Welcome to the World of Audio Computers | Jason Rugolo | TED

翻译人员: Yip Yan Yeung 校对人员: Yanyan Hong

You know, when I was a kid,

当我还是个孩子的时候，

dreaming about the amazing future that computers could bring,

梦想着计算机可以带来的美好未来，

I never thought it would look like this.

我从没想过它会是这样的。

I snapped this photo in line at a Chipotle, thinking,

我在 Chipotle 排队时 拍下了这张照片，心想：

"Man, what has the world come to?"

“兄弟，世界变成什么样了？”

You know, everyone's stuck in their phones all the time.

每个人无时无刻不被困在手机里。

And then I almost doubled over laughing at myself because there I am,

然后我差点加倍嘲讽自己， 因为看看我在干什么，

stuck in my phone being judgy about these people, stuck in theirs.

我被困在手机里，对这些人 被困在他们的手机里指指点点。

The truth is that we could all benefit from a little less screen time.

真相是我们都可以 从减少一点屏幕时间中受益。

And so how do we push back

我们该如何抵制诱惑，

and create a healthier relationship to our technology?

与我们的技术 建立更健康的关系呢？

I've been trying to figure out what comes next,

我一直在努力搞明白 接下来会发生什么，

what's the technology that we want to be using?

我们想使用什么技术？

I spent three years funding deep tech at ARPA-E,

我在 ARPA-E 花了三年时间 为深度科技提供资金，

and then I moved to Google X,

然后我跳槽到了谷歌的 “登月工厂”Google X，

Google's moonshot factory,

before creating a spin out called iyo.

然后打造了一款 名为 iyo 的衍生产品。

The last 10 years, diligently, and some may say obsessively,

在过去的 10 年里， 我勤勤恳恳地，也许是“如痴如醉地”

trying to peek beyond the curve.

试图瞥见未来的趋势。

What I think is next is that we need an entirely new kind of computer.

我认为接下来我们需要 一台全新的计算机。

One that speaks our language

它会说我们的语言，

instead of forcing us to speak their language of swipes and clicks.

而不是强迫我们说 它们的“滑动和点击”语言。

A computer that we can talk to.

一台我们可以与之交谈的计算机。

And not in the way that you speak at Siri

不是你在 Siri 上大声地 用机器人语音命令说话的方式，

with loud, robotic voice commands,

but in the way that we talk with each other.

而是以我们彼此交谈的方式。

So genuine, engaging conversation.

真情实感、引人入胜的对话。

So a new kind of natural language computing

一种基于新型计算硬件的 新型自然语言计算。

built on a new kind of compute hardware.

So we've been building this new kind of computer in secret for six years now,

我们秘密建造这种 新型计算机已经有六年了，

and today is the first time that we're talking about it publicly.

今天是我们第一次公开谈论它。

So this is a prototype “audio computer,”

这是一台我们称之为 “音频计算机”的原型机。

as we call them.

And the first thing that you'll see is they don't have a screen.

你会注意到的第一点就是 它们没有屏幕。

We've got perfectly good screens in our pockets,

我们的口袋里有非常好的屏幕，

and we're hoping that we could leave them there or even at home

我们希望我们能更频繁地 将它们丢在口袋里，

a little bit more often.

甚至放在家里。

So these are not just fancy-looking earbuds.

它们不只是外观酷炫的耳机。

They're an entire computer,

它们是一台完整的计算机，

and we think they're going to replace some of what we do

我们认为它们会取代 我们习惯使用视觉计算机

with the visual computers that we're used to.

完成的一部分工作。

We've created a new audio-based user interface,

我们打造了一个全新的 基于音频的用户界面，

which relies on natural language as the input

依赖自然语言作为输入，

and what we call “auditory space” as the output.

以我们所谓的“听觉空间”作为输出。

And we're hoping that these innovations will help us create

我们希望这些创新能够帮助我们

a new kind of relationship with computers,

与计算机建立一种新的关系，

where the iyo is more natural, more intuitive.

在这种关系中， iyo 更自然、更直观。

So you can just talk,

你只需要说话，

have a natural-language conversation

与各种强大的音频应用程序 进行自然语言对话。

with all kinds of different powerful audio apps.

So here's a really simple example.

这是一个非常简单的例子。

Hey, Q.

嘿，Q。

Voice: Hi, Jason. How are you doing?

语音：嗨，杰森。你怎么样？

Jason Rugolo: Hey, I'm good, I'm doing the TED talk right now.

杰森·鲁戈洛（Jason Rugolo）： 嘿，我很好，我现在正在 TED 演讲。

Voice: You got this, Jason, how can I support you?

语音：没问题，杰森。 我该怎么帮助你呢？

JR: I don't know, you know I'm nervous talking in front of this many people.

JR：我不知道， 我在这么多人面前说话很紧张。

Voice: Yeah, I know you don't like public speaking.

语音：是啊，我知道 你不喜欢公开演讲。

How can I help?

我能帮上什么忙呢？

JR: I guess I could use a little bit of encouragement.

JR：来点鼓励吧。

Voice: Do you want me to share a positive message from a loved one?

语音：你想让我分享一条 来自你挚爱之人的加油消息吗？

JR: Yeah, that'd be great.

JR：好啊，太棒了。

Maybe just one, I don't want to bore people.

就说一条吧， 我可不想让观众听烦了。

Voice: Your dad texted,

语音：你爸爸发短信说：

"Good luck today, Jason. I'm so proud of you."

“祝你今天好运，杰森。 我真为你感到骄傲。”

JR: That's awesome, thanks, Q.

JR：太棒了，谢谢，Q。

Can you tell him I said thanks and tell him I'm up on stage right now?

你能告诉他我说了谢谢， 再告诉他我现在在台上吗？

He'll get a kick out of that.

他会懂的。

Voice: No problem.

语音：没问题。

JR: So any suggestions for how I could loosen up a little bit?

JR：那有什么建议 可以让我放松一下吗？

Voice: Do you want to hear a joke to lighten the mood?

语音：你想听个笑话， 放松一下心情吗？

JR: OK, yeah, maybe if it's on theme.

JR：好的，来说个切合主题的。

Do you know any tech start-up jokes?

你知道什么科技创业公司的笑话吗？

Voice: How many start-up founders does it take to change a light bulb?

语音：换个灯泡要多少位 创业公司创始人？

JR: I don't know, tell me.

JR：不知道，你说吧。

Voice: Only one.

语音：就一个。

He just holds it in place, and the world revolves around him.

他只要拿着它， 世界就会围着他转的。

(Laughter)

（笑声）

JR: OK, sick burn.

JR：好吧，膝盖中箭。

I'm going to get back to the talk here.

说回演讲。

So do you see the difference between those robotic voice commands

你看出机器人语音命令 和对话之间的区别了吗？

and just a conversation?

Natural language is just more natural.

自然语言更自然。

It's intuitive, it's better.

它很直观、更好。

So what makes this kind of conversation possible?

那这种对话是怎么做到的呢？

It's the single most revolutionary breakthrough

这是信息技术历史上 最具革命性的突破。

in the history of information technology.

And don't worry, I'm not talking about our devices,

别担心，我说的不是我们的设备，

although they're pretty cool.

尽管它们很酷。

And I don't mean LLMs either, which are obviously a building block here.

我也不是在说 LLM， 它显然是基石。

What I'm talking about is something much older.

我说的是更古老的东西。

It's the first uniquely human form of communication,

是第一种人类特有的沟通形式，

the one that we naturally learn as children

是我们小时候自然学会的，

and has its structure built into our brain.

它的结构已融入我们的大脑。

It's the very thing I'm doing right now.

这正是我现在正在做的事情。

Talking.

说话。

Spoken language emerged in tandem with the evolution of human consciousness,

口语是随着人类意识的演变而出现的，

and to this day, it remains our most efficient

时至今日，它仍然是我们最有效

and emotionally robust form of communication.

和情感上最强大的沟通形式。

Conversation is not just transmitting ideas from one person to another.

对话不仅仅是将想法 从一个人传递给另一个人。

It's more like thinking together.

这更像是一起思考。

Modern neuroscientists have pioneered a whole new approach to the brain.

现代神经科学家 开创了一种研究大脑的新方法。

It's called second-person neuroscience,

它叫做“第二人称神经科学”，

and it’s built on the notion that how we think is not isolated.

基于“我们的思考方式不是孤立的” 这一概念。

It’s collective, and it happens out loud.

它是共同的，而且是说出来的。

Not just through words,

不只是通过文字，

but through subtle signals of tone and prosody,

还要通过微妙的信号， 来自语气、韵律、

your timbre and your pitch and intensity.

音色、音调和强度。

And neuroscience is just not complete

神经科学在你将第二个人加入 社交动态之前，都是不完整的。

until you add a second person into this full social dynamic.

So why can't we have a computer that we can talk with in that way?

那为什么我们不能让计算机 这样和我们说话呢？

With that kind of natural language.

说着自然语言。

A computer that has superhuman processing speed.

一台具有过人处理速度的计算机。

And it has access to the internet.

还可以联网。

And it’s been trained on the entire written record of human thought.

而且它是根据人类思想的 全部书面记录进行训练的。

But engages with you like a person would,

但像人一样和你互动，

that understands your intention

它能理解你的意图，

and that taps into the superpower of human natural language understanding.

发挥人类自然语言理解的超能力。

That's the promise of audio computing.

这就是音频计算的前景。

So think about not just how it can replace many of the things

不仅要考虑它如何取代 你在手机上所做的许多事情，

that you do on your phone

but actually make them better.

还要考虑如何真正让它们变得更好。

So take email, for example.

以电子邮件为例。

We pull out our phones, we swipe, we scroll,

我们拿出手机，滑动、滚动，

we furiously type with our thumbs.

用大拇指疯狂地打字。

Wouldn't it be better to just sit back with a cup of coffee

坐下来喝杯咖啡，

and to be briefed in a conversation?

在谈话中听取总结不是更好吗？

Or search, search is a big one.

或者搜索，搜索是个大问题。

It's an incredible technology

这是一项神奇的技术，

that made the world a radically better place.

它让世界变得更加美好。

But with these audio computers,

但是有了这些音频计算机，

you can just talk out loud about anything that you want to know.

你可以说出你想知道的任何事情。

It just feels so normal.

感觉很正常。

So there's a big difference between giving a voice command

这和给五大语音助手中的 任意一个下语音指令

to one of the big five voice assistants,

有很大的区别，

which are these structured, predefined choose-your-own-adventure dialogue models

它们是结构化、预定义的、 剧情游戏的对话模型，

that we all have to learn,

我们都得学习，

and I'm sure have all felt that frustration,

而且我敢肯定， 所有人都有过恼火的感觉，

and just having a real conversation.

只是进行一次真正的对话。

These natural-language applications can get to know you

这些自然语言应用程序

可以用我们互相了解的方式了解你。

in the same way that we get to know each other.

They build context about our lives just through us talking over time.

它们只是通过持续和我们交流 建立了我们生活的背景信息。

So later, take out your phone, look at all those apps,

之后请拿出手机， 看看这些应用程序，

all those candy-colored icons,

所有这些鲜艳的图标，

and think about how could you accomplish the same thing

想想我们怎么能仅仅通过对话 达到同样的目的？

but through conversation?

Or how could you make it better?

或者你怎么能让它变得更好？

You won't be able to do Instagram or TikTok,

你将无法使用 Instagram 或抖音，

those apps whose content is mostly visual.

这些应用程序的内容主要是视觉的。

But wouldn't it be better to spend a little bit less time in those apps,

但是在这些应用程序上少花点时间，

or just to need your screen a little bit less?

或者降低一些对屏幕的需要 不是更好吗？

So our goal is to be heads-up and hands-free

我们的目标是在一天中 多抬抬头，多释放双手一些。

for a little bit more of the day.

You know, just get back into the world.

回到世界中去。

Of course, if the auditory user interface or the AUI, as we call it,

当然，如果听觉用户界面 或我们所说的“AUI”

is going to really integrate into your life,

要真正融入你的生活，

it has to feel private and convenient to use.

它必须让人感到私密 且易于使用。

So that's why we built it as an all-day wearable for the ear.

这就是我们将其打造成 一款全天候可佩戴耳机的原因。

But your ears are for hearing first and foremost.

但是你的耳朵首先是用来听的。

And so if you're going to wear a computer on them all day,

如果整天都得在耳朵上 戴着一台计算机，

we can't mess that up.

我们可不能搞砸了。

In fact, we should probably make that better too.

我们也应该把它做得更好。

So these audio computers,

在过去六年的艰苦研发中， 这些音频计算机

over the last six grueling years of R and D,

became a sort of mixed-reality device.

变成了一种混合现实设备。

It's like the Apple Vision Pro, but for audio,

就像 Apple Vision Pro 那样， 但是是音频版本的，

where we can pass through and we can modify your ambient acoustics,

我们可以穿过， 也可以修改你的环境声，

giving you an unprecedented control over your personal soundscape.

让你以前所未有的方式 掌控自己的个人音景。

In order to do mixed-audio reality,

为了实现混合音频现实，

we sort of had to hack the auditory system

我们必须破解听觉系统，

to be able to display sound in ultra-high fidelity, spatially,

才能在空间上以超高保真度播放声音，

as if it's all around you.

就像声音全方位环绕着你。

So there's this whole field of research.

这是一整个研究领域。

It’s called psychoacoustics, which we’ve led on for years.

它被称为“心理声学”， 我们已经引领了这项研究多年。

We built this giant audio structure.

我们打造了这个巨大的音频结构。

It's a dome with 128 custom speakers coming from all directions,

这是一个圆顶，有 128 个 来自各个方向的定制扬声器，

so we could create virtual soundscapes.

这样我们就可以创建虚拟音景。

It's sort of like the Star Trek holodeck, but for audio.

它有点像《星际迷航》的 全息甲板，但用于音频。

And if you're standing in the middle of this and you close your eyes,

如果你站在中间，闭上眼睛，

we can transport you auditorily to anywhere that we want.

我们可以借助声音 将你传送到任何地方。

So we can render a virtual football game,

我们可以营造一场虚拟的足球比赛，

and you feel the energy.

你能感受到比赛的热烈。

Or we can make it sound like you’re in the middle of a bustling city street.

或者我们可以让你感觉 像身处于熙熙攘攘的城市街道之中。

And if you’re me, you feel the anxiety.

如果你是我，你会感到焦虑。

Or standing on a beach with the crashing waves,

或者站在惊涛拍岸的海滩上，

and you feel the peace.

感受宁静。

And so it's super cool.

太酷了。

I wish everyone could be inside there.

多希望每个人都能进来试试。

Then we ran countless experiments to figure out all the complicated ways

我们进行了无数次实验 试图找出你的大脑

that your brain positions sounds in space.

定位空间中的声音的 各种复杂方式。

Also, we could reverse-engineer those neural algorithms

此外，我们可以对这些神经算法 进行逆向工程

and code them into our software.

并将其编码到我们的软件中。

So our goal has been to create this experience

因此，创造这种体验，

but right here.

但是是在这里。

Us psychoacousticians call this “virtual auditory space”

我们的心理声学家称之为 “虚拟听觉空间”，

to distinguish from our real auditory space,

是为了区分于我们的真实听觉空间，

which is, you know, the sounds that are all around us.

也就是我们周围的空间。

And this is what's necessary to create a compelling mixed audio-reality device.

这是制作强大的 混合音频现实设备所必需的。

So it's actually impossible to demonstrate this experience

如果你没有亲耳听过， 是不可能展示这种体验的。

until you hear it with your ears yourself.

But to give you an idea, we have tried to simulate it for you.

但是为了让你感受一下， 我们会试着为你模拟一下。

So imagine that you’re sitting in a noisy restaurant,

想象一下， 你坐在一家嘈杂的餐厅里，

and you're having trouble hearing your friends across the table.

听不到桌子对面的朋友的声音。

(Overlapping voices, music and noise)

（重叠的声音、音乐和噪音）

Hey, can you enhance the sounds that are right in front of me?

嘿，你能增强我正前方的声音吗？

(People speaking)

（人们说话声）

(Baby crying)

（婴儿哭声）

And can you turn that baby down?

你能把婴儿的哭声调低吗？

(People talking)

（人们说话声）

That’s better.

好点了。

I'm still having a little trouble hearing Pedro.

我还是听不太清 佩德罗（Pedro）的声音。

Can you isolate Pedro for me?

你能只留下佩德罗的声音吗？

Pedro: (Speaking in Spanish)

佩德罗：（说西班牙语）

JR: That's perfect.

JR：太完美了。

And, you know, my Spanish is a little rusty.

我的西班牙语有点生疏了。

Can I hear Pedro but in English?

能让我听见佩德罗， 但是让他说英语吗？

Pedro: And at the end of the trip,

佩德罗：在旅行结束时，

we came back to the city to visit the historic center.

我们回到了市里， 参观历史中心。

JR: Hey Shell, close all programs.

JR：嘿 Shell，关闭所有程序。

(Noise enhances)

（噪音增强）

Ah, it’s so much worse.

啊，差了好多。

That's pretty cool, right?

太酷了，对吧？

It's pretty cool.

太酷了。

(Applause)

（掌声）

So what you just heard was a beamforming app,

你刚才听到的是 波束赋形应用程序、

the computational auditory scene analysis app,

计算听觉场景分析应用程序、

a machine-learning denoising app,

机器学习降噪应用程序、

an AI transcription and translation

AI 转录和翻译

and text-to-speech with style transfer app.

以及带有风格迁移功能的 文字转语音应用程序。

The point is that all those audio transformations are done by software.

关键是所有这些音频转换 都是由软件完成的。

So we think the possibilities for these audio computers

我们认为这些音频计算机的可能性

are pretty much endless,

几乎是无限的，

and we can't wait to see

我们迫不及待地想看看

what the world's developers are going to do here.

全世界的开发人员 将在这个领域做些什么。

Like imagine an education app that knows your personal learning style

想象一个教育应用程序， 它了解你的个人学习风格，

and can teach you with the quality of a world-class professor,

可以随时随地 以世界一流教授的水平教你。

on-call anytime.

Or like a fitness coach you can summon all day

或者是一位健身教练， 你可以随时问它

about your diet and exercise,

有关饮食和运动的信息，

who can also motivate you through conversation

它还可以通过对话来激励你，

and even gamify your workout with some auditory cues.

甚至用一些听觉提示 让你的健身更有趣味性。

Or, hey, K?

或者，嘿，K？

Voice: Hi, Jason. What's up?

语音：嗨，杰森。怎么了？

Hey, if you were going to make an audio app that could be anything,

嘿，如果你可以任意 做一款音频应用程序，

what would it be?

你会做什么？

Voice: How about a whoopee cushion

语音：坐下时会播放放屁声的 恶作剧放屁坐垫怎么样？

that plays a fart sound whenever you sit down?

(Laughter)

（笑声）

JR: Hey, K, if you were going to make an audio app

JR：嘿，K，如果你要做一款

that didn't have anything to do with farts, what would it be?

与放屁无关的音频应用程序， 你会做什么？

Voice: Maybe an app that generates personalized soundscapes

语音：也许是一款可以生成个性化音景 让人放松和集中注意力的应用程序？

for relaxation and focus?

JR: That's much better.

JR：好多了。

Alright, it looks like we still have a little fine-tuning to do here.

好吧，看来我们还要做一些微调。

So the point is,

关键是，

imagination is the only limit to what you can do here.

想象力是你在这个领域 能做的事情的唯一限制。

Our goal is not just to create the world's first audio computer,

我们的目标不仅是创造 世界上第一台音频计算机，

it's to create a truly intuitive computing experience

还要创造一种真正直观的计算体验，

where we're not monetizing your attention

我们不会用你的注意力赚钱

or making you captive to a new kind of device,

或让你对一种新的设备上瘾，

but instead interfacing machines with us

而是以我们与生俱来的方式 让机器与我们交互。

in the way that we were born to.

So I think it's time for a computer that speaks our language.

我认为是时候创造一台 能说我们语言的计算机了。

Thank you.

谢谢。

(Applause)

（掌声）