Torsten Reil: Using biology to make better animation

翻译人员: Xiaoqiao Xie 校对人员: Felix Chen

I'm going to talk about a technology that we're developing at Oxford now,

我想谈谈一个我们正在牛津开发的技术，

that we think is going to change the way that

我们认为这个技术可以改变

computer games and Hollywood movies are being made.

电脑游戏和好莱坞电影的制作方法

That technology is simulating humans.

这个技术就是模拟人类。

It's simulated humans with a simulated body

这个模拟人类有电脑模拟的身体，

and a simulated nervous system to control that body.

和电脑模拟的神经系统来控制身体。

Now, before I talk more about that technology,

现在，在我开始讲解这个技术之前，

let's have a quick look at what human characters look like

让我们先快速浏览一下在现今的电脑游戏中

at the moment in computer games.

游戏人物角色是什么样子的。

This is a clip from a game called "Grand Theft Auto 3."

这是从一个名为“侠盗飞车3”的游戏中摘出的片断。

We already saw that briefly yesterday.

我们昨天已经见过了。

And what you can see is -- it is actually a very good game.

可以看出，这是个非常精美的游戏。

It's one of the most successful games of all time.

这是有史以来最成功的电脑游戏之一。

But what you'll see is that all the animations in this game are very repetitive.

但是你也可以发现这个游戏中所有的模拟动作都有很大的重复性。

They pretty much look the same.

所有的动作看起来都一样。

I've made him run into a wall here, over and over again.

我让这个人物一次又一次地撞到墙上，

And you can see he looks always the same.

你可以发现每次它（的反应）看起来都一样。

The reason for that is that these characters

这是因为这些动画人物

are actually not real characters.

并不是真正的“人”，

They are a graphical visualization of a character.

它们是“人”的虚拟图像。

To produce these animations, an animator at a studio has to anticipate

设计师做出这些动画人物的时候

what's going to happen in the actual game,

得要猜想游戏中将会发生什么情境，

and then has to animate that particular sequence.

然后还得根据这些特定的情境画出（动画人物的反应）。

So, he or she sits down, animates it, and tries to anticipate what's going to happen,

所以设计师得坐下来，画出动画，还要猜想将会发生什么，

and then these particular animations are just played back

所以这些人物的反应都是重复播放的设定动作，

at appropriate times in the computer game.

只是根据电脑游戏中特定的情境播放罢了。

Now, the result of that is that you can't have real interactivity.

结果游戏中并没有真的互动。

All you have is animations that are played back

有的仅仅是重复播放的动画，

at more or less the appropriate times.

关键只是选择合适的时机罢了。

It also means that games aren't really going to be as surprising as they could be,

这也说明现在的电脑游戏并不具备它们应有的新鲜度，

because you only get out of it, at least in terms of the character,

因为所有你能得到的动画效果，至少在人物动作方面，

what you actually put into it.

仅仅是设计者储存在游戏中的那些。

There's no real emergence there.

并不会有真正的效果。

And thirdly, as I said, most of the animations are very repetitive because of that.

第三点，如我所说，绝大部分的动画重复性都很大，也是因为都是预存的。

Now, the only way to get around that

唯一的解决办法

is to actually simulate the human body

就是实实在在地模拟人体的反应，

and to simulate that bit of the nervous system of the brain that controls that body.

模拟人的神经系统是怎样控制身体的运动的。

And maybe, if I could have you for a quick demonstration

这样吧，如果我能借你做个演示

to show what the difference is --

来展示（真人的反应）的不同 —

because, I mean, it's very, very trivial.

因为，要知道，有的差异是很微妙的。

If I push Chris a bit, like this, for example, he'll react to it.

好比我轻轻推克里斯一下，就像这样，他就会给出反应。

If I push him from a different angle, he'll react to it differently,

如果我从不同的角度推他，他的身体反应会不同。

and that's because he has a physical body,

那是因为他有一个真正的身体，

and because he has the motor skills to control that body.

他有那些运动技能，能够控制他的身体。

It's a very trivial thing.

这是非常微妙的。

It's not something you get in computer games at the moment, at all.

这不是你能从现今的电脑游戏中得到的反应。

Thank you very much. Chris Anderson: That's it?

谢了克里斯。（克里斯·安德森 问：“这样就行了？”）

Torsten Reil: That's it, yes.

是的，就这样。

So, that's what we're trying to simulate --

这就是我们试图模拟的 —

not Chris specifically, I should say, but humans in general.

不是单单模拟 Chris 这个人，而是，所有的真人身体的反应。

Now, we started working on this a while ago at Oxford University,

在牛津大学我们已经开始一段时间的研究了，

and we tried to start very simply.

我们一开始做得很简单，

What we tried to do was teach a stick figure how to walk.

仅仅是训练这个火柴人走路。

That stick figure is physically stimulated. You can see it here on the screen.

这个火柴人的身体是能接受信号刺激的。你在屏幕上能看见。

So, it's subject to gravity, has joints, etc.

所以它能对重力有反应，它也有关节，等等。

If you just run the simulation, it will just collapse, like this.

如果你开始模拟程序，它会跌倒，就像这样。

The tricky bit is now to put an AI controller in it

现在难就难在怎样把人工智能控制器放进去，

that actually makes it work.

让它顺利工作。

And for that, we use the neural network, which we based on

我们用的是模拟的神经系统，

that part of the nervous system that we have in our spine

它是根据真的人体的脊髓中

that controls walking in humans.

控制走路的那部分神经系统设计的。

It's called the central pattern generator.

名字叫中枢模式发生器。

So, we simulated that as well, and then the really tricky bit

我们模拟了这个神经系统之后，真正的难点

is to teach that network how to walk.

就是教会这个神经系统如何控制身体来走路。

For that we used artificial evolution -- genetic algorithms.

为了解决这个问题我们使用的是人工进化系统 — 基因模拟算法。

We heard about those already yesterday,

我们昨天也听说过了。

and I suppose that most of you are familiar with that already.

这里我就当你们大部分人都知道那是怎么回事。

But, just briefly, the concept is that

但是，简短地说，这个概念是

you create a large number of different individuals --

你首先制造出一大批个体来，各个不同，

neural networks, in this case --

这里我们做的是一大批神经系统。

all of which are random at the beginning.

开始时，每个的工作模式都是随机的。

You hook these up -- in this case, to the virtual muscles

你把这些接到这个火柴人上 —这里我们把这些神经系统

of that two-legged creature here --

接到这些两腿人的模拟肌肉系统上 —

and hope that it does something interesting.

然后就等着它们随机工作。

At the beginning, they're all going to be very boring.

一开始火柴人的反应都挺没劲的。

Most of them won't move at all,

绝大部分都不动，

but some of them might make a tiny step.

有些动起来了，也仅仅是迈一小步。

Those are then selected by the algorithm,

这些让火柴人动起来的神经系统会被挑出来，

reproduced with mutation and recombinations to introduce sex as well.

用来生产下一代，其中加入了基因变异和基因重组，也加入了性别。

And you repeat that process over and over again,

然后你不断重复这些步骤，

until you have something that walks --

直到你得到真的能够让火柴人走路的神经系统 —

in this case, in a straight line, like this.

就像这里这个能走一条直线的。

So that was the idea behind this.

这就是这个实验想做的。

When we started this, I set up the simulation one evening.

当我们开始做时，我在一天晚上启动了一个实验，

It took about three to four hours to run the simulation.

整个模拟试验一般要三四个小时。

I got up the next morning, went to the computer and looked at the results,

第二天早上起来我跑去看结果，

and was hoping for something that walked in a straight line,

满怀希望能得到一个走直线的动画人，

like I've just demonstrated,

就像我给你们看的那个，

and this is what I got instead.

结果我看到的是这个。

(Laughter)

（笑声）

So, it was back to the drawing board for us.

我们又得从头来过。

We did get it to work eventually,

最后我们当然做成了，

after tweaking a bit here and there.

改改这里改改那里。

And this is an example of a successful evolutionary run.

这里是个进化成功的例子。

So, what you'll see in a moment is a very simple biped

你马上将要看到的是一个非常简单的两足动物

that's learning how to walk using artificial evolution.

用人工进化的方法学习如何走路。

At the beginning, it can't walk at all,

一开始它一点也走不了，

but it will get better and better over time.

但是走的越来越好。

So, this is the one that can't walk at all.

这里是那个一点也不会走路的。

(Laughter)

（笑声）

Now, after five generations of applying evolutionary process,

现在，当运用了人工进化，五代之后

the genetic algorithm is getting a tiny bit better.

基因算法变得好些了。

(Laughter)

（笑声）

Generation 10 and it'll take a few steps more --

第十代，这个两足动物能多走几步了。

still not quite there.

但是还是不太行。

But now, after generation 20, it actually walks in a straight line without falling over.

现在，二十代后，它能真的走出一条直线来，也不会跌到。

That was the real breakthrough for us.

这对我们来说是个真正的突破。

It was, academically, quite a challenging project,

这是个学术上非常有挑战性的项目，

and once we had reached that stage, we were quite confident

而且当我们能做到这一步的时候，我们就有信心

that we could try and do other things as well with this approach --

能够挑战其他的目标，比如用这个方法 —

actually simulating the body

真的来模拟人体，

and simulating that part of the nervous system that controls it.

模拟用来控制身体运动的那部分神经。

Now, at this stage, it also became clear that this could be very exciting

并且那个阶段我们确信这个技术将会

for things like computer games or online worlds.

对电脑游戏或者网络世界有很大意义。

What you see here is the character standing there,

现在你看到的是一个模拟人物，站在那里，

and there's an obstacle that we put in its way.

前面我们放了一个障碍物。

And what you see is, it's going to fall over the obstacle.

你将会看到，它会被障碍物绊倒。

Now, the interesting bit is, if I move the obstacle a tiny bit to the right,

现在，有趣的是，如果我把障碍物向右边挪一点，

which is what I'm doing now, here,

就像这样，

it will fall over it in a completely different way.

它会以一种非常不同的方式跌倒。

And again, if you move the obstacle a tiny bit, it'll again fall differently.

再来一次，如果你把障碍物再挪一点，它跌到的方式又会不同。

(Laughter)

（笑声）

Now, what you see, by the way, at the top there,

现在，你看到的是，喔，顺便一提，在屏幕的上面，

are some of the neural activations being fed into the virtual muscles.

是我们传入肌肉的一些神经的活动，

Okay. That's the video. Thanks.

好的，这里就是这个片段。谢谢。

Now, this might look kind of trivial, but it's actually very important

好的，这看起来微不足道但是其是非常重要，

because this is not something you get at the moment

因为这是现今你在任何的

in any interactive or any virtual worlds.

虚拟世界或者互动游戏中能够看到的。

Now, at this stage, we decided to start a company and move this further,

现在，从这个阶段，我们决定成立一个公司，开始进一步的研究，

because obviously this was just a very simple, blocky biped.

因为很明显这只是个非常简单的块状两足动物，

What we really wanted was a full human body.

我们真的想做的是一个完整的模拟人体。

So we started the company.

这样我们成立了公司。

We hired a team of physicists, software engineers and biologists

我们吸收了一些物理学家，软件工程师和生物学家

to work on this, and the first thing we had to work on

一起工作，我们做的第一件事是

was to create the human body, basically.

做出一个模拟人体。

It's got to be relatively fast, so you can run it on a normal machine,

这个模拟人需要很轻巧，这样你才能在一般的电脑上运作。

but it's got to be accurate enough, so it looks good enough, basically.

但是也需要非常地准确，才会好看。

So we put quite a bit of biomechanical knowledge into this thing,

所以我们夹了不少生化科技进去，

and tried to make it as realistic as possible.

使它看起来像是真的。

What you see here on the screen right now

这里你在屏幕上看到的，

is a very simple visualization of that body.

是这个模拟人体的一个简单版，

I should add that it's very simple to add things like hair, clothes, etc.,

值得一提的是给它加上头发，衣服之类的是非常简单的。

but what we've done here is use a very simple visualization,

但是我们决定做得很简洁

so you can concentrate on the movement.

这样大家的注意力才会集中在它的动作上。

Now, what I'm going to do right now, in a moment,

现在，我想做的是，

is just push this character a tiny bit and we'll see what happens.

推这个模拟人一下下，看看会发生什么。

Nothing really interesting, basically.

没什么有趣的。

It falls over, but it falls over like a rag doll, basically.

它跌到了，像个假娃娃。

The reason for that is that there's no intelligence in it.

这是因为我们没有放人工智能进去。

It becomes interesting when you put artificial intelligence into it.

当我们放人工智能进去就有趣多了。

So, this character now has motor skills in the upper body --

现在这个模拟人的上半身有运动技能。

nothing in the legs yet, in this particular one.

下半身没有，在这个人体里。

But what it will do -- I'm going to push it again.

但是马上你会看到 — 我现在在推它一下，

It will realize autonomously that it's being pushed.

它有自主意识被推了，

It's going to stick out its hands.

它会张开双臂，

It's going to turn around into the fall, and try and catch the fall.

向跌倒的方向转身，试图撑住。

So that's what you see here.

这就是你在这里看到的。

Now, it gets really interesting

如果下半身也加上人工智能，

if you then add the AI for the lower part of the body as well.

就更有趣了。

So here, we've got the same character.

这里看到的是同一个模拟人，

I'm going to push it a bit harder now,

我现在推得更狠一些，

harder than I just pushed Chris.

比刚才我推克里斯更用力些。

But what you'll see is -- it's going to receive a push now from the left.

你将看到我从左边推它。

What you see is it takes steps backwards,

你将会看到它向后退 —

it tries to counter-balance,

试图平衡自己，

it tries to look at the place where it thinks it's going to land.

试图向下看。

I'll show you this again.

我马上播放给你看。

And then, finally hits the floor.

最后它倒到地上。

Now, this becomes really exciting

现在当你从不同角度推它的时候，

when you push that character in different directions, again, just as I've done.

它的反应就有趣多了，就像我刚做的。

That's something that you cannot do right now.

这是你现在得不到的。

At the moment, you only have empty computer graphics in games.

现在你从游戏里只能看到空洞的动画图像。

What this is now is a real simulation. That's what I want to show you now.

这里我给你看的是真的模拟人生。

So, here's the same character with the same behavior I've just shown you,

这里是同一个模拟人，能作相同的反应。

but now I'm just going to push it from different directions.

现在我从不同角度推它，

First, starting with a push from the right.

先从右边退，

This is all slow motion, by the way, so we can see what's going on.

这是慢动作这样我们能看得更清楚。

Now, the angle will have changed a tiny bit,

现在我换个角度，

so you can see that the reaction is different.

你看一看到它的反应是不同的。

Again, a push, now this time from the front.

再来一次，推一下，从前面

And you see it falls differently.

你看它跌到的方式又变了。

And now from the left --

现在从左边。

and it falls differently.

又不一样。

That was really exciting for us to see that.

我们看到这个结果很激动，

That was the first time we've seen that.

这是第一次我们看到成果。

This is the first time the public sees this as well,

今天是第一次向公众发表。

because we have been in stealth mode.

因为我们还在保密阶段的左，

I haven't shown this to anybody yet.

我还没给人看过。

Now, just a fun thing:

现在，只是好玩，

what happens if you put that character --

看看如果我们把它放在不同情境会发生什么 —

this is now a wooden version of it, but it's got the same AI in it --

这是个木头人的版本，放了人工智能进去 —

but if you put that character on a slippery surface, like ice.

我们把它放在光滑的表面上，比如冰面。

We just did that for a laugh, just to see what happens.

我们这么做只是为了看看会发生什么滑稽的事。

(Laughter)

（笑声）

And this is what happens.

结果是这样。

(Laughter)

（笑声）

(Applause)

（鼓掌）

It's nothing we had to do about this.

我们没有人为的加东西进去。

We just took this character that I just talked about,

我们只是用了类似的模拟人，

put it on a slippery surface, and this is what you get out of it.

把它放在光滑的表面上，结果就是这样。

And that's a really fascinating thing about this approach.

这正是这个研究的过人之处，

Now, when we went to film studios and games developers

当我们把这个结果给电影工作者和游戏的设计者看的时候，

and showed them that technology, we got a very good response.

向他们展示这项技术，我们得到了很好的反响。

And what they said was, the first thing they need immediately is virtual stuntmen.

他们最需要的是虚拟替身。

Because stunts are obviously very dangerous, they're very expensive,

因为替身工作非常危险，雇替身的花费也很大。

and there are a lot of stunt scenes that you cannot do, obviously,

很多使用替身的片断还不能用真人，

because you can't really allow the stuntman to be seriously hurt.

因为这些特技太危险了，真人会受伤。

So, they wanted to have a digital version of a stuntman

所以他们很需要一个数码版的替身，

and that's what we've been working on for the past few months.

这就是我们过去几个月里做的项目。

And that's our first product that we're going to release in a couple of weeks.

这是我们的第一个产品，几周后会上市。

So, here are just a few very simple scenes of the guy just being kicked.

这里是一些非常简单的情境，比如这个人刚被踢了一脚。

That's what people want. That's what we're giving them.

这是电影公司需要的，我们只是按他们的需要做。

(Laughter)

（笑声）

You can see, it's always reacting.

你可以看到，模拟人总会反应，

This is not a dead body. This is a body who basically, in this particular case,

这不像是没有生命的个体。这是一个能够感觉到施力的身体，

feels the force and tries to protect its head.

在这个特定情境下，能够自我保护。

Only, I think it's quite a big blow again.

我认为这个看起来很真实，

You feel kind of sorry for that thing,

所以人们开始觉得不忍。

and we've seen it so many times now that

我们这些人看得次数太多了，

we don't really care any more.

已经漠不关心了。

(Laughter)

（笑声）

There are much worse videos than this, by the way, which I have taken out, but ...

很多片断更糟糕，我都不能给你们看。

Now, here's another one.

这里是另一个片断，

What people wanted as a behavior was to have an explosion,

他们要求的是一个对爆炸情境的反应，

a strong force applied to the character,

好像有一个强大的力量施加到这个人物身上，

and have the character react to it in midair.

这个人物需要在空中作反应。

So that you don't have a character that looks limp,

你不会想要一个软绵绵无生气的模拟人，

but actually a character that you can use in an action film straight away,

你要的是一个能够在动作电影中用的，

that looks kind of alive in midair as well.

在空中的反应看起来是活的模拟人。

So this character is going to be hit by a force,

现在这个模拟人会被重重一击，

it's going to realize it's in the air,

它会意识到自己飞到空中，

and it's going to try and, well,

它会试图，

stick out its arm in the direction where it's landing.

向它跌落的方向伸出手臂。

That's one angle; here's another angle.

这里是从一个角度看，这里是从另一个角度看。

We now think that the realism we're achieving with this

我们现在认为我们可以成功地

is good enough to be used in films.

使用这些电脑替身。

And let's just have a look at a slightly different visualization.

然我们看看一些不同的情境，

This is something I just got last night

这里是我昨晚上拿到的结果，

from an animation studio in London, who are using our software

伦敦的一个动画公司用我们的软件做的，

and experimenting with it right now.

他们正在实验把这个做到游戏中去。

So this is exactly the same behavior that you saw,

这是同一个模拟人，

but in a slightly better rendered version.

但是稍微精美一些的版本。

So if you look at the character carefully,

如果你仔细地看，

you see there are lots of body movements going on,

你可以看到很多的肢体动作，

none of which you have to animate like in the old days.

这些动作不需要你画出来，

Animators had to actually animate them.

用老方法动画人需要真的把它们画出来，

This is all happening automatically in the simulation.

这里都是自然而然在模拟人身上发生的。

This is a slightly different angle,

这是另一个角度，

and again a slow motion version of this.

还是慢动作。

This is incredibly quick. This is happening in real time.

这一切都发生得很快，实时发生的。

You can run this simulation in real time, in front of your eyes,

你可以实时跑这个程序，这些动作就会发生在你眼前。

change it, if you want to, and you get the animation straight out of it.

你可以改变这个程序，然后得到不同的动作。

At the moment, doing something like this by hand

目前，手工绘制出这些动作，

would take you probably a couple of days.

恐怕要花好几天。

This is another behavior they requested.

这是另一个他们需要的动作。

I'm not quite sure why, but we've done it anyway.

我不知道为什么，但是我们按需作活。

It's a very simple behavior that shows you the power of this approach.

这是个非常简单的特技但是你可以看到这个技术的潜力。

In this case, the character's hands

这里模拟人的手

are fixed to a particular point in space,

被固定在空中一个特定的点上，

and all we've told the character to do is to struggle.

然后我们让这个模拟人挣扎。

And it looks organic. It looks realistic.

它的动作看起来很自然，很真实。

You feel kind of sorry for the guy.

你甚至会为“他”难过。

It's even worse -- and that is another video I just got last night --

这个更残忍 — 这是另一个昨晚我才拿到的片断 —

if you render that a bit more realistically.

如果你把人物作得更真实一点。

Now, I'm showing this to you just to show you

现在我给你们看这个片断，

how organic it actually can feel, how realistic it can look.

你可以感受到这些动作有多自然，看起来多真实。

And this is all a physical simulation of the body,

而这全是因为我们有这个模拟的身体，

using AI to drive virtual muscles in that body.

我们用了人工智能来驾驭这些肌肉。

Now, one thing which we did for a laugh was

现在，我想给你们看一个只是做了好玩的片断

to create a slightly more complex stunt scene,

我们只是想做一个更复杂的特技，

and one of the most famous stunts is the one where James Bond

而一个非常著名的特技动作是当零零七

jumps off a dam in Switzerland and then is caught by a bungee.

在瑞士跳入一个大水库，然后被蹦极绳子救了。

Got a very short clip here.

这里是这个真的电影片断。

Yes, you can just about see it here.

你可以看到。

In this case, they were using a real stunt man. It was a very dangerous stunt.

这里他们请了一个真的特技演员，但是这是个非常危险的特技。

It was just voted, I think in the Sunday Times, as one of the most impressive stunts.

我想在“周日报导”上这个特技刚被评为最惊人的特技之一。

Now, we've just tried and -- looked at our character and asked ourselves,

这里，我们试着用我们的模拟人，然后问自己

"Can we do that ourselves as well?"

“我们的模拟特技人能做到么？”

Can we use the physical simulation of the character,

我们能不能用这个模拟的人物，

use artificial intelligence,

用我们的人工智能方法，

put that artificial intelligence into the character,

把人工智能放在模拟人身体里，

drive virtual muscles, simulate the way he jumps off the dam,

让控制肌肉，模拟真人跳入水库，

and then skydive afterwards,

跳进去，

and have him caught by a bungee afterwards?

然后被蹦极绳拉回来？

We did that. It took about altogether just two hours,

我们真的做到了。仅仅两个小时，

pretty much, to create the simulation.

就作出了这个模拟情境。

And that's what it looks like, here.

这里是它看起来的样子。

Now, this could do with a bit more work. It's still very early stages,

我们可以做得更好一点，这个研究还在初级阶段，

and we pretty much just did this for a laugh,

我们仅仅是作了好玩的。

just to see what we'd get out of it.

只是看看我们能做出什么来。

But what we found over the past few months

我们过去几个月的工作证明了

is that this approach -- that we're pretty much standard upon --

我们的技术是非常非常

is incredibly powerful.

有潜力的。

We are ourselves surprised what you actually get out of the simulations.

对于我们能够从这些模拟人身上得到的结果，我们自己也惊讶了。

There's very often very surprising behavior that you didn't predict before.

我们常常得到非常出乎意料的结果。

There's so many things we can do with this right now.

我们的技术有很多应用。

The first thing, as I said, is going to be virtual stuntmen.

第一个，像我说的，是数码特技替身，

Several studios are using this software now to produce virtual stuntmen,

一些电影工作室已经在用这个软件做数码替身了。

and they're going to hit the screen quite soon, actually,

这些特技镜头会在很短的时间内上映，

for some major productions.

很多大的制作都用它们。

The second thing is video games.

第二个应用就是电脑游戏。

With this technology, video games will look different and they will feel very different.

用这个技术电脑游戏看起来会很不同，感觉上也会很不同。

For the first time, you'll have actors that really feel very interactive,

历史上第一次你会看到演员的非常真实的互动，

that have real bodies that really react.

真的身体，真的反应。

I think that's going to be incredibly exciting.

我认为这非常令人激动。

Probably starting with sports games,

最先我们会用在运动性的游戏上，

which are going to become much more interactive.

比起平常的游戏它们更需要互动性。

But I particularly am really excited

但是我本人最感到激动的应用

about using this technology in online worlds,

是将这个技术用在网络世界里

like there, for example, that Tom Melcher has shown us.

比如像汤姆·梅尔彻给我们展示的那样。

The degree of interactivity you're going to get

我们能够得到的互动性

is totally different, I think, from what you're getting right now.

是非常不同的，我认为，和你能从老方法中得到的相比。

A third thing we are looking at and very interested in is simulation.

这是第三个我们非常感兴趣的应用。

We've been approached by several simulation companies,

已经有一些公司联系我们了，

but one project we're particularly excited about, which we're starting next month,

但是有一个特别的项目我们感到特别的激动，我们下个月会开始做，

is to use our technology -- and in particular, the walking technology --

是将我们的技术，特别是模拟走路这方面，

to help aid surgeons who work on children with cerebral palsy,

用在帮助那些治疗儿童脑瘫的医生

to predict the outcome of operations on these children.

来预计手术的愈后效果。

As you probably know,

如你们所知，

it's very difficult to predict what the outcome of an operation is

通常脑瘫病人手术后的行走能力的结果

if you try and correct the gait.

是非常难以预料的，

The classic quote is, I think, it's unpredictable at best,

经典说法是，我认为，人们目前认为

is what people think right now, is the outcome.

最难预料的，就是愈后结果。

Now, what we want to do with our software is allow our surgeons to have a tool.

现在，我们想做的是用我们的软件，帮助医生预测。

We're going to simulate the gait of a particular child

我们想模拟每个孩子的步态，

and the surgeon can then work on that simulation

医生能够用这个模拟人物，

and try out different ways to improve that gait,

试不同的手术方案，来帮助调整步态，

before he actually commits to an actual surgery.

在他们真正做手术之前。

That's one project we're particularly excited about,

这个项目我们感到特别激动，

and that's going to start next month.

下个月会启动。

Just finally, this is only just the beginning.

最后，我想说这只是个起步，

We can only do several behaviors right now.

我们现在只能做一些特定的动作，

The AI isn't good enough to simulate a full human body.

我们用的人工智能还不能模拟整个人体的神经系统。

The body yes, but not all the motor skills that we have.

我们能模拟整个身体，但是不能模拟所有的动作功能。

And, I think, we're only there if we can have something like ballet dancing.

我希望能做出像芭蕾舞这类的动作来，

Right now, we don't have that

现在还做不到。

but I'm very sure that we will be able to do that at some stage.

但是我确信我们将来可以。

We do have one unintentional dancer actually,

我们碰巧做出了一个舞者，

the last thing I'm going to show you.

这是我想给你们看的最后一个人物。

This was an AI contour that was produced and evolved --

这个人工智能我们做出来 —

half-evolved, I should say -- to produce balance, basically.

部分上是 — 保持平衡的。

So, you kick the guy and the guy's supposed to counter-balance.

你踢这个人一脚，他应该会平衡自己。

That's what we thought was going to come out of this.

这也是我们原本预计的结果。

But this is what emerged out of it, in the end.

结果这是我们最后看到的东西。

(Music)

（音乐起）

Bizarrely, this thing doesn't have a head. I'm not quite sure why.

看起来怪怪的，这个舞者没有头，我不知道为什么，

So, this was not something we actually put in there.

但是我们通常不用头。

He just started to create that dance himself.

他自顾自地开始跳舞了。

He's actually a better dancer than I am, I have to say.

我得说，他比我还强些。

And what you see after a while --

你将会看到的 —

I think he even goes into a climax right at the end.

这个舞最后还来了一个高潮呢。

And I think -- there you go.

我想，就是这里。

(Laughter)

（笑声）

So, that all happened automatically. We didn't put that in there.

这都是自动产生的，我们没有加这个舞进去。

That's just the simulation creating this itself, basically.

这都是这个模拟人自己产生的。

So it's just --

这样 —

(Applause)

（掌声）

Thanks.

谢谢。

Not quite John Travolta yet, but we're working on that as well,

他不像约翰·特拉沃尔塔那样帅，但我们在向这个方向努力。

so thanks very much for your time.

谢谢你们给我时间。

Thanks.

谢谢

(Applause)

（掌声）

CA: Incredible. That was really incredible.

克里斯·安德森 ：“令人惊叹，这真是太令人惊叹了。”

TR: Thanks.

托斯腾·雷尔：“谢谢。”