Gero Miesenboeck reengineers a brain

翻译人员: Jenny Yang 校对人员: Yi You

I have a doppelganger.

我有个翻版。

(Laughter)

(笑声)

Dr. Gero is a brilliant

Gero 博士是动画片<<七龙珠Z-机器人传奇>>

but slightly mad scientist

里一个非常聪明但有点疯疯颠颠

in the "Dragonball Z: Android Saga."

的科学家。

If you look very carefully,

如果你仔细观察，

you see that his skull has been replaced

你会发现他的脑壳被换成了一个

with a transparent Plexiglas dome

透明的有机玻璃圆顶盖子，

so that the workings of his brain can be observed

你可以透过盖子观察他的大脑活动

and also controlled with light.

而且可以用光线去控制它。

That's exactly what I do --

这就是我的工作——

optical mind control.

光学精神控制。

(Laughter)

(笑声)

But in contrast to my evil twin

但是我跟我那邪恶的孪生兄弟不同，

who lusts after world domination,

他企图控制全世界，

my motives are not sinister.

而我没有险恶的动机。

I control the brain

我控制大脑的原因是

in order to understand how it works.

想了解它是怎样运作的。

Now wait a minute, you may say,

你也许会说，等一下，

how can you go straight to controlling the brain

你不了解大脑

without understanding it first?

怎么能控制大脑呢？

Isn't that putting the cart before the horse?

那不是本末倒置吗？

Many neuroscientists agree with this view

很多神经学科学家都同意这个观点

and think that understanding will come

认为对大脑的了解需要通过

from more detailed observation and analysis.

很多详细的观察和分析来完成。

They say, "If we could record the activity of our neurons,

他们说：“如果我们可以记录我们的神经元的活动

we would understand the brain."

我们就能了解大脑了。”

But think for a moment what that means.

但是想一想这意味着什么。

Even if we could measure

即便我们可以侧量

what every cell is doing at all times,

每个细胞每时每刻的运作，

we would still have to make sense

我们还是需要弄明白

of the recorded activity patterns,

记录下来的那些活动到底有什么规律，

and that's so difficult,

而这是很困难的。

chances are we'll understand these patterns

结果很可能是我们只能了解这些规律的一小部分

just as little as the brains that produce them.

因为我们对产生它们的大脑知之甚少。

Take a look at what brain activity might look like.

请看这个大脑活动模拟图。

In this simulation, each black dot

在这个模拟中，每个黑点

is one nerve cell.

代表一个神经细胞。

The dot is visible

每当细胞放射一条

whenever a cell fires an electrical impulse.

电脉冲时, 对应的点就会显示出来。

There's 10,000 neurons here.

这里有一万个神经细胞。

So you're looking at roughly one percent

所以你现在看到的相当于

of the brain of a cockroach.

一只蟑螂大脑的百分之一。

Your brains are about 100 million times

你们的大脑比这复杂

more complicated.

一亿倍。

Somewhere, in a pattern like this,

在这个图案中， 有一部分

is you,

代表你，

your perceptions,

代表你的感知，

your emotions, your memories,

你的情绪， 记忆，

your plans for the future.

你对将来的计划。

But we don't know where,

但是我们不知道到底哪块地方代表了这些，

since we don't know how to read the pattern.

因为我们不懂如何解读这些图案。

We don't understand the code used by the brain.

我们不了解大脑所用的密码。

To make progress,

要有进展，

we need to break the code.

我们需要破解这个密码。

But how?

但怎样解码呢?

An experienced code-breaker will tell you

一个富有经验的密码破译员会告诉你

that in order to figure out what the symbols in a code mean,

要解开密码中各个符号的真正意义，

it's essential to be able to play with them,

我们必须能够使用它们，

to rearrange them at will.

以我们的意愿重新组合它们。

So in this situation too,

在这个情况下也是如此，

to decode the information

要取得蕴含在这样的图案中

contained in patterns like this,

的信息，

watching alone won't do.

关靠观察是不行的；

We need to rearrange the pattern.

我们需要重新组合这些图案。

In other words,

也就是说，

instead of recording the activity of neurons,

仅仅记录神经细胞的活动并不可取，

we need to control it.

我们需要控制它们。

It's not essential that we can control

我们不用控制

the activity of all neurons in the brain, just some.

脑部中所有神经细胞的活动，一些就足够。

The more targeted our interventions, the better.

我们对那些神经细胞活动的干预愈具定向性，效果愈佳。

And I'll show you in a moment

我稍后会给你们展示

how we can achieve the necessary precision.

我们是怎样达到所需的精确度的。

And since I'm realistic, rather than grandiose,

因为我个人比较现实而不浮夸，

I don't claim that the ability to control the function of the nervous system

我不会宣称只要具备能够操控神经系统功能的能力

will at once unravel all its mysteries.

就会一次揭开所有关于神经细胞的奥秘。

But we'll certainly learn a lot.

但是我们一定会从中学到很多。

Now, I'm by no means

现在，我决不是历史上

the first person to realize

第一个意识到工具的发明是

how powerful a tool intervention is.

多么重要的人。

The history of attempts

试图

to tinker with the function of the nervous system

去改变神经系统机能的历史

is long and illustrious.

长久卓越。

It dates back at least 200 years,

这可以追溯到至少二百年前,

to Galvani's famous experiments

十八世纪初

in the late 18th century and beyond.

伽伐尼的著名实验。

Galvani showed that a frog's legs twitched

伽伐尼证明，只要他把青蛙的腰椎神经

when he connected the lumbar nerve

接上电流

to a source of electrical current.

青蛙的脚就会抽搐。

This experiment revealed the first, and perhaps most fundamental,

这个实验揭示了第一个，也是最根本的一个，

nugget of the neural code:

关于神经编码的事实：

that information is written in the form

这些信息是以

of electrical impulses.

电子脉冲的形式编写的。

Galvani's approach

伽伐尼的

of probing the nervous system with electrodes

以电极探究神经系统的方法

has remained state-of-the-art until today,

即使在今天，仍然是最先进的，

despite a number of drawbacks.

虽然它有一些很明显的缺陷。

Sticking wires into the brain is obviously rather crude.

将电线接到脑部显然是相当残忍的。

It's hard to do in animals that run around,

这样的设计也难于在移动的动物身上实践。

and there is a physical limit

当中亦有物理上的限制

to the number of wires

只有有限数量的电线

that can be inserted simultaneously.

能够同时安插在脑部。

So around the turn of the last century,

因此，在上世纪末，

I started to think,

我开始思考，

"Wouldn't it be wonderful if one could take this logic

如果我可以把这个逻辑

and turn it upside down?"

翻个个儿，那该多好啊。

So instead of inserting a wire

不把电线

into one spot of the brain,

接到大脑的某一点，

re-engineer the brain itself

而是改造大脑本身

so that some of its neural elements

使得某些神经系统的组件

become responsive to diffusely broadcast signals

可以感应到漫射讯号，

such as a flash of light.

比如说一道闪光。

Such an approach would literally, in a flash of light,

这样的方式能够，在电光一闪之间，

overcome many of the obstacles to discovery.

跨越很多科学发现的障碍。

First, it's clearly a non-invasive,

首先，这显然是个无创伤性的，

wireless form of communication.

无线的交流形式。

And second, just as in a radio broadcast,

其次，就像无线电广播一样，

you can communicate with many receivers at once.

你可以跟很多讯号接收器同时沟通。

You don't need to know where these receivers are,

你并不需要知道那些接收器所在的地点。

and it doesn't matter if these receivers move --

再者，讯号接收器移动也没关系——

just think of the stereo in your car.

就好象你车里的收音机。

It gets even better,

我们的情况更加理想，

for it turns out that we can fabricate the receivers

因为我们可以用DNA中所编码的材料

out of materials that are encoded in DNA.

来制造这些接收器。

So each nerve cell

每个具有合适的基因构造

with the right genetic makeup

的神经细胞

will spontaneously produce a receiver

都会自然产生出接收器

that allows us to control its function.

让我们能控制它的功能。

I hope you'll appreciate

我希望你们能够欣赏

the beautiful simplicity

这个概念中

of this concept.

美丽的简单朴素。

There's no high-tech gizmos here,

这里面没有安置高科技的小玩意儿

just biology revealed through biology.

只有用生物学揭示出的生物学。

Now let's take a look at these miraculous receivers up close.

现在，让我们来近距离看一下这些不可思异的接收器。

As we zoom in on one of these purple neurons,

当我们把镜头拉近到一个紫色的神经细胞，

we see that its outer membrane

我们可以看到它的细胞外膜

is studded with microscopic pores.

布满着微细的气孔。

Pores like these conduct electrical current

这些气孔能够让电流通过

and are responsible

以及负责

for all the communication in the nervous system.

为神经系统传递信息。

But these pores here are special.

但是你看到的是一些特别的气孔。

They are coupled to light receptors

它们结合了光感应受体

similar to the ones in your eyes.

这些光感应受体跟你眼睛里的很相似。

Whenever a flash of light hits the receptor,

只要一束闪光射在这些光感应受体上，

the pore opens, an electrical current is switched on,

那些气孔便会打开，电流亦会随之而接通，

and the neuron fires electrical impulses.

那神经细胞便会发出电子脉冲。

Because the light-activated pore is encoded in DNA,

因为那些光驱动气孔是编写在DNA里的，

we can achieve incredible precision.

我们可以达到无比的精确性。

This is because,

这是因为

although each cell in our bodies

尽管我们身体里的每一个细胞

contains the same set of genes,

都有同一组基因，

different mixes of genes get turned on and off

但是在不同的细胞里

in different cells.

不同组合的基因会被开启或关闭。

You can exploit this to make sure

你可以利用这个原理来确保

that only some neurons

只有部分的神经细胞

contain our light-activated pore and others don't.

具有我们设计的光驱动气孔，而其他的没有。

So in this cartoon, the bluish white cell

在这个动画中，蓝白色的细胞

in the upper-left corner

在左上角的那些

does not respond to light

并不会对光有反应

because it lacks the light-activated pore.

因为它们缺少了光驱动气孔。

The approach works so well

这个方法很管用

that we can write purely artificial messages

我们可以把完全人工的讯息

directly to the brain.

直接编写入大脑。

In this example, each electrical impulse,

在这个例子中，每一个电子脉冲，

each deflection on the trace,

轨迹的每一次偏斜，

is caused by a brief pulse of light.

都是由一束短暂的光脉冲造成的。

And the approach, of course, also works

同时，这个方法亦适用

in moving, behaving animals.

于移动中，表现良好的动物身上。

This is the first ever such experiment,

这是同类实验中的第一个，

sort of the optical equivalent of Galvani's.

可以说是光学版本的伽伐尼实验。

It was done six or seven years ago

这是在六七年前

by my then graduate student, Susana Lima.

由我的研究生苏珊娜.利马完成的。

Susana had engineered the fruit fly on the left

苏珊娜改变了左边那只果蝇的基因

so that just two out of the 200,000 cells in its brain

使它大脑中二十万神经细胞中的两个

expressed the light-activated pore.

出现光驱动气孔。

You're familiar with these cells

你应该相当熟悉这些神经细胞

because they are the ones that frustrate you

因为它们就是那些在你尝试拍打果蝇的时候

when you try to swat the fly.

让你倍感沮丧的神经细胞。

They trained the escape reflex that makes the fly jump into the air

它们造就了果蝇的逃跑反射，使果蝇

and fly away whenever you move your hand in position.

在你的手要靠近的时候飞到空中逃走。

And you can see here that the flash of light has exactly the same effect.

你可以看到，一阵闪光跟拍打的动作有着同样的效果。

The animal jumps, it spreads its wings, it vibrates them,

果蝇跳起，伸展翅膀，震动它们，

but it can't actually take off

但它们不能真的飞离地面，

because the fly is sandwiched between two glass plates.

因为这只果蝇被夹在两片玻璃板中间。

Now to make sure that this was no reaction of the fly

要保证果蝇如此反应

to a flash it could see,

并非因为它看到了那道闪光，

Susana did a simple

苏珊娜做了一个简单

but brutally effective experiment.

但直截了当而有效的实验。

She cut the heads off of her flies.

她把果蝇的头切掉。

These headless bodies can live for about a day,

这些无头的身体可以生存约一天。

but they don't do much.

但它们并不会有太多活动。

They just stand around

它们会站着

and groom excessively.

并不停的替自己梳理。

So it seems that the only trait that survives decapitation is vanity.

如此看来，断头之后，好像只有虚荣心这一个人格特征能够保存下来。

(Laughter)

（笑声）

Anyway, as you'll see in a moment,

无论如何，一会后你将会看到，

Susana was able to turn on the flight motor

苏珊娜能够启动果蝇的逃走运动神经

of what's the equivalent of the spinal cord of these flies

这相等于果蝇的脊椎

and get some of the headless bodies

并令一些无头的身体

to actually take off and fly away.

离地飞走。

They didn't get very far, obviously.

很明显，它们飞不了很远。

Since we took these first steps,

自从我们走出了第一步之后，

the field of optogenetics has exploded.

光基因技术这个领域百花齐放。

And there are now hundreds of labs

现在，已经有数以百计的实验室

using these approaches.

在使用这些方法。

And we've come a long way

从伽伐尼及苏珊娜的第一步

since Galvani's and Susana's first successes

令动物抽搐或跳动的成功到现在

in making animals twitch or jump.

我们取得了长足的进步。

We can now actually interfere with their psychology

现在我们能够

in rather profound ways,

彻底地干扰它们的心理。

as I'll show you in my last example,

在我最后的例子中，我将会为你展示，

which is directed at a familiar question.

一个老生常谈的问题。

Life is a string of choices

生命就是一连串的决择

creating a constant pressure to decide what to do next.

这就造成了一个常存的压力，迫使我们决定下一步的行动。

We cope with this pressure by having brains,

我们以大脑来应对这个压力。

and within our brains, decision-making centers

我们的大脑是我们的决策中心

that I've called here the "Actor."

我把它叫做“执行者”。

The Actor implements a policy that takes into account

这个“执行者”执行一个政策

the state of the environment

这政策会考虑到周遭环境的因素

and the context in which we operate.

以及我们生活的背景。

Our actions change the environment, or context,

我们的行为改变环境，或背景，

and these changes are then fed back into the decision loop.

而这些改变会反馈到我们的决策回路中。

Now to put some neurobiological meat

现在要把一些神经生物学的东西

on this abstract model,

添加到这个模型中，

we constructed a simple one-dimensional world

我们建造了一个简单的一维空间

for our favorite subject, fruit flies.

把我们最爱的实验对象，果蝇，放进去。

Each chamber in these two vertical stacks

两架玻璃管中的每一个小室

contains one fly.

都放有一只果蝇。

The left and the right halves of the chamber

小室的左右两边

are filled with two different odors,

弥漫着两种不同的气味，

and a security camera watches

闭路电视会一直观察着

as the flies pace up and down between them.

看着果蝇来回踱步。

Here's some such CCTV footage.

这里有一些闭路电视的片段。

Whenever a fly reaches the midpoint of the chamber

当一只果蝇到达小室的中间

where the two odor streams meet,

两种气味会交汇的地方时，

it has to make a decision.

果蝇必须作出决定。

It has to decide whether to turn around

它要决定是回头

and stay in the same odor,

留在同一种气味中，

or whether to cross the midline

还是跨过中线

and try something new.

去尝试新事物。

These decisions are clearly a reflection

这些决策显然是

of the Actor's policy.

根据“执行者”的政策来执行的。

Now for an intelligent being like our fly,

对我们的果蝇，这样有智慧的生物来说，

this policy is not written in stone

这个政策并非刻在石头上一成不变的，

but rather changes as the animal learns from experience.

它会根据自身的经验而转变。

We can incorporate such an element

我们可以将这样一个

of adaptive intelligence into our model

适应性智能的元素加进我们的模型里

by assuming that the fly's brain

我们要假设果蝇的大脑里

contains not only an Actor,

不只有一个“执行者”，

but a different group of cells,

而是也拥有多组不同的细胞，

a "Critic," that provides a running commentary

包括一个“批评家”，不断地

on the Actor's choices.

给“执行者”的决定提出意见。

You can think of this nagging inner voice

你可以把这个唠叨的内在声音

as sort of the brain's equivalent

当成是大脑里的

of the Catholic Church,

天主教教堂，

if you're an Austrian like me,

如果你跟我一样是奥地利人的话，

or the super-ego, if you're Freudian,

或者可以把它当成佛洛依德学说中所说的“超我”，

or your mother, if you're Jewish.

如果你是犹太人，或者可以把它当成你的“母亲大人”。

(Laughter)

（笑声)

Now obviously,

现在，明显的

the Critic is a key ingredient

这个“批评家”在我们的智力系统中

in what makes us intelligent.

是一个重要的组成部分

So we set out to identify

所以，我们希望可以认出

the cells in the fly's brain

这些在果蝇的大脑中

that played the role of the Critic.

扮演着“批评家”角色的细胞。

And the logic of our experiment was simple.

我们实验的逻辑很简单。

We thought if we could use our optical remote control

我们的假设是：如果我们能够用我们的光学遥控器

to activate the cells of the Critic,

来驱动那些“批评家”细胞，

we should be able, artificially, to nag the Actor

那么我们应该可以，人工地，不断烦扰“执行者”

into changing its policy.

使它改变它的政策。

In other words,

换句话说，

the fly should learn from mistakes

这飞蝇应该能够从错误中学习

that it thought it had made

它们会认为自己犯了错误，

but, in reality, it had not made.

即使它们其实并没有。

So we bred flies

我们培植了一些果蝇

whose brains were more or less randomly peppered

它们的脑部被多多少少地随机安置了

with cells that were light addressable.

一些可以用光驱动的细胞。

And then we took these flies

我们拿出这些果蝇

and allowed them to make choices.

给予它们做决策的机会。

And whenever they made one of the two choices,

每当它们作出两选一的决定，

chose one odor,

选到一种气味时，

in this case the blue one over the orange one,

在这个具体案例中，它选了蓝色而非橙色的那种，

we switched on the lights.

我们就亮灯。

If the Critic was among the optically activated cells,

如果“批评家”在光驱动细胞中，

the result of this intervention

这种干扰的结果会使

should be a change in policy.

果蝇改变它的政策。

The fly should learn to avoid

果蝇就应该学会去避开

the optically reinforced odor.

那种受光学加强的气味。

Here's what happened in two instances:

这里有两个不同案例中的情况。

We're comparing two strains of flies,

我们在比较两个不同品种的果蝇，

each of them having

它们的脑部分别有

about 100 light-addressable cells in their brains,

大概一百个可光驱动的细胞，

shown here in green on the left and on the right.

在这里以绿色来显示。

What's common among these groups of cells

两组细胞的相同之处

is that they all produce the neurotransmitter dopamine.

在于它们都会制造神经递质多巴胺。

But the identities of the individual

但是每个制造多巴胺的神经细胞

dopamine-producing neurons

的身份

are clearly largely different on the left and on the right.

在左右两边显然是不同的。

Optically activating

以光学来驱动

these hundred or so cells

大概一百多个细胞

into two strains of flies

在两个不同品种的果蝇里，

has dramatically different consequences.

会有极其不同的后果。

If you look first at the behavior

如果你先看看

of the fly on the right,

右边果蝇的行为，

you can see that whenever it reaches the midpoint of the chamber

你可以看到，当它到达小室的中心点

where the two odors meet,

两种气味的交汇的地方时，

it marches straight through, as it did before.

它会一直走过去，就像从前一样。

Its behavior is completely unchanged.

它的行为完全没有受光学驱动细胞的影响。

But the behavior of the fly on the left is very different.

但是左边的果蝇，情况就大不相同了

Whenever it comes up to the midpoint,

每当它来到中心点，

it pauses,

它会停下，

it carefully scans the odor interface

谨慎地检查气味的接合点，

as if it was sniffing out its environment,

就好像在打探周遭的环境，

and then it turns around.

然后它会回头。

This means that the policy that the Actor implements

这证明“执行者”实施的政策中

now includes an instruction to avoid the odor

现在包括了躲开那种气味的指令

that's in the right half of the chamber.

该种气味是从小室右边散发过来的。

This means that the Critic

这说明“批评家”

must have spoken in that animal,

一定对果蝇提过意见了，

and that the Critic must be contained

这说明在左边的果蝇中，“批评家”

among the dopamine-producing neurons on the left,

是那些能够制造多巴胺的神经细胞中的一个，

but not among the dopamine producing neurons on the right.

但在右边的果蝇中，“批评家”并不在那些能够制造多巴胺的神经细胞中。

Through many such experiments,

通过很多这样的实验，

we were able to narrow down

我们成功地把

the identity of the Critic

“批评家”的身分

to just 12 cells.

缩小到了十二个细胞。

These 12 cells, as shown here in green,

这十二个细胞，在这里以绿色显示，

send the output to a brain structure

它们对一个

called the "mushroom body,"

叫做“蘑菇体”的脑结构发出讯号，

which is shown here in gray.

在这里以灰色呈现出来。

We know from our formal model

我们从抽象模型中知道

that the brain structure

那个接收“评论员”批评的脑结构

at the receiving end of the Critic's commentary is the Actor.

就是“执行者”。

So this anatomy suggests

这个分析指出

that the mushroom bodies have something to do

蘑菇体和

with action choice.

行动的决策有着一定的关系。

Based on everything we know about the mushroom bodies,

基于我们对蘑菇体的了解，

this makes perfect sense.

这是完全能够想像的。

In fact, it makes so much sense

事实上，这非常合理，

that we can construct an electronic toy circuit

我们能够建构一个电子玩具电路

that simulates the behavior of the fly.

用以模拟果蝇的行为。

In this electronic toy circuit,

这个电子玩具电路中，

the mushroom body neurons are symbolized

蘑菇体神经细胞以

by the vertical bank of blue LEDs

在电路中心的竖排蓝色发光二极管

in the center of the board.

来表示。

These LED's are wired to sensors

这些发光二极管上接有感应器

that detect the presence of odorous molecules in the air.

用以探测空气中的气味分子。

Each odor activates a different combination of sensors,

不同的气味都会激活不同组合的感应器，

which in turn activates

而它们再驱动

a different odor detector in the mushroom body.

蘑菇体中另一个气味检测器。

So the pilot in the cockpit of the fly,

所以在果蝇驾驶舱中的飞行员，

the Actor,

“执行者”，

can tell which odor is present

要知道哪一种气味存在

simply by looking at which of the blue LEDs lights up.

只要看看哪颗蓝色发光二极管亮起来就行了。

What the Actor does with this information

“执行者”得到这个讯息之后的行为

depends on its policy,

取决于它的政策，

which is stored in the strengths of the connection,

这些政策都是根据

between the odor detectors

气味检测器

and the motors

与运动神经之间关联的强度来储存的

that power the fly's evasive actions.

这驱动了果蝇的逃亡行为。

If the connection is weak, the motors will stay off

如果关联性弱，运动神经会保持关上

and the fly will continue straight on its course.

那只果蝇会继续前进。

If the connection is strong, the motors will turn on

如果关联性强，运动神经就会启动

and the fly will initiate a turn.

那只果蝇就会作一个转身。

Now consider a situation

现在试想这样一个情况

in which the motors stay off,

就是当运动神经保持关上时，

the fly continues on its path

那只果蝇继续前行

and it suffers some painful consequence

它就会遭受一些痛苦的后果

such as getting zapped.

例如遭到电击。

In a situation like this,

在这样的情况下，

we would expect the Critic to speak up

我们可以预期“批评家”会发表意见

and to tell the Actor

并告诉“执行者”

to change its policy.

要它改变它的政策。

We have created such a situation, artificially,

我们人工地制造了这样的一个情境

by turning on the critic with a flash of light.

以一束光来启动“批评家”。

That caused a strengthening of the connections

这样就可以加强

between the currently active odor detector

正在起作用的气味检测细胞

and the motors.

与运动神经之间的关联。

So the next time

所以，下一次

the fly finds itself facing the same odor again,

当果蝇再次面对同样的气味时，

the connection is strong enough to turn on the motors

关联性将会有足够的强度去启动运动神经

and to trigger an evasive maneuver.

并引发果蝇进行回避。

I don't know about you,

我不知道你的想法，

but I find it exhilarating to see

但我觉得这样的实验非常令人兴奋

how vague psychological notions

虚无的心理学概念

evaporate and give rise

不见了，却引出了

to a physical, mechanistic understanding of the mind,

对思维在物理学上，机能上的理解，

even if it's the mind of the fly.

虽然这只是果蝇的思维。

This is one piece of good news.

这是一个好消息。

The other piece of good news,

另一个好消息，

for a scientist at least,

至少对于一个科学家来说，

is that much remains to be discovered.

就是世界上还有很多尚待发掘的东西。

In the experiments I told you about,

在我讲诉的这个实验里，

we have lifted the identity of the Critic,

我们发掘出了“批评家”的真正身分，

but we still have no idea

但是我们还不知道

how the Critic does its job.

“批评家”是怎样完成它的工作的。

Come to think of it, knowing when you're wrong

试想一下，在没有老师，或者你的母亲大人告诉你的情况下

without a teacher, or your mother, telling you,

要知道你自己犯了错，

is a very hard problem.

是一件非常不容易的事情。

There are some ideas in computer science

电脑科学领域

and in artificial intelligence

以及人工智能领域中的一些想法

as to how this might be done,

就是试图去解释这件事情是如何发生的，

but we still haven't solved

但是到现在我们还无法解释

a single example

任何一个实际例子中

of how intelligent behavior

智能行为是如何

springs from the physical interactions

在生命体的物理交互中

in living matter.

产生的。

I think we'll get there in the not too distant future.

我想在不久的将来， 我们会找到答案的。

Thank you.

谢谢。

(Applause)

(掌声)