ABOUT THE SPEAKER
Deborah Gordon - Ecologist
By studying how ant colonies work without any one leader, Deborah Gordon has identified striking similarities in how ant colonies, brains, cells and computer networks regulate themselves.

Why you should listen

Ecologist Deborah M. Gordon has learned that ant colonies can work without central control by using simple interactions like how often the insects touch antennae. Contrary to the notion that colonies are organized by efficient ants, she has instead discovered that evolution has produced “noisy” systems that tolerate accident and respond flexibly to the environment. When conditions are tough, natural selection favors colonies that conserve resources.

Her studies of ant colonies have led her and her Stanford colleagues to the discovery of the “Anternet,” which regulates foraging in ants in the same way the internet regulates data traffic. But as she said to Wired in 2013, "Insect behavior mimicking human networks ... is actually not what’s most interesting about ant networks. What’s far more interesting are the parallels in the other direction: What have the ants worked out that we humans haven’t thought of yet?" Her latest exploration: How do ants behave in space?

More profile about the speaker
Deborah Gordon | Speaker | TED.com
TED2003

Deborah Gordon: The emergent genius of ant colonies

黛博拉.高登 研究蚂蚁

Filmed:
1,170,189 views

黛博拉.高登与她的学生一起,拿着一把日产标记笔和一台满是尘土的挖土机,挖开了亚利桑那州荒漠地区的蚁群,试图研究复杂系统(是如何运作的)。
- Ecologist
By studying how ant colonies work without any one leader, Deborah Gordon has identified striking similarities in how ant colonies, brains, cells and computer networks regulate themselves. Full bio

Double-click the English transcript below to play the video.

00:13
I study研究 ants蚂蚁, and that's because I like to think about how organizations组织 work.
0
1000
5000
我研究蚂蚁,因为我喜欢去思考组织是如何运作的。
00:18
And in particular特定, how the simple简单 parts部分 of organizations组织
1
6000
6000
特别是组织中的简单个体
00:24
interact相互作用 to create创建 the behavior行为 of the whole整个 organization组织.
2
12000
4000
如何通过互动来创造整体的行为。
00:29
So, ant蚂蚁 colonies群落 are a good example of an organization组织 like that,
3
17000
6000
所以,蚁群是一个很好的例子。
00:35
and there are many许多 others其他. The web卷筒纸 is one.
4
23000
2000
还有很多其他的(例子),比如网络。
00:37
There are many许多 biological生物 systems系统 like that --
5
25000
2000
比如许多生物系统 ——
00:39
brains大脑, cells细胞, developing发展 embryos胚胎.
6
27000
3000
像大脑,细胞群,发育中的胚胎(都是很好的例子)。
00:42
There are about 10,000 species种类 of ants蚂蚁.
7
30000
3000
世上有大约一万种蚂蚁。
00:45
They all live生活 in colonies群落 consisting of one or a few少数 queens皇后,
8
33000
5000
他们都以蚁群的方式聚居,每个蚁群有一只或者若干只蚁后。
00:50
and then all the ants蚂蚁 you see walking步行 around are sterile无菌 female workers工人.
9
38000
5000
所有那些你看到在忙碌的是没有生育能力的雌性工蚁。
00:55
And all ant蚂蚁 colonies群落 have in common共同 that there's no central中央 control控制.
10
43000
4000
所有蚁群有一个共同的特点,那就是没有中央控制。
00:59
Nobody没有人 tells告诉 anybody任何人 what to do.
11
47000
2000
没有哪只蚂蚁是被命令去做事的
01:01
The queen女王 just lays乐事 the eggs. There's no management管理.
12
49000
4000
蚁后只负责产卵。蚁群中没有所谓的管理机制。
01:05
No ant蚂蚁 directs指导 the behavior行为 of any other ant蚂蚁.
13
53000
4000
没有蚂蚁指挥其他蚂蚁的行为。
01:09
And I try to figure数字 out how that works作品.
14
57000
2000
所以我试图了解蚁群运作的原理。
01:13
And I've been working加工 for the past过去 20 years年份
15
61000
2000
我在过去的20年都在研究
01:15
on a population人口 of seed-eating种子为食 ants蚂蚁 in southeastern东南 Arizona亚利桑那.
16
63000
4000
一种生活在亚利桑那州东南部食种子的蚂蚁。
01:19
Here's这里的 my study研究 site现场. This is really a picture图片 of ants蚂蚁,
17
67000
2000
这是我的研究地点,照片上满是蚂蚁,
01:21
and the rabbit兔子 just happens发生 to be there.
18
69000
3000
这只兔子只是凑巧在那儿罢了。
01:25
And these ants蚂蚁 are called harvester收割机 ants蚂蚁 because they eat seeds种子.
19
73000
4000
这些蚂蚁被称作收获蚁因为它们吃植物的种子为生。
01:29
This is the nest of the mature成熟 colony殖民地, and there's the nest entrance入口.
20
77000
4000
这是成熟蚁群的巢穴,那是蚁穴的入口
01:33
And they forage饲料 maybe for about 20 meters away,
21
81000
3000
他们可能会去20米以外的地方觅食
01:36
gather收集 up the seeds种子 and bring带来 them back to the nest, and store商店 them.
22
84000
4000
收集种子,并把种子带回蚁穴加以储存。
01:40
And every一切 year I go there and make a map地图 of my study研究 site现场.
23
88000
2000
每年我去我的研究地点,都会画张蚁群的分布图。
01:42
This is just a road. And it's not very big:
24
90000
3000
这(副图)是一条路,这地方并不大,
01:45
it's about 250 meters on one side, 400 on the other.
25
93000
3000
一边大约250米,另一边是400米。
01:48
And every一切 colony殖民地 has a name名称, which哪一个 is a number,
26
96000
2000
我给每个蚁群都标了号,
01:50
which哪一个 is painted on a rock. And I go there every一切 year
27
98000
3000
并把号码标记在石头上。我每年都会过去
01:53
and look for all the colonies群落 that were alive the year before,
28
101000
3000
寻找前一年生存的蚁群
01:56
and figure数字 out which哪一个 ones那些 have died死亡, and put all the new ones那些 on the map地图.
29
104000
4000
看看哪些已经死了,并把新的蚁群在分布图上标记出来。
02:00
And by doing this I know how old they all are.
30
108000
2000
这样做我就知道它们各自的年龄。
02:02
And because of that, I've been able能够 to study研究 how their behavior行为 changes变化
31
110000
4000
也就是因为这样,随着当蚁群越来越成熟,越来越壮大,
02:06
as the colony殖民地 gets得到 older旧的 and larger.
32
114000
3000
我可以研究它们行为的变化。
02:09
So I want to tell you about the life cycle周期 of a colony殖民地.
33
117000
2000
(首先)我想告诉你们蚁群的生命周期。
02:11
Ants蚂蚁 never make more ants蚂蚁; colonies群落 make more colonies群落.
34
119000
4000
蚂蚁是不能生更多的蚂蚁的,蚁群却会变多。
02:15
And they do that by each year sending发出 out the reproductives繁殖 --
35
123000
3000
每年每个蚁群都会派出可繁殖的蚂蚁——
02:18
those are the ones那些 with wings翅膀 -- on a mating交配 flight飞行.
36
126000
3000
这些蚂蚁有翅膀——进行婚飞。
02:22
So every一切 year, on the same相同 day -- and it's a mystery神秘 exactly究竟 how that happens发生 --
37
130000
4000
所以每年,并在同一天——很神奇每年都是这同一天——
02:26
each colony殖民地 sends发送 out its virgin处女, unmated未交配 queens皇后 with wings翅膀, and the males男性,
38
134000
6000
每个蚁群都会派出有翅膀的没有交配过的蚁后,还有雄蚁,
02:32
and they all fly to a common共同 place地点. And they mate伴侣.
39
140000
4000
它们飞到一个共同的地方交配。
02:36
And this shows节目 a recently最近 virgin处女 queen女王. Here's这里的 her wings翅膀.
40
144000
6000
这是一只刚开始交配的蚁后。那是它的翅膀。
02:42
And she's in the process处理 of mating交配 with this male,
41
150000
4000
它正在和这只雄蚁交配,
02:47
and there's another另一个 male on top最佳 waiting等候 his turn.
42
155000
3000
还有一只雄蚁在上面排队。
02:50
Often经常 the queens皇后 mate伴侣 more than once一旦.
43
158000
2000
通常蚁后交配不止一次,
02:52
And after that, the males男性 all die. That's it for them.
44
160000
3000
交配之后雄蚁就死了。竟落得如此下场。
02:55
(Laughter笑声)
45
163000
2000
02:57
And then the newly mated配合 queens皇后 fly off somewhere某处, drop下降 their wings翅膀,
46
165000
5000
然后刚交配的蚁后飞到某个地方,脱下它们的翅膀
03:02
dig a hole and go into that hole and start开始 laying铺设 eggs.
47
170000
4000
挖一个洞,进入洞中产卵。
03:06
And they will live生活 for 15 or 20 years年份, continuing继续 to lay铺设 eggs
48
174000
5000
它们会在里面生存15或20年,不停地
03:11
using运用 the sperm精子 from that original原版的 mating交配.
49
179000
2000
利用那次交配得到的精子产卵。
03:14
So the queen女王 goes down in there.
50
182000
2000
蚁后就这样在洞穴深处生存着,
03:16
She lays乐事 eggs, she feeds供稿 the larvae幼虫 -- so an ant蚂蚁 starts启动 as an egg, then it's a larva幼虫.
51
184000
5000
它产卵,给幼蚁喂食——蚂蚁开始是卵,然后成为幼蚁
03:21
She feeds供稿 the larvae幼虫 by regurgitating反刍 from her fat脂肪 reserves储量.
52
189000
4000
它把身体储存的脂肪反刍出来给幼蚁喂食。
03:25
Then, as soon不久 as the ants蚂蚁 -- the first group of ants蚂蚁 -- emerge出现,
53
193000
4000
渐渐地,第一窝蚂蚁就出现了
03:29
they're larvae幼虫. Then they're pupae. Then they come out as adult成人 ants蚂蚁.
54
197000
3000
他们是幼蚁,接着成了蛹,接着长成了成蚁。
03:32
They go out, they get the food餐饮, they dig the nest,
55
200000
3000
然后他们出洞,觅食,再把巢穴挖大。
03:35
and the queen女王 never comes out again.
56
203000
2000
但蚁后永远待在洞中不出来了。
03:37
So this is a one-year-old一岁 colony殖民地 -- this happens发生 to be 536.
57
205000
4000
这是一岁大的蚁穴,536号
03:41
There's the nest entrance入口, there's a pencil铅笔 for scale规模.
58
209000
2000
那是蚁穴的入口,那是笔,作为比例尺。
03:43
So this is the colony殖民地 founded成立 by a queen女王 the previous以前 summer夏季.
59
211000
4000
这是一只蚁后在去年夏天建的蚁群。
03:47
This is a three-year-old三十岁 colony殖民地.
60
215000
2000
这是三岁大的蚁群,
03:49
There's the nest entrance入口, there's a pencil铅笔 for scale规模.
61
217000
4000
那是蚁穴的入口,那是笔,作为比例尺。
03:53
They make a midden粪堆, a pile of refuse垃圾 -- mostly大多 the husks of the seeds种子 that they eat.
62
221000
5000
那些蚂蚁制造了一大堆垃圾---大多是吃种子留下的壳。
03:58
This is a five-year-old五十岁 colony殖民地. This is the nest entrance入口, here's这里的 a pencil铅笔 for scale规模.
63
226000
5000
这是五岁大的蚁群。这是蚁穴的入口,那是笔,作为比例尺。
04:03
This is about as big as they get, about a meter仪表 across横过.
64
231000
3000
这是最大的蚁穴,入口大约一米宽。
04:07
And then this is how colony殖民地 size尺寸 and numbers数字 of worker工人 ants蚂蚁 changes变化 --
65
235000
5000
这(张图)显示了蚁群的规模和工蚁数量的变化规律——
04:12
so this is about 10,000 worker工人 ants蚂蚁 --
66
240000
3000
这是大约一万个工蚁——
04:15
changes变化 as a function功能 of colony殖民地 age年龄, in years年份.
67
243000
3000
在几年内,随着蚁群生存年龄的变化而变化。
04:18
So it starts启动 out with zero ants蚂蚁, just the founding创建 queen女王,
68
246000
5000
一开始没有蚂蚁,只是蚁后
04:23
and it grows成长 to a size尺寸 of about 10 or 12 thousand ants蚂蚁 when the colony殖民地 is five.
69
251000
4000
五年后,这个蚁群就壮大到1万或1万2千的蚂蚁。
04:27
And it stays入住 that size尺寸 until直到 the queen女王 dies
70
255000
3000
并且这个数量能保持到蚁后死亡,
04:30
and there's nobody没有人 to make more ants蚂蚁, when she's about 15 or 20 years年份 old.
71
258000
5000
也就是当蚁后达到15或20岁的时候,当它没办法繁殖出更多的蚂蚁了。
04:35
And it's when they reach达到 this stable稳定 size尺寸, in numbers数字 of ants蚂蚁,
72
263000
4000
就在蚂蚁的数量稳定下来的时候(大约5岁)。
04:39
that they start开始 to reproduce复制.
73
267000
2000
蚂蚁就会再开始繁殖,
04:41
That is, to send发送 more winged queens皇后 and males男性 to that year's年份 mating交配 flight飞行.
74
269000
6000
也就是派更多的有翅膀的蚁后和雄蚁去交配。
04:47
And I know how colony殖民地 size尺寸 changes变化 as a function功能 of colony殖民地 age年龄
75
275000
5000
我之所以知道蚁群规模随着生存年龄增长的变化规律,
04:52
because I've dug up colonies群落 of known已知 age年龄 and counted all the ants蚂蚁. (Laughter笑声)
76
280000
5000
是因为我挖开了我知道生存时间的蚁群,并一只只数了蚂蚁。
04:59
So that's not the most fun开玩笑 part部分 of this research研究, although虽然 it's interesting有趣.
77
287000
5000
虽然这很有趣,但并不是研究过程中最好玩的
05:04
(Laughter笑声)
78
292000
1000
05:05
Really the question that I think about with these ants蚂蚁 is what I call task任务 allocation分配.
79
293000
5000
在研究蚂蚁过程中我所考虑的真正问题叫做“任务分配”。
05:10
That's not just how is the colony殖民地 organized有组织的,
80
298000
3000
这不单单是蚁群组织的原理,
05:13
but how does it change更改 what it's doing?
81
301000
2000
而是蚁群怎么改变工作内容。
05:15
How is it that the colony殖民地 manages管理 to adjust调整
82
303000
4000
蚁群到底是如何根据环境的变化设法调整
05:19
the numbers数字 of workers工人 performing执行 each task任务 as conditions条件 change更改?
83
307000
4000
实施每项工作的工蚁的数量的呢?
05:23
So, things happen发生 to an ant蚂蚁 colony殖民地.
84
311000
2000
比如,蚁群总是会遇上变故的,
05:25
When it rains降雨 in the summer夏季, it floods洪水 in the desert沙漠.
85
313000
3000
夏天下雨了,雨水淹没了沙漠
05:28
There's a lot of damage损伤 to the nest,
86
316000
2000
对蚁穴造成了极大的破坏,
05:30
and extra额外 ants蚂蚁 are needed需要 to clean清洁 up that mess食堂.
87
318000
3000
就需要更多的蚂蚁收拾残局。
05:33
When extra额外 food餐饮 becomes available可得到 --
88
321000
2000
而当外面有更多食物时——
05:35
and this is what everybody每个人 knows知道 about picnics野餐 --
89
323000
2000
大家出去野餐时都知道(蚂蚁会来)——
05:37
then extra额外 ants蚂蚁 are allocated分配 to collect搜集 the food餐饮.
90
325000
3000
也需要派更多的蚂蚁去收集食物。
05:40
So, with nobody没有人 telling告诉 anybody任何人 what to do, how is it that
91
328000
3000
但既然没有哪只蚂蚁被告知去做什么事,那整个蚁群是如何
05:43
the colony殖民地 manages管理 to adjust调整 the numbers数字 of workers工人 performing执行 each task任务?
92
331000
4000
协调完成某项工作的工蚁数量呢?
05:47
And that's the process处理 that I call task任务 allocation分配.
93
335000
3000
我称这个协调的过程为“任务分配”。
05:51
And in harvester收割机 ants蚂蚁, I divide划分 the tasks任务
94
339000
2000
在这些收获蚁中,我把
05:53
of the ants蚂蚁 I see just outside the nest
95
341000
2000
在蚁穴外面看到的蚂蚁的工作
05:55
into these four categories类别: where an ant蚂蚁 is foraging觅食,
96
343000
3000
分为四类: 这蚂蚁在觅食
05:58
when it's out along沿 the foraging觅食 trail落后, searching搜索 for food餐饮 or bringing使 food餐饮 back.
97
346000
5000
它走在觅食的路上,寻找食物,将食物搬回
06:03
The patrollers巡逻 -- that's supposed应该 to be a magnifying放大 glass玻璃 --
98
351000
3000
巡逻员---那应该是个放大镜---
06:06
are an interesting有趣 group that go out early in the morning早上
99
354000
2000
是一群有趣的家伙,它们一早就出去了
06:08
before the foragers征粮 are active活性.
100
356000
2000
在觅食蚂蚁开始活跃起来之前
06:10
They somehow不知何故 choose选择 the direction方向 that the foragers征粮 will go,
101
358000
3000
它们设法为觅食的蚂蚁确定方向,
06:13
and by coming未来 back -- just by making制造 it back --
102
361000
3000
凭着(它们)能回巢——就只凭着(它们)能会得了巢——
06:16
they tell the foragers征粮 that it's safe安全 to go out.
103
364000
3000
它们告诉觅食的蚂蚁可以安全外出。
06:19
Then the nest maintenance保养 workers工人 work inside the nest,
104
367000
3000
负责修理的蚂蚁在蚁穴中工作,
06:22
and I wanted to say that the nests look a lot like Bill法案 Lishman'sLishman的 house.
105
370000
5000
我想说的是蚁穴与比尔里斯曼的房子特别相似
06:27
That is, that there are chambers inside,
106
375000
2000
因为有很多小房间在里面。
06:29
they line线 the walls墙壁 of the chambers with moist湿 soil
107
377000
4000
它们用潮湿的土壤塑起了小房间的墙
06:33
and it dries干了 to a kind of an adobe-like土坯样 surface表面 in it.
108
381000
3000
干了之后就成了类似土坯的东西。
06:36
It also looks容貌 very similar类似 to some of the cave洞穴 dwellings住宅
109
384000
3000
蚁穴也很像在那个地区
06:39
of the Hopi赫必族 people that are in that area.
110
387000
3000
霍皮人居住的洞穴。
06:42
And the nest maintenance保养 workers工人 do that inside the nest,
111
390000
4000
负责维修的蚂蚁在蚁穴内工作
06:46
and then they come out of the nest carrying携带 bits of dry soil in their mandibles下颌骨.
112
394000
3000
然后背着细碎的干燥土壤爬出蚁穴,
06:49
So you see the nest maintenance保养 workers工人 come out with a bit of sand,
113
397000
3000
所以你可以看到这些蚂蚁出来的时候大颚里带着一些土壤,
06:52
put it down, turn around, and go back in.
114
400000
2000
然后把土壤放下,转身回到蚁穴。
06:54
And finally最后, the midden粪堆 workers工人 put some kind of territorial领土的 chemical化学 in the garbage垃圾.
115
402000
8000
最后,负责垃圾的蚂蚁把蚁穴特有的化学物质放入垃圾箱。
07:02
So what you see the midden粪堆 workers工人 doing is making制造 a pile of refuse垃圾.
116
410000
3000
所以你看到的是这些蚂蚁在堆一堆垃圾,
07:05
On one day, it'll它会 all be here, and then the next下一个 day
117
413000
2000
这一天,垃圾还全在,第二天
07:07
they'll他们会 move移动 it over there, and then they'll他们会 move移动 it back.
118
415000
2000
它们就把垃圾搬到别处,然后再搬回来
07:09
So that's what the midden粪堆 workers工人 do.
119
417000
4000
这就是负责垃圾的蚂蚁的工作。
07:13
And these four groups are just the ants蚂蚁 outside the nest.
120
421000
4000
这四类只是在蚁穴外工作的蚂蚁
07:17
So that's only about 25 percent百分 of the colony殖民地, and they're the oldest最老的 ants蚂蚁.
121
425000
4000
这些蚂蚁只占整个蚁群的25%,并且它们是最资深的蚂蚁。
07:21
So, an ant蚂蚁 starts启动 out somewhere某处 near the queen女王.
122
429000
3000
所以蚂蚁是从靠近蚁后的地方开始(生活历程)的,
07:24
And when we dig up nests we find they're about as deep
123
432000
3000
当我们挖开蚁穴,我们发现这些蚁穴
07:27
as the colony殖民地 is wide, so about a meter仪表 deep for the big old nests.
124
435000
4000
和蚁群的宽度一样,时间较长的蚁穴深达一米。
07:31
And then there's another另一个 long tunnel隧道 and a chamber, where we often经常 find the queen女王,
125
439000
5000
然后有一条很长的通道和一个小房间,在那里我们通常能发现蚁后。
07:36
after eight hours小时 of hacking黑客 away at the rock with pickaxes.
126
444000
4000
那时我们用手镐刨了八小时的岩石之后发现的。
07:40
I don't think that chamber has evolved进化 because of me and my backhoe反铲
127
448000
4000
我知道那个小房间这么隐蔽不是为了防我和我的挖沟机,
07:44
and my crew船员 of students学生们 with pickaxes,
128
452000
2000
也不是为了防我那些拿着手镐的学生,
07:46
but instead代替 because when there's flooding洪水,
129
454000
2000
而是因为(夏天)有水会淹进来。
07:48
occasionally偶尔 the colony殖民地 has to go down deep.
130
456000
3000
所以常常蚁群要潜入地下。
07:51
So there's this whole整个 network网络 of chambers.
131
459000
2000
这是所有小房间的网状图。
07:53
The queen's女王 in there somewhere某处; she just lays乐事 eggs.
132
461000
3000
蚁后在差不多这里,它只负责产卵。
07:56
There's the larvae幼虫, and they consume消耗 most of the food餐饮.
133
464000
3000
那是幼蚁,他们吃掉了大多数食物,
07:59
And this is true真正 of most ants蚂蚁 --
134
467000
2000
这反映了大多数蚂蚁的情况——
08:01
that the ants蚂蚁 you see walking步行 around don't do much eating.
135
469000
2000
你看到的那些忙碌的工蚁不怎么吃。
08:03
They bring带来 it back and feed饲料 it to the larvae幼虫.
136
471000
2000
它们把食物带回来给幼蚁吃。
08:05
When the foragers征粮 come in with food餐饮, they just drop下降 it into the upper chamber,
137
473000
5000
觅食的蚂蚁带着食物进入蚁穴也只是把食物放在上面的小房间里
08:10
and other ants蚂蚁 come up from below下面, get the food餐饮,
138
478000
3000
其他的蚂蚁从下面走到上面的小房间拿走食物,
08:13
bring带来 it back, husk the seeds种子, and pile them up.
139
481000
3000
把食物放进巢里,剥去种子的壳,然后把种子堆起来。
08:16
There are nest maintenance保养 workers工人 working加工 throughout始终 the nest.
140
484000
3000
而那些负责维修的蚂蚁在整个蚁穴中忙碌。
08:19
And curiously好奇, and interestingly有趣, it looks容貌 as though虽然 at any time
141
487000
5000
挺奇怪的,也挺有趣的是,在任何时候似乎
08:24
about half the ants蚂蚁 in the colony殖民地 are just doing nothing.
142
492000
3000
大约一半的蚂蚁无事可作。
08:27
So, despite尽管 what it says in the Bible圣经,
143
495000
2000
所以,虽然如圣经中所说的
08:29
about, you know, "Look to the ant蚂蚁, thou sluggard懒汉,"
144
497000
4000
“看这些蚂蚁(多勤劳),懒鬼们”,
08:33
in fact事实, you could think of those ants蚂蚁 as reserves储量.
145
501000
3000
事实上,你可以把那些蚂蚁看成是预备队。
08:36
That is to say, if something happened发生 -- and I've never seen看到 anything like this happen发生,
146
504000
3000
也就是说,如果有事发生了---但我从来没有发现有这样的事发生,
08:39
but I've only been looking for 20 years年份 --
147
507000
2000
但我也只关注了20年---
08:41
if something happened发生, they might威力 all come out if they were needed需要.
148
509000
4000
如果有事发生,它们可以根据所需都会出去工作
08:45
But in fact事实, mostly大多 they're just hanging around in there.
149
513000
2000
但实际上,大多时候他们都在闲逛
08:47
And I think it's a very interesting有趣 question --
150
515000
2000
我认为这是非常有趣的问题——
08:49
what is there about the way the colony殖民地 is organized有组织的
151
517000
3000
蚁群是如何组织工作的这个问题
08:52
that might威力 give some function功能 to a reserve保留 of ants蚂蚁 who are doing nothing?
152
520000
7000
可能会让我们明白那些无所事事的蚂蚁的作用。
08:59
And they sort分类 of stand as a buffer缓冲 in between之间
153
527000
4000
它们像是夹在蚁穴深处工作的蚂蚁
09:03
the ants蚂蚁 working加工 deep inside the nest and the ants蚂蚁 working加工 outside.
154
531000
3000
和在外面工作的蚂蚁的一个缓冲带。
09:06
And if you mark标记 ants蚂蚁 that are working加工 outside, and dig up a colony殖民地,
155
534000
4000
如果你把在蚁穴外面工作的蚂蚁标记出来,然后挖一个蚁穴
09:10
you never see them deep down.
156
538000
2000
你永远也不会在深处发现它们。
09:12
So what's happening事件 is that the ants蚂蚁 work inside the nest when they're younger更年轻.
157
540000
5000
所以其实当蚂蚁幼小的时候它们是在蚁穴里面工作的
09:17
They somehow不知何故 get into this reserve保留.
158
545000
2000
然后进入预备队
09:19
And then eventually终于 they get recruited应征 to join加入 this exterior外观 workforce劳动力.
159
547000
4000
最后通过选拔参与了外部工作的队伍。
09:23
And once一旦 they belong属于 to the ants蚂蚁 that work outside, they never go back down.
160
551000
4000
一旦在外部工作了,它们就永远不会再回巢深处了。
09:28
Now ants蚂蚁 -- most ants蚂蚁, including包含 these, don't see very well.
161
556000
3000
蚂蚁---绝大多数蚂蚁,包括这些收获蚁,视力并不好。
09:31
They have eyes眼睛, they can distinguish区分 between之间 light and dark黑暗,
162
559000
3000
它们有眼睛,可以分辨光和黑暗,
09:34
but they mostly大多 work by smell.
163
562000
2000
但它们通常是用嗅觉来工作的
09:36
So just to reinforce加强 that what you might威力 have thought
164
564000
4000
我还要强调的是大家通常对蚁后的认识
09:40
about ant蚂蚁 queens皇后 isn't true真正 --
165
568000
2000
并非是正确的——
09:42
you know, even if the queen女王 did have the intelligence情报
166
570000
3000
你们知道,即使蚁后有智力
09:45
to send发送 chemical化学 messages消息 through通过 this whole整个 network网络 of chambers
167
573000
4000
通过房间网络去发送化学信息
09:49
to tell the ants蚂蚁 outside what to do,
168
577000
2000
给外面工作的蚂蚁分配工作,
09:51
there is no way that such这样 messages消息 could make it in time to see
169
579000
4000
但这些信息也不可能及时传达,使得
09:55
the shifts转变 in the allocation分配 of workers工人 that we actually其实 see outside the nest.
170
583000
5000
外面蚂蚁的工作分配变化。
10:00
So that's one way that we know the queen女王 isn't directing导演 the behavior行为 of the colony殖民地.
171
588000
4000
所以从这方面来说,我们知道蚁后没有在指挥蚁群的工作。
10:05
So when I first set out to work on task任务 allocation分配,
172
593000
3000
所以当我一开始研究任务分配时
10:08
my first question was, "What's the relationship关系
173
596000
3000
我的第一个问题是:到底
10:11
between之间 the ants蚂蚁 doing different不同 tasks任务?
174
599000
3000
不同工作的蚂蚁之间是什么关系?
10:14
Does it matter to the foragers征粮 what the nest maintenance保养 workers工人 are doing?
175
602000
3000
负责维修的蚂蚁对觅食的蚂蚁有影响吗?
10:17
Does it matter to the midden粪堆 workers工人 what the patrollers巡逻 are doing?"
176
605000
3000
巡逻的蚂蚁对负责垃圾的蚂蚁有影响吗?
10:20
And I was working加工 in the context上下文 of a view视图 of ant蚂蚁 colonies群落 in which哪一个 each ant蚂蚁
177
608000
6000
我一直持这样的观点:就是蚁群中的每只蚂蚁
10:26
was somehow不知何故 dedicated专用 to its task任务 from birth分娩
178
614000
3000
从出生开始就一直致力于其工作,
10:29
and sort分类 of performed执行 independently独立地 of the others其他,
179
617000
3000
而且似乎是独立地工作,
10:32
knowing会心 its place地点 on the assembly部件 line线.
180
620000
2000
也明白自己在生产线上处于什么位置。
10:34
And instead代替 I wanted to ask, "How are the different不同 task任务 groups interdependent相互依存?"
181
622000
4000
相反,我想问的是:不同的工作种类之间是怎么相互联系的呢?
10:38
So I did experiments实验 where I changed one thing.
182
626000
2000
所以我做了个实验,只改变了(关系网中的)一个元素。
10:40
So for example, I created创建 more work for the nest maintenance保养 workers工人
183
628000
4000
比如说,我把一捆牙签放在蚁穴入口处,
10:44
by putting out a pile of toothpicks牙签 near the nest entrance入口,
184
632000
4000
给负责维修的蚂蚁更多工作。
10:48
early in the morning早上 when the nest maintenance保养 workers工人 are first active活性.
185
636000
4000
每天一大早负责维修的蚂蚁是最早开始活动起来的。
10:52
This is what it looks容貌 like about 20 minutes分钟 later后来.
186
640000
2000
这张图展示的是20分钟之后的情况,
10:54
Here it is about 40 minutes分钟 later后来.
187
642000
2000
这是大约40分钟之后。
10:56
And the nest maintenance保养 workers工人 just take all the toothpicks牙签
188
644000
2000
这些蚂蚁把所有牙签都搬到
10:58
to the outer edge边缘 of the nest mound and leave离开 them there.
189
646000
3000
离蚁穴较远的地方。
11:01
And what I wanted to know was, "OK,
190
649000
3000
然后我想要知道的是:好
11:04
here's这里的 a situation情况 where extra额外 nest maintenance保养 workers工人 were recruited应征 --
191
652000
3000
这种情况下蚁群需要用更多负责维修的蚂蚁
11:07
is this going to have any effect影响 on the workers工人 performing执行 other tasks任务?"
192
655000
4000
这是否会对作其他工作的蚂蚁有影响呢?
11:11
Then we repeated重复 all those experiments实验 with the ants蚂蚁 marked.
193
659000
4000
下一步我们拿那些标记的蚂蚁做这个实验,
11:15
So here's这里的 some blue蓝色 nest maintenance保养 workers工人.
194
663000
3000
这是一些蓝色的负责维修的蚂蚁,
11:18
And lately最近 we've我们已经 gotten得到 more sophisticated复杂的
195
666000
2000
接着我们做得更复杂一点,
11:20
and we have this three-color三色 system系统.
196
668000
2000
我们用三种不同颜色来标记蚂蚁,
11:22
And we can mark标记 them individually个别地 so we know which哪一个 ant蚂蚁 is which哪一个.
197
670000
3000
这样我们可以把每一只蚂蚁都单独标记出来,这样我们就知道哪只是哪只了。
11:25
We started开始 out with model模型 airplane飞机 paint涂料
198
673000
2000
我们开始用飞机模型的涂料
11:27
and then we found发现 these wonderful精彩 little Japanese日本 markers标记,
199
675000
2000
然后找到这些不错的小型日产标记笔。
11:29
and they work really well.
200
677000
2000
结果显示这些工具非常好用。
11:31
And so just to summarize总结 the result结果,
201
679000
2000
在此我只总结一下实验的结果,
11:33
well it turns out that yes, the different不同 tasks任务 are interdependent相互依存.
202
681000
3000
那就是的确不同的工作的相互有联系的。
11:36
So, if I change更改 the numbers数字 performing执行 one task任务,
203
684000
3000
所以如果你改变了作一类工作的蚂蚁数量,
11:39
it changes变化 the numbers数字 performing执行 another另一个.
204
687000
2000
就改变了作另一类工作的蚂蚁数量。
11:41
So for example, if I make a mess食堂
205
689000
2000
比如说,如果你(把巢)弄乱,
11:43
that the nest maintenance保养 workers工人 have to clean清洁 up,
206
691000
2000
负责维修的蚂蚁就要清扫干净,
11:45
then I see fewer ants蚂蚁 out foraging觅食.
207
693000
2000
我就发现出去觅食的蚂蚁变少了。
11:47
And this was true真正 for all the pair-wise成对 combinations组合 of tasks任务.
208
695000
3000
这个(此消彼长的)结论适用于所有成对组合的工作。
11:50
And the second第二 result结果, which哪一个 was surprising奇怪 to a lot of people,
209
698000
5000
(我们)得出的第二个结论可能会让很多人吃惊,就是
11:55
was that ants蚂蚁 actually其实 switch开关 tasks任务.
210
703000
2000
蚂蚁其实在交换着工作。
11:57
The same相同 ant蚂蚁 doesn't do the same相同 task任务 over and over its whole整个 life.
211
705000
4000
一只蚂蚁不会一生一世只做同一项工作。
12:01
So for example, if I put out extra额外 food餐饮, everybody每个人 else其他 --
212
709000
3000
比如说,如果我放点食物在那边,(不光是迷失的蚂蚁,)其他蚂蚁——
12:04
the midden粪堆 workers工人 stop doing midden粪堆 work and go get the food餐饮,
213
712000
2000
原来只负责垃圾的蚂蚁停止手头的工作,出来搬走食物
12:06
they become成为 foragers征粮.
214
714000
2000
它们成了觅食的蚂蚁啦。
12:08
The nest maintenance保养 workers工人 become成为 foragers征粮.
215
716000
2000
负责维修的蚂蚁也成了觅食的蚂蚁。
12:10
The patrollers巡逻 become成为 foragers征粮.
216
718000
2000
巡逻的蚂蚁也是。
12:12
But not every一切 transition过渡 is possible可能. And this shows节目 how it works作品.
217
720000
4000
但不是任何(两个工作)都能互变的。这张图显示着(蚁群)如何转变(他们的工作)。
12:16
Like I just said, if there is more food餐饮 to collect搜集, the patrollers巡逻, the midden粪堆 workers工人,
218
724000
4000
正如我说的,如果有更多的食物要搬,那些巡逻的蚂蚁,负责垃圾的蚂蚁
12:20
the nest maintenance保养 workers工人 will all change更改 to forage饲料.
219
728000
2000
负责维修的蚂蚁都可以成为觅食蚂蚁。
12:22
If there's more patrolling巡逻 to do --
220
730000
2000
如果有更多巡逻的任务要做
12:24
so I created创建 a disturbance骚乱, so extra额外 patrollers巡逻 were needed需要 --
221
732000
4000
比如我给它们制造了点麻烦,所以需要更多的巡逻蚂蚁,
12:28
the nest maintenance保养 workers工人 will switch开关 to patrol巡逻.
222
736000
2000
负责维修的蚂蚁可以转变为巡逻蚂蚁。
12:30
But if more nest maintenance保养 work is needed需要 --
223
738000
3000
但如果需要更多的负责维修的蚂蚁,
12:33
for example, if I put out a bunch of toothpicks牙签 --
224
741000
2000
比如说我放了一捆牙签——
12:35
then nobody没有人 will ever switch开关 back to nest maintenance保养,
225
743000
3000
没有其他有工作的蚂蚁会回来为负责维修的蚂蚁,
12:38
they have to get nest maintenance保养 workers工人 from inside the nest.
226
746000
3000
它们必须从蚁穴里面把负责维修的蚂蚁叫出来。
12:41
So foraging觅食 acts行为 as a sink水槽, and the ants蚂蚁 inside the nest act法案 as a source资源.
227
749000
6000
所以觅食谁都可以做,而在巢里面工作的蚂蚁就是人力储备。
12:48
And finally最后, it looks容貌 like each ant蚂蚁 is deciding决定
228
756000
2000
最后,看起来似乎每只蚂蚁都
12:50
moment时刻 to moment时刻 whether是否 to be active活性 or not.
229
758000
3000
每时每刻都在作决定该去做什么事。
12:53
So, for example, when there's extra额外 nest maintenance保养 work to do,
230
761000
4000
比如说,如果需要做维修工作,
12:57
it's not that the foragers征粮 switch开关 over. I know that they don't do that.
231
765000
4000
觅食的蚂蚁不会转变。我知道它们不会这样做。
13:01
But the foragers征粮 somehow不知何故 decide决定 not to come out.
232
769000
2000
但觅食的蚂蚁是因为某种原因能下决定不来帮忙。
13:03
And here was the most intriguing奇妙 result结果: the task任务 allocation分配.
233
771000
4000
于是就产生了最神奇的结果:任务分配。
13:07
This process处理 changes变化 with colony殖民地 age年龄, and it changes变化 like this.
234
775000
4000
这个过程随着蚁群年龄的增长而像这样变化。
13:11
When I do these experiments实验 with older旧的 colonies群落 --
235
779000
3000
当我对比较年长的蚁群做实验时——
13:14
so ones那些 that are five years年份 or older旧的 --
236
782000
2000
那些生存五年或更长的蚁群——
13:16
they're much more consistent一贯 from one time to another另一个
237
784000
3000
它们从一种情况到另一种情况的转变更稳定,
13:19
and much more homeostatic稳态. The worse更差 things get,
238
787000
2000
更能自我平衡。遇到的事越糟糕,
13:21
the more I hassle争论 them, the more they act法案 like undisturbed不受干扰 colonies群落.
239
789000
3000
我越打扰它们,他们越表现得不被打扰一样。
13:24
Whereas the young年轻, small colonies群落 --
240
792000
2000
然后那些比较年轻的小蚁群——
13:26
the two-year-old二十岁 colonies群落 of just 2,000 ants蚂蚁 -- are much more variable变量.
241
794000
4000
只有两千只蚂蚁的两年蚁群——显得更多变。
13:30
And the amazing惊人 thing about this is that an ant蚂蚁 lives生活 only a year.
242
798000
5000
更令人惊奇的是,蚂蚁只活一年。
13:35
It could be this year, or this year.
243
803000
2000
可能是这一年或是那一年。
13:37
So, the ants蚂蚁 in the older旧的 colony殖民地 that seem似乎 to be more stable稳定
244
805000
4000
所以,在年长的蚁群中那些更稳重的蚂蚁们,
13:41
are not any older旧的 than the ants蚂蚁 in the younger更年轻 colony殖民地.
245
809000
3000
并不比年轻的蚁群中的蚂蚁们老多少。
13:44
It's not due应有 to the experience经验 of older旧的, wiser聪明 ants蚂蚁.
246
812000
3000
(所以)这不是因为较老的蚂蚁有经验,更聪明,
13:47
Instead代替, something about the organization组织
247
815000
2000
而是因为当蚁群变老时,
13:49
must必须 be changing改变 as the colony殖民地 gets得到 older旧的.
248
817000
3000
整个组织里有什么别的东西变化了。
13:52
And the obvious明显 thing that's changing改变 is its size尺寸.
249
820000
3000
而变化最大的就是蚁群的规模。
13:55
So since以来 I've had this result结果, I've spent花费 a lot of time trying to figure数字 out
250
823000
6000
自从我得出了这个结论,我用了很长时间尝试着去弄明白
14:01
what kinds of decision决定 rules规则 -- very simple简单, local本地, probably大概 olfactory嗅觉, chemical化学
251
829000
6000
蚂蚁们用的是一种什么规律来作决定的——(这种规律必须是)简单的,小范围内的,极有可能是用嗅觉或化学信息传导的。
14:07
rules规则 could an ant蚂蚁 could be using运用, since以来 no ant蚂蚁 can assess评估 the global全球 situation情况 --
252
835000
5000
既然没有蚂蚁可以纵观全局,每一只蚂蚁都在使用这个规律——
14:12
that would have the outcome结果 that I see,
253
840000
3000
这样才会导致我所看到的现象,
14:15
these predictable可预测 dynamics动力学, in who does what task任务.
254
843000
3000
这些可预测的变化,关于哪只蚂蚁做什么工作,
14:18
And it would change更改 as the colony殖民地 gets得到 larger.
255
846000
3000
而且这个规律还会随蚁群的壮大而发生变化。
14:21
And what I've found发现 out is that ants蚂蚁 are using运用 a network网络 of antennal触角 contact联系.
256
849000
7000
更进一步,我发现蚂蚁正在用触觉网络。
14:28
So anybody任何人 who's谁是 ever looked看着 at ants蚂蚁 has seen看到 them touch触摸 antennae天线.
257
856000
3000
看到过蚂蚁的人可以会看到它们触碰着触角,
14:31
They smell with their antennae天线.
258
859000
2000
它们靠触角来嗅。
14:33
When one ant蚂蚁 touches触摸 another另一个, it's smelling it,
259
861000
2000
当一只蚂蚁用触角在接触另一只蚂蚁,就说明它在嗅它。
14:35
and it can tell, for example, whether是否 the other ant蚂蚁 is a nest mate伴侣
260
863000
4000
并且蚂蚁可以分辨气味。比如说,对方是不是同一个蚁穴的。
14:39
because ants蚂蚁 cover themselves他们自己 and each other, through通过 grooming梳洗,
261
867000
4000
因为蚂蚁区分其他蚂蚁是通过
14:43
with a layer of grease润滑脂, which哪一个 carries携带 a colony-specific集落特定 odor气味.
262
871000
4000
涂抹一层带有蚁穴特殊气味的油脂。
14:47
And what we're learning学习 is that an ant蚂蚁 uses使用 the pattern模式 of its antennal触角 contacts往来,
263
875000
5000
于是我们就知道蚂蚁是通过分析它用触角接触过的
14:52
the rate at which哪一个 it meets符合 ants蚂蚁 of other tasks任务, in deciding决定 what to do.
264
880000
5000
其它工种的蚂蚁的几率和频率,来决定做它自己该作什么。
14:57
And so what the message信息 is, is not any message信息
265
885000
3000
所以蚂蚁间传递的信息,并不是一般的(无关紧要的)信息,
15:00
that they transmit发送 from one ant蚂蚁 to another另一个, but the pattern模式.
266
888000
4000
它们相互传达的,其实是概率。
15:04
The pattern模式 itself本身 is the message信息.
267
892000
2000
概率本身就是信息。
15:06
And I'll tell you a little bit more about that.
268
894000
2000
我想稍微详细地再讲解一下(这个概率的问题)。
15:08
But first you might威力 be wondering想知道:
269
896000
2000
但首先你们可能想知道,
15:10
how is it that an ant蚂蚁 can tell, for example, I'm a forager觅食.
270
898000
5000
蚂蚁是怎么分辨(其他工种的)——比如说,我是觅食蚂蚁,
15:15
I expect期望 to meet遇到 another另一个 forager觅食 every一切 so often经常.
271
903000
2000
我预料到我会经常碰到其他觅食蚂蚁,
15:17
But if instead代替 I start开始 to meet遇到 a higher更高 number of nest maintenance保养 workers工人,
272
905000
4000
但如果我碰到更多的是负责维修的蚂蚁,
15:21
I'm less likely容易 to forage饲料.
273
909000
2000
我就不太乐意去觅食。
15:23
So it has to know the difference区别 between之间
274
911000
2000
所以它必须知道
15:25
a forager觅食 and a nest maintenance保养 worker工人.
275
913000
2000
觅食蚂蚁与负责维修的蚂蚁的区别。
15:27
And we've我们已经 learned学到了 that, in this species种类 --
276
915000
3000
我们已经知道,在蚂蚁这类生物——
15:30
and I suspect疑似 in others其他 as well --
277
918000
3000
我怀疑对其它生物也是这样-——
15:33
these hydrocarbons碳氢化合物, this layer of grease润滑脂 on the outside of ants蚂蚁,
278
921000
4000
这些碳氢化合物,这层涂在蚂蚁外表的油脂,
15:37
is different不同 as ants蚂蚁 perform演出 different不同 tasks任务.
279
925000
3000
会因不同的工种而不同。
15:40
And we've我们已经 doneDONE experiments实验 that show显示 that
280
928000
2000
我们做的实验表明,
15:42
that's because the longer an ant蚂蚁 stays入住 outside,
281
930000
2000
那是因为在蚁穴外面待的时间越长,
15:44
the more these simple简单 hydrocarbons碳氢化合物 on its surface表面 change更改,
282
932000
4000
这层碳氢化合物的表面变化越大,
15:48
and so they come to smell different不同 by doing different不同 tasks任务.
283
936000
3000
所以它们就因不同的工种而有不同的气味。
15:51
And they can use that task-specific任务的具体 odor气味 in cuticular表皮 hydrocarbons碳氢化合物 --
284
939000
4000
于是它们就可以利用这种在表层的由于工种不同而不同的气味——
15:55
they can use that in their brief简要 antennal触角 contacts往来 to somehow不知何故
285
943000
4000
它们在短暂的触角接触之后可以利用这些气味
15:59
keep track跟踪 of the rate at which哪一个 they're meeting会议 ants蚂蚁 of certain某些 tasks任务.
286
947000
4000
对碰见其他工种的蚂蚁的频率做到心中有数。
16:03
And we've我们已经 just recently最近 demonstrated证明 this
287
951000
3000
我们最近展示了这个原理。
16:06
by putting extract提取 of hydrocarbons碳氢化合物 on little glass玻璃 beads,
288
954000
4000
我们把浓缩的碳氢化合物涂在小型玻璃珠上,
16:10
and dropping落下 the beads gently平缓 down into the nest entrance入口 at the right rate.
289
958000
4000
以正确的频率轻轻地丢在蚁穴入口,
16:14
And it turns out that ants蚂蚁 will respond响应 to the right rate of contact联系
290
962000
4000
结果是蚂蚁会以正确的频率去接触
16:18
with a glass玻璃 bead珠子 with hydrocarbon extract提取 on it,
291
966000
3000
这些涂有浓缩碳氢化合物的小型玻璃珠,
16:21
as they would to contact联系 with real真实 ants蚂蚁.
292
969000
3000
就好像它们与真的蚂蚁接触一样。
16:26
So I want now to show显示 you a bit of film电影 --
293
974000
3000
所以我想给你们看一小段视频——
16:33
and this will start开始 out, first of all, showing展示 you the nest entrance入口.
294
981000
5000
一开始,你们看到的是蚁穴的入口
16:38
So the idea理念 is that ants蚂蚁 are coming未来 in and out of the nest entrance入口.
295
986000
3000
蚂蚁在入口进进出出。
16:41
They've他们已经 gone走了 out to do different不同 tasks任务, and the rate at which哪一个 they meet遇到
296
989000
5000
它们去做不同的工作,当它们进出入口的时候,
16:46
as they come in and out of the nest entrance入口 determines确定, or influences影响,
297
994000
4000
它们接触的频率决定了,或影响了
16:50
each ant's蚂蚁的 decision决定 about whether是否 to go out, and which哪一个 task任务 to perform演出.
298
998000
4000
每只蚂蚁关于是否出去,去做什么工作的决定。
16:54
This is taken采取 through通过 a fiber纤维 optics光学 microscope显微镜. It's down inside the nest.
299
1002000
4000
这是由光纤显微镜拍摄的。这是在蚁穴里面。
16:58
In the beginning开始 you see the ants蚂蚁
300
1006000
2000
一开始你可以看到蚂蚁们
17:00
just kind of engaging with the fiber纤维 optics光学 microscope显微镜.
301
1008000
2000
似乎在玩光纤显微镜,
17:02
But the idea理念 is that the ants蚂蚁 are in there,
302
1010000
4000
但关键是这些蚂蚁呆在那里,
17:06
and each ant蚂蚁 is experiencing经历 a certain某些 flow of ants蚂蚁 past过去 it --
303
1014000
5000
每一只都会“感觉”其他蚂蚁们经过——
17:11
a stream of contacts往来 with other ants蚂蚁.
304
1019000
3000
通过与其他蚂蚁一系列的接触。
17:14
And the pattern模式 of these interactions互动 determines确定
305
1022000
3000
而这些交流的频率就决定了
17:17
whether是否 the ant蚂蚁 comes back out, and what it does when it comes back out.
306
1025000
5000
(这些呆在显微镜前的)蚂蚁是否出来,出来后做什么。
17:22
You can also see this in the ants蚂蚁 just outside the nest entrance入口 like these.
307
1030000
7000
你可以看到这些在蚁穴入口的蚂蚁也是这样。
17:29
Each ant蚂蚁, then, as it comes back in, is contacting联系 other ants蚂蚁.
308
1037000
4000
每只蚂蚁,当它回来时,都在接触其它蚂蚁。
17:33
And the ants蚂蚁 that are waiting等候 just inside the nest entrance入口
309
1041000
3000
那些等在蚁穴入口里面的蚂蚁,
17:36
to decide决定 whether是否 to go out on their next下一个 trip,
310
1044000
3000
为了决定是否要再出去,
17:39
are contacting联系 the ants蚂蚁 coming未来 in.
311
1047000
2000
正在和进来的蚂蚁接触。
17:41
So, what's interesting有趣 about this system系统 is that it's messy.
312
1049000
4000
所以,这个系统最有趣的地方是乱中有序。
17:45
It's variable变量. It's noisy嘈杂. And, in particular特定, in two ways方法.
313
1053000
5000
这是多变的,而且有很多干扰。特别在两个方面:
17:50
The first is that the experience经验 of the ant蚂蚁 -- of each ant蚂蚁 -- can't be very predictable可预测.
314
1058000
5000
一是蚂蚁们的体验——每只蚂蚁的体验——是是没法精确预测的。
17:55
Because the rate at which哪一个 ants蚂蚁 come back depends依靠 on
315
1063000
3000
因为蚂蚁回来的频率决定于
17:58
all the little things that happen发生 to an ant蚂蚁 as it goes out and does its task任务 outside.
316
1066000
5000
每只蚂蚁在出去工作时所发生的小事。
18:03
And the second第二 thing is that an ant's蚂蚁的 ability能力 to assess评估 this pattern模式
317
1071000
5000
二是蚂蚁评估这种几率的能力
18:08
must必须 be very crude原油 because no ant蚂蚁 can do any sophisticated复杂的 counting数数.
318
1076000
6000
肯定是很粗糙的,因为没有蚂蚁可以做精确的计算。
18:15
So, we do a lot of simulation模拟 and modeling造型, and also experimental试验 work,
319
1083000
4000
所以我们做了很多模拟试验的工作
18:19
to try to figure数字 out how those two kinds of noise噪声 combine结合 to,
320
1087000
6000
去弄明白当这两个干扰共同作用,来
18:25
in the aggregate骨料, produce生产 the predictable可预测 behavior行为 of ant蚂蚁 colonies群落.
321
1093000
4000
作为一个整体,导致蚁群可预测的行为。
18:31
Again, I don't want to say that this kind of haphazard偶然 pattern模式 of interactions互动
322
1099000
5000
同时,我不认为这类偶然的交流方式,
18:36
produces产生 a factory that works作品 with the precision精确 and efficiency效率 of clockwork发条装置.
323
1104000
7000
可以建起工厂里一样机械化的精确度和效率。
18:43
In fact事实, if you watch ants蚂蚁 at all, you end结束 up trying to help them
324
1111000
3000
事实上,如果你看着这些蚂蚁,你最终会想去帮它们。
18:46
because they never seem似乎 to be doing anything
325
1114000
2000
因为无论它们做什么,总是看起来
18:48
exactly究竟 the way that you think that they ought应该 to be doing it.
326
1116000
3000
和你觉得应该他们应该用的方式有出入。
18:51
So it's not really that out of these haphazard偶然 contacts往来, perfection完美 arises出现.
327
1119000
7000
所以由于这些偶然的接触,这种机制是不完美的。
18:58
But it works作品 pretty漂亮 well.
328
1126000
2000
但它却运行良好。
19:00
Ants蚂蚁 have been around for several一些 hundred million百万 years年份.
329
1128000
3000
蚂蚁已经生存了几亿年了
19:03
They cover the earth地球, except for Antarctica南极洲.
330
1131000
3000
除了南极洲,它们到处都有。
19:06
Something that they're doing is clearly明确地 successful成功 enough足够
331
1134000
4000
它们这种偶然接触的沟通方式
19:10
that this pattern模式 of haphazard偶然 contacts往来, in the aggregate骨料,
332
1138000
3000
总体来说, 必定是非常成功的,才能
19:13
produces产生 something that allows允许 ants蚂蚁 to make a lot more ants蚂蚁.
333
1141000
4000
使蚂蚁的数量变得更多。
19:17
And one of the things that we're studying研究 is how natural自然 selection选择
334
1145000
3000
另一件我们要研究的是自然选择
19:20
might威力 be acting演戏 now to shape形状 this use of interaction相互作用 patterns模式 --
335
1148000
7000
如何改变蚂蚁的这种交流方式——
19:27
this network网络 of interaction相互作用 patterns模式 --
336
1155000
2000
这种交流的网络——
19:29
to perhaps也许 increase增加 the foraging觅食 efficiency效率 of ant蚂蚁 colonies群落.
337
1157000
5000
去提高蚁群觅食的效率。
19:34
So the one thing, though虽然, that I want you to remember记得 about this
338
1162000
3000
但有件事我想大家要记住的是
19:37
is that these patterns模式 of interactions互动
339
1165000
3000
这些交流的方式,
19:40
are something that you'd expect期望 to be closely密切 connected连接的 to colony殖民地 size尺寸.
340
1168000
5000
是与蚁群的规模密切相关的。
19:45
The simplest简单 idea理念 is that when an ant蚂蚁 is in a small colony殖民地 --
341
1173000
5000
最简单的解释是当一只蚂蚁在小的蚁群中——
19:50
and an ant蚂蚁 in a large colony殖民地 can use the same相同 rule规则,
342
1178000
4000
当然在大的蚁群的蚂蚁同样的规则也适用
19:54
like "I expect期望 to meet遇到 another另一个 forager觅食 every一切 three seconds."
343
1182000
4000
会说:“我想我每三秒钟就会碰见另一只觅食蚂蚁。”
19:58
But in a small colony殖民地, it's likely容易 to meet遇到 fewer foragers征粮,
344
1186000
3000
但在小的蚁群中,这只蚂蚁碰到觅食蚂蚁的可能性较小,
20:01
just because there are fewer other foragers征粮 there to meet遇到.
345
1189000
4000
仅仅因为觅食蚂蚁少,所以不常碰到。
20:05
So this is the kind of rule规则 that, as the colony殖民地 develops发展 and gets得到 older旧的 and larger,
346
1193000
6000
所以当蚁群发展地越大,变得越老,
20:11
will produce生产 different不同 behavior行为 in an old colony殖民地 and a small young年轻 one.
347
1199000
5000
这个规则会在大小蚁群身上有不同的结果。
20:16
Thank you.
348
1204000
2000
谢谢各位。
20:18
(Applause掌声)
349
1206000
2000
鼓掌
Translated by Bin Xie
Reviewed by Alison Xiaoqiao Xie

▲Back to top

ABOUT THE SPEAKER
Deborah Gordon - Ecologist
By studying how ant colonies work without any one leader, Deborah Gordon has identified striking similarities in how ant colonies, brains, cells and computer networks regulate themselves.

Why you should listen

Ecologist Deborah M. Gordon has learned that ant colonies can work without central control by using simple interactions like how often the insects touch antennae. Contrary to the notion that colonies are organized by efficient ants, she has instead discovered that evolution has produced “noisy” systems that tolerate accident and respond flexibly to the environment. When conditions are tough, natural selection favors colonies that conserve resources.

Her studies of ant colonies have led her and her Stanford colleagues to the discovery of the “Anternet,” which regulates foraging in ants in the same way the internet regulates data traffic. But as she said to Wired in 2013, "Insect behavior mimicking human networks ... is actually not what’s most interesting about ant networks. What’s far more interesting are the parallels in the other direction: What have the ants worked out that we humans haven’t thought of yet?" Her latest exploration: How do ants behave in space?

More profile about the speaker
Deborah Gordon | Speaker | TED.com