The Transformative Potential of AGI — and When It Might Arrive | Shane Legg and Chris Anderson | TED

翻译人员: Yip Yan Yeung 校对人员: Yanyan Hong

克里斯·安德森（Chris Anderson）：

Chris Anderson: Shane, give us a snapshot of you growing up

沙恩（Shane），请你 简短介绍一下你的成长历程，

and what on Earth led you to get interested in artificial intelligence?

你到底为什么 会对人工智能感兴趣呢？

Shane Legg: Well, I got my first home computer on my 10th birthday,

沙恩·莱格（Shane Legg）： 我在十岁生日那天

收到了我的第一台家用电脑，

and I --

而我——

this was before the internet and everything.

那时互联网和这一切还没出现。

So you couldn't just go and surf the web and so on.

你不能就这么上网冲浪等等。

You had to actually make stuff yourself and program.

你得真的自己动手做东西、写程序。

And so I started programming,

于是我开始了编程，

and I discovered that in this computer there was a world,

发现这台计算机里有着一个世界，

I could create a world,

我可以创造出一个世界，

I could create little agents that would run around

我可以做出跑来跑去、 互相追逐、干活等等的小智能体。

and chase each other and do things and so on.

And I could sort of, bring this whole universe to life.

某种程度上我可以实现整个宇宙。

And there was sort of that spark of creativity that really captivated me

创造力的火花让我无法自拔，

and sort of, I think that was really the seeds

我认为这就埋下了我兴趣的种子，后来

of my interest that later grew

into an interest in artificial intelligence.

长成了对人工智能的兴趣。

CA: Because in your standard education, you had some challenges there.

CA：听说你在传统教育中遇到了一些困难。

SL: Yeah, I was dyslexic as a child.

SL：是的，我小时候有阅读障碍。

And so they were actually going to hold me back a year

所以在我十岁的时候， 他们让我推后一年，

when I was 10 years old,

and they sent me off to get my IQ tested to sort of,

把我送去测智商，

you know, assess how bad the problem was.

评估一下问题有多严重。

And they discovered I had an exceptionally high IQ.

然后他们发现我的智商格外地高。

And then they were a little bit confused about what was going on.

他们有点疑惑这是什么情况。

And fortunately, at that time,

所幸当时

there was somebody in the town I lived in

我居住的小镇里

who knew how to test for dyslexia.

有人知道如何测试阅读障碍。

And it turns out I wasn't actually of limited intelligence.

结果我并不是智力有限。

I was dyslexic, and that was the issue.

而是患有阅读障碍，这就是问题所在。

CA: You had reason from an early age to believe

CA：你从小就有理由相信

that our standard assumptions about intelligence might be off a bit.

我们对智力的普遍认知可能是错的。

SL: Well, I had reason, from an early age, to sometimes doubt authority.

SL：嗯，我从小就有理由质疑权威。

(Laughter)

（笑声）

You know, if the teacher thinks you're dumb, maybe it's not true.

如果老师认为你很傻， 可能并不是如此。

Maybe there are other things going on.

也许还有其他情况。

But I think it also created in me an interest in intelligence

但我认为这样的儿时经历

也在我心中建立起了对智力的兴趣。

when I sort of had that experience as a child.

CA: So you're credited by many

CA：许多人

as coining the term “artificial general intelligence,” AGI.

认为你创造了 “通用人工智能”一词，即 AGI。

Tell us about 2001, how that happened.

说说 2001 年吧，那是怎么发生的。

SL: Yeah, so I was approached by someone called Ben Goertzel

SL：是的，有个叫 本·戈策尔（Ben Goertzel）的人

who I'd actually been working with,

找到了我，我之后也在与他共事，

and he was going to write a book,

当时他打算写一本书，

and he was thinking about a book on AI systems

他在考虑写一本关于 AI 系统的书，

that would be much more general and capable,

它会更通用、更强大，

rather than focusing on very narrow things.

而不是专注于非常狭窄的领域。

And he was thinking about a title for the book.

他正在思考这本书的标题。

So I suggested to him, "If you're interested in very general systems,

我向他建议：“如果你 对很通用的系统感兴趣，

call it artificial general intelligence."

那就叫它通用人工智能吧。”

And so he went with that.

于是他就接受了。

And then him and various other people started using the term online

然后他和好多人开始在线上

and the internet,

和互联网上使用这个词，

and then it sort of became popularized from there.

从此风靡了起来。

后来我们发现有个叫 迈克·加罗德（Mike Garrod）的人，

We later discovered there was someone called Mike Garrod,

他在 97 年的一份安全纳米技术 期刊上发表了一篇论文。

who published a paper in a security nanotech journal in '97.

So he is actually the first person to have used the term.

他其实是第一个使用该术语的人。

But it turns out he pretty much meant the same thing as us anyway.

但他想表达的意思和我们的一样。

CA: It was kind of an idea whose time had come,

CA：发现其中潜力的时代到了。

to recognize the potential here.

I mean, you made an early prediction that many people thought was bonkers.

你在很多人认为是痴人说梦的时候 就先人一步做出了预测。

What was that?

预测了什么？

SL: Well, in about 2001,

SL：大约在 2001 年，

a similar time to when I suggested this term artificial general intelligence,

差不多是我提出“通用人工智能” 这个术语的时候，

I read a book by Ray Kurzweil, actually, "Age of Spiritual Machines,"

我读了雷·库兹韦尔（Ray Kurzweil）的 一本书《机器之心》，

and I concluded that he was fundamentally right,

认为他说的很对，

that computation was likely to grow exponentially for at least a few decades,

计算可能会在至少几十年内 呈指数级增长，

and the amount of data in the world would grow exponentially

世界上的数据量 将在几十年内呈指数级增长。

for a few decades.

And so I figured that if that was going to happen,

我想，如果真的出现了这种情况，

then the value of extremely scalable algorithms

那么利用这些数据和计算的 高度可扩展算法

that could harness all this data and computation

将会有非常高的价值。

were going to be very high.

And then I also figured that in the mid 2020s,

我还发现，到了 2020 年代中期，

it would be possible then,

那时候就可以成真了，

if we had these highly scalable algorithms,

我们可以用这些高度可扩展的算法

to train artificial intelligence systems

训练人工智能系统，

on far more data than a human would experience in a lifetime.

所用的数据远远超过 人类一生所能经历的数据。

And so as a result of that,

因此，我从 2009 年左右 就开始在我的博客中说到这点，

you can find it on my blog from about 2009

I think it's the first time I publicly talked about it,

我觉得那是我第一次 公开谈论这一点，

I predicted a 50 percent chance of AGI by 2028.

我预计 2028 年实现 AGI 的 可能性将达到 50％。

I still believe that today.

时至今日我依旧如此认为。

CA: That's still your date.

CA：你觉得还是这个时间点。

How did you define AGI back then, and has your definition changed?

以前你是怎么定义 AGI 的？ 你的定义改变了吗？

SL: Yeah, I didn't have a particularly precise definition at the beginning.

SL：嗯，一开始 我没有特别精准的定义。

It was really just an idea of systems that would just be far more general.

只是“系统”的概念， 通用得多的系统。

So rather than just playing Go or chess or something,

不仅仅是下围棋、下象棋等等，

rather than actually be able to do many, many different things.

不仅仅是做各种各样的事。

The definition I use now is that it's a system

我现在使用的定义是一个系统，

that can do all the cognitive kinds of tasks

可以完成人类可以完成的 所有认知任务，可能还要更多，

that people can do, possibly more,

but at least it can do the sorts of cognitive tasks

但至少它可以完成人类 通常能完成的那种认知任务。

that people can typically do.

CA: So talk about just the founding of DeepMind

CA：谈谈 DeepMind 的成立

and the interplay between you and your cofounders.

还有你和你的联合创始人之间的互动。

SL: Right. So I went to London to the place called the Gatsby Unit,

SL：好。我去了伦敦，去了一个叫做 “盖茨比计算神经科学中心”的地方，

which studies theoretical neuroscience and machine learning.

那里研究理论神经科学和机器学习。

And I was interested in learning the relationships

我感兴趣的是研究

between what we understand about the brain

我们对大脑的理解

and what we know from machine learning.

和我们从机器学习中的所得 之间的关系。

So that seemed like a really good place.

所以应该是来对了。

And I met Demis Hassabis there.

我在此遇到了戴密斯·哈萨比斯 （Demis Hassabis）。

He had the same postdoc supervisor as me,

我和他的博士后导师是同一位，

and we got talking.

所以就聊了起来。

And he convinced me

他说服我是时候创办一家公司了。

that it was the time to start a company then.

That was in 2009 we started talking.

我们聊起来的时候是 2009 年。

And I was a little bit skeptical.

我有点怀疑。

I thought AGI was still a bit too far away,

我认为 AGI 还远在天边，

but he thought the time was right, so we decided to go for it.

但他认为时机已经成熟， 所以我们决定开干。

And then a friend of his was Mustafa Suleyman.

他的朋友穆斯塔法·苏莱曼 （Mustafa Suleyman）也加入了。

CA: And specifically, one of the goals of the company

CA：具体地说，公司的目标之一

was to find a pathway to AGI?

就是找到通往 AGI 的途径吗？

SL: Absolutely.

SL：没错。

On our first business plan that we were circulating

我们在 2010 年为寻求投资 发布的第一份商业企划

when we were looking for investors in 2010,

it had one sentence on the front cover and it said,

封面上有一句话：

"Build the world's first artificial general intelligence."

“打造世界上第一个通用人工智能。”

So that was right in from the beginning.

从一开始就是如此。

CA: Even though you knew

CA：即使你知道

that building that AGI might actually have

打造这样的 AGI 可能会

apocalyptic consequences in some scenarios?

在某种情况下导致灾难性后果？

SL: Yeah.

SL：是的。

So it's a deeply transformative technology.

它是一项极具颠覆性的技术。

I believe it will happen.

我相信它会成真。

I think that, you know,

我认为，

these algorithms can be understood and they will be understood at the time.

这些算法可以被理解， 届时也真的会被理解。

And I think that intelligence is fundamentally

我认为智能

an incredibly valuable thing.

是一个相当有价值的东西。

Everything around us at the moment -- the building we’re in,

那时我们周围的一切—— 我们所在的大楼、

the words I’m using, the concepts we have, the technology around us --

我使用的文字、我们拥有的概念、 我们身边的技术，

you know, all of these things are being affected by intelligence.

这些东西都会受到智能的影响。

So having intelligence in machines

让机器拥有智能

is an incredibly valuable thing to develop.

是非常值得开发的。

And so I believe it is coming.

所以我相信它会来的。

Now when a very, very powerful technology arises,

当一项非常非常强大的技术出现时，

there can be a range of different outcomes.

可能会产生一系列不同的结果。

Things could go very, very well,

事情可能进展得非常非常顺利，

but there is a possibility things can go badly as well.

但也有可能会翻车。

And that was something I was aware of also from about 20 years ago.

这也是我在大约 20 年前 就意识到的一点。

CA: So talk about, as DeepMind developed,

CA：谈谈随着 DeepMind 的发展，

was there a moment where you really felt,

有没有哪一刻你真的觉得：

"My goodness, we're onto something unbelievably powerful?"

“天啊，我们正在研究强大得惊人的东西？”

Like, was it AlphaGo, that whole story, or what was the moment for you?

是做 AlphaGo 的时候吗？ 还是哪个时刻？

SL: Yeah, there were many moments over the years.

SL：是的，这些年来有很多时刻。

One was when we did the Atari games.

一个是我们做 Atari 游戏的时候。

Have you seen those videos

你有没有看过那些视频，

where we had an algorithm that could learn to play multiple games

我们有一个算法， 可以学会玩多款游戏，

without being programmed for any specific game?

而无需为其中 任意一款游戏专门编程？

There were some exciting moments there.

那时有一些激动人心的时刻。

Go, of course, was a really exciting moment.

AlphaGo 当然也是 一个激动人心的时刻。

But I think the thing that's really captured my imagination,

但我认为，真正吸引我的想象力、

a lot of people's imagination,

很多人的想象力的是

is the phenomenal scaling of language models in recent years.

近年来语言模型的惊人扩展。

I think we can see they're systems

我们可以看出，它们这些系统

that really can start to do some meaningful fraction

确实可以开始完成一部分

of the cognitive tasks that people can do.

人类可以完成的有意义的认知任务。

CA: Now, you were working on those models,

CA：你正在研究这些模型，

but were you, to some extent, blindsided

但在某种程度上，你有没有

by OpenAI's, sort of, sudden unveiling of ChatGPT?

被 OpenAI 突然推出的 ChatGPT 打得措手不及？

SL: Right.

SL：对。

We were working on them and you know,

我们正在研究它们，

the transformer model was invented in Google,

Transformer 模型是谷歌研发的，

and we had teams who were building big transformer language models and so on.

我们有团队正在构建 大型 Transformer 语言模型。

CA: Google acquired DeepMind at some point in this journey.

CA：谷歌在这段时间里 收购了 DeepMind。

SL: Yeah, exactly.

SL：是的，没错。

And so what I didn't expect

我没想到的是

was just how good a model could get training purely on text.

单凭文本训练得出的模型能有多好。

I thought you would need more multimodality.

我以为需要多模态的数据。

You'd need images, you'd need sound, you'd need video and things like that.

需要图像、声音、视频等等。

But due to the absolutely vast quantities of text,

但由于文本的海量存在，

it can sort of compensate for these things to an extent.

可以从某种程度上弥补这些问题。

I still think you see aspects of this.

我觉得这是可以看到一些迹象的。

I think language models tend to be weak in areas

我认为语言模型在不易 以文本形式表达的领域相对薄弱。

that are not easily expressed in text.

But I don’t think this is a fundamental limitation.

但我不认为这是一个根本性缺陷。

I think we're going to see these language models expanding into video

我们会见证这些语言模型 扩展至视频、

and images and sound and all these things,

图像、声音等等，

and these things will be overcome in time.

这些问题迟早会被攻克。

CA: So talk to us, Shane,

CA：和我们讲讲，沙恩，

about the things that you, at this moment,

谈谈你此刻

passionately feel that the world needs to be thinking about more cogently.

迫切认为这个世界 该更审慎地思考的是什么。

SL: Right.

SL：好。

So I think that very, very powerful,

我认为非常、非常强大、

very intelligent artificial intelligence is coming.

非常智慧的人工智能即将到来。

I think that this is very, very likely.

我认为很有可能。

I don't think it's coming today.

我觉得不会今天到来。

I don't think it's coming next year or the year after.

我觉得不会明年或者后年到来。

It's probably a little bit further out than that.

有可能更久一点。

CA: 2028?

CA：2028？

SL: 2028, that's a 50 percent chance.

SL：2028，这是 50% 的可能。

So, you know, if it doesn't happen in 2028,

如果 2028 年没有发生，

I'm not going to be surprised, obviously.

我显然也不会感到惊讶。

CA: And when you say powerful,

CA：你说“强大”，

I mean there's already powerful AI out there.

现在已经有强大的 AI 了。

But you're saying basically a version

但你的意思是 某种通用人工智能即将到来。

of artificial general intelligence is coming.

SL: Yeah.

SL：是的。

CA: So give us a picture of what that could look like.

CA：给我们描绘一下这样的图景吧。

SL: Well, if you had an artificial general intelligence,

SL：嗯，如果你有了通用人工智能，

you could do all sorts of amazing things.

你可以做各种奇妙的事情。

Just like human intelligence is able to do many, many amazing things.

就像人类智能 能做很多奇妙的事情一样。

So it's not really about a specific thing,

这不是某一件事，

that's the whole point of the generality.

而是“通用”的意义。

But to give you one example,

但举个例子，

we developed the system AlphaFold,

我们开发了 AlphaFold 系统，

which will take a protein and compute, basically, the shape of that protein.

它能根据一种蛋白质 计算出它的形状。

And that enables you to do all sorts of research

这样你就能进行各种研究，

into understanding biological processes,

了解生物学过程、开发药物等等。

developing medicines and all kinds of things like that.

Now, if you had an AGI system,

如果你有了一个 AGI 系统，

instead of requiring what we had at DeepMind,

而不是像 DeepMind 这样的配置，

about roughly 30 world-class scientists

大约 30 位世界一流的科学家

working for about three years to develop that,

花了差不多三年的时间 开发这个系统，

maybe you could develop that with just a team

也许你可以在一年内 以一个由少数科学家组成的团队

of a handful of scientists in one year.

开发出这个系统。

So imagine these, sort of, AlphaFold-level developments

想象一下世界各地定期

taking place around the world on a regular basis.

进行 AlphaFold 级别的开发。

This is the sort of thing that AGI could enable.

这就是 AGI 可以实现的东西。

CA: So within months of AGI being with us, so to speak,

CA：我们一旦用上几个月的 AGI，

it's quite possible that some of the scientific challenges

很有可能人类纠缠了几十年、

that humans have wrestled with for decades, centuries, if you like,

也许几百年的科学挑战，

will start to tumble in rapid succession.

就会被迅速逐个击破。

SL: Yeah, I think it'll open up all sorts of amazing possibilities.

SL：是的，我认为它会 开辟各种奇妙的可能。

And it could be really a golden age of humanity

这可能是一个真正的人类黄金时代，

where human intelligence,

人类智能，

which is aided and extended with machine intelligence,

在机器智能的辅助和扩展下，

enables us to do all sorts of fantastic things

使我们能够做到各种神奇的事情，

and solve problems that previously were just intractable.

解决以前难以解决的问题。

CA: So let's come back to that.

CA：我们之后再回来聊这点。

But I think you also,

但我认为，

you're not like, an irredeemable optimist only,

你不像一个无可救药的乐观主义者，

you see a potential for it to go very badly in a different direction.

你还是看到了它跑偏翻车的可能。

Talk about what that pathway could look like.

谈谈那条路会是什么样子。

SL: Well, yeah, I want to explain.

SL：嗯，是的，我想解释一下。

I don't believe the people

我不相信那些坚信一切顺利的人，

who are sure that it's going to go very well,

and I don't believe the people

我也不相信

who are sure that it’s going to go very, very badly.

那些坚信一定会翻车的人。

Because what we’re talking about is an incredibly profound transition.

因为我们讨论的 是一个非常重大的转变。

It's like the arrival of human intelligence in the world.

这就像人类智能降临世界一样。

This is another intelligence arriving in the world.

另一个智能来到了这个世界。

And so it is an incredibly deep transition,

这是一个极其深刻的转变，

and we do not fully understand all the implications

我们并不能完全了解 其中含义和后果。

and consequences of this.

And so we can't be certain

因此，我们无法确定

that it's going to be this, that or the other thing.

它会这样、那样还是怎样。

So we have to be open-minded about what may happen.

所以我们必须以开放的心态 看待可能发生的情况。

I have some optimism because I think

我有点乐观，因为我认为，

that if you want to make a system safe,

如果你想确保系统的安全，

you need to understand a lot about that system.

你必须得非常了解这个系统。

You can't make an airplane safe

如果你不知道飞机是如何运作的，

if you don't know about how airplanes work.

你就无法保证飞机的安全。

So as we get closer to AGI,

随着我们越来越接近 AGI，

we will understand more and more about these systems,

我们会越来越了解这些系统，

and we will see more ways to make these systems safe,

我们会看到更多 保障这些系统安全的方式，

make highly ethical AI systems.

打造高度符合伦理的 AI 系统。

But there are many things we don't understand about the future.

但是我们对未来 还是有很多不理解的地方。

So I have to accept that there is a possibility that things may go badly

所以我必须接受 有偏离轨道的可能性，

because I don't know what's going to happen.

因为我不知道会发生什么。

I can't know that about the future in such a big change.

在这场巨变中， 我不知道未来会发生什么。

And even if the probability of something going bad is quite small,

即使出岔子的可能性很小，

we should take this extremely seriously.

我们也应该非常慎重地看待这个问题。

CA: Paint a scenario of what going bad could look like.

CA：想象一个跑偏的场景是什么样的。

SL: Well, it's hard to do

SL：嗯，很难，

because you're talking about systems

因为我们说的是

that potentially have superhuman intelligence, right?

可能拥有超人类智能的系统，对吧？

So there are many ways in which things would go bad in the world.

在这世上跑偏的方式有很多。

People sometimes point to, I don't know, engineered pathogens, right?

人们有时会点名 人工培养病原体，对吧？

Maybe a superintelligence could design an engineered pathogen.

也许超级智能可以 设计出一种人工培养病原体。

It could be much more mundane things.

可能是更平平无奇的事情。

Maybe with AGI, you know,

也许有了 AGI，

it gets used to destabilize democracy in the world,

它会被用于破坏世界民主的稳定，

with, you know, propaganda or all sorts of other things like that.

通过宣传等等方式。

We don't know --

我们不知道……

CA: That one might already have happened.

CA：可能已经发生了。

SL: There might be happening a bit already.

SL：可能已经发生了一些了。

But, you know, there may be a lot more of this

但是，如果我们有了更强大的系统， 可能会出现更多这样的情况。

if we have more powerful systems.

So there are many ways in which societies can be destabilized.

破坏社会稳定的方式有很多。

And you can see that in the history books.

你可以在历史书中看到这一点。

CA: I mean, Shane, if you could have asked all humans,

CA：沙恩，如果你问每个人，

say, 15 years ago, OK, we can open a door here,

假设 15 年前， 我们可以在这里打开一扇门，

and opening this door could lead to the best-ever outcomes for humanity.

打开这扇门可以为人类带来 有史以来最好的结果。

But there's also a meaningful chance,

但是，也有相当大的可能，

let's say it's more than five percent,

比方说，超过 5%，

that we could actually destroy our civilization.

我们可能会摧毁我们的文明。

I mean, isn't there a chance that most people would have actually said,

难道大多数人不会这么说：

"Don't you dare open that damn door.

“不许打开那扇门。

Let's wait."

我们等着吧。”

SL: If I had a magic wand and I could slow things down,

SL：如果我有一根魔杖，能减慢速度，

I would use that magic wand, but I don't.

我会用那根魔杖，但我没有。

There are dozens of companies,

有几十家公司，

well, there's probably 10 companies in the world now

现在全球可能有 10 家公司，

that can develop the most cutting-edge models, including, I think,

可以开发出最前沿的模型，包括

some national intelligence services who have secret projects doing this.

一些国家情报机构 通过秘密项目进行研发。

And then there's, I don't know,

还有一大堆公司在开发 落后一代的东西。

dozens of companies that can develop something that's a generation behind.

And remember, intelligence is incredibly valuable.

请记住，智能非常宝贵。

It's incredibly useful.

非常有用。

We're doing this

我们之所以这样做，

because we can see all kinds of value that can be created in this

是因为我们可以看到 出于各种原因在此开发过程中

for all sorts of reasons.

能够创造的价值。

How do you stop this process?

你要如何停下这个进程？

I don't see any realistic plan that I've heard of,

我没有听说过任何切合实际的计划

of stopping this process.

能够停下这个进程。

Maybe we can --

也许我们可以——

I think we should think about regulating things.

我认为我们应该考虑加以监管。

I think we should do things like this as we do with every powerful technology.

我认为面临任何强大科技 都应该这么做。

There's nothing special about AI here.

AI 并不是什么特例。

People talk about, oh, you know, how dare you talk about regulating this?

人们会说， 你怎么敢说要监管它们啊？

No, we regulate powerful technologies all the time in the interests of society.

不，为了社会的利益， 我们一直在监管强大的技术。

And I think this is a very important thing that we should be looking at.

我认为这是我们应该考虑的 一件非常重要的事情。

CA: It's kind of the first time we have this superpowerful technology out there

CA：这好像是第一次 我们有了超级强大的技术，

that we literally don't understand in full how it works.

但我们并不完全了解 它是如何运作的。

Is the most single, most important thing we must do, in your view,

在你看来，我们该做的 最重要的那一件事，

to understand, to understand better what on Earth is going on,

是去理解、更好地了解 到底发生了一些什么吗？

so that we least have a shot at pointing it in the right direction?

让我们至少还有一线希望 将其导向正确的方向。

SL: There is a lot of energy behind capabilities at the moment

SL：目前功能背后消耗了大量的精力，

because there's a lot of value in developing the capabilities.

因为开发功能有很大的价值。

I think we need to see a lot more energy going into actual science,

我认为我们得将更多的精力 投入到实际的科学中，

understanding how these networks work,

了解这些网络是如何运作的，

what they're doing, what they're not doing,

它们在做些什么，它们没有做什么，

where they're likely to fail,

它们可能在哪里崩溃，

and understanding how we put these things together

理解我们如何 将这些东西整合在一起，

so that we're able to find ways

让我们可以找到

to make these AGI systems profoundly ethical and safe.

让这些 AGI 系统 高度符合伦理又安全的方式。

I believe it's possible.

我相信这是可能的。

But we need to mobilize more people's minds and brains

但是，我们需要鼓动 更多人的思想和大脑

to finding these solutions

寻找这些解决方案，

so that we end up in a future

这样我们才能在未来

where we have incredibly powerful machine intelligence

拥有极其强大的机器智能，

that's also profoundly ethical and safe,

而它们也高度符合伦理又安全，

and it enhances and extends humanity

将人类增强、延伸至

into, I think, a new golden period for humanity.

一个全新的人类黄金时代。

CA: Shane, thank you so much for sharing that vision

CA：沙恩，非常感谢你分享这个愿景，

and coming to TED.

也感谢你来到 TED。

(Applause)

（掌声）