ABOUT THE SPEAKER
Sean Gourley - Physicist and military theorist
Sean Gourley, trained as a physicist, has turned his scientific mind to analyzing data about a messier topic: modern war and conflict. He is a TED Fellow.

Why you should listen

Sean Gourley's twin passions are physics (working on nanoscale blue-light lasers and self-assembled quantum nanowires) and politics (he once ran for a national elected office back home in New Zealand).

A Rhodes scholar, he's spent the past five years working at Oxford on complex adaptive systems and collective intelligent systems -- basically, using data to understand the nature of human conflict. As he puts it, "This research has taken me all over the world from the Pentagon, to the House of Lords, the United Nations and most recently to Iraq". Originally from New Zealand, he now lives in San Francisco, where he is the co-founder and CTO of Quid which is building a global intelligence platform. He's a 2009 TED Fellow.

In December 2009, Gourley and his team's research was published in the scientific journal Nature. He is co-founder and CTO of Quid.

More profile about the speaker
Sean Gourley | Speaker | TED.com
TED2009

Sean Gourley: The mathematics of war

Sean Gourley 戰爭的數學模式

Filmed:
1,030,256 views

藉由分析伊拉克戰爭和其他戰爭中的死傷數據,Sean Gourley和他的團隊宣稱,在襲擊的頻率與死傷數之間,存在有一個驚人的數學模式。
- Physicist and military theorist
Sean Gourley, trained as a physicist, has turned his scientific mind to analyzing data about a messier topic: modern war and conflict. He is a TED Fellow. Full bio

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

00:12
We look around the media媒體, as we see on the news新聞 from Iraq伊拉克,
0
0
3000
媒體上報導有關伊拉克、
00:15
Afghanistan阿富汗, Sierra內華達 Leone塞拉利昂,
1
3000
3000
阿富汗、獅子山共和國的戰爭,
00:18
and the conflict衝突 seems似乎 incomprehensible費解的 to us.
2
6000
4000
對我們來說似乎是個難解的問題,
00:22
And that's certainly當然 how it seemed似乎 to me when I started開始 this project項目.
3
10000
4000
尤其是當我決定要開始進行這個專案的時候。
00:26
But as a physicist物理學家,
4
14000
2000
但是身為一個物理學家,
00:28
I thought, well if you give me some data數據,
5
16000
3000
我想只要你能提供一些資料,
00:31
I could maybe understand理解 this. You know, give us a go.
6
19000
2000
我或許能找到解答,就讓我試一試也無妨。
00:33
So as a naive幼稚 New Zealander新西蘭人
7
21000
2000
我這個天真的紐西蘭人就想,
00:35
I thought, well I'll go to the Pentagon五角大樓.
8
23000
2000
我該去找五角大廈的人談談,
00:37
Can you get me some information信息?
9
25000
2000
看他們能不能給我一些資訊?
00:39
(Laughter笑聲)
10
27000
3000
(笑聲)
00:42
No. So I had to think a little harder更難.
11
30000
4000
當然不行。所以我得再想別的辦法。
00:46
And I was watching觀看 the news新聞 one night in Oxford牛津.
12
34000
3000
有天晚上我在牛津看新聞的時候,
00:49
And I looked看著 down at the chattering抖動 heads on my channel渠道 of choice選擇.
13
37000
3000
我發現在螢幕下方有一行跑馬字,
00:52
And I saw that there was information信息 there.
14
40000
2000
那就是我要的資訊了!
00:54
There was data數據 within the streams of news新聞 that we consume消耗.
15
42000
3000
我們每天在看新聞的時候,就有大量的資訊流過,
00:57
All this noise噪聲 around us actually其實 has information信息.
16
45000
4000
環繞在我們四週的噪音裡,其實隱藏了重要的資訊。
01:01
So what I started開始 thinking思維 was,
17
49000
2000
所以我開始想,
01:03
perhaps也許 there is something like open打開 source資源 intelligence情報 here.
18
51000
3000
或許有一些公開的資訊我可以取得,
01:06
If we can get enough足夠 of these streams of information信息 together一起,
19
54000
3000
只要我取得足夠的資訊,
01:09
we can perhaps也許 start開始 to understand理解 the war戰爭.
20
57000
3000
或許我們就可以開始瞭解戰爭。
01:12
So this is exactly究竟 what I did. We started開始 bringing使 a team球隊 together一起,
21
60000
3000
我所做的,是組成一個團隊,
01:15
an interdisciplinary跨學科 team球隊 of scientists科學家們,
22
63000
2000
由各種學者專家組成,包含科學家、
01:17
of economists經濟學家, mathematicians數學家.
23
65000
3000
經濟學家和數學家。
01:20
We brought these guys together一起 and we started開始 to try and solve解決 this.
24
68000
3000
我們把這些人集合起來,開始嚐試解決這個問題。
01:23
We did it in three steps腳步.
25
71000
2000
我們的工作分成三個步驟,
01:25
The first step we did was to collect蒐集. We did 130 different不同 sources來源 of information信息 --
26
73000
4000
第一步是蒐集資料,我們的資料來源有130個,
01:29
from NGO非政府組織 reports報告 to newspapers報紙 and cable電纜 news新聞.
27
77000
3000
包含非營利組織的報告、報紙和有線電視新聞。
01:32
We brought this raw生的 data數據 in and we filtered過濾 it.
28
80000
3000
我們把資料蒐集起來,然後做初步的篩選,
01:35
We extracted提取 the key bits on information信息 to build建立 the database數據庫.
29
83000
3000
我們把最精華的資訊輸入,以建立資料庫。
01:38
That database數據庫 contained
30
86000
2000
資料庫裡包含有
01:40
the timing定時 of attacks攻擊,
31
88000
2000
襲擊的時間、
01:42
the location位置, the size尺寸 and the weapons武器 used.
32
90000
2000
地點、規模及使用的武器。
01:44
It's all in the streams of information信息 we consume消耗 daily日常,
33
92000
3000
這些資訊都在我們每天所吸收的新聞裡,
01:47
we just have to know how to pull it out.
34
95000
2000
問題只在於如何擷取。
01:49
And once一旦 we had this we could start開始 doing some cool stuff東東.
35
97000
2000
所以當我們建立了資料庫,我們就可以開始做一些很酷的事了。
01:51
What if we were to look at the distribution分配 of the sizes大小 of attacks攻擊?
36
99000
3000
我們可不可以看看襲擊規模的分佈狀況?
01:54
What would that tell us?
37
102000
2000
那代表什麼意義?
01:56
So we started開始 doing this. And you can see here
38
104000
2000
所以我們開始進行分析。
01:58
on the horizontal axis
39
106000
2000
你們可以在水平軸上
02:00
you've got the number of people killed殺害 in an attack攻擊
40
108000
2000
看到每次襲擊的死亡人數,
02:02
or the size尺寸 of the attack攻擊.
41
110000
2000
也可稱之為襲擊的規模,
02:04
And on the vertical垂直 axis you've got the number of attacks攻擊.
42
112000
3000
垂直軸上有襲擊的次數,
02:07
So we plot情節 data數據 for sample樣品 on this.
43
115000
2000
當我們把資料點畫上去後,
02:09
You see some sort分類 of random隨機 distribution分配 --
44
117000
2000
你會看到類似隨機分佈的圖形 --
02:11
perhaps也許 67 attacks攻擊, one person was killed殺害,
45
119000
3000
像有67次襲擊裡都有一個人死亡,
02:14
or 47 attacks攻擊 where seven people were killed殺害.
46
122000
3000
或是有47次襲擊裡有七個人死亡。
02:17
We did this exact精確 same相同 thing for Iraq伊拉克.
47
125000
2000
我們為伊拉克戰爭做了同樣的分析,
02:19
And we didn't know, for Iraq伊拉克 what we were going to find.
48
127000
3000
我們一開始並不知道伊拉克戰爭的分析結果會是什麼,
02:22
It turns out what we found發現 was pretty漂亮 surprising奇怪.
49
130000
4000
但結果非常令人驚訝。
02:26
You take all of the conflict衝突,
50
134000
2000
當我們把所有衝突、
02:28
all of the chaos混沌, all of the noise噪聲,
51
136000
2000
混亂及噪音
02:30
and out of that
52
138000
2000
都排除在外,
02:32
comes this precise精確 mathematical數學的 distribution分配
53
140000
2000
我們畫出了這個精確的數學分佈圖,
02:34
of the way attacks攻擊 are ordered有序 in this conflict衝突.
54
142000
3000
告訴我們其實戰爭的襲擊是有規則的。
02:37
This blew自爆 our mind心神.
55
145000
2000
這讓我們大吃一驚。
02:39
Why should a conflict衝突 like Iraq伊拉克 have this
56
147000
4000
為什麼像伊拉克這種戰爭
02:43
as its fundamental基本的 signature簽名?
57
151000
2000
會產生這種圖形?
02:45
Why should there be order訂購 in war戰爭?
58
153000
2000
戰爭怎麼會有規則可循?
02:47
We didn't really understand理解 that.
59
155000
2000
我們真的不瞭解,
02:49
We thought maybe there is something special特別 about Iraq伊拉克.
60
157000
4000
我們認為或許伊拉克戰爭有其特殊之處,
02:53
So we looked看著 at a few少數 more conflicts衝突.
61
161000
2000
所以我們又研究了其他幾個戰爭,
02:55
We looked看著 at Colombia哥倫比亞, we looked看著 at Afghanistan阿富汗,
62
163000
2000
我們研究了哥倫比亞、阿富汗
02:57
and we looked看著 at Senegal塞內加爾.
63
165000
2000
和塞內加爾的戰事。
02:59
And the same相同 pattern模式 emerged出現 in each conflict衝突.
64
167000
2000
每一個戰事都有相同的圖形,
03:01
This wasn't supposed應該 to happen發生.
65
169000
2000
不應該是這樣啊...
03:03
These are different不同 wars戰爭, with different不同 religious宗教 factions派系,
66
171000
3000
不同的戰爭有不同的宗教背景、
03:06
different不同 political政治 factions派系, and different不同 socioeconomic社會經濟 problems問題.
67
174000
3000
不同的政治背景,就連社會經濟問題也不同,
03:09
And yet然而 the fundamental基本的 patterns模式 underlying底層 them
68
177000
4000
但是他們所呈現出來的圖形
03:13
are the same相同.
69
181000
3000
卻是一樣!
03:16
So we went a little wider更寬的.
70
184000
2000
我們將範圍再度擴大,
03:18
We looked看著 around the world世界 at all the data數據 we could get our hands on.
71
186000
3000
我們研究世界上所有的戰事,利用手邊所有可取得的資料,
03:21
From Peru秘魯 to Indonesia印度尼西亞,
72
189000
3000
從祕魯到印尼,
03:24
we studied研究 this same相同 pattern模式 again.
73
192000
2000
我們又看到一樣的圖形,
03:26
And we found發現 that not only
74
194000
3000
而且他們
03:29
were the distributions分佈 these straight直行 lines,
75
197000
2000
不只具有相同的這條分佈直線,
03:31
but the slope of these lines, they clustered集群 around
76
199000
2000
就連這條線的斜率 α ,
03:33
this value of alphaα equals等於 2.5.
77
201000
3000
都集中在2.5這個值上下。
03:36
And we could generate生成 an equation方程
78
204000
2000
因此我們得出一條方程式,
03:38
that could predict預測 the likelihood可能性 of an attack攻擊.
79
206000
3000
讓我們得以預測某次襲擊的可能結果。
03:41
What we're saying here
80
209000
2000
我們所說的結果是指
03:43
is the probability可能性 of an attack攻擊 killing謀殺 X number of people
81
211000
4000
在像伊拉克這種戰爭中,
03:47
in a country國家 like Iraq伊拉克
82
215000
2000
某次襲擊所會造成 X 人死亡的機率,
03:49
is equal等於 to a constant不變, times the size尺寸 of that attack攻擊,
83
217000
3000
這個值相當於某個常數乘上該次襲擊的規模
03:52
raised上調 to the power功率 of negative alphaα.
84
220000
3000
的負 α 次方,
03:55
And negative alphaα is the slope of that line I showed顯示 you before.
85
223000
6000
這個負 α 就是先前提到的斜率。
04:01
So what?
86
229000
2000
這又代表什麼?
04:03
This is data數據, statistics統計. What does it tell us about these conflicts衝突?
87
231000
3000
這些是資料、統計數據,他們能夠為我們解釋戰爭嗎?
04:06
That was a challenge挑戰 we had to face面對 as physicists物理學家.
88
234000
3000
這是身為物理學家所要面對的挑戰,
04:09
How do we explain說明 this?
89
237000
3000
我們該怎麼解讀這些數據?
04:12
And what we really found發現 was that alphaα,
90
240000
2000
我們最後瞭解到,
04:14
if we think about it, is the organizational組織
91
242000
2000
α 就代表了
04:16
structure結構體 of the insurgency暴動.
92
244000
3000
叛軍的組織結構。
04:19
AlphaΑ is the distribution分配 of the sizes大小 of attacks攻擊,
93
247000
3000
雖然 α 所代表的是襲擊規模的分佈狀況,
04:22
which哪一個 is really the distribution分配
94
250000
2000
但實際上卻代表
04:24
of the group strength強度 carrying攜帶 out the attacks攻擊.
95
252000
2000
發動襲擊一方的組織力量。
04:26
So we look at a process處理 of group dynamics動力學:
96
254000
2000
所以我們來看看組織的生態 --
04:28
coalescence合併 and fragmentation碎片,
97
256000
3000
集中或分散。
04:31
groups coming未來 together一起, groups breaking破壞 apart距離.
98
259000
2000
這些組織合久必分,分久必合。
04:33
And we start開始 running賽跑 the numbers數字 on this. Can we simulate模擬 it?
99
261000
3000
所以我們開始分析這些數據。我們是否能夠進行模擬?
04:36
Can we create創建 the kind of patterns模式 that we're seeing眼看
100
264000
3000
我們是否能複製出類似
04:39
in places地方 like Iraq伊拉克?
101
267000
3000
伊拉克戰爭的圖形?
04:42
Turns out we kind of do a reasonable合理 job工作.
102
270000
2000
漸漸地我們看到了一些眉目,
04:44
We can run these simulations模擬.
103
272000
2000
我們可以進行模擬了,
04:46
We can recreate重建 this using運用 a process處理 of group dynamics動力學
104
274000
3000
我們可以依據組織生態來解釋
04:49
to explain說明 the patterns模式 that we see
105
277000
2000
世界各地的戰爭
04:51
all around the conflicts衝突 around the world世界.
106
279000
5000
所呈現出的圖形了。
04:56
So what's going on?
107
284000
2000
接下來呢?
04:58
Why should these different不同 -- seemingly似乎 different不同 conflicts衝突
108
286000
3000
為什麼這些看似不同的戰事,
05:01
have the same相同 patterns模式?
109
289000
2000
卻會呈現相同的圖形?
05:03
Now what I believe is going on is that
110
291000
3000
我相信這是因為
05:06
the insurgent叛亂 forces軍隊, they evolve發展 over time. They adapt適應.
111
294000
4000
叛軍的組織,會隨時間不斷地改變,
05:10
And it turns out there is only one solution
112
298000
2000
這是面對一個強而有力的敵人
05:12
to fight鬥爭 a much stronger enemy敵人.
113
300000
2000
所必須採取的策略。
05:14
And if you don't find that solution as an insurgent叛亂 force,
114
302000
3000
若身為一個叛軍組織而不瞭解這一點,
05:17
you don't exist存在.
115
305000
2000
這個叛軍組織就不會存活。
05:19
So every一切 insurgent叛亂 force that is ongoing不斷的,
116
307000
2000
所以只要叛軍還存在,
05:21
every一切 conflict衝突 that is ongoing不斷的,
117
309000
2000
只要戰爭還在打,
05:23
it's going to look something like this.
118
311000
2000
就會出現這個圖形,
05:25
And that is what we think is happening事件.
119
313000
3000
這是我們的想法。
05:28
Taking it forward前鋒, how do we change更改 it?
120
316000
2000
更進一步來說,我們該怎麼改變這個戰爭?
05:30
How do we end結束 a war戰爭 like Iraq伊拉克?
121
318000
2000
我們該怎麼終止像伊拉克這種戰爭?
05:32
What does it look like?
122
320000
2000
這種戰爭裡的 α 值是多少?
05:34
AlphaΑ is the structure結構體. It's got a stable穩定 state at 2.5.
123
322000
3000
α 就是叛軍的組織結構,當戰事持續的時候,
05:37
This is what wars戰爭 look like when they continue繼續.
124
325000
4000
它趨向一個穩定值2.5。
05:41
We've我們已經 got to change更改 that.
125
329000
2000
我們得改變這個值,
05:43
We can push it up:
126
331000
2000
我們可以把它往上推升,
05:45
the forces軍隊 become成為 more fragmented支離破碎;
127
333000
2000
叛軍的組織就會分裂,
05:47
there is more of them, but they are weaker較弱.
128
335000
4000
叛軍組織變多了,但每一個都很弱;
05:51
Or we push it down:
129
339000
2000
或是我們可以把它往下降,
05:53
they're more robust強大的; there is less groups;
130
341000
2000
叛軍就會變強,但是組織數目會減少,
05:55
but perhaps也許 you can sit and talk to them.
131
343000
4000
你可能可以和他們坐下來進行談判。
05:59
So this graph圖形 here, I'm going to show顯示 you now.
132
347000
2000
所以,我要給你們看一張圖,
06:01
No one has seen看到 this before. This is literally按照字面
133
349000
3000
以前還未曾發表過,
06:04
stuff東東 that we've我們已經 come through通過 last week.
134
352000
2000
這是我們上星期才畫出來的圖。
06:06
And we see the evolution演化 of AlphaΑ through通過 time.
135
354000
4000
我們可以在圖上看到 α 值的演變,
06:10
We see it start開始. And we see it grow增長 up to the stable穩定 state
136
358000
3000
從開始逐漸增長到穩定值,
06:13
the wars戰爭 around the world世界 look like.
137
361000
2000
就像所有的戰爭一樣,
06:15
And it stays入住 there through通過 the invasion侵入 of Fallujah費盧杰
138
363000
3000
就算在Falusia入侵時期也還算穩定,
06:18
until直到 the Samarra薩馬拉 bombings爆炸 in the
139
366000
2000
直到06年伊拉克大選時,
06:20
Iraqi伊拉克人 elections選舉 of '06.
140
368000
3000
Samarra炸彈攻擊事件
06:23
And the system系統 gets得到 perturbed忐忑. It moves移動 upwards向上
141
371000
2000
才開始讓系統變得紊亂,往上發展
06:25
to a fragmented支離破碎 state.
142
373000
2000
成為分裂的叛軍組織,
06:27
This is when the surge浪湧 happens發生.
143
375000
2000
接著美軍增援部隊進入伊拉克。
06:29
And depending根據 on who you ask,
144
377000
2000
不管你問誰,
06:31
the surge浪湧 was supposed應該 to push it up even further進一步.
145
379000
3000
大家都認為美軍增援會讓 α 值更往上推升,
06:34
The opposite對面 happened發生.
146
382000
2000
但是 α 值往下降了,
06:36
The groups became成為 stronger.
147
384000
2000
叛軍變得更強大,
06:38
They became成為 more robust強大的.
148
386000
2000
戰力也提升了。
06:40
And so I'm thinking思維, right, great, it's going to keep going down.
149
388000
3000
這時我在想,很好,時機對了! α 值會持續往下降,
06:43
We can talk to them. We can get a solution. The opposite對面 happened發生.
150
391000
3000
便能進行談判,共同找出解決的方法。但天不從人願,
06:46
It's moved移動 up again. The groups are more fragmented支離破碎.
151
394000
3000
α 值又往上推升了,叛軍組織又分裂了。
06:49
And this tells告訴 me one of two things.
152
397000
2000
這讓我發現,
06:51
Either we're back where we started開始
153
399000
3000
要不是美軍增援毫無幫助,
06:54
and the surge浪湧 has had no effect影響;
154
402000
2000
讓我們退回原點;
06:56
or finally最後 the groups have been fragmented支離破碎 to the extent程度
155
404000
4000
要不然就是叛軍已經分裂到不具破壞力,
07:00
that we can start開始 to think about maybe moving移動 out.
156
408000
4000
美軍可以撤出伊拉克了。
07:04
I don't know what the answer回答 is to that.
157
412000
2000
我不知道答案是哪一個,
07:06
But I know that we should be looking at the structure結構體 of the insurgency暴動
158
414000
3000
但我知道我們該持續觀察叛軍的組織,
07:09
to answer回答 that question.
159
417000
2000
以便找出正確的答案。
07:11
Thank you.
160
419000
2000
謝謝各位。
07:13
(Applause掌聲)
161
421000
5000
(掌聲)
Translated by Marie Wu
Reviewed by Terry Lin

▲Back to top

ABOUT THE SPEAKER
Sean Gourley - Physicist and military theorist
Sean Gourley, trained as a physicist, has turned his scientific mind to analyzing data about a messier topic: modern war and conflict. He is a TED Fellow.

Why you should listen

Sean Gourley's twin passions are physics (working on nanoscale blue-light lasers and self-assembled quantum nanowires) and politics (he once ran for a national elected office back home in New Zealand).

A Rhodes scholar, he's spent the past five years working at Oxford on complex adaptive systems and collective intelligent systems -- basically, using data to understand the nature of human conflict. As he puts it, "This research has taken me all over the world from the Pentagon, to the House of Lords, the United Nations and most recently to Iraq". Originally from New Zealand, he now lives in San Francisco, where he is the co-founder and CTO of Quid which is building a global intelligence platform. He's a 2009 TED Fellow.

In December 2009, Gourley and his team's research was published in the scientific journal Nature. He is co-founder and CTO of Quid.

More profile about the speaker
Sean Gourley | Speaker | TED.com