Ray Kurzweil: Get ready for hybrid thinking

翻译人员: Minging Zhang 校对人员: Yuanqing Edberg

Let me tell you a story.

让我来给你们讲个故事。

It goes back 200 million years.

要追溯到两亿年前。

It's a story of the neocortex,

是个关于新大脑皮层的故事，

which means "new rind."

讲的就是大脑的表层。

So in these early mammals,

对于早期的哺乳类动物，

because only mammals have a neocortex,

由于只有他们有新大脑皮层，

rodent-like creatures.

就像啮齿类的生物。

It was the size of a postage stamp and just as thin,

皮质尺寸像邮票一样而且很薄，

and was a thin covering around

这个薄的皮质包裹着他们

their walnut-sized brain,

像核桃大小的头脑。

but it was capable of a new type of thinking.

但它可以产生新的思维方式。

Rather than the fixed behaviors

不像那些非哺乳类动物，

that non-mammalian animals have,

只有固定的行为，

it could invent new behaviors.

它可以创造新的行为。

So a mouse is escaping a predator,

例如一只老鼠在逃避捕食者，

its path is blocked,

它的路被堵住了，

it'll try to invent a new solution.

就想想出一个新的解决方案。

That may work, it may not,

那方案可能成功也可能失败，

but if it does, it will remember that

但如果成功了，它就会记住，

and have a new behavior,

于是就有了一种新的行为，

and that can actually spread virally

同时那个方案会迅速传遍

through the rest of the community.

到其余的团体。

Another mouse watching this could say,

比如另一只老鼠看到这会说，

"Hey, that was pretty clever, going around that rock,"

“噢，绕过那块岩石，真是高明，”

and it could adopt a new behavior as well.

于是它也会采取那种行为。

Non-mammalian animals

非哺乳类动物

couldn't do any of those things.

不能做这些事情。

They had fixed behaviors.

因为他们有固定的行为方式。

Now they could learn a new behavior

现在他们能学习新的行为

but not in the course of one lifetime.

但不是在一个生命的过程中。

In the course of maybe a thousand lifetimes,

也许在几千个生命周期内，

it could evolve a new fixed behavior.

它可以衍生出一个新的固定的行为。

That was perfectly okay 200 million years ago.

那对两亿年前来讲是好极了。

The environment changed very slowly.

那时的环境变化很慢。

It could take 10,000 years for there to be

那时可能要过一万年才会

a significant environmental change,

发生一次巨大的环境变化，

and during that period of time

在那期间，

it would evolve a new behavior.

可能进化一种新的行为。

Now that went along fine,

那样发展似乎还不错，

but then something happened.

但有些事情发生了。

Sixty-five million years ago,

六千五百万年前，

there was a sudden, violent change to the environment.

发生了一个突然的,剧烈的环境变化。

We call it the Cretaceous extinction event.

我们称之为白垩纪灭绝事件。

That's when the dinosaurs went extinct,

那是恐龙走向灭绝的时候，

that's when 75 percent of the

是百分之七十五的动植物

animal and plant species went extinct,

走向灭绝的时候，

and that's when mammals

也是哺乳类动物

overtook their ecological niche,

取代生态位，

and to anthropomorphize, biological evolution said,

而达到人格化，生物进化学说道，

"Hmm, this neocortex is pretty good stuff,"

“嗯，这个新大脑皮层是个好东西，”

and it began to grow it.

于是开始发展。

And mammals got bigger,

哺乳类动物逐渐变大，

their brains got bigger at an even faster pace,

他们的大脑变大的速度更快，

and the neocortex got bigger even faster than that

新大脑皮层同时变大的速度也更快，

and developed these distinctive ridges and folds

发展出明显的隆起和褶皱

basically to increase its surface area.

来增加它的表面积。

If you took the human neocortex

如果你有一个人的新大脑皮层

and stretched it out,

然后把它伸展开，

it's about the size of a table napkin,

大概有一方餐巾那么大，

and it's still a thin structure.

它也是一个很薄的构造。

It's about the thickness of a table napkin.

就像餐巾那么薄。

But it has so many convolutions and ridges

但它有很多的隆起和褶皱。

it's now 80 percent of our brain,

现在它占据我们的大脑有百分之八十

and that's where we do our thinking,

那也是我们用来思考的地方,

and it's the great sublimator.

所以那是个很棒的升华。

We still have that old brain

我们仍旧还是有那个

that provides our basic drives and motivations,

提供基本动力和动机的大脑，

but I may have a drive for conquest,

但也许我会有一个要去征服的想法，

and that'll be sublimated by the neocortex

那就要新大脑皮层

into writing a poem or inventing an app

借由写首诗或发明一个程序

or giving a TED Talk,

或来一个TED演讲而达到升华，

and it's really the neocortex that's where

它的确是在新的大脑皮层

有了行动。

the action is.

五十年前，我写了一篇论文，

Fifty years ago, I wrote a paper

describing how I thought the brain worked,

描述我对大脑如何运作的想法，

and I described it as a series of modules.

我描述说大脑就像一系列模块。

Each module could do things with a pattern.

每个模块可以用一种方式做事情。

It could learn a pattern. It could remember a pattern.

每个模块可以学习和记住一种方式。

It could implement a pattern.

也可以执行一种方式。

And these modules were organized in hierarchies,

然后这些模块被分派到统治集团中，

and we created that hierarchy with our own thinking.

我们用我们自己的想法创造了统治集团。

And there was actually very little to go on

后来我的这个想法就没怎么继续了。

那还是50年前。

50 years ago.

It led me to meet President Johnson.

它让我去见了约翰逊总统。

I've been thinking about this for 50 years,

我已经思考了五十年，

and a year and a half ago I came out with the book

一年半前我出了本书，

"How To Create A Mind,"

”如何创造思想，“

which has the same thesis,

这本书和那篇论文有着相同的主题，

but now there's a plethora of evidence.

但现在有了更多的证据支撑。

The amount of data we're getting about the brain

我们从神经科学得到

from neuroscience is doubling every year.

关于大脑的数据每年都成倍增加。

Spatial resolution of brainscanning of all types

各类脑扫描的空间分辨率也是。

is doubling every year.

每年双倍增加。

We can now see inside a living brain

我们现在可以看到一个活大脑的内部

and see individual interneural connections

看到个别神经元间的连接，

connecting in real time, firing in real time.

实时连接，实时放电

我们可以看到你大脑创造思维。

We can see your brain create your thoughts.

We can see your thoughts create your brain,

我们可以看到你的思维也在创造你的大脑，

which is really key to how it works.

这对了解大脑如何运作很重要。

So let me describe briefly how it works.

让我简单描述一下大脑如何工作的。

I've actually counted these modules.

我算过这些单位的数量。

We have about 300 million of them,

我们大约有三亿，

and we create them in these hierarchies.

我们在大脑层里创造他们。

I'll give you a simple example.

给你们简单举例。

I've got a bunch of modules

我有一堆模块，

that can recognize the crossbar to a capital A,

它们可以认知A的一横，

and that's all they care about.

那是它们关心的全部。

A beautiful song can play,

一首动人的歌在播放，

a pretty girl could walk by,

一个美丽的姑娘经过，

they don't care, but they see a crossbar to a capital A,

它们都不在意，但当它们看见A的一横，

they get very excited and they say "crossbar,"

它们就会很兴奋的说“横，”

and they put out a high probability

然后他们

on their output axon.

从输出轴突输出一个高度的可能性，

That goes to the next level,

那就到了下一个等级，

and these layers are organized in conceptual levels.

这些层次被分布在概念性等级中。

Each is more abstract than the next one,

每一个都比下一个更抽象，

so the next one might say "capital A."

所以下一个可能说“字母A。”

That goes up to a higher level that might say "Apple."

去到更高一个等级可能说“apple”

Information flows down also.

信息也这样流动。

If the apple recognizer has seen A-P-P-L,

如果那个认出apple的看到 a-p-p-l，

it'll think to itself, "Hmm, I think an E is probably likely,"

它就会想，“嗯，我觉得接下来是e,”

and it'll send a signal down to all the E recognizers

然后它就会把信号传送个所有认知e的

saying, "Be on the lookout for an E,

说，“看住e，

I think one might be coming."

我觉得它就要来了。”

The E recognizers will lower their threshold

e的认知这就会降低警觉

and they see some sloppy thing, could be an E.

它们可能粗心的看到一些东西觉得就是E。

Ordinarily you wouldn't think so,

通常你不会这样想，

but we're expecting an E, it's good enough,

但我们在期待一个E， 那就够了，

and yeah, I've seen an E, and then apple says,

于是我看到了E，然后认知的apple说，

"Yeah, I've seen an Apple."

“太好了，我看到了apple。”

Go up another five levels,

再往上五个等级，

and you're now at a pretty high level

现在你就在一个很高的水平，

of this hierarchy,

的这种大脑层，

and stretch down into the different senses,

于是延伸到不同的感官，

and you may have a module that sees a certain fabric,

你可能有一个模块看到了一个特殊东西，

hears a certain voice quality, smells a certain perfume,

听到一个声音，闻到到某个特殊的香水，

and will say, "My wife has entered the room."

它就会说，“我老婆进来房间了。”

Go up another 10 levels, and now you're at

往上十个等级，现在你在

a very high level.

一个非常高的等级。

You're probably in the frontal cortex,

你可能在大脑额叶，

and you'll have modules that say, "That was ironic.

然后你有模块说，“那很讽刺。

That's funny. She's pretty."

那很有趣。她很美。”

You might think that those are more sophisticated,

你可能觉得那些模块很复杂，

but actually what's more complicated

实际上更复杂的是

is the hierarchy beneath them.

在他们之下的大脑层集团。

There was a 16-year-old girl, she had brain surgery,

有一个十六岁的女孩，她做了一个大脑手术，

and she was conscious because the surgeons

她依然是清醒的，因为外科医生

wanted to talk to her.

要和她谈话。

You can do that because there's no pain receptors

手术可以做是因为大脑里没有疼痛的感觉器官。

在大脑里，

in the brain.

And whenever they stimulated particular,

当他们刺激到某个部分，

very small points on her neocortex,

在她大脑皮层的很小的点，

shown here in red, she would laugh.

这里显示红色的，她就会笑。

So at first they thought they were triggering

所以一开始他们以为他们触碰到

some kind of laugh reflex,

某个笑神经，

but no, they quickly realized they had found

但不是，他们很快意识到他们发现

the points in her neocortex that detect humor,

那些在新大脑皮层的小点能探测到幽默，

and she just found everything hilarious

然后她发现一切都很可笑，

whenever they stimulated these points.

每当刺激到那些点的时候。

"You guys are so funny just standing around,"

“你们站在这里真是太好笑了，”

was the typical comment,

这是她典型的言论，

and they weren't funny,

但

not while doing surgery.

实际上他们在做手术时并不有趣。

So how are we doing today?

所以我们现在怎么样？

Well, computers are actually beginning to master

事实上电脑逐渐开始

human language with techniques

通过科技掌握人类语言，

that are similar to the neocortex.

这和新大脑皮层类似。

I actually described the algorithm,

我实际上描述了运算法则，

which is similar to something called

这和

a hierarchical hidden Markov model,

脑层隐藏的马尔可夫模型类似，

something I've worked on since the '90s.

这是一些我从90年代就开始研究的事。

"Jeopardy" is a very broad natural language game,

"Jeopardy"是一个很广泛的语言游戏，

and Watson got a higher score

Watson得了一个

than the best two players combined.

比两人加在一起还高的分数。

It got this query correct:

它纠正了这个问题：

"A long, tiresome speech

"一段很长很无聊的演讲

delivered by a frothy pie topping,"

就像馅饼上的装饰。“

and it quickly responded, "What is a meringue harangue?"

于是有了很快的回复，”什么是长篇大论？“

And Jennings and the other guy didn't get that.

没人理解这个问题。

It's a pretty sophisticated example of

那是个很复杂的例子

computers actually understanding human language,

关于电脑理解人类语言，

and it actually got its knowledge by reading

它实际得到了自己的知识通过

Wikipedia and several other encyclopedias.

维基百科和一些其他百科。

Five to 10 years from now,

从现在起五到十年，

search engines will actually be based on

搜索引擎会不仅仅基于

not just looking for combinations of words and links

对文字、链接组合的寻找，

but actually understanding,

而是真正的理解，

reading for understanding the billions of pages

通过阅读来理解

on the web and in books.

网络和书中成千上万页。

So you'll be walking along, and Google will pop up

那么当你郁郁独行，古狗会跳出来

and say, "You know, Mary, you expressed concern

说，”你知道吗，Mary， 你一个月前

to me a month ago that your glutathione supplement

你向我述说的你补充的谷胱甘肽

wasn't getting past the blood-brain barrier.

没有通过血脑屏障。

Well, new research just came out 13 seconds ago

十三秒前刚出了个新研究，

that shows a whole new approach to that

显示有一个全新方法的来解决这个问题。

and a new way to take glutathione.

Let me summarize it for you."

一个服用谷胱甘肽的新方法，

Twenty years from now, we'll have nanobots,

让我来为你总结一下。“

从现在起二十年， 我们会有纳米机器人，

because another exponential trend

因为另一个指数趋势

is the shrinking of technology.

显示科技的收缩。

They'll go into our brain

他们会通过

毛细血管进入我们的大脑

through the capillaries

然后把我们的大脑皮层连接

and basically connect our neocortex

to a synthetic neocortex in the cloud

到枢纽里的合成大脑皮层。

providing an extension of our neocortex.

Now today, I mean,

而提供一个大脑皮层的延伸

you have a computer in your phone,

现在，我的意思是，

but if you need 10,000 computers for a few seconds

你手机里有一个电脑，

但如果你需要一万台电脑用几秒钟

to do a complex search,

来做一个复杂的研究，

you can access that for a second or two in the cloud.

你可以进入那个枢纽一两秒。

In the 2030s, if you need some extra neocortex,

在2030年，如果你需要一些额外的大脑皮层，

you'll be able to connect to that in the cloud

你可以在枢纽里

直接与你大脑连接。

directly from your brain.

So I'm walking along and I say,

所以当我走过我会说，

"Oh, there's Chris Anderson.

“哦，这是Chris Anderson。”

He's coming my way.

他在向我走来。

I'd better think of something clever to say.

我最好想一个聪明的方式来说。

I've got three seconds.

我有三秒钟来想。

My 300 million modules in my neocortex

我大脑皮层里的三亿个模块

还不够

isn't going to cut it.

我需要另外一亿个。“

I need a billion more."

I'll be able to access that in the cloud.

于是我可以在枢纽里实现。

And our thinking, then, will be a hybrid

那时我们的思考，会像一个

of biological and non-biological thinking,

生物和非生物思考的混合，

but the non-biological portion

但非生物的部分

is subject to my law of accelerating returns.

取决于我加速回收的原则。

它会成指数增长。

It will grow exponentially.

And remember what happens

还记得

the last time we expanded our neocortex?

上次我们伸展我们大脑皮层发生了什么吗？

That was two million years ago

那是两亿年前

when we became humanoids

当我们成为变类人

and developed these large foreheads.

进化这些大前额时。

Other primates have a slanted brow.

其他灵长目动物有倾斜的额。

They don't have the frontal cortex.

他们没有前大脑皮层。

But the frontal cortex is not really qualitatively different.

但那不是真正的不同。

It's a quantitative expansion of neocortex,

不同在于大脑皮层的伸展，

but that additional quantity of thinking

但那额外的思考

was the enabling factor for us to take

是我们能够飞跃的启动因子，

并因此发明了语言、艺术、科学

a qualitative leap and invent language

艺术，科学和科技

and art and science and technology

还有TED会议。

and TED conferences.

No other species has done that.

没有其他物种可以做到这样。

And so, over the next few decades,

在接下来的几十年，

we're going to do it again.

我们要再一次。

We're going to again expand our neocortex,

我们会再次伸展我们的新大脑皮层，

only this time we won't be limited

只有这样我们不会

by a fixed architecture of enclosure.

被固定的框架结构所限制。

It'll be expanded without limit.

它会无限伸展。

那个额外的量

That additional quantity will again

会再次成为一个启动因子

使我们在文化科技中有一个质的飞跃。

be the enabling factor for another qualitative leap

in culture and technology.

Thank you very much.

非常感谢！

(Applause)

（鼓掌）