与摩根·弗里曼一起穿越虫洞（2010）

Through The Wormhole With Morgan Freeman S04E07

Part 4 April 19, 2021, 2:53 p.m.

2006年
In 2006,
迈克尔着手开♥发♥智能计算机软件
Michael began developing intelligent computer software
这个软件可以观察复杂的自然系统
that could observe complex natural systems
从看起来像是混乱的事件中得到意义
and derive meaning from what seems like chaos.
这是一个双摆
So, what I have here is a double pendulum.
你可以看到它只有两个臂
If you look at it, it consists of two arms.
一个臂的顶部可绕轴运动
One arm swings along the top axis,
第二个臂的一端连接在第一个臂的尾部
and the second arm is attached to the bottom of the first arm,
这两个摆臂挂在一起
and it's two pendulums that are hooked together,
一个位于另一个的末端
one pendulum at the end of the other.
现在这个双摆就是复杂性的很好的例子
Now, the pendulum is a great example of complexity
因为它能够表现出一些最复杂的行为
because it exhibits some of the most complex behavior
我们称之为混沌
that we're aware of, which is called chaos.
所以当你从这类设备中收集数据时
So, when you collect data from this sort of device,
它的行为看起来几乎是完全随机的
it looks almost completely random,
不会出现任何类型的固定模式
and there doesn't appear to be any sort of pattern.
但由于这是一个确定的物理系统
But because this is a physical deterministic system,
就存在一个运动规则
a pattern does exist.
在双摆的混沌运动中寻找固定模式
Finding a pattern amidst the chaos of the double pendulum
几十年来一直困扰着科学家
has stumped scientists for decades.
但这时迈克尔突现灵感
But then Michael had a flash of inspiration.
为什么不以大自然创造我们的方式来想办法呢
Why not grow new ideas the same way nature created us,
通过进化？
using evolution?
他称自己的程序为“找到了”
He called his program Eureka.
“找到了”开始于充斥着随机公式的原始汤
Eureka starts with a primordial soup of random equations
然后检查这些公式看哪些
and checks how closely they fit
更接近于描述双摆的运动方式
the behavior of the double pendulum.
不合理的公式就会被电脑“杀掉”
If they don't fit, the computer kills them.
如果接近，计算机就允许它们进入下一代
If they do, the computer moves them into the next generation,
在这一代中它们发生突变，并试图变得更符合对双摆运动的描述
where they mutate and try to get an even closer fit.
最终，就会产生一个公式赢家
Eventually, a winning equation emerges,
阿基米德也会引以为傲
one that Archimedes would be proud of.
“找到了！”
Eureka!
现在我正在进行着我们的算法
And I'm running our algorithm now.
左边窗口中的是公式列表
On the left pane are the lists of the equations
都是“找到了”想出的解决双摆运动的公式
that Eureka has thought up for this double pendulum.
再往前，我们能看到公式变得复杂了
Walking up, we can see we increase the complexity,
同时我们也增加了与收集到的数据的一致性
and we're also increasing the agreement with the data.
随着进化，我们往上看
And eventually, as you go up,
你就能得到与数据极为一致的公式
you start to get an extremely close agreement with the data,
最终你就能得到一个真♥相♥
and eventually you snap on to a truth
它在描述双摆运动的准确性上有了很大的提升
where you get a large improvement in the accuracy.
我们就能看到这里产生的一个公式
And we can actually look in here and see exactly what pops out.
比方说在这里你能看到一个9.8
For example here, you might notice we have a 9.8,
如果你还记得你的物理课的内容
and if you remember from physics courses,
这个数字正是地球的重力系数（重力加速度）
that is the coefficient of gravity on earth.
重要的是这里的
What's very important is the difference
双摆的两个不同的摆动角度
between the two angles of the double pendulum.
这样就产生了
This pops out.
实质上，我们通过这个软件
Essentially, we've used this software
以及为了模拟混沌来收集的数据
and the data we've collected to model chaos,
直接从这些数据中梳理得到了解答
and we've teased out the solution directly from the data.
“找到了”不仅发现了
Eureka has not only discovered
描述双摆运动的公式
a single equation to explain how a double pendulum moves.
它还为看起来像是混沌的事件赋予了含义
It has found meaning in what looks like chaos --
在这之前从未有人类或机器人成功过
something no human or machine has done before.
我们可以收集整个数据集
So, we could collect an entirely new data set,
再次运行这个程序
run this process again,
尽管这些数据完全不同
and even though the data is completely different --
我们看到的运动方式是不同的
we could have different observations --
我们仍然可以辨认出隐藏的真♥相♥
we can still identify the underlying truth,
隐藏的模式，就是这个公式
the underlying pattern, which is this equation.
对于迈克尔来说，未来的科学探索
To Michael, the future of scientific exploration
并不在于我们的头脑中
isn't inside our heads.
而在于机器中
It's inside machines.
它们能否看着一些数据
Whether they're looking at patterns of data
来自于遗传学、粒子物理以及气象学
from genetics, particle physics, or meteorology,
像“找到了”这样的程序就能按需要发展出灵感
programs like Eureka can evolve inspiration on demand,
让它们发现大自然的基本真理
finding basic truths about nature
而这些人类都没能做到
that no human ever could.
我们即将到达这种
We're gonna reach a point
我们可以决定我们想要发现什么的水平
where we decide what we want to discover
我们让机器帮我们搞清事实
and we let the machines figure this out for us.
“找到了”能够取得这些发现
Eureka can find these relationships
而不会带有人类的偏见与限制
without human bias and without human limitations.
我们创造了为我们服务的机器人
We created robots to serve us.
随着它们学着自己的运动、感受和思考方式
As the machines learn their own ways to move, feel, and think,
它们最终会突破这种角色
they will eventually grow out of that role.
那么机器人能否共同工作呢？
What if they start working together?
它们能否建立一个社会
Could they build their own society,
机器人为机器人自己建立起来的社会？
one made by the robots for the robots?
地球上没有比我们更成功的物种
There's no species on earth more successful than us.
这是由于
We owe that success
我们头脑中有台强大的“计算机”
to the powerful computer inside our heads.
但征服地球用到的不仅是一个个体的脑
But it takes more than one brain to conquer a planet.
人类能够繁荣是因为我们学会了
Homo sapiens thrive because we have learned
将这些“计算机”组成“社会”共同工作
to make those computers work together as a society.
如果机器人也共同思考会什么样呢？
What will happen when robots put their heads together?
白天他是机器人学家，晚上则是美食大厨
Roboticist by day and gourmet chef by night,
弗吉尼亚理工学院的丹尼斯·洪教授
Professor Dennis Hong of Virginia tech
是制♥造♥可合作机器人的专家
is a specialist in building cooperative robots.
但他也在实验室外搜寻合作的例子
But he also sees cooperation outside the lab.
我们通常不会专门考虑这些
So, we don't really think about it,
但日常生活中处处都涉及了合作
but everything in our daily lives involves cooperation.
比方说，烹饪，通常认为是个体行为
For example, cooking oftentimes is thought of as a solo act,
但你是否考虑过这涉及了很多人
but if you think about it, a lot of people are involved
而且需要大量的细心协调
and a lot of careful coordination is required
才能做出美食
to make it happen.
哦，查理，谢谢你
Oh, thank you, Charli.
拿个西红柿作为例子
Take this tomato as an example.
这个西红柿的产生开始于一粒种子
This tomato most likely started its life as a seed,
这需要育种人员
where a group of breeders
选择携带优势基因的种子
need to choose the right sequence of genes
这样的西红柿滚圆多汁美味
for a plump, juicy, tasty tomato.
然后这些种子会被种植，自身生长，最后收获
The seeds needed to be planted, grown, harvested,
接着西红柿就要送去市场
then the tomatoes needed to get to the market.
粮食生产是一个复杂的合作网络
Food production is a complex web of coordination.
虽然这样已经很好了
But as good as it is,
但人类的合作是有限度的
human cooperation has its limits.
哎呀
Oops.
每天，就像大多数人一样
Every day, like most of us,
丹尼斯就要面对人类
Dennis has to contend with the prime example
错误合作的第一个例子
of human cooperation gone wrong --
交通
traffic.
问题在于，我们作为人类
The problem is, us being human,
我们都需要、想要尽快到达我们的目的地
we all need to, want to get to our destination
这样就造成了交通堵塞
as quick as possible, thus we have traffic jams.
如果没有交通灯
If it wasn't for traffic lights,
实际生活中它们也是很简单的机器人
which are, in reality, very simple robots,
我们就几乎不可能到达任何地方
it will be almost impossible to get anywhere.
这些交通灯之间互相交谈
These traffic lights, they talk to each other.
它们与其他信♥号♥♥交叉口的
They communicate with other traffic lights
交通灯交流
at other intersections.
另外它们有摄像头
And they have cameras,
这样就能看到交通模式
so they actually see the traffic patterns
并为我们人类做决定
and make decisions for us, for humans.
哦，是这样啊
Oh, there you go.
谢谢你，交通灯
Thank you, traffic light.
交通是一个烦人的事
Traffic is a nuisance.
不过人类其他的失败合作
But other failures of human cooperation