ABOUT THE SPEAKER
Bruce Bueno de Mesquita - Political scientist
A consultant to the CIA and the Department of Defense, Bruce Bueno de Mesquita has built an intricate computer model that can predict the outcomes of international conflicts with bewildering accuracy.

Why you should listen

Every motive has a number, says Bruce Bueno de Mesquita. A specialist in foreign policy, international relations and state building, he is also a leading -- if controversial -- scholar of rational choice theory, which says math underlies the nation-scale consequences of individuals acting for personal benefit. He created forecasting technology that has, time and again, exceeded the accuracy of old-school analysis, even with thorny quarrels charged by obscure contenders, and often against odds. (One example: He called the second Intifada two years in advance.)

Bueno de Mesquita's company, Mesquita & Roundell, sells his system's predictions and analysis to influential government and private institutions that need heads-ups on policy. He teaches at NYU and is a senior fellow at the Hoover Institution.

More profile about the speaker
Bruce Bueno de Mesquita | Speaker | TED.com
TED2009

Bruce Bueno de Mesquita: A prediction for the future of Iran

布魯斯·布恩諾·德·梅斯奎塔預測伊朗局勢的未來

Filmed:
1,045,301 views

布鲁斯·布恩·德·梅斯奎塔運用數學分析來預測一些人類突發事件,像戰爭,政權轉變,巴勒斯坦起義等等(經常他是對了)。經過一番對他預測方法簡潔的解釋後,梅斯奎塔對伊朗未來局勢提出三個预言。
- Political scientist
A consultant to the CIA and the Department of Defense, Bruce Bueno de Mesquita has built an intricate computer model that can predict the outcomes of international conflicts with bewildering accuracy. Full bio

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

00:18
What I'm going to try to do is explain說明 to you
0
0
3000
我將向各位說明
00:21
quickly很快 how to predict預測,
1
3000
2000
如何預測,
00:23
and illustrate說明 it with some predictions預測
2
5000
2000
並用一些實例說明,
00:25
about what Iran伊朗 is going to do in the next下一個 couple一對 of years年份.
3
7000
5000
關於伊朗在未來一兩年將會怎麼做,
00:30
In order訂購 to predict預測 effectively有效,
4
12000
3000
為了預測正確
00:33
we need to use science科學.
5
15000
3000
我們需要運用科學
00:36
And the reason原因 that we need to use science科學
6
18000
3000
需要運用科學的原因,
00:39
is because then we can reproduce複製 what we're doing;
7
21000
2000
是因為這樣才可以重複驗證我們做的,
00:41
it's not just wisdom智慧 or guesswork猜測.
8
23000
3000
不只是靠智慧或是猜測。
00:44
And if we can predict預測,
9
26000
3000
如果我們可以預測,
00:47
then we can engineer工程師 the future未來.
10
29000
2000
那麼我們就可以掌握未來,
00:49
So if you are concerned關心 to influence影響 energy能源 policy政策,
11
31000
4000
所以如果你想影響能源政策,
00:53
or you are concerned關心 to influence影響 national國民 security安全 policy政策,
12
35000
5000
或你想影響國家安全政策,
00:58
or health健康 policy政策, or education教育,
13
40000
3000
或是健保,或教育,
01:01
science科學 -- and a particular特定 branch of science科學 -- is a way to do it,
14
43000
4000
科學,與科學中特別的一支,是正確的方法,
01:05
not the way we've我們已經 been doing it,
15
47000
2000
與我們從前的方法不同,
01:07
which哪一個 is seat-of-the-pants座椅的最褲子 wisdom智慧.
16
49000
2000
從前用的是憑感覺經驗這種智慧,
01:09
Now before I get into how to do it
17
51000
2000
現在,在我說明如何做它之前,
01:11
let me give you a little truth真相 in advertising廣告,
18
53000
3000
先告訴你們一些事實,
01:14
because I'm not engaged訂婚 in the business商業 of magic魔法.
19
56000
3000
因為我做的不是魔術,
01:17
There are lots of thing that the approach途徑 I take can predict預測,
20
59000
4000
我用的方法可以預測許多事情,
01:21
and there are some that it can't.
21
63000
2000
但有些事沒有辦法,
01:23
It can predict預測 complex複雜 negotiations談判
22
65000
3000
它可以預測複雜的談判,
01:26
or situations情況 involving涉及 coercion強迫 --
23
68000
3000
或有用到威脅手段的情形,
01:29
that is in essence本質 everything that has to do with politics政治,
24
71000
4000
幾乎可以應用於任何與政治有關的事情,
01:33
much of what has to do with business商業,
25
75000
2000
也可以用在多數與商業有關的事上,
01:35
but sorry, if you're looking to speculate推測 in the stock股票 market市場,
26
77000
6000
但很抱歉,如果你想在股票市場投機,
01:41
I don't predict預測 stock股票 markets市場 -- OK,
27
83000
2000
我不預測股票市場--嗯,
01:43
it's not going up any time really soon不久.
28
85000
3000
它不會在短期內上漲,
01:46
But I'm not engaged訂婚 in doing that.
29
88000
3000
我不從事這種預測,
01:49
I'm not engaged訂婚 in predicting預測 random隨機 number generators發電機.
30
91000
3000
我不預測會出現什麼隨機數目,
01:52
I actually其實 get phone電話 calls電話 from people
31
94000
2000
真的有人打電話問我,
01:54
who want to know what lottery抽獎 numbers數字 are going to win贏得.
32
96000
3000
想知道什麼樂透數字會中獎?
01:57
I don't have a clue線索.
33
99000
3000
我毫無線索,
02:00
I engage從事 in the use of game遊戲 theory理論, game遊戲 theory理論 is a branch of mathematics數學
34
102000
4000
我採用的是博弈理論,它是數學的一支,
02:04
and that means手段, sorry, that even in the study研究 of politics政治,
35
106000
4000
就是說,就是在政治研究當中,
02:08
math數學 has come into the picture圖片.
36
110000
3000
數學也來軋一腳,
02:11
We can no longer pretend假裝 that we just speculate推測 about politics政治,
37
113000
4000
我們對政治不能再只靠猜測,
02:15
we need to look at this in a rigorous嚴格 way.
38
117000
3000
我們該用嚴肅的方法看待它,
02:18
Now, what is game遊戲 theory理論 about?
39
120000
3000
那麼什麼是博弈理論?
02:21
It assumes假設 that people are looking out for what's good for them.
40
123000
5000
它假設人是追尋他們的利益,
02:26
That doesn't seem似乎 terribly可怕 shocking觸目驚心 --
41
128000
2000
這看來並不很離譜--
02:28
although雖然 it's controversial爭論的 for a lot of people --
42
130000
2000
雖然對不少人這是有得爭議的--
02:30
that we are self-interested自利.
43
132000
4000
就是我們是追求自利的,
02:34
In order訂購 to look out for what's best最好 for them
44
136000
2000
為了追求他們最大的利益,
02:36
or what they think is best最好 for them,
45
138000
2000
或者他們認為什麼是對他們最好的,
02:38
people have values -- they identify鑑定 what they want, and what they don't want.
46
140000
4000
人們有價值觀--他們認清自己想要的是什麼,不想要的是什麼,
02:42
And they have beliefs信仰 about what other people want,
47
144000
3000
也相信別人想要的是什麼,
02:45
and what other people don't want, how much power功率 other people have,
48
147000
3000
還有別人不想要什麼,他們有多大的權力,
02:48
how much those people could get in the way of whatever隨你 it is that you want.
49
150000
4000
還有你算計的他們最多可以得到多少
02:52
And they face面對 limitations限制, constraints限制,
50
154000
4000
以及他們面對的限制,
02:56
they may可能 be weak, they may可能 be located位於 in the wrong錯誤 part部分 of the world世界,
51
158000
3000
他們可能柔弱,可能處於這世界不對的地方,
02:59
they may可能 be Einstein愛因斯坦, stuck卡住 away farming農業
52
161000
3000
他們也可能像愛因斯坦那樣有本事,但困在農村,
03:02
someplace某個地方 in a rural鄉村 village in India印度 not being存在 noticed注意到,
53
164000
4000
住在印度一個鄉下,沒人注意到,
03:06
as was the case案件 for Ramanujan拉馬努金 for a long time,
54
168000
3000
就像Ramanujan那樣很長一段時間,
03:09
a great mathematician數學家 but nobody沒有人 noticed注意到.
55
171000
3000
一位了不起的數學家,但沒人注意到,
03:12
Now who is rational合理的?
56
174000
2000
那麼誰是理性的呢?
03:14
A lot of people are worried擔心 about what is rationality理性 about?
57
176000
3000
許多人擔心理性是什麼?
03:17
You know, what if people are rational合理的?
58
179000
2000
人們都是理性的,
03:19
Mother母親 Theresa有一個, she was rational合理的.
59
181000
3000
德瑞沙修女,她是理性的,
03:22
Terrorists恐怖分子, they're rational合理的.
60
184000
4000
恐怖分子,他們也是理性的,
03:26
Pretty漂亮 much everybody每個人 is rational合理的.
61
188000
3000
一般來說每個人都是理性的,
03:29
I think there are only two exceptions例外 that I'm aware知道的 of --
62
191000
3000
我所知的例外只有兩種情形--
03:32
two-year-olds兩歲的孩子, they are not rational合理的,
63
194000
2000
兩歲的小孩,他們不是理性的,
03:34
they have very fickle薄情 preferences優先,
64
196000
3000
他們想要的變來變去,
03:37
they switch開關 what they think all the time,
65
199000
2000
他們的想法不時的改變,
03:39
and schizophrenics精神分裂症患者 are probably大概 not rational合理的,
66
201000
3000
還有精神分裂病患者大概也不理性,
03:42
but pretty漂亮 much everybody每個人 else其他 is rational合理的.
67
204000
2000
但其他多數人都是理性的,
03:44
That is, they are just trying to do
68
206000
2000
就是說,他們只想做
03:46
what they think is in their own擁有 best最好 interest利益.
69
208000
5000
他們認為對他們最有利的事。
03:51
Now in order訂購 to work out what people are going to do
70
213000
2000
為了推算出人們下一步會怎麼做
03:53
to pursue追求 their interests利益,
71
215000
2000
以追求他們的利益,
03:55
we have to think about who has influence影響 in the world世界.
72
217000
2000
我們必須考慮這世界上誰有影響力,
03:57
If you're trying to influence影響 corporations公司 to change更改 their behavior行為,
73
219000
5000
如果你想影響公司改變它們的行為,
04:02
with regard看待 to producing生產 pollutants污染物,
74
224000
3000
關於產生污染的問題,
04:05
one approach途徑, the common共同 approach途徑,
75
227000
2000
一個做法,較通常的做法,
04:07
is to exhort勸告 them to be better,
76
229000
2000
是勸告他們改善,
04:09
to explain說明 to them what damage損傷 they're doing to the planet行星.
77
231000
3000
讓他們知道他們 對地球造成怎樣的傷害,,
04:12
And many許多 of you may可能 have noticed注意到 that doesn't have
78
234000
2000
你們可能知道這效果有限,
04:14
as big an effect影響, as perhaps也許 you would like it to have.
79
236000
4000
不如你們期望的那麼有用,
04:18
But if you show顯示 them that it's in their interest利益,
80
240000
3000
但如果你能顯示這樣對他們有利,
04:21
then they're responsive響應.
81
243000
2000
他們就會有反應。
04:23
So, we have to work out who influences影響 problems問題.
82
245000
3000
所以,我們必須找出誰有影響力,
04:26
If we're looking at Iran伊朗, the president主席 of the United聯合的 States狀態
83
248000
2000
例如看伊朗問題,美國總統
04:28
we would like to think, may可能 have some influence影響 --
84
250000
3000
我們認為,可能有些影響力--
04:31
certainly當然 the president主席 in Iran伊朗 has some influence影響 --
85
253000
4000
自然,伊朗總統有些影響力--
04:35
but we make a mistake錯誤 if we just pay工資 attention注意
86
257000
3000
但我們可能犯錯,如果只注意
04:38
to the person at the top最佳 of the power功率 ladder階梯
87
260000
3000
在權力階梯頂端的人,
04:41
because that person doesn't know much about Iran伊朗,
88
263000
3000
因為那個人對伊朗所知不多,
04:44
or about energy能源 policy政策,
89
266000
2000
或關於能源政策,
04:46
or about health健康 care關心,
90
268000
2000
或關於健康保險,
04:48
or about any particular特定 policy政策.
91
270000
2000
或任何特定的政策,
04:50
That person surrounds圍繞著 himself他自己 or herself她自己 with advisers顧問.
92
272000
5000
決策者周圍有許多顧問。
04:55
If we're talking about national國民 security安全 problems問題,
93
277000
2000
如果我們談的是國家安全問題,
04:57
maybe it's the Secretary秘書 of State,
94
279000
2000
那麼可能 是國務卿
04:59
maybe it's the Secretary秘書 of Defense防禦,
95
281000
2000
也許是國防部長,
05:01
the Director導向器 of National國民 Intelligence情報,
96
283000
2000
或情報局長,
05:03
maybe the ambassador大使 to the United聯合的 Nations國家, or somebody else其他
97
285000
2000
或駐聯合國大使,或其他的
05:05
who they think is going to know more about the particular特定 problem問題.
98
287000
4000
他們認為對某個問題知道較多的人,
05:09
But let's face面對 it, the Secretary秘書 of State doesn't know much about Iran伊朗.
99
291000
3000
讓我們想想看,國務卿對伊朗所知不多,
05:12
The secretary秘書 of defense防禦 doesn't know much about Iran伊朗.
100
294000
3000
國防部長也不是伊朗問題專家。
05:15
Each of those people in turn
101
297000
3000
他們每個人
05:18
has advisers顧問 who advise勸告 them,
102
300000
2000
都各有自己的顧問,
05:20
so they can advise勸告 the president主席.
103
302000
3000
因此他們可以對總統提意見。
05:23
There are lots of people shaping成型 decisions決定
104
305000
3000
這是許多人共同影響決策的,
05:26
and so if we want to predict預測 correctly正確地
105
308000
2000
所以如果我們想正確的預測,
05:28
we have to pay工資 attention注意 to everybody每個人
106
310000
3000
就需注意到每一個人,
05:31
who is trying to shape形狀 the outcome結果,
107
313000
2000
每個對結果有影響的人,
05:33
not just the people at the pinnacle巔峰
108
315000
3000
而不是只在尖端的人,
05:36
of the decision-making做決定 pyramid金字塔.
109
318000
4000
決策金字塔的,
05:40
Unfortunately不幸, a lot of times we don't do that.
110
322000
2000
不幸的是,常常我們不是這樣做。
05:42
There's a good reason原因 that we don't do that,
111
324000
2000
不這樣做是有原因的,
05:44
and there's a good reason原因 that using運用 game遊戲 theory理論 and computers電腦,
112
326000
3000
使用博弈理論與電腦是有很好的理由,
05:47
we can overcome克服 the limitation局限性
113
329000
3000
我們可以超越限制,
05:50
of just looking at a few少數 people.
114
332000
2000
不是只看少數幾個人。
05:52
Imagine想像 a problem問題 with just five decision-makers決定者.
115
334000
4000
假設一個問題只涉及五名決策人士,
05:56
Imagine想像 for example
116
338000
2000
想像一下,
05:58
that Sally出擊 over here,
117
340000
2000
這裡是Sally,
06:00
wants to know what Harry掠奪, and Jane,
118
342000
3000
她想知道Harry,Jane
06:03
and George喬治 and Frank坦率 are thinking思維,
119
345000
3000
George,Frank怎麼想,
06:06
and sends發送 messages消息 to those people.
120
348000
2000
並把意見傳達給他們,
06:08
Sally'sSally的 giving her opinion意見 to them,
121
350000
2000
Sally把她的看法傳給他們,
06:10
and they're giving their opinion意見 to Sally出擊.
122
352000
3000
他們也把自己的意見傳給Sally。
06:13
But Sally出擊 also wants to know
123
355000
2000
但Sally也想知道
06:15
what Harry掠奪 is saying to these three,
124
357000
3000
Harry對其他三人說了什麼,
06:18
and what they're saying to Harry掠奪.
125
360000
2000
以及他們對Harry說了什麼,
06:20
And Harry掠奪 wants to know
126
362000
2000
而Harry又想知道
06:22
what each of those people are saying to each other, and so on,
127
364000
3000
其他每個人間互相說了什麼,
06:25
and Sally出擊 would like to know what Harry掠奪 thinks those people are saying.
128
367000
3000
Sally想知道Harry對其他人說法的感覺如何。
06:28
That's a complicated複雜 problem問題; that's a lot to know.
129
370000
3000
這是個複雜的問題,需要知道的很多。
06:31
With five decision-makers決定者
130
373000
3000
當有五名決策人士時
06:34
there are a lot of linkages聯繫 --
131
376000
2000
就有許多的關係聯繫--
06:36
120, as a matter of fact事實,
132
378000
2000
實際上是120條連線,
06:38
if you remember記得 your factorials階乘.
133
380000
2000
如果你記得代數裡的階乘關係,
06:40
Five factorial階乘 is 120.
134
382000
2000
5的階乘是120,
06:42
Now you may可能 be surprised詫異 to know
135
384000
2000
現在你可能驚異的發現
06:44
that smart聰明 people can keep 120 things straight直行
136
386000
3000
聰明人可以清楚記得120件事
06:47
in their head.
137
389000
2000
在他們的腦中,
06:49
Suppose假設 we double the number of influencers影響力
138
391000
2000
假如我們把有影響的人數加倍,
06:51
from five to 10.
139
393000
2000
從5變成10,
06:53
Does that mean we've我們已經 doubled翻倍 the number of pieces of information信息
140
395000
4000
是資訊的數目也加倍嗎?
06:57
we need to know, from 120 to 240?
141
399000
2000
我們需要知道的,從120 變成240嗎 ?
06:59
No. How about 10 times?
142
401000
2000
不是,那是10倍嗎?
07:01
To 1,200? No.
143
403000
3000
成為1200?不是,
07:04
We've我們已經 increased增加 it to 3.6 million百萬.
144
406000
3000
我們需要增加到360萬,
07:07
Nobody沒有人 can keep that straight直行 in their head.
145
409000
2000
沒有人的腦袋可以記得這麼多,
07:09
But computers電腦,
146
411000
3000
但是電腦,
07:12
they can. They don't need coffee咖啡 breaks休息,
147
414000
3000
它們可以,他們不需要休息喝咖啡,
07:15
they don't need vacations休假,
148
417000
3000
它們不需要休假,
07:18
they don't need to go to sleep睡覺 at night,
149
420000
2000
它們不需要在晚上睡覺,
07:20
they don't ask for raises加薪 either.
150
422000
3000
它們也不要求加薪,
07:23
They can keep this information信息 straight直行
151
425000
2000
它們能夠保持資訊正確,
07:25
and that means手段 that we can process處理 the information信息.
152
427000
3000
這表示我們可以處理這些資訊,
07:28
So I'm going to talk to you about how to process處理 it,
153
430000
2000
我將談談如何處理它,
07:30
and I'm going to give you some examples例子 out of Iran伊朗,
154
432000
3000
我將以伊朗為例,
07:33
and you're going to be wondering想知道,
155
435000
2000
你們也許會奇怪,
07:35
"Why should we listen to this guy?
156
437000
2000
“為什麼要聽這個人的?
07:37
Why should we believe what he's saying?"
157
439000
3000
我們為什麼要相信他所說的?”
07:40
So I'm going to show顯示 you a factoid仿真陳述.
158
442000
4000
我給你們看這證詞,
07:44
This is an assessment評定 by the Central中央 Intelligence情報 Agency機構
159
446000
3000
這是一個中央情報局的評估報告,
07:47
of the percentage百分比 of time
160
449000
2000
是正確的百分比
07:49
that the model模型 I'm talking about
161
451000
2000
我將談的這個模型
07:51
is right in predicting預測 things whose誰的 outcome結果 is not yet然而 known已知,
162
453000
3000
在預測還未發生的事情時,
07:54
when the experts專家 who provided提供 the data數據 inputs輸入
163
456000
4000
在提供數據的專家
07:58
got it wrong錯誤.
164
460000
2000
預測錯誤時,
08:00
That's not my claim要求, that's a CIA中央情報局 claim要求 -- you can read it,
165
462000
3000
這不是我說的,是中情局說的--你們可以讀它,
08:03
it was declassified解密 a while ago. You can read it in a volume edited編輯 by
166
465000
3000
不久前它才解密,你們可以在
08:06
H. Bradford布拉德福德 Westerfield韋斯特, Yale耶魯 University大學 Press.
167
468000
3000
威斯特菲編輯,耶魯大學出版這書上看到,
08:09
So, what do we need to know
168
471000
2000
我們需要知道什麼,
08:11
in order訂購 to predict預測?
169
473000
2000
才可以預測?
08:13
You may可能 be surprised詫異 to find out we don't need to know very much.
170
475000
3000
你也許會因發現我們不需要知道很多而嚇一跳,
08:16
We do need to know who has a stake賭注
171
478000
3000
我們需要知道誰有利害關係,
08:19
in trying to shape形狀 the outcome結果 of a decision決定.
172
481000
5000
要試圖影響一個決定的結果。
08:24
We need to know what they say they want,
173
486000
3000
我們需要知道他們說他們想要什麼,
08:27
not what they want in their heart of hearts心中,
174
489000
3000
而不是他們心中想要什麼,
08:30
not what they think they can get,
175
492000
2000
不是他們認為他們可以得到什麼,
08:32
but what they say they want, because that is a strategically戰略性 chosen選擇 position位置,
176
494000
3000
而是他們說他們想要的是什麼,因為這是策略性選擇的位置,
08:35
and we can work backwards向後 from that
177
497000
2000
我們可以從這倒退推算,
08:37
to draw inferences推論 about important重要 features特徵 of their decision-making做決定.
178
499000
4000
由他們決策時的重要特質去推理。
08:41
We need to know how focused重點 they are
179
503000
2000
我們需要知道他們有多關注,
08:43
on the problem問題 at hand.
180
505000
2000
在手上的問題。
08:45
That is, how willing願意 are they to drop下降 what they're doing when the issue問題 comes up,
181
507000
3000
就是在問題發生時,他們有多願意放下手上正在做的事,
08:48
and attend出席 to it instead代替 of something else其他 that's on their plate盤子 --
182
510000
4000
會關注這件事而不是身邊其他的事--
08:52
how big a deal合同 is it to them?
183
514000
2000
這事對他們有多重要?
08:54
And how much clout影響力 could they bring帶來 to bear
184
516000
3000
他們有多大的能耐可以影響決策,
08:57
if they chose選擇 to engage從事 on the issue問題?
185
519000
5000
如果他們選擇關注這問題,
09:02
If we know those things
186
524000
2000
如果我們有這些資訊,
09:04
we can predict預測 their behavior行為 by assuming假設 that everybody每個人
187
526000
3000
我們就可以預測他們的行為,
09:07
cares管它 about two things on any decision決定.
188
529000
5000
假設每個人在任何決定上關心兩件事,
09:12
They care關心 about the outcome結果. They'd他們會 like an outcome結果 as close to
189
534000
2000
他們關心結果,他們希望結果與他們
09:14
what they are interested有興趣 in as possible可能.
190
536000
3000
的利益盡量的接近,
09:17
They're careerists野心家, they also care關心 about getting得到 credit信用 --
191
539000
3000
他們有自己的職志,關心信譽的建立,
09:20
there's ego自我 involvement參與,
192
542000
2000
這裡有個人自我
09:22
they want to be seen看到 as important重要 in shaping成型 the outcome結果,
193
544000
4000
他們希望被認為在形成結果上有重要性,
09:26
or as important重要, if it's their druthers能選擇的話, to block an outcome結果.
194
548000
5000
或者也很重要的,是他們的反對而阻止了某些結果的發生,
09:31
And so we have to figure數字 out how they balance平衡 those two things.
195
553000
3000
所以我們需要辨認清楚他們是如何平衡那兩件事的,
09:34
Different不同 people trade貿易 off
196
556000
2000
這因人而異,
09:36
between之間 standing常設 by their outcome結果,
197
558000
3000
人們有不同的立場不同的追求,
09:39
faithfully忠實 holding保持 to it, going down in a blaze火焰 of glory榮耀,
198
561000
3000
有的人堅持,追求榮耀,
09:42
or giving it up, putting their finger手指 in the wind,
199
564000
3000
或者放棄,見風使舵,
09:45
and doing whatever隨你 they think is going to be a winning勝利 position位置.
200
567000
3000
而做任何他們認為最有利的,
09:48
Most people fall秋季 in between之間, and if we can work out where they fall秋季
201
570000
3000
多數人介於兩者之間,如果我們能夠認清他們處於那裡,
09:51
we can work out how to negotiate談判 with them
202
573000
2000
我們就可以認清如何與他們談判,
09:53
to change更改 their behavior行為.
203
575000
2000
以改變他們的行為。
09:55
So with just that little bit of input輸入
204
577000
3000
所以只要這麼一點輸入的資訊,
09:58
we can work out what the choices選擇 are that people have,
205
580000
3000
我們就能知道別的人有些什麼可能的選擇,
10:01
what the chances機會 are that they're willing願意 to take,
206
583000
3000
他們願意接受的機會有多大,
10:04
what they're after, what they value, what they want,
207
586000
3000
他們追求的是什麼,他們的價值觀如何,他們期望什麼,
10:07
and what they believe about other people.
208
589000
3000
他們眼中的別人是如何,
10:10
You might威力 notice注意 what we don't need to know:
209
592000
4000
你也許也注意到一些我們不需要知道的,
10:14
there's no history歷史 in here.
210
596000
2000
這裡不需要歷史記錄,
10:16
How they got to where they are
211
598000
2000
關於他們是如何成為現在這樣的,
10:18
may可能 be important重要 in shaping成型 the input輸入 information信息,
212
600000
2000
對如何形成所輸入的資訊可能有重要性,
10:20
but once一旦 we know where they are
213
602000
2000
但一旦我們知道他們現狀是如何後,
10:22
we're worried擔心 about where they're going to be headed當家 in the future未來.
214
604000
3000
我們就會想知道他們未來會朝那個方向發展,
10:25
How they got there turns out not to be terribly可怕 critical危急 in predicting預測.
215
607000
4000
他們如何變成現狀的,在預測上就不是這麼重要,
10:29
I remind提醒 you of that 90 percent百分 accuracy準確性 rate.
216
611000
4000
提醒你們這是90%的準確性,
10:33
So where are we going to get this information信息?
217
615000
2000
所以我們要從那裡獲得這些資訊?
10:35
We can get this information信息
218
617000
3000
我們可以獲得這些資訊,
10:38
from the Internet互聯網, from The Economist經濟學家,
219
620000
3000
從互聯網,從《經濟學人》雜誌、
10:41
The Financial金融 Times, The New York紐約 Times,
220
623000
3000
從《金融時報》、《紐約時報》,
10:44
U.S. News新聞 and World世界 Report報告, lots of sources來源 like that,
221
626000
3000
《美國新聞》與《世界報導》,還有很多類似的來源,
10:47
or we can get it from asking experts專家
222
629000
2000
或者可以從請教專家而得到,
10:49
who spend their lives生活 studying研究 places地方 and problems問題,
223
631000
3000
這些專家長期研究各地方的問題,
10:52
because those experts專家 know this information信息.
224
634000
3000
因為這些專家知道這些資訊,
10:55
If they don't know, who are the people trying to influence影響 the decision決定,
225
637000
3000
如果他們不知道,那麼是誰會能影響決策,
10:58
how much clout影響力 do they have,
226
640000
2000
他們有多大的影響力,
11:00
how much they care關心 about this issue問題, and what do they say they want,
227
642000
3000
他們對這問題有多關心,他們對所希望的怎麼說,
11:03
are they experts專家? That's what it means手段 to be an expert專家,
228
645000
3000
他們是專家嗎?專家應該知道這些,
11:06
that's the basic基本 stuff東東 an expert專家 needs需求 to know.
229
648000
4000
那是一個專家需要知道的基本資料,
11:10
Alright好的, lets讓我們 turn to Iran伊朗.
230
652000
2000
好,讓我們看伊朗問題,
11:12
Let me make three important重要 predictions預測 --
231
654000
3000
我在這做三個重要的預測--
11:15
you can check this out, time will tell.
232
657000
3000
你可以檢驗這些預測的正確性,時間會證明,
11:18
What is Iran伊朗 going to do about its nuclear weapons武器 program程序?
233
660000
8000
伊朗的核武計劃將來會如何?
11:26
How secure安全 is the theocratic政教合一 regime政權 in Iran伊朗?
234
668000
3000
伊朗的神權政府是多穩定?
11:29
What's its future未來?
235
671000
2000
未來如何?
11:31
And everybody's每個人的 best最好 friend朋友,
236
673000
3000
還有大家的好朋友,
11:34
Ahmadinejad內賈德. How are things going for him?
237
676000
3000
艾瑪迪-內賈德,他將來會如何?
11:37
How are things going to be working加工 out for him in the next下一個 year or two?
238
679000
6000
在未來一兩年他的地位會怎樣演變?
11:43
You take a look at this, this is not based基於 on statistics統計.
239
685000
3000
你看這裡,這不是依據統計,
11:46
I want to be very clear明確 here. I'm not projecting突出 some past過去 data數據 into the future未來.
240
688000
5000
在這裡我要說清楚,我不是以過去的數據延伸來推測未來,
11:51
I've taken採取 inputs輸入 on positions位置 and so forth向前,
241
693000
3000
我是以各個人的立場為輸入資料,
11:54
run it through通過 a computer電腦 model模型
242
696000
2000
經電腦模型運算,
11:56
that had simulated模擬 the dynamics動力學 of interaction相互作用,
243
698000
3000
模擬動態的相互作用,
11:59
and these are the simulated模擬 dynamics動力學,
244
701000
2000
這些經過模擬的動態,
12:01
the predictions預測 about the path路徑 of policy政策.
245
703000
3000
就是政策變化過程的預測,
12:04
So you can see here on the vertical垂直 axis,
246
706000
3000
所以你可以在縱軸上看到,
12:07
I haven't沒有 shown顯示 it all the way down to zero,
247
709000
2000
我沒有顯示它一直向下走到零
12:09
there are lots of other options選項, but here I'm just showing展示 you the prediction預測,
248
711000
3000
有很多可能的選項,但這裡我只向各位展示預測,
12:12
so I've narrowed收窄 the scale規模.
249
714000
2000
我將比例縮窄,
12:14
Up at the top最佳 of the axis, "Build建立 the Bomb炸彈."
250
716000
3000
在軸線的頂端,是要建造核彈,
12:17
At 130, we start開始 somewhere某處 above以上 130,
251
719000
4000
那是130,我們從高於130開始,
12:21
between之間 building建造 a bomb炸彈, and making製造 enough足夠 weapons-grade武器級 fuel汽油
252
723000
3000
在建造核彈與製造足夠的核武等級燃料之間,
12:24
so that you could build建立 a bomb炸彈.
253
726000
2000
這樣就可以建造核彈,
12:26
That's where, according根據 to my analyses分析,
254
728000
3000
這是依據我的分析,
12:29
the Iranians伊朗人 were at the beginning開始 of this year.
255
731000
3000
伊朗今年初的位置,
12:32
And then the model模型 makes品牌 predictions預測 down the road.
256
734000
3000
然後這模型預測後來的變化,
12:35
At 115 they would only produce生產 enough足夠 weapons武器 grade年級 fuel汽油
257
737000
4000
在115表示伊朗能生產足夠的核彈等級的燃料,
12:39
to show顯示 that they know how, but they wouldn't不會 build建立 a weapon武器:
258
741000
2000
展現他們有此能力,但不去建造武器,
12:41
they would build建立 a research研究 quantity數量.
259
743000
2000
他們只建立了足夠研究的數量,
12:43
It would achieve實現 some national國民 pride自豪,
260
745000
2000
達到國家榮譽的程度,
12:45
but not go ahead and build建立 a weapon武器.
261
747000
3000
但不再往前建造武器,
12:48
And down at 100 they would build建立 civilian平民 nuclear energy能源,
262
750000
2000
到100表示他們建立民生用的核能發電,
12:50
which哪一個 is what they say is their objective目的.
263
752000
4000
這是他們所宣稱的目標,
12:54
The yellow黃色 line shows節目 us the most likely容易 path路徑.
264
756000
3000
這黃線代表最可能的發展路徑,
12:57
The yellow黃色 line includes包括 an analysis分析
265
759000
1000
這黃線包含了分析
12:58
of 87 decision決定 makers製造商 in Iran伊朗,
266
760000
3000
87名伊朗的決策人士,
13:01
and a vast廣大 number of outside influencers影響力
267
763000
3000
以及大量的外部有力人士
13:04
trying to pressure壓力 Iran伊朗 into changing改變 its behavior行為,
268
766000
3000
他們對伊朗施壓 要改變它的行為,
13:07
various各個 players玩家 in the United聯合的 States狀態, and Egypt埃及,
269
769000
3000
許多在美國以及埃及的人士,
13:10
and Saudi沙特 Arabia阿拉伯, and Russia俄國, European歐洲的 Union聯盟,
270
772000
2000
還有沙烏地阿拉伯、蘇俄、歐盟、
13:12
Japan日本, so on and so forth向前.
271
774000
2000
日本,等等
13:14
The white白色 line reproduces再現 the analysis分析
272
776000
4000
這白線是再一次的分析,
13:18
if the international國際 environment環境
273
780000
2000
假如國際環境
13:20
just left Iran伊朗 to make its own擁有 internal內部 decisions決定,
274
782000
3000
讓伊朗自己內部做決定,
13:23
under its own擁有 domestic國內 political政治 pressures壓力.
275
785000
2000
在它自己國內的政治壓力下--
13:25
That's not going to be happening事件,
276
787000
2000
那是不會發生的--
13:27
but you can see that the line comes down faster更快
277
789000
4000
但你能看到這條線下降得快一些,
13:31
if they're not put under international國際 pressure壓力,
278
793000
3000
如果沒有國際壓力,
13:34
if they're allowed允許 to pursue追求 their own擁有 devices設備.
279
796000
2000
如果他們能夠以自己的方式追求,
13:36
But in any event事件, by the end結束 of this year,
280
798000
3000
但任何情況下,到今年底,
13:39
beginning開始 of next下一個 year, we get to a stable穩定 equilibrium平衡 outcome結果.
281
801000
3000
從明年開始,我們會達到平衡穩定的結果,
13:42
And that equilibrium平衡 is not what the United聯合的 States狀態 would like,
282
804000
4000
這平衡點不是美國所盼望的,
13:46
but it's probably大概 an equilibrium平衡 that the United聯合的 States狀態 can live生活 with,
283
808000
3000
但可能是美國能夠接受的平衡狀態,
13:49
and that a lot of others其他 can live生活 with.
284
811000
2000
還有很多其他國家也可以接受,
13:51
And that is that Iran伊朗 will achieve實現 that nationalist民族主義者 pride自豪
285
813000
4000
那是伊朗能夠達成國家榮譽的程度,
13:55
by making製造 enough足夠 weapons-grade武器級 fuel汽油, through通過 research研究,
286
817000
4000
並經由研究而製作出足夠的武器等級燃料,
13:59
so that they could show顯示 that they know how to make weapons-grade武器級 fuel汽油,
287
821000
4000
這樣他們可以炫耀他們知道如何製造武器等級的燃料,
14:03
but not enough足夠 to actually其實 build建立 a bomb炸彈.
288
825000
5000
但不足以實際製造核彈,
14:08
How is this happening事件?
289
830000
2000
這怎麼發生的?
14:10
Over here you can see this is the distribution分配
290
832000
4000
這裡你可以看到這是權利的分配
14:14
of power功率 in favor偏愛 of civilian平民 nuclear energy能源 today今天,
291
836000
5000
它有利於今天民生用途的核能發電,
14:19
this is what that power功率 block is predicted預料到的 to be like
292
841000
3000
這是依據權利分配而預測出來的,
14:22
by the late晚了 parts部分 of 2010, early parts部分 of 2011.
293
844000
6000
在2010年底2011年初,
14:28
Just about nobody沒有人 supports支持 research研究 on weapons-grade武器級 fuel汽油 today今天,
294
850000
4000
還沒有人支持武器等級核燃料的研究,
14:32
but by 2011 that gets得到 to be a big block,
295
854000
3000
但到2011年,那成為很大的力量,
14:35
and you put these two together一起, that's the controlling控制 influence影響 in Iran伊朗.
296
857000
4000
你把這兩個加在一起,對伊朗有控制性的影響力,
14:39
Out here today今天, there are a bunch of people --
297
861000
3000
今天,有許多人--
14:42
Ahmadinejad內賈德 for example --
298
864000
2000
像艾瑪迪內賈德--
14:44
who would like not only to build建立 a bomb炸彈,
299
866000
2000
他不僅希望建造核彈,
14:46
but test測試 a bomb炸彈.
300
868000
2000
還測試這炸彈,
14:48
That power功率 disappears消失 completely全然;
301
870000
2000
這個力量完全消失了,
14:50
nobody沒有人 supports支持 that by 2011.
302
872000
3000
到2011年已經沒人支持這觀點。
14:53
These guys are all shrinking萎縮,
303
875000
2000
這群人數在萎縮,
14:55
the power功率 is all drifting漂流 out here,
304
877000
3000
權力流到這邊,
14:58
so the outcome結果 is going to be the weapons-grade武器級 fuel汽油.
305
880000
3000
所以結果是武器等級的燃料,
15:01
Who are the winners獲獎者 and who are the losers失敗者 in Iran伊朗?
306
883000
3000
誰是伊朗的贏家與輸家呢?
15:04
Take a look at these guys, they're growing生長 in power功率,
307
886000
3000
看這些人,他們的權力在增加,
15:07
and by the way, this was doneDONE a while ago
308
889000
3000
順便說一下,這是不久前做的,
15:10
before the current當前 economic經濟 crisis危機,
309
892000
2000
是在現在這經濟危機之前,
15:12
and that's probably大概 going to get steeper陡峭的.
310
894000
2000
現在可能會更為傾斜,
15:14
These folks鄉親 are the moneyed富有的 interests利益 in Iran伊朗,
311
896000
2000
這些人是伊朗有權勢的人,
15:16
the bankers銀行家, the oil people, the bazaariesbazaaries.
312
898000
4000
銀行家,油人,商家,
15:20
They are growing生長 in political政治 clout影響力,
313
902000
3000
他們的政治影響力都在成長,
15:23
as the mullahs毛拉 are isolating隔離 themselves他們自己 --
314
905000
3000
在穆斯林神職人員自我隔絕時--
15:26
with the exception例外 of one group of mullahs毛拉,
315
908000
2000
除了有一群教士外,
15:28
who are not well known已知 to Americans美國人.
316
910000
2000
美國人一般不知道他們,
15:30
That's this line here, growing生長 in power功率,
317
912000
2000
只有他們的權力在擴張,
15:32
these are what the Iranians伊朗人 call the quietists靜修主義.
318
914000
4000
伊朗人稱他們為Quietists教派,
15:36
These are the Ayatollahs阿亞圖拉, mostly大多 based基於 in Qom庫姆,
319
918000
3000
他們被稱為阿亞圖拉,主要存在於庫姆這地區,
15:39
who have great clout影響力 in the religious宗教 community社區,
320
921000
4000
他們在宗教圈裡很有影響力,
15:43
have been quiet安靜 on politics政治 and are going to be getting得到 louder,
321
925000
3000
以往對政治沉默,現在聲音愈來愈大,
15:46
because they see Iran伊朗 going in an unhealthy不良 direction方向,
322
928000
2000
因為他們覺得伊朗走向不健康的方向,
15:48
a direction方向 contrary相反
323
930000
2000
一個反方向
15:50
to what Khomeini霍梅尼 had in mind心神.
324
932000
4000
與柯梅尼心中希望的。
15:54
Here is Mr先生. Ahmadinejad內賈德.
325
936000
2000
這位是艾馬迪內賈德總統,
15:56
Two things to notice注意: he's getting得到 weaker較弱,
326
938000
3000
兩件事值得注意,他變弱了,
15:59
and while he gets得到 a lot of attention注意 in the United聯合的 States狀態,
327
941000
2000
當他在美國受到許多注意時,
16:01
he is not a major重大的 player播放機 in Iran伊朗.
328
943000
2000
他在伊朗卻不是很重要的人,
16:03
He is on the way down.
329
945000
2000
他在走下坡,
16:05
OK, so I'd like you to take a little away from this.
330
947000
4000
好的,我要你們注意力從這移開一下,
16:09
Everything is not predictable可預測: the stock股票 market市場
331
951000
2000
不是所有事情都能預測,股票市場
16:11
is, at least最小 for me, not predictable可預測,
332
953000
3000
至少對我是不能預測的,
16:14
but most complicated複雜 negotiations談判 are predictable可預測.
333
956000
5000
但多數複雜的談判是可以預測的,
16:19
Again, whether是否 we're talking health健康 policy政策, education教育,
334
961000
4000
不管是談健保政策、教育、
16:23
environment環境, energy能源,
335
965000
3000
環境、能源、
16:26
litigation訴訟, mergers兼併,
336
968000
2000
訴訟、併購,
16:28
all of these are complicated複雜 problems問題
337
970000
2000
這些都是複雜的問題,
16:30
that are predictable可預測,
338
972000
2000
它們是可以預測的,
16:32
that this sort分類 of technology技術 can be applied應用的 to.
339
974000
4000
可以應用這類的技術,
16:36
And the reason原因 that being存在 able能夠 to predict預測 those things is important重要,
340
978000
5000
而能夠預測這些事情的原因是重要的,
16:41
is not just because you might威力 run a hedge樹籬 fund基金 and make money off of it,
341
983000
3000
不是因為你可能經營一個避險基金,要從它賺錢,
16:44
but because if you can predict預測 what people will do,
342
986000
3000
而是因為假如你能夠預測人們將怎麼做,
16:47
you can engineer工程師 what they will do.
343
989000
3000
你就能夠控制他們將做什麼,
16:50
And if you engineer工程師 what they do you can change更改 the world世界,
344
992000
2000
如果你能控制他們要做什麼,你就能改變這世界,
16:52
you can get a better result結果.
345
994000
2000
你可能得到較好的結果,
16:54
I would like to leave離開 you with one thought, which哪一個 is
346
996000
3000
我想留個問題給你們思考,
16:57
for me, the dominant優勢 theme主題 of this gathering蒐集,
347
999000
5000
對我,這聚會的重要議題,
17:02
and is the dominant優勢 theme主題 of this way of thinking思維 about the world世界.
348
1004000
3000
也是這樣思考世界問題的主要論點就是,
17:05
When people say to you,
349
1007000
3000
當人們對你說,
17:08
"That's impossible不可能,"
350
1010000
2000
“那個不可能,”
17:10
you say back to them,
351
1012000
2000
你回應說,
17:12
"When you say 'That's'那是 impossible不可能,'
352
1014000
2000
“當你說‘那個不可能,’
17:14
you're confused困惑 with,
353
1016000
2000
你真的意思是,
17:16
'I don't know how to do it.'"
354
1018000
3000
”我不知道怎麼做“
17:19
Thank you.
355
1021000
2000
謝謝
17:21
(Applause掌聲)
356
1023000
4000
(掌聲)
17:25
Chris克里斯 Anderson安德森: One question for you.
357
1027000
2000
安德森:請教個問題,
17:27
That was fascinating迷人.
358
1029000
3000
這個很棒,
17:30
I love that you put it out there.
359
1032000
3000
很高興你提出來,
17:33
I got very nervous緊張 halfway through通過 the talk though雖然,
360
1035000
2000
但聽講到一半時,我變得緊張,
17:35
just panicking恐慌 whether是否 you'd included包括 in your model模型, the possibility可能性 that
361
1037000
3000
不知在你的模型中,會包含可能性
17:38
putting this prediction預測 out there might威力 change更改 the result結果.
362
1040000
4000
做這些預測可能改變結果,
17:42
We've我們已經 got 800 people in Tehran德黑蘭 who watch TEDTalksTED演講.
363
1044000
3000
我們在德黑蘭有800人觀看這演講,
17:45
Bruce布魯斯 Bueno布埃諾 de Mesquita梅斯基塔: I've thought about that,
364
1047000
2000
我想到這個,
17:47
and since以來 I've doneDONE a lot of work for the intelligence情報 community社區,
365
1049000
4000
我們為精英社群做了許多工作,
17:51
they've他們已經 also pondered沉吟 that.
366
1053000
2000
他們也考慮
17:53
It would be a good thing if
367
1055000
3000
那將是好事如果
17:56
people paid支付 more attention注意, took seriously認真地,
368
1058000
3000
人們更加注意,嚴肅對待,
17:59
and engaged訂婚 in the same相同 sorts排序 of calculations計算,
369
1061000
2000
也進行相似的算計,
18:01
because it would change更改 things. But it would change更改 things in two beneficial有利 ways方法.
370
1063000
4000
因為這會改變事情,它會在兩個有利的方面改變事情,
18:05
It would hasten how quickly很快 people arrive到達 at an agreement協議,
371
1067000
6000
他們提高 人們達成協議的速度,
18:11
and so it would save保存 everybody每個人 a lot of grief哀思 and time.
372
1073000
3000
所以它會節省每個人許多痛苦與時間。
18:14
And, it would arrive到達 at an agreement協議 that everybody每個人 was happy快樂 with,
373
1076000
4000
它也會達成每個人都喜歡的協議,
18:18
without having to manipulate操作 them so much --
374
1080000
3000
而不需要非常操控他們--
18:21
which哪一個 is basically基本上 what I do, I manipulate操作 them.
375
1083000
3000
這也就是我現在做的,我操控他們,
18:24
So it would be a good thing.
376
1086000
2000
這將是個好事,
18:26
CACA: So you're kind of trying to say, "People of Iran伊朗, this is your destiny命運, lets讓我們 go there."
377
1088000
4000
安德森:所以你是在客氣的說 ,“伊朗人,這是你們命定的,我們往這前進。”
18:30
BBMBBM: Well, people of Iran伊朗, this is what many許多 of you are going to evolve發展 to want,
378
1092000
6000
BBM:伊朗人,這是你們許多人將會轉變而期望的,
18:36
and we could get there a lot sooner,
379
1098000
2000
我們可以快一點達到那裡,
18:38
and you would suffer遭受 a lot less trouble麻煩 from economic經濟 sanctions制裁,
380
1100000
3000
這樣你也少受一些經濟制裁帶來的困難,
18:41
and we would suffer遭受 a lot less fear恐懼 of the use of military軍事 force on our end結束,
381
1103000
6000
我們也會少受點軍事威脅的恐懼,
18:47
and the world世界 would be a better place地點.
382
1109000
2000
世界也變得更可愛些,
18:49
CACA: Here's這裡的 hoping希望 they hear it that way. Thank you very much Bruce布魯斯.
383
1111000
3000
安德森:這是期望他們所聽到的,謝謝你。
18:52
BBMBBM: Thank you.
384
1114000
2000
BBM:謝謝。
18:54
(Applause掌聲)
385
1116000
5000
(掌聲)
Translated by chufung dou
Reviewed by Shelley Krishna R. TSANG

▲Back to top

ABOUT THE SPEAKER
Bruce Bueno de Mesquita - Political scientist
A consultant to the CIA and the Department of Defense, Bruce Bueno de Mesquita has built an intricate computer model that can predict the outcomes of international conflicts with bewildering accuracy.

Why you should listen

Every motive has a number, says Bruce Bueno de Mesquita. A specialist in foreign policy, international relations and state building, he is also a leading -- if controversial -- scholar of rational choice theory, which says math underlies the nation-scale consequences of individuals acting for personal benefit. He created forecasting technology that has, time and again, exceeded the accuracy of old-school analysis, even with thorny quarrels charged by obscure contenders, and often against odds. (One example: He called the second Intifada two years in advance.)

Bueno de Mesquita's company, Mesquita & Roundell, sells his system's predictions and analysis to influential government and private institutions that need heads-ups on policy. He teaches at NYU and is a senior fellow at the Hoover Institution.

More profile about the speaker
Bruce Bueno de Mesquita | Speaker | TED.com