

and I thought I was going to go and live in the Islamic Disney World."

我以為我是要去住到 伊斯蘭的迪士尼樂園。」

That's what she saw in ISIS.

這是她看到的伊斯蘭國。

She thought she'd meet and marry a jihadi Brad Pitt

她以為她要去見一位聖戰士中的 布萊德彼得並嫁給他，

and go shopping in the mall all day and live happily ever after.

整天都能去購物， 從此幸福快樂地生活。

ISIS understands what drives people,

伊斯蘭國知道什麼能驅使人，

and they carefully craft a message for each audience.

他們會為每一位觀眾 精心策劃一則訊息。

光是去看看他們把他們的行銷素材

Just look at how many languages

they translate their marketing material into.

翻譯成多少語言，就能了解了。

They make pamphlets, radio shows and videos

他們會製作小冊子、 廣播節目，和影片，

in not just English and Arabic,

不只用英語和阿拉伯語，

but German, Russian, French, Turkish, Kurdish,

還有德語、俄語、法語、 土耳其語、庫德語、

Hebrew,

希伯來語、

Mandarin Chinese.

華語（中文）。

I've even seen an ISIS-produced video in sign language.

我甚至看過一支伊斯蘭國 製作的影片是用手語的。

Just think about that for a second:

花點時間思考一下：

ISIS took the time and made the effort

伊斯蘭國投入時間和心力，

to ensure their message is reaching the deaf and hard of hearing.

來確保他們的訊息 能夠傳達給聽障人士。

It's actually not tech-savviness

伊斯蘭國能贏得人心和人信，

that is the reason why ISIS wins hearts and minds.

並不是因為他們很精通科技。

It's their insight into the prejudices, the vulnerabilities, the desires

因為他們有洞見，了解 他們試圖接觸的人有什麼

of the people they're trying to reach

偏見、脆弱、慾望，

that does that.

才能做到這一點。

That's why it's not enough

那就說明了為什麼

for the online platforms to focus on removing recruiting material.

線上平台只把焦點放在 移除召募素材是不足的。

If we want to have a shot at building meaningful technology

如果我想要有機會建立 一種有意義的技術，

that's going to counter radicalization,

用它來對抗極端化，

we have to start with the human journey at its core.

我們就得要從它核心的 人類旅程開始著手。

So we went to Iraq

所以，我們去了伊拉克，

to speak to young men who'd bought into ISIS's promise

去和年輕人交談， 我們找的對象曾相信伊斯蘭國

of heroism and righteousness,

所做的關於英雄主義與公正的承諾，

who'd taken up arms to fight for them

曾拿起武器為他們作戰，

and then who'd defected

接著，在目擊了

after they witnessed the brutality of ISIS's rule.

伊斯蘭國統治的殘酷之後選擇變節。

And I'm sitting there in this makeshift prison in the north of Iraq

我坐在北伊拉克的一間臨時監獄裡，

with this 23-year-old who had actually trained as a suicide bomber

會見一位在變節前 受過訓練的自殺炸彈客，

before defecting.

年僅 23 歲。

And he says,

他說：

"I arrived in Syria full of hope,

「我到敘利亞時滿懷著希望，

and immediately, I had two of my prized possessions confiscated:

可一下子我兩項最重要的 東西就被沒收了：

my passport and my mobile phone."

我的護照和我的手機。」

The symbols of his physical and digital liberty

在他抵達時， 這兩樣象徵他實體自由

were taken away from him on arrival.

與數位自由的東西被奪去了。

And then this is the way he described that moment of loss to me.

接著，他這樣向我描述迷失的時刻。

He said,

他說：

"You know in 'Tom and Jerry,'

「在《湯姆貓與傑利鼠》中，

when Jerry wants to escape, and then Tom locks the door

當傑利想要逃脫時，湯姆把門鎖住，

and swallows the key

把鑰匙吞掉，

and you see it bulging out of his throat as it travels down?"

還可以從外表形狀看到鑰匙 延著喉嚨下滑，記得嗎？」

And of course, I really could see the image that he was describing,

當然，我能看見他所描述的畫面，

and I really did connect with the feeling that he was trying to convey,

我真的能和他試圖傳達的 這種感受產生連結，

which was one of doom,

一種在劫難逃的感受，

when you know there's no way out.

你知道沒有路可逃了。

And I was wondering:

而我很納悶：

What, if anything, could have changed his mind

在他離家的那一天，如果有的話，

the day that he left home?

什麼能改變他的心意？

So I asked,

於是，我問：

"If you knew everything that you know now

「如果你當時知道 你現在知道的這些

about the suffering and the corruption, the brutality --

關於苦難、腐敗、殘酷的狀況——

that day you left home,

在離家的那天就知道，

would you still have gone?"

你還會選擇離開嗎？」

And he said, "Yes."

他說：「會。」

And I thought, "Holy crap, he said 'Yes.'"

我心想：「老天爺，他說『會』。」

And then he said,

接著他說：

"At that point, I was so brainwashed,

「在那個時點，我完全被洗腦了，

I wasn't taking in any contradictory information.

我不會接受任何有所矛盾的資訊。

I couldn't have been swayed."

我當時不可能被動搖。」

"Well, what if you knew everything that you know now

「那麼如果你在你離開前 六個月就已經知道

six months before the day that you left?"

你現在知道的這些，結果會如何？」

"At that point, I think it probably would have changed my mind."

「若在那個時點， 我想我可能會改變心意。」

Radicalization isn't this yes-or-no choice.

極端化並不是關於是非的選擇。

It's a process, during which people have questions --

它是一個過程，在這過程中， 人們會有問題——

about ideology, religion, the living conditions.

關於意識型態、宗教、 生活條件的問題。

And they're coming online for answers,

他們會上網尋求答案，

which is an opportunity to reach them.

這就是一個接觸他們的機會。

And there are videos online from people who have answers --

有答案的人會在網路提供影片——

defectors, for example, telling the story of their journey

比如，叛逃者訴說他們

投入和脫離暴力的心路歷程；

into and out of violence;

stories like the one from that man I met in the Iraqi prison.

就像我在伊拉克監獄見到的 那名男子告訴我的故事。

There are locals who've uploaded cell phone footage

有當地人會上傳手機影片，

of what life is really like in the caliphate under ISIS's rule.

呈現在伊斯蘭國統治之下 穆斯林國的真實生活樣貌。

There are clerics who are sharing peaceful interpretations of Islam.

有教會聖職人員分享 關於伊斯蘭的和平詮釋。

But you know what?

但你們知道嗎？

These people don't generally have the marketing prowess of ISIS.

這些人通常都沒有 伊斯蘭國的高超行銷本領。

They risk their lives to speak up and confront terrorist propaganda,

他們冒著生命危險說出來， 和恐怖主義的宣傳對質，

and then they tragically don't reach the people

但很不幸的是，他們無法接觸到

who most need to hear from them.

最需要聽到他們聲音的人。

And we wanted to see if technology could change that.

我們想試看看， 科技是否能改變這一點。

So in 2016, we partnered with Moonshot CVE

2016 年，我們和 Moonshot CVE （信息泄露組織）合作，

to pilot a new approach to countering radicalization

試驗一種對抗極端化的新方法，

called the "Redirect Method."

叫做「重新定向法」。

It uses the power of online advertising

它用線上廣告的力量，

to bridge the gap between those susceptible to ISIS's messaging

在容易受到伊斯蘭國訊息影響的人

與揭發那些訊息真面目的 可靠聲音之間搭起橋樑。

and those credible voices that are debunking that messaging.

And it works like this:

它的運作方式如下：

someone looking for extremist material --

有人在尋找極端主義的素材——

say they search for "How do I join ISIS?" --

比如他們搜尋 「如何加入伊斯蘭國？」——

will see an ad appear

就會有一則廣告出現，

that invites them to watch a YouTube video of a cleric, of a defector --

邀請他們上 YouTube， 看聖職人員、變節者的影片——

someone who has an authentic answer.

有真實答案的人所拍的影片。

And that targeting is based not on a profile of who they are,

這個方法鎖定目標對象的方式 不是依據個人資料，

but of determining something that's directly relevant

而是由與他們的查詢或問題有直接

to their query or question.

相關的東西來決定。

During our eight-week pilot in English and Arabic,

我們用英語和阿拉伯語 做了八週的測試，

we reached over 300,000 people

接觸到了超過三十萬人，

who had expressed an interest in or sympathy towards a jihadi group.

他們都是對聖戰團體 表示感興趣或同情的人。

These people were now watching videos

現在這些人在看的影片

that could prevent them from making devastating choices.

能預防他們做出毀滅性的選擇。

And because violent extremism isn't confined to any one language,

因為暴力極端主義 不侷限於任何一種語言、

religion or ideology,

宗教，或意識形態，

the Redirect Method is now being deployed globally

「重新定向法」現已在全球實施，

to protect people being courted online by violent ideologues,

保護大家上網時不會受到 暴力意識形態的引誘，

whether they're Islamists, white supremacists

不論是伊斯蘭教的、 白人至上主義的，

or other violent extremists,

或其他暴力極端主義的，

with the goal of giving them the chance to hear from someone

我們的目標是要 給他們機會去聽聽看

on the other side of that journey;

在那旅程另一端的人怎麼說；

to give them the chance to choose a different path.

給他們機會去選擇不同的路。

It turns out that often the bad guys are good at exploiting the internet,

事實證明， 壞人通常擅長利用網際網路，

not because they're some kind of technological geniuses,

並不是因為他們是什麼科技天才，

but because they understand what makes people tick.

而是因為他們了解人的癢處。

I want to give you a second example:

我再舉個例子說明：

online harassment.

線上騷擾。

Online harassers also work to figure out what will resonate

線上騷擾者也在致力於 找出什麼能讓

with another human being.

另一個人產生共鳴。

But not to recruit them like ISIS does,

但他們的目的不像 伊斯蘭國是要招募人，

but to cause them pain.

而是造成別人痛苦。

Imagine this:

想像這個狀況：

you're a woman,

你是一名女子，

you're married,

已婚，

you have a kid.

有一個孩子。

You post something on social media,

你在社交媒體上發了一篇文，

and in a reply, you're told that you'll be raped,

你得到一則回應，說你會被強暴，

that your son will be watching,

你的兒子會被監視，

details of when and where.

還有時間和地點的細節資訊。

In fact, your home address is put online for everyone to see.

事實上，在網路上大家 都能看到你家的地址。

That feels like a pretty real threat.

那威脅感覺十分真實。

Do you think you'd go home?

你認為你會回家嗎？

Do you think you'd continue doing the thing that you were doing?

你認為你會繼續做你正在做的事嗎？

Would you continue doing that thing that's irritating your attacker?

你會繼續做那件惹惱了 攻擊你的人的那件事嗎？

Online abuse has been this perverse art

線上虐待一直都是種 刻意作對的藝術，

of figuring out what makes people angry,

找出什麼能讓人生氣，

what makes people afraid,

什麼能讓人害怕，

what makes people insecure,

什麼能讓人沒有安全感，

and then pushing those pressure points until they're silenced.

接著去壓那些對壓力敏感之處， 直到它們被壓制下來。

When online harassment goes unchecked,

當線上騷擾不受管束時，

free speech is stifled.

自由言論就會被扼殺。

And even the people hosting the conversation

即使主持對話的人

throw up their arms and call it quits,

棄械並宣佈到此為止，

closing their comment sections and their forums altogether.

把他們的留言區以及論壇都關閉。

That means we're actually losing spaces online

那意味著，我們其實正在失去線上

to meet and exchange ideas.

可以碰面交換點子的空間。

And where online spaces remain,

在還有線上空間的地方，

we descend into echo chambers with people who think just like us.

我們會陷入到迴音室當中， 只和相同想法的人聚在一起。

But that enables the spread of disinformation;

但那會造成錯誤訊息被散佈出去；

that facilitates polarization.

那會促成兩極化。

What if technology instead could enable empathy at scale?

但如果反之能用科技 來大量產生同理心呢？

This was the question that motivated our partnership

就是這個問題

驅使我們和谷歌的反虐待小組、

with Google's Counter Abuse team,

Wikipedia

維基，

and newspapers like the New York Times.

以及像紐約時報這類報紙合作。

We wanted to see if we could build machine-learning models

我們想要看看我們 是否能建立出能夠了解

that could understand the emotional impact of language.

語言會帶來什麼情緒影響的 機器學習模型，

Could we predict which comments were likely to make someone else leave

我們能否預測什麼樣的意見 有可能會讓另一個人

the online conversation?

離開線上對談？

And that's no mean feat.

那不是容易的事。

That's no trivial accomplishment

人工智慧要做到

for AI to be able to do something like that.

那樣的事，並不是理所當然。

I mean, just consider these two examples of messages

我是指，想想這兩個例子，

都是在上週我有可能 會收到的訊息。

that could have been sent to me last week.

「祝在 TED 順利！」 （直譯：在 TED 斷一條腿。）

"Break a leg at TED!"

... and

以及

"I'll break your legs at TED."

「我會在 TED 打斷你一條腿。」

(Laughter)

（笑聲）

You are human,

你們是人，

that's why that's an obvious difference to you,

那就是為何你們能明顯看出，

even though the words are pretty much the same.

用字幾乎相同的兩個句子有何差別。

But for AI, it takes some training to teach the models

但對人工智慧來說， 要透過訓練來教導模型

to recognize that difference.

去辨視那差別。

The beauty of building AI that can tell the difference

建立能夠分辨出那差別的 人工智慧，有個美好之處，

is that AI can then scale to the size of the online toxicity phenomenon,

就是人工智慧能處理 線上毒素現象的規模，

and that was our goal in building our technology called Perspective.

為此目的，我們建立了 一種出名為「觀點」的技術。

With the help of Perspective,

在「觀點」的協助下，

the New York Times, for example,

以紐約時報為例，

has increased spaces online for conversation.

他們增加了線上交談的空間。

Before our collaboration,

在與我們合作之前，

they only had comments enabled on just 10 percent of their articles.

他們的文章只有 大約 10% 有開放留言。

With the help of machine learning,

在機器學習的協助下，

they have that number up to 30 percent.

這個數字提升到了 30%。

So they've tripled it,

增加了三倍，

and we're still just getting started.

且我們才剛開始合作而已。

But this is about way more than just making moderators more efficient.

這絕對不只是讓版主更有效率。

Right now I can see you,

現在，我可以看見你們，

and I can gauge how what I'm saying is landing with you.

我可以估量我所說的話 會如何對你們產生影響。

You don't have that opportunity online.

在線上沒有這樣的機會。

Imagine if machine learning could give commenters,

想像一下， 當留言者在打字的時候，

如果機器學習能夠

as they're typing,

real-time feedback about how their words might land,

即使給他們回饋， 說明他們的文字會造成什麼影響，

just like facial expressions do in a face-to-face conversation.

就像在面對面交談時， 面部表情的功能。

Machine learning isn't perfect,

機器學習並不完美，

and it still makes plenty of mistakes.

它仍然會犯許多錯誤。

But if we can build technology

但如果我們能建立出

that understands the emotional impact of language,

能了解語言有什麼 情緒影響力的技術，

we can build empathy.

我們就能建立同理心。

That means that we can have dialogue between people

那就表示，我們能讓兩個人對話，

with different politics,

即使他們政治立場不同，

different worldviews,

世界觀不同，

different values.

價值觀不同。

And we can reinvigorate the spaces online that most of us have given up on.

我們能讓大部分人已經放棄的 線上空間再度復甦。

When people use technology to exploit and harm others,

當人們用科技來利用和傷害他人時，

they're preying on our human fears and vulnerabilities.

他們靠的是我們人類的恐懼和脆弱。

If we ever thought that we could build an internet

如果我們認為我們能夠建立一個完全

insulated from the dark side of humanity,

沒有人性黑暗面的網際網路，

we were wrong.

我們就錯了。

If we want today to build technology

如果現今我們想要建立技術

that can overcome the challenges that we face,

來克服我們面臨的挑戰，

we have to throw our entire selves into understanding the issues

我們就得把自己全心全意投入， 並去了解這些議題，

and into building solutions

去建立人類的解決方案，

that are as human as the problems they aim to solve.

來解決人類的問題。

Let's make that happen.

讓我們來實現它吧。

Thank you.

謝謝。

(Applause)

（掌聲）