Deborah Gordon: What ants teach us about the brain, cancer and the Internet

翻译人员: Wei Wu 校对人员: Fang Yi

I study ants

我研究各种蚂蚁

in the desert, in the tropical forest

沙漠中的、热带雨林里的、

and in my kitchen,

我厨房里的蚂蚁，

and in the hills around Silicon Valley where I live.

以及我硅谷的家周边山上的蚂蚁。

I've recently realized that ants

最近我注意到

are using interactions differently

蚂蚁在不同的环境下

in different environments,

交互方式也是不同的，

and that got me thinking that we could learn from this

这让我想到或许我们能从中学到些什么

about other systems,

用到其它系统上。

like brains and data networks that we engineer,

例如大脑结构或者我们的数据网络

and even cancer.

甚至是癌症。

So what all these systems have in common

这些系统的共同点在于

is that there's no central control.

没有一个中央控制结构。

An ant colony consists of sterile female workers --

蚁群的工蚁由不育的雌性构成—

those are the ants you see walking around —

工蚁就是你能看到的蚂蚁—

and then one or more reproductive females

而能够生育的雌性蚂蚁（蚁后）

who just lay the eggs.

只负责产卵。

They don't give any instructions.

蚁后不会发号指令。

Even though they're called queens,

虽然它们叫做蚁后，

they don't tell anybody what to do.

但是它们不会指挥其它工蚁。

So in an ant colony, there's no one in charge,

所以任何蚁群都没有一个最高负责人，

and all systems like this without central control

所有这些系统都是没有中央控制的，

are regulated using very simple interactions.

仅仅通过简单的交互方式进行运作.

Ants interact using smell.

蚂蚁的交互是通过嗅觉进行的.

They smell with their antennae,

它们用触角（antennae）去嗅.

and they interact with their antennae,

用触角来交流。

so when one ant touches another with its antennae,

所以当一只蚂蚁的触角碰到另一只蚂蚁的触角

it can tell, for example, if the other ant

它就知道另一个蚂蚁

is a nestmate

是不是同一个巢穴的

and what task that other ant has been doing.

以及这只蚂蚁正要做什么事情.

So here you see a lot of ants moving around

现在你看到的这个蚂蚁的活动场所

and interacting in a lab arena

通过玻璃管子跟另外两个场所连接着

that's connected by tubes to two other arenas.

蚂蚁在这些活动场所里走来走去.

So when one ant meets another,

当一只蚂蚁遇到了另外一支蚂蚁,

it doesn't matter which ant it meets,

遇见的是哪只蚂蚁并不重要,

and they're actually not transmitting

它们也没通过触角传递

any kind of complicated signal or message.

任何复杂的信号或消息.

All that matters to the ant is the rate

唯一传递的是两只蚂蚁

at which it meets other ants.

相互遇见的频率.

And all of these interactions, taken together,

这些交互信息汇总起来后,

produce a network.

我们就得到了一个网络.

So this is the network of the ants

这就是刚才你看到的蚂蚁

that you just saw moving around in the arena,

四处移动之后生成的网络图,

and it's this constantly shifting network

正是这张不断变化中的网络,

that produces the behavior of the colony,

塑造了这个蚁群的行为,

like whether all the ants are hiding inside the nest,

像是有多少蚂蚁躲在巢穴里,

or how many are going out to forage.

多少蚂蚁出去寻食之类的信息.

A brain actually works in the same way,

大脑差不多也是这么工作的,

but what's great about ants is

相比起来观察蚂蚁吸引人的地方之一,

that you can see the whole network as it happens.

是你可以看到整个网路是如何运作的.

There are more than 12,000 species of ants,

蚂蚁的种类超过一万两千种,

in every conceivable environment,

你能想象到的环境里都有蚂蚁存在,

and they're using interactions differently

而且不同环境下的蚁群会使用

to meet different environmental challenges.

不同的交流方式以适应环境特点.

So one important environmental challenge

例如不同环境下不同蚁群

that every system has to deal with

普遍面临的问题之一

is operating costs, just what it takes

是如何控制"运营开支", 即需要花多大成本

to run the system.

才能生存下来.

And another environmental challenge is resources,

另一个环境带来的挑战是,

finding them and collecting them.

如何去搜寻和收集资源.

In the desert, operating costs are high

在沙漠中, 水非常的稀少,

because water is scarce,

所以运营开支很大,

and the seed-eating ants that I study in the desert

我研究的一种生活在沙漠中以植物种子为食的蚂蚁

have to spend water to get water.

寻找水源的同时需要消耗水.

So an ant outside foraging,

所以当一只蚂蚁外出觅食的时候,

searching for seeds in the hot sun,

在火辣辣的太阳底下找种子的时候,

just loses water into the air.

它体内的水分会被蒸发.

But the colony gets its water

而蚁群

by metabolizing the fats out of the seeds

可以通过消化种子富含的脂肪

that they eat.

产生需要的水.

So in this environment, interactions are used

所以在这种环境下, 蚂蚁之间的交互

to activate foraging.

主要用来决定是否外出觅食.

An outgoing forager doesn't go out unless

一个准备外出的觅食者不会轻易外出,

it gets enough interactions with returning foragers,

除非得到了足够的归巢的觅食者的反馈,

and what you see are the returning foragers

你现在看到的是回来的觅食者,

going into the tunnel, into the nest,

在通过蚁巢的管道进入蚁穴时,

and meeting outgoing foragers on their way out.

跟沿路准备外外出的蚂蚁进行交流.

This makes sense for the ant colony,

这对蚁群来说很重要,

because the more food there is out there,

因为外面的食物越多,

the more quickly the foragers find it,

觅食的蚂蚁找到食物的速度越快,

the faster they come back,

它们回来的就更快,

and the more foragers they send out.

那么就会有更多的蚂蚁出去觅食.

The system works to stay stopped,

这个系统默认的行为是按兵不动,

unless something positive happens.

除非看到了足够的好处.

So interactions function to activate foragers.

所以在这里交互是为了决定是否出去觅食.

And we've been studying the evolution of this system.

我们已经研究这种系统演化有一段时间了.

First of all, there's variation.

首先, 这种演化各不相同.

It turns out that colonies are different.

不同的蚁群的行为是不一样的.

On dry days, some colonies forage less,

在旱季, 有些蚁群觅食的少,

so colonies are different in how

不同蚁群之间的差异

they manage this trade-off

就体现在它们如何做权衡

between spending water to search for seeds

如何在消耗更多水分去寻找食物

and getting water back in the form of seeds.

以及获得更多食物和水之间权衡

And we're trying to understand why

我们尝试将蚁群

some colonies forage less than others

类比成神经细胞组织

by thinking about ants as neurons,

基于脑神经科学的相关理论

using models from neuroscience.

来理解蚁群觅食行为的差异。

So just as a neuron adds up its stimulation

所以就像是一个神经元是否触发，

from other neurons to decide whether to fire,

取决于相连的神经元触发强度之和，

an ant adds up its stimulation from other ants

蚂蚁的行为也由其它蚂蚁决定，

to decide whether to forage.

是否要出去觅食。

And what we're looking for is whether there might be

于是我们就希望能够找到

small differences among colonies

觅食行为存在差异的蚁群之间

in how many interactions each ant needs

是否蚂蚁在觅食前交互的其它蚂蚁数量

before it's willing to go out and forage,

也是存在对应差异的。

because a colony like that would forage less.

因为像那样的蚁群会更少外出觅食。

And this raises an analogous question about brains.

这个问题也可以用大脑来进行类比。

We talk about the brain,

我们提到的大脑

but of course every brain is slightly different,

当然也是每个大脑都有些许不同的

and maybe there are some individuals

肯定有一些个体

or some conditions

在某些环境下

in which the electrical properties of neurons are such

他们的神经元的电特性决定了

that they require more stimulus to fire,

需要接受更多的刺激才会激发。

and that would lead to differences in brain function.

而这会导致脑的功能差异。

So in order to ask evolutionary questions,

而为了解答之前系统演化的问题，

we need to know about reproductive success.

我们首先需要研究下后代繁殖率。

This is a map of the study site

这张图显示的是我的研究站附近的蚁群图

where I have been tracking this population

我在这个地方研究收获蚂蚁（一种西方蚁）

of harvester ant colonies for 28 years,

种群演化已经超过28年了。

which is about as long as a colony lives.

这大概也是一个种群能够延续的时间。

Each symbol is a colony,

每一个圆圈都表示一个种群，

and the size of the symbol is how many offspring it had,

圆圈的大小表示后代的规模，

because we were able to use genetic variation

我们可以通过基因变化分析(genetic variation)

to match up parent and offspring colonies,

来确认种群之间的父子关系，

that is, to figure out which colonies

也就是能够确认每个蚁群

were founded by a daughter queen

里面的蚁后来自于

produced by which parent colony.

哪个父代蚁群。

And this was amazing for me, after all these years,

研究这么多年之后我有了一些

to find out, for example, that colony 154,

迷人的发现，例如，154号种群，

whom I've known well for many years,

我研究很多年的这个，

is a great-grandmother.

算是祖母级别的。

Here's her daughter colony,

这是她的女儿种群，

here's her granddaughter colony,

这是她的孙女种群，

and these are her great-granddaughter colonies.

这是重孙女种群。

And by doing this, I was able to learn

分析这些种群使我能够

that offspring colonies resemble parent colonies

发现后代种群（的多少）体现了

in their decisions about which days are so hot

父代种群在炎热天气下

that they don't forage,

觅食的策略差异，

and the offspring of parent colonies

而且考虑到父代种群

live so far from each other that the ants never meet,

与后代种群之间距离很远，不可能遇见，

so the ants of the offspring colony

所以后代种群中的蚂蚁

can't be learning this from the parent colony.

不会从父代种群那里学习到什么。

And so our next step is to look

于是第二步就是看看

for the genetic variation underlying this resemblance.

这种相似性的基因学变异根源。

So then I was able to ask, okay, who's doing better?

然后我就可以提出这个问题：哪群蚂蚁的策略更好？

Over the time of the study,

在研究进行中的那些年里，

and especially in the past 10 years,

尤其是最近的十年，

there's been a very severe and deepening drought

实验所在的美国西南部

in the Southwestern U.S.,

经历了非常严重和持久的干旱，

and it turns out that the colonies that conserve water,

结果是那些更注重保持水分的蚁群，

that stay in when it's really hot outside,

那些大热天不出门的蚁群，

and thus sacrifice getting as much food as possible,

也就是那些失去了更多觅食机会的蚁群，

are the ones more likely to have offspring colonies.

反而是更有可能有后代蚁群的。

So all this time, I thought that colony 154

我曾经一度认为154号种群

was a loser, because on really dry days,

是进化的失败者， 因为在旱季，

there'd be just this trickle of foraging,

它们很少出去觅食，

while the other colonies were out

反之其它的种群

foraging, getting lots of food,

会出去寻找更多的食物，

but in fact, colony 154 is a huge success.

但是结果是，154号种群非常的成功。

She's a matriarch.

她是事实上的统领。

She's one of the rare great-grandmothers on the site.

她是这个研究点非常少见的有重孙后代的蚁群

To my knowledge, this is the first time

就我所知，这还是第一次

that we've been able to track

我们人类能够追踪到

the ongoing evolution of collective behavior

自然界中野生生物群体的

in a natural population of animals

集体行为进化

and find out what's actually working best.

以及找到最适合环境的生存方式.

Now, the Internet uses an algorithm

现在, 互联网使用的算法

to regulate the flow of data

用来分配数据流动的算法

that's very similar to the one

与这些蚂蚁使用的算法

that the harvester ants are using to regulate

即如何安排工蚁外出觅食的算法

the flow of foragers.

非常相似.

And guess what we call this analogy?

你们猜我们如何称呼这种相似性?

The anternet is coming.

蚁群互联网(Anternet)的到来.

(Applause)

(掌声)

So data doesn't leave the source computer

所以发送数据的电脑

unless it gets a signal that there's enough bandwidth

在得到信号确认带宽足够之前

for it to travel on.

不会将数据发送出去.

In the early days of the Internet,

在互联网的早期,

when operating costs were really high

发送和接收数据的成本非常高,

and it was really important not to lose any data,

所以任何形式的数据丢失都是不可以接受的,

then the system was set up for interactions

所以网络系统被设计利用相互之间的交互

to activate the flow of data.

来决定何时发送数据.

It's interesting that the ants are using an algorithm

发现蚂蚁跟我们人类最近才发明的算法

that's so similar to the one that we recently invented,

有这么大的相似性是很叫人惊喜的,

but this is only one of a handful of ant algorithms

而且现在我们只发现了蚂蚁使用的算法中

that we know about,

一小部分的算法,

and ants have had 130 million years

蚂蚁已经有了1.3亿年的历史

to evolve a lot of good ones,

已经演化出很多好的算法,

and I think it's very likely

因此我相信有可能

that some of the other 12,000 species

另外尚未研究的1.2万蚂蚁种类中

are going to have interesting algorithms

也有很多有意思的算法,

for data networks

可以用于数据网络

that we haven't even thought of yet.

这些算法甚至超过了我们的想象.

So what happens when operating costs are low?

例如, 当运营成本很低的时候呢?

Operating costs are low in the tropics,

热带雨林里, 蚁群觅食的成本很低,

because it's very humid, and it's easy for the ants

因为那里非常的湿润, 对于蚁群来说

to be outside walking around.

外出觅食也非常容易.

But the ants are so abundant

但是蚂蚁的种类是如此的繁多

and diverse in the tropics

数量也非常庞大

that there's a lot of competition.

因此蚂蚁之间的竞争非常激烈.

Whatever resource one species is using,

一个蚁群需要用到的任何资源

another species is likely to be using that

基本上都有竞争者

at the same time.

与之争夺.

So in this environment, interactions are used

所以在这样的环境下, 相互接触的用途

in the opposite way.

完全反了过来.

The system keeps going

蚁群的系统不断的扩张,

unless something negative happens,

直到一些不好的事情发生,

and one species that I study makes circuits

我研究的一种蚁群会在丛林里

in the trees of foraging ants

构建自己的觅食网络,

going from the nest to a food source and back,

在蚁穴和食物时间不断的来回,

just round and round,

一圈一圈的觅食,

unless something negative happens,

直到一些不好的事情发生,

like an interaction

例如遇到了

with ants of another species.

别的种类的蚂蚁.

So here's an example of ant security.

这是蚂蚁安防的一个例子.

In the middle, there's an ant

中间的位置, 一只蚂蚁

plugging the nest entrance with its head

在跟另外的种群的蚂蚁触碰了触角之后

in response to interactions with another species.

将蚁穴的入口用自己的头挡住了.

Those are the little ones running around

这些小的、腹部朝上的蚂蚁

with their abdomens up in the air.

正在这周围走动.

But as soon as the threat is passed,

但是一旦危险解除,

the entrance is open again,

入口就会重新开启,

and maybe there are situations

或许我们也可以联想到

in computer security

在计算机安全领域

where operating costs are low enough

这个领域的运营成本也低到

that we could just block access temporarily

我们可以临时的中断网络访问

in response to an immediate threat,

以应对临时的威胁,

and then open it again,

稍后继续开放,

instead of trying to build

而不是现在的做法

a permanent firewall or fortress.

尝试构造一个永久的防火墙.

So another environmental challenge

另一个环境带来的挑战

that all systems have to deal with

所有的蚁群系统都需要面对的

is resources, finding and collecting them.

是如何寻找和搜集资源.

And to do this, ants solve the problem

蚁群为了解决这个问题, 采用了

of collective search,

集体搜索(collective search)的方法,

and this is a problem that's of great interest

而这个问题现在已经引起了

right now in robotics,

机器人研究人员的极大兴趣,

because we've understood that,

因为我们都知道,

rather than sending a single,

与其用一个单一的

sophisticated, expensive robot out

复杂且昂贵的机器人

to explore another planet

去探索另外的星球

or to search a burning building,

或去火场搜救,

that instead, it may be more effective

或许有更好的方式

to get a group of cheaper robots

就是造一堆便宜的机器人

exchanging only minimal information,

相互之间仅仅交换简单的信息,

and that's the way that ants do it.

就像是蚂蚁所做的那样.

So the invasive Argentine ant

这种外来的阿根廷蚂蚁

makes expandable search networks.

很擅长扩大自己的搜索网络.

They're good at dealing with the main problem

它们非常善于解决集体搜索中的

of collective search,

主要问题,

which is the trade-off between

即如何在两个不同的目标之间权衡

searching very thoroughly

既要能够搜索的彻底

and covering a lot of ground.

又要搜索的范围广.

And what they do is,

when there are many ants in a small space,

它们是这么做的,

当搜索空间小而蚂蚁很多时,

then each one can search very thoroughly

它们会搜寻的非常彻底

because there will be another ant nearby

因为它们知道临近的区域

searching over there,

有别的蚂蚁在搜索,

but when there are a few ants

但是当搜索面积很大

in a large space,

且蚂蚁很少时,

then they need to stretch out their paths

它们会扩张自己的搜索路径

to cover more ground.

去覆盖更大的面积.

I think they use interactions to assess density,

我想它们之间的接触主要交换的是蚂蚁的密度信息,

so when they're really crowded,

当它们的密度很大时,

they meet more often,

它们碰见的就越多,

and they search more thoroughly.

搜寻的也就越仔细.

Different ant species must use different algorithms,

不同种类的蚂蚁使用的算法应该是不同的,

because they've evolved to deal with

因为随着一代代的演化

different resources,

它们需要的资源不同.

and it could be really useful to know about this,

知道这些差异真的很有用.

and so we recently asked ants

所以最近我们把蚂蚁

to solve the collective search problem

放在微重力的极端环境中

in the extreme environment

希望能够帮助

of microgravity

国际空间站

in the International Space Station.

解决集体搜索的难题.

When I first saw this picture, I thought,

当我第一次看到这张照片, 我想,

Oh no, they've mounted the habitat vertically,

呀, 他们把蚁穴竖起来放着了,

but then I realized that, of course, it doesn't matter.

但是马上意识到, 其实横竖都一样的.

So the idea here is that the ants

这个实验的想法是

are working so hard to hang on

蚂蚁要花大力气把自己挂在墙上

to the wall or the floor or whatever you call it

或者也可以说是地板上, 你怎么看都行

that they're less likely to interact,

这样它们就没有精力去交互了,

and so the relationship between

所以关于蚂蚁密度的信息

how crowded they are and how often they meet

以及它们相互遇见的频率

would be messed up.

都会乱掉.

We're still analyzing the data.

我们还在分析这些数据.

I don't have the results yet.

我还没有结论.

But it would be interesting to know

但如果我们能够知道地球上的

how other species solve this problem

其它物种如何解决此类问题

in different environments on Earth,

这一定非常的有意思,

and so we're setting up a program

所以我们创建了一个活动

to encourage kids around the world

鼓励全世界的小朋友们

to try this experiment with different species.

用不同的蚂蚁种类重复我们的实验.

It's very simple.

非常简单.

It can be done with cheap materials.

做起来也不需要多少成本.

And that way, we could make a global map

这样, 我们就能够绘制一张

of ant collective search algorithms.

蚂蚁集体搜索算法的"世界地图".

And I think it's pretty likely that the invasive species,

我想那些外来的蚂蚁种类,

the ones that come into our buildings,

那些混进我们大楼的蚂蚁,

are going to be really good at this,

对于集体搜索非常在行,

because they're in your kitchen

因为它们已经跑到你的厨房

because they're really good at finding food and water.

非常地善于找到食物和水.

So the most familiar resource for ants

对于蚂蚁而言最为相似的资源

is a picnic,

是野餐的地方,

and this is a clustered resource.

是一个集中的资源.

When there's one piece of fruit,

当一块水果掉在地上,

there's likely to be another piece of fruit nearby,

周围很可能还有更多的水果渣,

and the ants that specialize on clustered resources

因此生活在集中资源多的地方的蚂蚁

use interactions for recruitment.

通过相互接触来召集伙伴.

So when one ant meets another,

所以当一只蚂蚁遇见另一只蚂蚁,

or when it meets a chemical deposited

或是另一只蚂蚁沿路留下的

on the ground by another,

化学气味,

then it changes direction to follow

然后它就会改变自己的方向

in the direction of the interaction,

冲着接触方提供的方向去搜寻

and that's how you get the trail of ants

这就是为什么能够有一只蚂蚁大军

sharing your picnic.

与你分享野餐的原因.

Now this is a place where I think we might be able

现在, 我觉得我们或许可以

to learn something from ants about cancer.

从蚂蚁身上获得治疗癌症的一些启发.

I mean, first, it's obvious that we could do a lot

我是说, 首先, 我们可以做很多事情

to prevent cancer

来阻止癌症

by not allowing people to spread around

例如禁止有人向其他人销售

or sell the toxins that promote

可能增加我们身体患癌症风险的

the evolution of cancer in our bodies,

有毒有害商品,

but I don't think the ants can help us much with this

但是我不认为在这点上蚂蚁能够帮助我们什么,

because ants never poison their own colonies.

因为它们从来不会毒害同类.

But we might be able to learn something from ants

但是我们或许可以从蚂蚁那里学到一些方法

about treating cancer.

来治疗癌症.

There are many different kinds of cancer.

癌症有很多不同的种类.

Each one originates in a particular part of the body,

每一种癌症一开始都附着在身体的特定部位.

and then some kinds of cancer will spread

然后一些类型的癌症(癌细胞)

or metastasize to particular other tissues

会扩散或传播到其它特定的组织结构中

where they must be getting resources that they need.

它们需要在那里获得自己需要的资源.

So if you think from the perspective

现在如果你从这个角度

of early metastatic cancer cells

去看待早期癌细胞

as they're out searching around

它们也是在体内搜寻

for the resources that they need,

寻找他们需要的资源,

if those resources are clustered,

如果这些资源是集中的,

they're likely to use interactions for recruitment,

那么它们很可能通过相互接触来召唤更多的癌细胞,

and if we can figure out how cancer cells are recruiting,

那么如果我们能够破解癌细胞相互召唤的机制

then maybe we could set traps

我们或许就能够设置陷阱

to catch them before they become established.

在癌细胞聚集之前捕获它们.

So ants are using interactions in different ways

所以蚂蚁在不同的环境下

in a huge variety of environments,

使用了完全不同的交互算法.

and we could learn from this

我们能够从中学习

about other systems that operate

并将结果用于那些没有

without central control.

中央控制的系统.

Using only simple interactions,

仅仅通过简单的接触,

ant colonies have been performing

蚂蚁已经创造了

amazing feats for more than 130 million years.

长达1.3亿年的伟大历史.

We have a lot to learn from them.

我们还有很多需要向它们学习.

Thank you.

感谢大家.

(Applause)

(掌声)