创作者经济如何让你说话像互联网一样

嗯哼。

Mhmm.

然后我们就继续前进。

And then we move on.

而且总有新的笑话在不断涌现。

And there's constantly new jokes percolating.

我还认为关注那些不太明显或一开始甚至不像笑话的笑话很重要。

I also think it's important to pay attention to the jokes that don't seem as obvious or don't even appear as jokes in the first place.

我认为这些词汇将更持久地改变我们的语言。

I think these are the words that are actually going to change our language more permanently.

那些通常来自少数群体或边缘现象的词汇，最终反而会对我们的语言产生影响。

Words that are percolating from usually minority groups or from more fringe phenomena, those are the ones that actually end up influencing our language.

所以我要谈谈'skibbity'和'low key'之间的区别。

So I I talk about the distinction between skibbity and low key.

'low key'是个副词。

So low key is this adverb.

这个词自19世纪以来就存在，意思是'低调的'之类的。

It's been around since the eighteen hundreds as a word meaning like muted or something.

嗯。

Mhmm.

但最近它作为副词用法流行起来，这源自非裔美国人英语，人们会说'那个派对低调地好'。

But it was recently popularized in an adverbial sense, and this is from African American English, where people would say, oh, that party was low key good.

意思是像'放松'那样吗？

Meaning like chill?

对。

Yeah.

就是'low key'。

Well, low key.

没错。

Yeah.

它有点像是'漫不经心'的意思。

It kinda means like, you know, nonchalant.

我不想过多提及，想保持低调。

I don't wanna say it too much being more low key about it.

对吧？

Right?

它是被消音的。

It's it's muted.

是一种消音状态。

It's mutedness.

是有所保留的。

It's reservedness.

还有，你那种漫不经心的用法，跟我理解的漫不经心不一样。

Also, the way you just use nonchalant, not how I use nonchalant.

哦，确实。

Oh, yeah.

‘漫不经心’这个词的语义也在演变。

The word nonchalant is evolving as well.

自2023年左右就有网络梗在讨论人们怎么使用这个词。

There's been an Internet meme since like 2023 about how people are using this.

所以我觉得当译者自己也深陷算法推送时，工作会变得很困难。

So I guess it's tough being a translator when you're also deep in the kind of algorithm yourself.

不过确实。

But yeah.

所以‘低调’这个词，以及我们看到许多其他词汇的定义变化都在发生，只是没有‘skibbidi’那么引人注目。

So low key, I think, and these changing definitions that we see with a lot of other words are also happening, but don't stick out as much as skibbidi.

我认为正是那些不显眼的词汇，将在数百年后对我们的语言产生更深远的影响。

And I think it's those words that don't stick out that are going to have a greater impact on our language hundreds of years from now.

所以你刚才说的这些，其实解释了我15岁女儿的很多行为。

So what you're just saying actually is explaining a lot of my 15 year old daughter's behavior.

我是说，青少年嘛，从来都是这样。

I mean, teenagers, whatever, it was ever this.

每一代人都有自己独特的叛逆方式。

Every generation has its own way of rebelling.

但冷漠态度绝对是个典型特征。

But nonchalance is definitely a thing.

哦，完全同意。

Oh, absolutely.

关于15岁青少年我还想强调一个重要观点。

I also an important point about the 15 year olds.

我认为这些孩子是语言发展走向的风向标。

I think these are these kids are the bellwethers of where language is gonna go.

他们才是语言行为最具可塑性的人群。

They are the ones that are actually most flexible with their linguistic behaviors.

小学生还不太能把握社会潮流，也不那么急于塑造自我身份。

The elementary schoolers aren't quite as tapped into social trends yet or as searching to create identities.

而年长者更固守语言应该怎样或词汇该如何发音的观念。

And older people are more entrenched in our ideas of how language should be or what words are supposed to sound like.

初中生则正处于身份认同的形成阶段。

And middle schoolers are forming their identity.

他们试图将自己与成年人和前几代人区分开来。

They're trying to differentiate themselves from adults and from previous generations.

他们正试图为自己构建一种集体认同。

They're trying to build a collective kind of identity for themselves.

因此他们在语言接纳上更为灵活。

And so they are more flexible with adopting language.

所以如果你想真正了解英语的现状，就该多关注15岁的孩子们。

So you should be paying attention to your 15 year old if you wanna really understand what the state of the English language is right now.

我猜有人会说，是啊。

So I guess someone would say like, yeah.

但你知道，在我的学校，我记得有些特定的词。

But in my, you know, at my school, I remember there was like certain words.

这有点恶心——四年级时有个小孩会用舌下腺体朝人吐口水，他管这叫‘舌射’。

This is kinda gross, but in fourth grade, there was a kid who would spit at people but only using his glands under his tongue, and he called it a gleek.

超级怪。

Super weird.

但这个行为在四年级突然流行起来。

But then it just took off in the fourth grade.

男生们会朝喜欢的女生‘舌射’

It was like boys would gleek the girl

那个

that

他们喜欢的。

they liked.

恶心。

Gross.

但显然，这也是个好词，因为它多年来一直萦绕在我心头。

But also, clearly, good word because it stuck with me for all these years.

但但情况一直如此。

But but that's always been the case.

比如，它与社交群体创造自己的语言，或者印刷机、广播这些媒介有什么不同？

Like, how is it different than, you know, social groups coming up with their own language or for that matter, like, the printing press, the the radio?

很高兴我们提到了印刷机和广播，因为媒介即信息。

I'm so glad we're bringing up the printing press and the radio because the medium is the message.

我们的交流方式将塑造趋势传播的路径。

The way we're talking is going to shape the way that trends diffuse.

我并不是说人类用语言交流或创造社会趋势是新鲜事。

And I'm not going to say, oh, it's new that humans are using language to communicate or come up with social trends.

不是的。

No.

这本质上就是人类的活动。

That is a fundamentally human activity.

我们会一直这样做，并且总会找到新的方式。

And we will always do it, and we will always find new ways to do it.

但现在算法正在调节这种互动，就像教室对'gleek'这个词的影响一样。

But the algorithm is now mediating that interaction in the same way that the classroom did with the word gleek.

语言的整个故事是由我们使用的交流媒介所讲述的。

The story of language as a whole is told by what mediums we're using to communicate.

在我们拥有羊皮纸、莎草纸等一切之前，我们更多依赖口头交流。

Before we had parchment and papyrus and everything, we relied a lot more on oral communication.

故事曾以押韵和格律传颂，后来我们转向羊皮纸和纸张，当时人们对此深感忧虑。

Stories would be told through rhyme and meter, and then we moved to parchment and paper, and people were really concerned about that at the time.

我认为柏拉图担心的是人们不再大量记忆会导致脑力衰退。

I think Plato was concerned about brain rot from people not memorizing things as much.

随后我们开始用纸张以不同方式分割故事。

Then we start segmenting our stories differently with paper.

对吧？

Right?

我们可以划分章节。

We can we can have chapters.

我们能在词与词之间留出空格，而不像石碑上所有文字都挤在一起。

We can actually put spaces between words as opposed to stone tablets where all the words were smooshed together.

接着我们开始用印刷机大规模生产。

And then we we start mass producing things with the printing press.

印刷机使得方言得以复制传播，同时也巩固了传统话语权掌控者的地位。

And the printing press now allows for vernacular language to reproduce and kind of solidify also at the same time the traditional gatekeepers of who gets to control how we talk.

因此有段时间，印刷机是主导的传播方式。

And so for a while, the printing press is the dominant mode of communication.

大量标准语言在各种语言中被复制，而不仅限于教会拉丁语。

You have a lot of standard language being replicated in all these different languages rather than just church Latin.

最终互联网为非正式口语的书面复制创造了机会。

And the Internet finally creates this opportunity for the written replication of informal speech.

人们开始使用俚语。

You have people using slang words.

人们开始用小写字母书写并使用缩写。

You have people writing in lower case and using abbreviations.

这某种程度上是一场关于谁有权发言、谁有权决定语言形态的革命。

And this is sort of this revolution in who can talk and who can vote on what language is supposed to be like.

这为我认为的新转折点奠定了基础。

And that sort of lays the foundation for what I think is the new inflection point.

M to the b, some to the b, some

And m to the b, some to the b, some

算法掌控着内容的分发。

Algorithms govern the distribution of content.

我还不知道要在这里发布什么类型的视频。

I don't know what type of videos I'm gonna post on here yet.

好的。

Okay.

但如果你做的一切都对，为什么你的内容没有爆火？

But why is your content not blowing up if you're doing everything right?

但让我来指点你

But let me put you on

除非符合算法机制，否则内容不会病毒式传播。

Things don't go viral unless they work for the algorithm.

这意味着我们使用的语言某种程度上也始终在为算法服务。

And what that means is that the language we're using is always sort of performing for the algorithm as well.

而这种媒介始终存在，不断塑造着思想在网络上的传播方式。

And this is this medium that is always there, that is always shaping how ideas diffuse online.

这意味着任何传播开来的思想要么对算法有利，要么就是为算法而生的。

And that means any idea that diffuses is either good for the algorithm or it's created for the algorithm.

这就导致了像'非活体'这类词汇的使用。

That means using words like unalive.

这就催生了诸如'Skibbidi'这类趋势的跟风。

That means playing into trends like skibbidi.

这意味着许多创作者被激励去创造新词或使用流行语，因为算法将这些语言视为元数据。

That means a lot of creators are just incentivized to come up with new words or use trendy language because the algorithm sees that language as metadata.

过去，元数据更像是标签这类能提供内容信息的东西。

In the past, metadata was something like a hashtag, something that gives you information about the content.

现在算法能洞察一切。

Now the algorithm sees everything.

自然语言处理、计算机视觉组件等所有子算法都在分析这些数据，从而生成视频内容的数字化表征。

The natural language processing, computer vision components, all sub algorithms analyze that, and they create this numerical representation of what a video is.

嗯。

Mhmm.

它们早已洞悉视频内容。

And they already know what the video is about.

而词汇正是其中的组成部分。

And but the word is part of that.

因此我们被激励使用那些有助于视频传播的词汇。

So we're incentivized to use words that aid in the distribution of our own video.

只是

Just

跟进一下，你知道的，就是马歇尔·麦克卢汉那个经典理论——媒介即讯息。

to follow-up on, you know, the classic Marshall McEwen idea of the medium is the message.

算法才是真正塑造语言的力量。

The algorithm is the one that's shaping language.

但需要明确的是，算法是由人编写的，这些人为想要变现的科技巨头工作。

But to be clear, the algorithm is programmed by people, people who work for very large tech companies who want to monetize.

完全正确。

Absolutely.

他们建立的这些机制会被后来的网红们效仿，比如强调需要互动、需要留存率（即观众观看视频的时长）、需要点赞。

And they create those structures that are then replicated by other influencers down the line that, oh, you need engagement, you need retention, which is how long people watch a video, you need likes.

所以现在作为网红，我制作视频时就要想方设法让观众看完，还要获得点赞。

And so now, as a influencer, I'm trying to make all my videos to get people to watch the whole time, and I wanna get likes.

否则视频就不会被推荐，我只能顺应这套规则。

Otherwise, the video is not gonna be distributed, so I have to play into that.

我还想指出，有些社会现象并非完全由工程师编程决定。

I also want to note that there's emergent social phenomena that aren't programmed just by these engineers.

并不是嗯...

It's not Mhmm.

纯粹自上而下的过程。

Purely like this top down process.

每种媒介都伴随着相应的社会期待。

There is social expectations that come with each medium.

这里是NPR新闻华盛顿分台现场报道。

Live from NPR News in Washington.

我是戴夫·马丁利。

I'm Dave Mattingly.

在NPR，你会听到一种经过专业训练的播音腔调。

On NPR, you have, like, an NPR voice that's been studied.

特朗普总统正展开他上任后的首次正式出访

President Trump is making his first official trip of

显然，我并不擅长适应这种风格。

Clearly, I am not good at adapting into this.

不过我觉得他们多年来一直想摆脱这种腔调。

Oh, I think they've been trying to get rid of that for years, though.

但与此同时

But at the

我就在这里。

same time Here I am.

嗯哼。

Uh-huh.

你现在用的可能是更符合广播要求的发声方式。

You're speaking in a maybe a more radio friendly voice.

我正用着算法更易识别的发音方式，因为这是我们进行这类交流时学会的说话方式。

I'm speaking in a more algorithm friendly voice because that's what we've we've learned to talk when we're kind of communicating like this.

等等。

Wait.

有人可能会问，你说的算法友好型声音是什么意思？

Somebody might be like, what do you mean an algorithm friendly voice?

解释一下。

Define that.

实际上网红口音有很多种类型，但你可能最熟悉那种著名的生活方式类网红口音。

There's actually a bunch of different types of influencer accents, but you're probably familiar with the most famous sort of lifestyle influencer accent.

就是那种'嗨，大家好'的调调。

The hey, guys.

欢迎收听NPR。

Welcome to NPR.

今天我们要讨论算法。

Today, we're gonna be talking about algorithms.

带着那种上扬的尾音。

It has that uptalk.

语调总是往上扬。

It has the rising tone.

呃，你可能没

Well, could you're not

收到任何人的回应。

getting any response from anyone.

所以你就会

So you're like

呃，它

Well, it

试图与自己对话。

trying to have a conversation with yourself.

确实如此。

Exactly.

这是一种单向交流，虽然对塑造它很重要，但远不止于此。

This is a one-sided communication that is an important part of shaping it, but that's not all.

要知道，如果我只是一直面无表情地对着镜头说话，这种视频是不会爆红的，因为人们对说话方式是有期待的。

You know, I could just continuously monotone deadpan into the camera and that wouldn't go viral because there's an expectation of how people are supposed to talk.

这类生活方式博主的腔调源自早年的YouTube口音，而YouTube口音又源自'山谷女孩'口音。

And this sort of lifestyle influencer accent came out of the previous YouTube accent, which came out of the Valley Girl accent.

所以我们说话方式里存在社会声望的层级结构，人们对生活方式博主应该怎么说话是有既定期待的。

So there's layers of social prestige in how we talk, and there's an expectation of what the lifestyle influencer is supposed to sound like.

等等。

Now, wait.

等一下。

Wait.

稍等。

Wait.

你是说人类大脑会被这种声音吸引，从而获得更多收听量，所以算法就持续推送它？

Are you saying that the human brain is attracted to those and therefore it gets more listens and therefore the algorithm continues to push it?

还是说算法能识别这些语调特征，然后更大力地向人们推送？

Or are you saying that the algorithm is hearing those intonations and pushing it harder on people?

不是。

No.

我认为算法没有捕捉到口音语调。

I don't think the algorithm is picking up on accent intonations.

我认为人类确实存在对社会声望的偏见。

I think there is actual human bias toward social prestige.

我们通过自己的视角评估各种社交情境，比如判断某件事是否值得倾听。

We evaluate any kind of social situation through our lens of like, is this something worth listening to?

然后我们的轻微滑动习惯会通过算法推荐被复制传播。

And then, like, our slight scroll patterns get replicated through what the algorithm recommends.

算法甚至不需要分析这个人是否在用生活方式影响者口音说话。

The algorithm doesn't even need to analyze whether or not this person is speaking in a lifestyle influence or accent.

它只是表明：'哦，这个观众快速滑走了，因为他们认为说话者没有使用社会认可的口音，这意味着我要减少推荐这个视频的用户量'。

It's merely telling that, oh, this this viewer scrolled away quickly because this person wasn't speaking in a socially approved accent in their minds, and that means I'm gonna recommend this video to fewer users.

因此某种程度上形成了对社会认可口音的筛选复制机制。

And so there's sort of a selection that happens for socially desirable accents to replicate.

我还想澄清这仅是其中一种最刻板印象化的口音，因为女性在语言表达上总是受到更严苛的审视。

And I also wanna clarify that's only one type of accent that's the most stereotyped because, you know, women are scrutinized with language.

我用的是自称为'教育类影响者口音'的说话方式。

I speak in a what I call an educational influencer accent.

我会说得非常快。

I'll talk very quickly.

我会加重更多词语的发音来吸引你继续观看我的视频。

I'll stress more words to keep you watching my video.

不过我感觉你一直都是个语速很快的人。

I get the sense though that you were a fast talker always.

不是？

No?

我确实语速很快，而且在这里我确实暴露了自己。

I definitely am a fast talker, and I'm definitely giving away myself here.

但更重要的是，我在视频中追求完美，如果语调不对或措辞不当，或者语速不够快，我会重拍某些片段。

But also, there's a perfection that I strive for in the videos where I'll retake certain clips if it seems like I don't have the right intonation on the right word, or if I stumble on my words, if I don't talk fast enough.

所以现在对我来说更像是一种更自然的说话方式，如果在更大的群体对话中，我可能会说得不一样，那种场合不需要这样，或者更随意，更含蓄。

And so what now is still like a kind of more natural way of speaking for me, and I, again, would probably speak differently if I was in a larger group conversation where you don't need to or it's like more laid back, you know, it's more demure.

你真是太淡定了，亚当。

You're so nonchalant, Adam.

天啊。

My god.

我不想总是显得漫不经心。

I don't wanna be shalant all the time.

有时需要漫不经心，有时则需要淡定。

There's a time and a place to be shalant, and there's a time to be nonchalant.

等等。

Wait.

稍等一下。

There's a hold on.

漫不经心和淡定有区别吗？

There's a difference between shalant and nonchalant?

嗯，淡定是指保持克制。

Well, nonchalant is being reserved.

然后

And then

但'shalant'这个词本身并不存在

But shalant is not its own word.

什么？

What?

但它确实存在

But it is.

现在它存在了

It is now.

这是一个通过社交媒体流行起来的俚语

It's a slang word that was popularized through, like, social media.

意思是？

Meaning?

人们用它表示与非冷淡相反的意思

And people talk meaning the opposite of nonchalant.

意思是，你对某事表现得很明显

Meaning, you are overt about something.

就像'兴奋'那样？

Like jazzed?

对

Yeah.

你对某事感到兴奋

You're excited about something.

我对语言很随意，你懂的。

I'm chalant about language, you know.

当我说兴奋时，你是那种'哇，87岁吗'之类的反应吗？

When I say jazzed, are you like, woah, 87 or what?

我是说，我可能直接用'随意'这个词。

I mean, I probably would just use the word shalant.

没错。

Exactly.

稍后回来，我们将讨论亚当对社交媒体用户（尤其是年轻用户）的关切与乐观看法。

When we come back, Adam's concern and optimism for social media users, younger ones in particular.

另外，亚当和我要火了。

Plus, Adam and I go viral.

也许吧。

Maybe.

哦，你居然不知道

Oh, you didn't know about the

我们该怎么开始？

How do we begin?

哦，抱歉。

Oh, sorry.

糟透了。

Terrible.

好吧。

Okay.

本期节目将探讨创作者经济及其对语言的影响。

Today on the show, the creator economy and its effects on language.

我是马努莎·扎莫罗迪，您正在收听的是NPR出品的TED广播时间。

I'm Manousha Zamorodi, and you're listening to the TED Radio Hour from NPR.

请继续关注我们。

Stay with us.

这里是NPR出品的TED广播时间。

It's the TED Radio Hour from NPR.

我是马努莎·扎莫罗迪。

I'm Manouche Zamorodi.

我们一直在与亚当·阿莱克西克讨论创作者经济。

We've been talking to Adam Alexic about the creator economy.

亚当自称是社交媒体语言学家。

Adam calls himself a social media linguist.

他还是《算法语言：社交媒体如何重塑语言未来》一书的作者。

He's also the author of Algo Speak, how social media is transforming the future of language.

他的职业生涯致力于阐释词汇使用中体现的文化与代际鸿沟。

And he has built his career explaining the cultural and generational divide on how we use words.

在19世纪时我们还没有'社交世代'这个概念。

We didn't have the idea of a social generation in the eighteen hundreds.

这算是新鲜事物。

This is like a new thing.

当时人们正从第一次世界大战归来。

People are coming back from World War one.

我们称他们为迷惘的一代。

We called them the lost generation.

哦，那些从二战归来的人们。

Oh, the people coming back from World War two.

那是大兵一代。

That's the GI generation.

接着我们有了婴儿潮一代。

And we have baby boomers.

然后我们其实不知道该怎么称呼你们，就是之后的那一代人。

And then we actually don't know what to call you, the the next generation after that.

我们就想，好吧，也许他们不想被定义，所以就叫他们X世代吧。

We're like, well, I guess they don't have a desire to be defined, so we'll call them Gen X.

然后我们就继续沿用世代存在的这个概念。

And then we just keep running with the idea that generations exist.

再后来就有了千禧一代，对，因为是2000年嘛。

And then we have like, millennial, yeah, because it's 2,000.

然后呢，我们也不知道该怎么称呼Z世代。

And then, well, we don't know what to call Gen Z either.

所以我们就随便用X世代之后的第二个字母来命名。

So we'll just do whatever was two after Gen X.

现在字母都用完了，所以就叫阿尔法世代吧。

And now, we just ran out of the alphabet, so we'll call it Gen Alpha.

但我认为网红们助长了这些标签，因为人类社交中存在着对被贴标签的迷恋。

But I think influencers play into these labels because there is a human social fascination with being labeled.

我们喜欢MBTI、星座这类事物，也喜欢归属于某个群体和世代的概念。

We love things like MBTIs and zodiacs and the idea of belonging to a bucket and with a generation.

还有简写。

And shorthand.

对吧？

Right?

确实如此。

It is.

这就是简写。

It is shorthand.

显然这过于简化了，我想人们也意识到了这点。

Obviously, it's reductive, and I think people realize that.

但我们比以往更频繁地使用世代标签，因为它确实有效。

But we're using more generational language than we were before because it works.

比如用'Z世代手指心形'和'千禧一代心形'来区分就很有效。

It works to talk about the Gen Z finger heart versus the millennial heart.

或者在X世代之前的人拍短视频时，他们会先做个小小的吸气动作。

Or before Gen X people will start a short form video, they'll they'll do, like, a little breath in.

嗯哼。

Uh-huh.

我们这样把自己归类实在太奇怪了，而这些标签实际上成了算法趋势的诱饵，用来宣称'哦这是Z世代的特征'。

It's it's so strange to me that we put ourselves in buckets like that, and they actually serve as algorithmic trend bait to say that you are, oh, this is a Gen z thing.

这是...嗯哼。

This is a Uh-huh.

Z世代的凝视表情最近在网上疯传。

The Gen z stare went viral recently.

我做过一个关于婴儿潮一代用省略号的视频。

I did a video on the boomer ellipses.

他们的消息中间总会出现随意的点点点。

There's always a random dot dot dot in the middle of their messages.

哦，我就这样。

Oh, I do that.

我故意那么说是因为知道会火。

I phrased it that way because I knew it'd go viral.

比如，为什么婴儿潮一代要在文字里加句点？年轻人可不会这么做。

Like, why do boomers put dots in their text, which is not a normal thing that young people do?

所以分开表达想法其实更高效

So it's simply more efficient to separate ideas

但这就像你说话时声音渐弱。

But it's like trailing off in your language.

是啊。

Yeah.

对啊。

Yeah.

只是不同世代有不同的短信礼仪。

It's just it there's different texting etiquette among different generations.

再说用'世代'这个词本身，其实所有语言都在把事物分类。

And again, to use that word generation already feels like I mean, all of language is putting things into categories.

但我认为某些分类方式比其他方式更具简化性。

But I think some categories are more reductive than others.

例如，后缀'core'最近在Z世代俚语中非常流行，用来描述特定的美学风格，比如'cottage core'、'goblin core'或'angel core'。

For example, the suffix core has recently gotten very popular in Gen z slang to describe specific aesthetics, like cottage core or goblin core or angel core.

从表面上看，这些都很可爱。

And on the surface level, these are cute.

你看了个田园风视频，觉得喜欢。

You watch a cottage core video, you like it.

之后你会收到更多田园风内容。

Later on, you get more cottage core content.

你甚至可能开始认同田园风美学。

You might even start to identify with the cottage core aesthetic.

但问题是。

But here's the thing.

这一切都是假的。

It's all fake.

这些美学风格存在的全部原因，是因为TikTok算法认定'cottagecore'这类词汇属于热门元数据。

The entire reason these aesthetics exist is because TikTok algorithm has decided that words like cottagecore qualify as trending metadata.

于是创作者们就制作更多田园风内容来传播这个词，更多人互动后又让这个词更热门。

So creators respond by making more cottagecore content that propagates the word, and then more people interact with it, which makes the word trendier.

这种现象源于社交媒体算法希望你认同高度细分化的标签，这样它们就能为你提供针对该身份的极度精准的商业化内容。

And this happens because social media algorithms wants to make you identify with hypercompartmentalized labels since they can then give you extremely specific commercialized content catering to that identity.

现在你成了田园风爱好者，每次收到田园风视频都会觉得自己很特别。

Now that you're a cottagecore person, you feel special every time you get a cottagecore video.

你是那种田园风的人吗？

You're like, cottagecore?

算法真的很了解我。

Well, the algorithm really knows me.

算法给了你这个身份。

The algorithm gave you that identity.

你甚至可能开始购买田园风服装或装饰来适应你作为田园风人士的新生活方式，而这正是他们想要的。

You might even start buying cottagecore clothing or cottagecore decorations to fit your new lifestyle as a cottagecore person, and that's exactly what they want.

最疯狂的是他们甚至没打算隐藏这一点。

The craziest part is they're not even trying to hide this.

这正是资本主义创客经济的核心。

This is this is at the heart of the capitalist maker economy.

不是吗？

No?

这是理解算法运作原理的关键点。

This is such an important point to understand about how algorithms work.

他们想要更多分类。

They want more categories.

如果我是一个半性恋地精风Z世代霉粉，对他们来说就再好不过了。

It's really good for them if I'm a demisexual goblin core Gen Z Swiftie.

我完全听不懂你在说什么，不过你继续。

I don't even know what you just said, but yes, go on.

重点是，这些都是小标识符，是关于我作为用户的一小段元数据。

The point is, those are all little identifiers, little pieces of metadata about myself as a user.

算法实际上并不需要这些文字，但人类感觉到自己在为算法表演，于是创造出这些词汇，循环往复中自发形成了一个认同该标签的用户群体，这反过来又为算法提供了更多可奖励的内容。

The algorithm actually doesn't need the words, but humans sense that they are trying to perform for the algorithm, and they create these words, and they actually circularly, emergently make more of a cluster of users who identify with this label, and then it makes more for the algorithm to reward.

这一切都是监控资本主义体系的一部分，现在他们有更多关于我们的标签，能以更具体的方式锁定我们。

And it's all part of this surveillance capitalism system where now they have more things to label about us, more ways to target us more specifically.

如果你查看2021年TikTok的商业页面，上面说亚文化已成为新的人口统计特征，并为企业提供了如何从田园风、坏女孩美学等标签中获利的思路。

If you look at TikTok's business page in 2021, it says that subcultures are the new demographics, and it gives businesses ideas for how to profit off things like cottagecore and hashtag baddie aesthetic.

最让我震惊的是，他们居然在页面上使用了'人口统计'这个词。

And what's crazy to me about that page, they are using the word demographic.

嗯。

Mhmm.

因为过去的人口统计特征指的是种族、年龄、性别。

Because in the past, a demographic was like race, age, gender.

而现在还包括你是否属于田园风。

And now it's also whether you're cottagecore or not.

作为一个X世代的人，我内心有一部分在想：老兄，你这完全是在出卖自己啊。

So so as a Gen Xer, part of me is like, dude, you're totally selling out.

但我猜那是一种极具时效性的测试方式，用来检验你为何选择现在的职业。

But I'm guessing that that was an extremely time specific way of testing you on why you are in the career you're in.

我想我非常清楚这种矛盾——作为一个创作者谈论这些词汇并关心算法如何影响我们的文化时必须要面对的割裂感。

I think I'm extremely aware of the dissonance, I guess, that is required to be a creator talking about these words and caring about how algorithms are affecting our culture.

我认为颠覆性地使用这些算法是可能的，否则我根本不会这么做。

I I think it is possible to use these algorithms subversively, otherwise, I I would not.

另外，如果我们回归麦克卢汉的媒介理论，我认为混合多种媒介形式也非常重要。

It's also, I think, very important to mix forms of media if we're going back to McLuhan and media theory.

有位学者叫哈罗德·英尼斯，嗯。

There's a scholar called Harold Innis Mhmm.

他在加拿大的麦克卢汉之前。

Who came before McLuhan in Canada.

我认为他关于空间偏向与时间偏向传播的观点非常精辟。

I I think he has a fascinating point about space biased versus time biased communication.

时间偏向媒介与文化记录共存，比如书籍和口头传统。

And time biased sticks with the cultural record, like books and oral traditions.

虽然要触达同等规模的受众可能需要更长时间，但空间偏向媒介能迅速占据大量空间，却难以持久。

And so while it might take way longer to reach an audience that's the same size, space biased media fills up a lot of space really quickly, but doesn't stick around.

像报纸、大众媒体、电视，或者现在的算法，都具有极强的空间偏向性。

Something like a newspaper, or mass media, or TV, or right now algorithms are incredibly space biased.

它们充斥我们的意识，然后转瞬即逝。

They they fill our consciousness, and then they go away.

看来两种媒介类型各有利弊。

So it seems that both types of media have their pros and cons.

对吧？

Right?

空间偏向媒介擅长覆盖广泛受众，时间偏向媒介则利于保持文化延续性。

Space biased media is very good for communicating to a big audience, and time biased media is very good for maintaining some kind of cultural constancy or longevity.

因此我认为应该尽可能混合使用两种媒介，通过接触多元媒介来构建更全面的社会认知。

And so I think we should be mixing as much time biased and space biased media as possible, where we build a more holistic picture of society and reality by consuming and engaging with all these different types of media.

这就是我写书的原因。

So that's why I wrote a book.

是啊。

Yeah.

我确实想问你这个问题。

I did wanna ask you that.

写书在某种程度上似乎有些复古，但根据你刚才说的，我是否可以认为你很高兴创作了它？

Writing a book seems kind of retro in a way, but based on what you've just said, am I right to assume that you were happy to create it?

嗯，一是因为你拿到了报酬，二是能让更多人了解你的想法。

Well, a, because you got paid, and b, gets your ideas to more people.

但第三点更重要，它更有可能经受住时间的考验，不像你的一些TikTok视频？

But c, it's more tangible that it'll stand the test of time, maybe unlike some of your TikTok videos?

没错。

Yeah.

我觉得没人会看我2023年的TikTok视频了。

I don't think anybody's looking at my TikTok videos from 2023.

但两年后可能还会有人看这本书。

Somebody might look at this book two years from now though.

所以思想传播的方式确实存在差异，不同媒介会带来不同的传播效果。

So there is a difference to how ideas can diffuse, and ideas will diffuse differently depending on the medium.

这其中很多都是人类社会构建的层层规则。

So a lot of this is just built up layers of human social constructs.

比如NPR规定的谈话方式，和书籍要求的写作规范——顺便说句，这本书用的是非正式英语。

The NPR social construct of how we're supposed to talk versus the book social construct of how I'm supposed to write, which is informal English, mind you.

此外还有TikTok平台要求的表达方式。

And then there's the TikTok construct of how I'm supposed to talk.

每一种情境对我应如何沟通都有不同层次的尊重或理解。

And each of these has a different layer of respect or understanding of how I should be communicating.

那么亚当，你究竟是谁？

So who are you really, Adam?

我认为优秀的沟通者应该懂得适应不同媒介、不同类型人群和不同社交场合，我们时刻都在进行语境转换。

I think a good communicator should know to adapt to different media and to different types of people and to different social settings, and we are constantly code switching all around us.

你和祖母交谈时的状态与和挚友聊天时完全不同，这再正常不过。

You you are a different person talking to your grandmother than talking to your best friend, and that is super normal.

祖母家的氛围与你和朋友聚会的酒吧本就是不同的沟通媒介。

The your grandmother's house is a different medium than the bar that you're hanging out with your friend at.

所以根据媒介切换状态是人类天性的一部分，但某种程度上仍要保持概念上的一致性。

So this is a aspect of human nature to switch with the medium, but kind of maintain conceptual constancy, I guess.

始终坚守自己的道德体系和行动方向。

Still have your own moral system and your own direction with what you're trying to do.

我邀请亚当和我一起拍个视频。

I asked Adam to make a video with me.

对你并不了解。

Didn't know about you.

我们怎么开始？

How do we begin?

哦，抱歉。

Oh, sorry.

用他的爆款公式试试能否走红。

And use his formula to see if we could go viral.

灯光有点差。

The lighting's kinda bad.

迎合算法。

By playing to the algorithm.

是啊。

Yeah.

好吧。

Okay.

对啊。

Yeah.

我们的首要需求，一个时髦的新词。

Our first requirement, a trendy new word.

哦，你不知道'clanker'这个词吗？

Oh, you didn't know about the word clanker?

而且你总需要一个好噱头。

And you always need a good hook.

X世代对阵Z世代。

Gen X versus Gen Z.

不。

No.

我完全不知道。

I had no idea.

我是X世代。

I'm Gen X.

这算是一种对机器人的臆想性蔑称。

It's kind of a speculative slur for robots.

这个想法是它们未来会变得更有感知力，而现在就可以被当作人类少数群体对待。

So the idea that they would become more sentient in the future and they could be treated as a human minority today.

这个视频的重点是让人们思考：为什么我们要为人类创造的事物发明蔑称。

The point of this video, to get people thinking about why we would even invent a slur for something that humans created.

在这个案例中，我们正用'铁皮佬'这个蔑称来指代机器人。

In this case, robots, which we are now calling clankers as a slur.

没错。

Exactly.

所以这个想法是对机器人'哐当声'的贬义化演绎，同时也影射了N开头的那个词。

So the idea is a derogatory play on the idea of clanking as a robot, but also the n word itself.

这个概念源自《星球大战》，在那里机器人被视为次等种族，有很多关于'铁皮佬'作为底层阶级的设定，但某种程度上也是未来拟人化的公民。

It comes from Star Wars where robots are seen as sort of a secondary race, and there's a lot of ideas about clankers being a lower class, but sort of anthropomorphized citizen in the future.

从某些角度看确实合理。

In some ways, makes sense.

比如它们不是人类，难道不该被当作铁皮佬对待吗？

Like, they're not humans, so shouldn't they be treated as clankers?

你知道，这个想法是未来它们会产生意识，而我们却拒绝承认这点。

You know, it's just the idea that in the future they would become conscious, and we aren't recognizing that.

这有点像网络迷因，设想2060年你女儿带回家一个铁皮佬男友。

It's sort of an online meme that's speculating about your daughter bringing home a clanker in the year 2060.

所以就像，2060年我女儿会带回家一个对她特别特别好的机器人，比真人男友更体贴，但他是个铁皮佬。

So, like, 2060, my daughter's gonna bring home a robot who is really, really nice to her, therefore, a nicer boyfriend than a real human, but he's kind of a clanker.

对。

Right.

嗯，这个想法就是你不希望自己的女儿和机器人约会。

Well, the idea is you don't want your daughter dating a robot.

确实如此。

Amen to that.

嗯，到了2060年，这种想法可能就成种族歧视了。

Well, in 2060, that might be racist.

天啊。

Oh god.

一年后'clanker'这个词还会存在吗？

Will the word clanker be around a year from now?

也许不会。

Maybe not.

这些流行词和梗图算是当今世界最紧迫的问题吗？

Are these trendy words and memes among the world's most urgent issues?

当然不算。

Definitely not.

尽管它们有趣又新奇，但这个算法和梗图的世界也有其阴暗面。

But as much as they are fun and interesting, there is a darker side to this world of algorithms and memes.

我确实想确认一下，有些人正利用语言和算法运作方式来传播各种观点。

I do wanna make sure that I ask you, there are people who are using language and the way the algorithm works to get across all kinds of ideas.

最近就发生了令人不安的可怕事件——保守派网红查理·柯克被杀害了。

And, you know, most recently, the very upsetting and awful occurrence of Charlie Kirk, a conservative influencer, was killed.

但我很好奇，听说他在网上积累了大量粉丝，部分原因在于他对语言的独特运用。

But I guess I'm curious to hear he built a big following online in part through a very particular use of language.

对吧？

Right?

它简短有力，专为病毒式传播设计。

It was short, punchy, geared for viralness.

从你的角度来看，这如何说明社交媒体平台正在塑造公众人物沟通方式的现象？

From your perspective, how does that illustrate the way that social media platforms are shaping the way public figures communicate?

是的。

Yeah.

有种容易走红的'陷阱式'风格，你会在本·夏皮罗或查理·柯克过去的内容中看到。

There's a sort of gotcha style that goes viral that you'll see Ben Shapiro or Charlie Kirk, in the past use.

就是那种——保持简洁，做成适合剪辑传播的片段。

That's like, keep it concise, keep it clip farmed.

比如能作为30秒片段走红的尖酸语录。

So something that can go viral as a thirty second clip, some snarky sound bite.

谁更正确不重要，重要的是谁更能病毒式传播。

It doesn't matter who's more correct, just who can go more viral.

如果你的传播力比对手强，就能压制反驳或反对观点。

And if you can go more viral than the other person, you can outcompete like rebuttals or counterpoints.

所以重点在于抛出观点，而非给出深思熟虑的回应。

So it's more about just get your idea out there than have some measured response.

算法总是奖励极端内容。

And algorithms reward extreme things.

对吧？

Right?

当我们讨论参与度优化时，这些算法为了吸引我们的注意力所采取的手段。

When we were talking about engagement optimization, what these algorithms are doing to get our attention.

不幸的是，真正能吸引注意力的往往是极端的内容。

Things that are good at getting attention are unfortunately really extreme things.

极左和极右的观点与声音将会被放大。

So far left and far right opinions and voices are going to be amplified.

你知道的，AOC和玛乔丽·泰勒·格林总会比我家乡纽约奥尔巴尼的议员保罗·唐科更容易走红。

You know, AOC and Marjorie Taylor Greene are always gonna go viral than my congressman where I grew up in Albany, New York, Paul Tonko.

无意冒犯。

No offense.

他有点无聊。

He's he's kinda boring.

他不会走红。

He doesn't go viral.

对吧？

Right?

他的观点没有极端之处，所以在算法上传播不了那么远。

There's no kind of extreme point to his views, so it's not gonna go as far on the algorithm.

因此极端观点比温和观点更具竞争力，传播更广，可能导致更多人开始接受这些观点。

And so you have this ability of more extreme views to outcompete less extreme views, diffuse further, and then more people maybe start adopting that.

这为人们捐款提供了新机会，或者说我认为特朗普是第一位算法总统，就像肯尼迪是第一位电视总统一样。

It represents a new chance for people to donate money or I think Donald Trump is the first algorithmic president in the same way that Kennedy was the first TV president.

他之所以能战胜尼克松当选，很大程度上是因为他更上镜。

He was famously kind of elected over Nixon because he looked more photogenic.

而我认为唐纳德·特朗普是更具网络迷因特质的总统。

And I think Donald Trump is the more memeable president.

如果说上镜的候选人适合电视时代，那么具有迷因特质的总统就更适合算法时代。

And if photogenic candidates were good on TV, memeable presidents are good for algorithms.

举个例子来说明——还是回到查理·柯克的话题——这种现象在现实中如何体现。

I mean, one illustration I think of how just back to Charlie Kirk, how this can play out in real life.

我曾和一位当妈妈的朋友聊天，她不好意思地承认自己从没听说过查理·柯克。

I had a conversation with a friend, a mom, who said to me, I I hate to admit it, I had not heard of Charlie Kirk.

但我问她14岁的儿子时，他回答说：'拜托老妈，我当然知道他。'

But I asked my 14 year old son if he had, and he said, duh, mom, of course I have.

'我可是社交媒体上的白人14岁少年。'

I'm a white 14 year old kid on social media.

'他们投放的内容就是专门针对我的。'

I'm targeted exactly by them.

'当然看过他的视频。'

Of course, I've seen his videos.

关键在于他完全明白自己接收的内容及其推送原因。

The fact that he understood what he was getting and why he was getting it.

你认为年轻一代是否更精明？他们是否会因为思考媒介属性而质疑接收到的信息？

Do you think that this the younger generation is more savvy, and are they questioning what the information that they're getting because they are thinking about the medium?

是的。

Yes.

我对年轻一代确实充满乐观。

I actually have so much optimism for the younger generation.

实际上我认为老一辈人对新媒体可能更加陌生，这令人担忧。

I actually think older generations are maybe more concerningly illiterate of the new medium.

最早被低质内容欺骗的是脸书上的婴儿潮一代，比如那个病毒式传播的'虾仁耶稣'。

The first people to get tricked by slop were boomers on Facebook, and there was like a shrimp Jesus that went viral.

我认为年轻人普遍能较好地理解'哦，这可能是AI生成的'。

I think young people generally have a decent understanding of, oh, this could be AI.

这可能不是真实的。

This might not be real.

或者说，这是算法在作祟，算法不一定会展示我真正想看的内容。

Or like, this is the algorithm, and the algorithm is not necessarily showing me what I want.

当人们较少接触新媒体时，他们更难意识到某些内容是虚假或被过度呈现的。

When people engage less with new media, they are less aware of how some things are fake or overrepresented.

但没错，这确实仍是年轻人被激进化的渠道之一。

But yes, definitely, it is still a pipeline for younger people to get radicalized.

很多极端意识形态直接源自4chan论坛。

A lot of this comes directly out of extreme ideology built on 4chan.

算法形成的某些语言社群可能具有实际危害性。

Some of the linguistic communities created by the algorithm can be actively harmful.

许多年轻人开始用后缀'-pilled'表示'被说服接受某种生活方式'。

Many younger people have started using the suffix pilled to mean convinced into a lifestyle.

比如如果我最近发现自己特别爱吃墨西哥卷饼，就可以说'我完全墨西哥卷饼pilled了'。

If I recently discovered that I really like eating burritos, for example, I can say, I'm so burrito pilled.

但这个词是通过类比'black pilled'形成的，后者指被说服接受非自愿独身主义意识形态。

But that word was formed through analogy with black pilled, a term meaning convinced into incel ideology.

如今非自愿独身主义者是危险的厌女群体，但他们的词汇却通过算法提供的空间渗透进了Z世代俚语。

Now incels are a dangerous misogynistic group, and yet somehow the vocabulary is filtering into Gen Z slang because the algorithm gave these hate groups the space.

而且很多人根本不知道这些词的来源。

And again, many people don't even know where it came from.

但对少数可能对其核心理念感兴趣的人来说，网络俚语的传播方式让这些思想变得更易接触了。

But for the few people who might be interested in the underlying idea, it's now more accessible to them because of the way that slang spreads on the Internet.

你认为这意味着什么？

What do you think that means?

你刚才还说对年轻一代持乐观态度。

You just said you were an optimist for younger generation.

但听完你说的这些，我非常担忧。

But after hearing you say that, I I worry a lot.

我乐观的是年轻一代能认清这些东西的本质。

I'm an optimist for younger generations having literacy about this stuff.

但对算法及其文化影响方面，我并不乐观。

I am not an optimist in terms of algorithms and how they impact our culture.

我想把语言和文化区分开来，因为我经常同时讨论这两者。

I do wanna draw that distinction between language and culture because I sort of comment on both a lot.

我认为语言本身没问题，单个词汇从不会出问题。

And I think language is fine, and there's nothing wrong with individual words.

它们永远不会腐蚀大脑。

They're never brain rot.

我也认同基安蒂语所指向的文化变迁趋势。

I do think the same way as Chianti language points us towards greater cultural shifts.

你会发现很多词汇源自4chan男性圈层，我们应当关注这些词汇为何会从该圈层中产生。

You see a lot of words coming from that 4chan manosphere space, and we should be paying attention to why are these words coming out that space.

因为这个圈层擅长制造传播其意识形态的模因，有时其意识形态与极右翼思想和美学深度交织。

It's because that space is good at creating memes that export their ideology, and sometimes their ideology is very intertwined with far right ideas and aesthetics.

这某种程度上也是算法自身的问题——它们如何助推极端观点。

And, yeah, that's kind of a problem of algorithms themselves, how they push extreme perspectives.

先前谈到媒体融合时，我真心希望我们能形成一种文化共识：不应仅依赖算法获取新闻，当然主流媒体也存在自身偏见。

What I said earlier about mixing media, I really hope that we come to realize as a culture that we should not just be relying on algorithms for our news, but also, you know, mainstream news has its own biases.

从尽可能多的渠道获取新闻有助于构建更完整的现实图景。

It's good to get news from as many places as possible and build a greater picture of reality.

通过现实人际交往来了解他人也很重要，因为算法呈现的社会面貌是扁平化的。

It's good to build your stories of who people are through in person connections as well because algorithms are going to show you a flattened representation of what society is.

我认为我们确实需要保持警觉。

I do think we should be aware.

当我们的说话方式可能已被算法塑造时，我们应当有所觉察。

We should be aware when the way we're talking may have been conditioned by the algorithm.

当我们使用的词汇可能是被设计来推销商品时，我们应当有所觉察。

We should be aware when the words we're using may have been engineered to sell us things.

当我们的语言复读极端主义论调时，当这种语言可能伤害他人时，我们都应当有所觉察。

We should be aware when our language regurgitates extremist rhetoric, and we should be aware when that language can be used to harm other people.

我们总体上应当关注词源学，因为这能帮助我们更好地理解当下的自己。

We should be aware of etymology in general because it helps us better understand who we are today.

我们应该保持警觉。

We should be aware.

最后，我还要分享一个俚语给大家。

And with that, I've just one final piece of slang for you.

这是年轻人在结束冗长解释时常用的说法。

It's a common phrase used by younger people when we finish a long winded explanation of something.

感谢收听我的TED演讲。

Thanks for coming to my TED talk.

刚才发言的是亚当·亚历克西克。

That was Adam Alexic.

他是《算法语言：社交媒体如何改变语言未来》的作者。

He's the author of algo speak, how social media is transforming the future of language.

你可以在社交媒体上通过@etymologynerd找到他，也可以在ted.com观看他的演讲。

You can find him at etymology nerd on social media, and you can see his talk at ted.com.

非常感谢大家收听今天的节目。

Thank you so much for listening to our show today.

如果喜欢我们的节目，请留言评论。

If you enjoyed it, leave us a comment.

你正在使用哪些新词汇？

What words are you using?

我们很想知道。

We would love to know.

你可以在Spotify留言或发邮件至TedRadioHour@NPR.org告诉我们。

You can do that on Spotify Spotify or email us at TedRadioHour@NPR.org.

我们很高兴收到您的反馈。

We love hearing from you.

本期节目由马修·克卢捷制作，由萨娜兹·梅什坎普尔和我负责剪辑。

This episode was produced by Matthew Cloutier and edited by Sanaz Meshkampur and me.

NPR的制作团队成员还包括詹姆斯·德拉胡西、瑞秋·福克纳·怀特、凯蒂·蒙特莱昂、菲奥娜·吉隆、哈沙姆·内贾达和菲比·莱特。

Our production staff at NPR also includes James Dellahusi, Rachel Faulkner White, Katie Monteleone, Fiona Giron, Harsham Nejada, and Phoebe Lett.

我们的执行制片人是艾琳·野口。

Our executive producer is Irene Noguchi.

音频工程师是达米安·赫林和吉利·穆恩。

Our audio engineers were Damian Herring and Gilly Moon.

主题音乐由拉姆廷·阿拉布鲁伊创作。

Our theme music was written by Ramtine Arablui.

TED的合作伙伴包括克里斯·安德森、罗克珊·海拉什和丹妮拉·巴拉拉佐。

Our partners at TED are Chris Anderson, Roxanne Hailash, and Daniella Balarazzo.

我是马努什·扎莫罗迪，您正在收听的是NPR出品的TED广播时间。

I'm Manoush Zamorodi, and you've been listening to the TED Radio Hour from NPR.