ABOUT THE SPEAKER
Sebastian Seung - Computational neuroscientist
Sebastian Seung is a leader in the new field of connectomics, currently the hottest space in neuroscience, which studies, in once-impossible detail, the wiring of the brain.

Why you should listen

In the brain, neurons are connected into a complex network. Sebastian Seung and his lab at MIT are inventing technologies for identifying and describing the connectome, the totality of connections between the brain's neurons -- think of it as the wiring diagram of the brain. We possess our entire genome at birth, but things like memories are not "stored" in the genome; they are acquired through life and accumulated in the brain. Seung's hypothesis is that "we are our connectome," that the connections among neurons is where memories and experiences get stored.

Seung and his collaborators, including Winfried Denk at the Max Planck Institute and Jeff Lichtman at Harvard University, are working on a plan to thin-slice a brain (probably starting with a mouse brain) and trace, from slice to slice, each neural pathway, exposing the wiring diagram of the brain and creating a powerful new way to visualize the workings of the mind. They're not the first to attempt something like this -- Sydney Brenner won a Nobel for mapping all the 7,000 connections in the nervous system of a tiny worm, C. elegans. But that took his team a dozen years, and the worm only had 302 nerve cells. One of Seung's breakthroughs is in using advanced imagining and AI to handle the crushing amount of data that a mouse brain will yield and turn it into richly visual maps that show the passageways of thought and sensation.

More profile about the speaker
Sebastian Seung | Speaker | TED.com
TEDGlobal 2010

Sebastian Seung: I am my connectome

Sebastian Seung: 我就是我的聯結體

Filmed:
1,131,223 views

Sebastian Seung 正在製造一個野心十足的大腦聯結圖譜﹐它將會描繪聯繫所有神經元的聯結。他稱這是我們的“聯結體”(connectome)﹐那就和我們的基因體一樣獨特﹐理解它能開啟一個理解大腦和意識的新方法。
- Computational neuroscientist
Sebastian Seung is a leader in the new field of connectomics, currently the hottest space in neuroscience, which studies, in once-impossible detail, the wiring of the brain. Full bio

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

00:17
We live生活 in in a remarkable卓越 time,
0
2000
3000
我們生在一個偉大的時代
00:20
the age年齡 of genomics基因組學.
1
5000
3000
一個染色體組的時代。
00:23
Your genome基因組 is the entire整個 sequence序列 of your DNA脫氧核糖核酸.
2
8000
3000
你的染色體是你所有的DNA序列
00:26
Your sequence序列 and mine are slightly different不同.
3
11000
3000
你的序列和我的有些不同
00:29
That's why we look different不同.
4
14000
2000
因此我們長得不一樣
00:31
I've got brown棕色 eyes眼睛;
5
16000
2000
我的眼睛是褐色的
00:33
you might威力 have blue藍色 or gray灰色.
6
18000
3000
你的可能是藍或灰
00:36
But it's not just skin-deep膚淺的.
7
21000
2000
但這不只是表面
00:38
The headlines新聞頭條 tell us
8
23000
2000
新聞標題告訴我們
00:40
that genes基因 can give us scary害怕 diseases疾病,
9
25000
3000
基因裡可能隱藏著遺傳疾病
00:43
maybe even shape形狀 our personality個性,
10
28000
3000
甚至影響我們的個性
00:46
or give us mental心理 disorders障礙.
11
31000
3000
或給我們帶來精神異常
00:49
Our genes基因 seem似乎 to have
12
34000
3000
我們的基因似乎
00:52
awesome真棒 power功率 over our destinies命運.
13
37000
3000
對我們的命運有著極大的影響力
00:56
And yet然而, I would like to think
14
41000
3000
仍然﹐我希望
00:59
that I am more than my genes基因.
15
44000
3000
我不只是我的基因
01:04
What do you guys think?
16
49000
2000
你覺得呢﹖
01:06
Are you more than your genes基因?
17
51000
3000
你不只是你的基因吧﹖
01:09
(Audience聽眾: Yes.) Yes?
18
54000
2000
(觀眾﹕不只) 是嗎﹖
01:13
I think some people agree同意 with me.
19
58000
2000
我想觀眾中有些人同意我的說法
01:15
I think we should make a statement聲明.
20
60000
2000
我認為我們應該宣示一下
01:17
I think we should say it all together一起.
21
62000
2000
我認為我們應該一起宣示
01:20
All right: "I'm more than my genes基因" -- all together一起.
22
65000
3000
來吧﹕我不只是我的基因 -- 一起來
01:23
Everybody每個人: I am more than my genes基因.
23
68000
4000
眾人﹕我不只是我的基因
01:27
(Cheering打氣)
24
72000
2000
(歡呼)
01:30
Sebastian塞巴斯蒂安 Seung: What am I?
25
75000
2000
那我是什麼﹖
01:32
(Laughter笑聲)
26
77000
3000
(笑聲)
01:35
I am my connectome連接組.
27
80000
3000
我是我的聯結體。
01:40
Now, since以來 you guys are really great,
28
85000
2000
你們實在太棒了
01:42
maybe you can humor幽默 me and say this all together一起 too.
29
87000
2000
為了讓我開心﹐或許我們可以再宣示一次﹖
01:44
(Laughter笑聲)
30
89000
2000
(笑聲)
01:46
Right. All together一起 now.
31
91000
2000
好﹐一起來
01:48
Everybody每個人: I am my connectome連接組.
32
93000
3000
眾人﹕我是我的聯結體。
01:53
SSSS: That sounded滿面 great.
33
98000
2000
這實在太棒了。
01:55
You know, you guys are so great, you don't even know what a connectome連接組 is,
34
100000
2000
你們實在太棒了﹐你們甚至不知道聯結體是什麼
01:57
and you're willing願意 to play along沿 with me.
35
102000
2000
配合度這麼高
01:59
I could just go home now.
36
104000
3000
或許我現在就可以走了
02:02
Well, so far only one connectome連接組 is known已知,
37
107000
3000
現在﹐我們只知道一個聯結體
02:05
that of this tiny worm.
38
110000
3000
在這個小蟲裡
02:08
Its modest謙虛 nervous緊張 system系統
39
113000
2000
最小的神經系統
02:10
consists of just 300 neurons神經元.
40
115000
2000
裡面有300個神經元
02:12
And in the 1970s and '80s,
41
117000
2000
在1970和80年代
02:14
a team球隊 of scientists科學家們
42
119000
2000
有一組科學家
02:16
mapped映射 all 7,000 connections連接
43
121000
2000
畫出神經元中間
02:18
between之間 the neurons神經元.
44
123000
2000
的七千個聯繫
02:21
In this diagram, every一切 node節點 is a neuron神經元,
45
126000
2000
這個圖表裡的每個結都是一個神經元
02:23
and every一切 line is a connection連接.
46
128000
2000
每條線都是一個聯結。
02:25
This is the connectome連接組
47
130000
2000
這就是秀麗隱桿線蟲的
02:27
of the worm C. elegans線蟲.
48
132000
4000
聯結體
02:31
Your connectome連接組 is far more complex複雜 than this
49
136000
3000
你的聯結體比這個複雜多了
02:34
because your brain
50
139000
2000
因為你的腦子裡
02:36
contains包含 100 billion十億 neurons神經元
51
141000
2000
有一百億個神經元
02:38
and 10,000 times as many許多 connections連接.
52
143000
3000
以及一萬倍的聯結體
02:41
There's a diagram like this for your brain,
53
146000
2000
你的腦子也能做成像這樣的圖表
02:43
but there's no way it would fit適合 on this slide滑動.
54
148000
3000
只是不可能放得進這張投影片
02:47
Your connectome連接組 contains包含 one million百萬 times more connections連接
55
152000
3000
你聯結體的聯結是
02:50
than your genome基因組 has letters.
56
155000
3000
基因體編碼的一百萬倍
02:53
That's a lot of information信息.
57
158000
2000
裡面有很多資料
02:55
What's in that information信息?
58
160000
3000
這些資料裡包含了什麼﹖
02:59
We don't know for sure, but there are theories理論.
59
164000
3000
我們還不能確定﹐但有一些學說
03:02
Since以來 the 19th century世紀, neuroscientists神經學家 have speculated推測
60
167000
3000
從十九世紀以來﹐神經科學家推測
03:05
that maybe your memories回憶 --
61
170000
2000
你的記憶 -
03:07
the information信息 that makes品牌 you, you --
62
172000
2000
那些讓你之所以為你的資料 -
03:09
maybe your memories回憶 are stored存儲
63
174000
2000
說不定你的記憶就儲存在
03:11
in the connections連接 between之間 your brain's大腦的 neurons神經元.
64
176000
2000
腦子裡神經元的聯結裡
03:15
And perhaps也許 other aspects方面 of your personal個人 identity身分 --
65
180000
2000
或者你其他的個人特性
03:17
maybe your personality個性 and your intellect智力 --
66
182000
3000
你的個性和你的思維方式
03:20
maybe they're also encoded編碼
67
185000
2000
說不定它們都寫在
03:22
in the connections連接 between之間 your neurons神經元.
68
187000
3000
你的神經元的聯結裡
03:26
And so now you can see why I proposed建議 this hypothesis假設:
69
191000
3000
現在你可以了解為什麼我要提出這個假設﹕
03:29
I am my connectome連接組.
70
194000
3000
我就是我的聯結體。
03:32
I didn't ask you to chant it because it's true真正;
71
197000
3000
我要你和我一起吟誦不是因為它是真的
03:35
I just want you to remember記得 it.
72
200000
2000
我只是希望你記住它
03:37
And in fact事實, we don't know if this hypothesis假設 is correct正確,
73
202000
2000
事實上﹐我們不知道這個假設是否正確
03:39
because we have never had technologies技術
74
204000
2000
因為我們從來沒有如此強大的科技
03:41
powerful強大 enough足夠 to test測試 it.
75
206000
2000
足以測試這個假設
03:44
Finding查找 that worm connectome連接組
76
209000
3000
找到這個線蟲的聯結體
03:47
took over a dozen years年份 of tedious乏味 labor勞動.
77
212000
3000
花了12年的努力
03:50
And to find the connectomesconnectomes of brains大腦 more like our own擁有,
78
215000
3000
要找到我們腦子裡的這些聯結體
03:53
we need more sophisticated複雜的 technologies技術, that are automated自動化,
79
218000
3000
我們需要更精密的自動儀器
03:56
that will speed速度 up the process處理 of finding發現 connectomesconnectomes.
80
221000
3000
才能加速我們找尋聯結體的速度。
03:59
And in the next下一個 few少數 minutes分鐘, I'll tell you about some of these technologies技術,
81
224000
3000
接下來的幾分鐘﹐我會告訴你們這些
04:02
which哪一個 are currently目前 under development發展
82
227000
2000
在我和我的合作者的實驗室裡
04:04
in my lab實驗室 and the labs實驗室 of my collaborators合作者.
83
229000
3000
發展中的新科技。
04:08
Now you've probably大概 seen看到 pictures圖片 of neurons神經元 before.
84
233000
3000
你可能看過神經元的圖片
04:11
You can recognize認識 them instantly即刻
85
236000
2000
你可以從它們的姿態裡
04:13
by their fantastic奇妙 shapes形狀.
86
238000
3000
輕易的認出它來
04:16
They extend延伸 long and delicate精巧 branches分支機構,
87
241000
3000
它們延伸著長長的精密分支
04:19
and in short, they look like trees樹木.
88
244000
3000
簡單來說﹐看上去就像棵樹
04:22
But this is just a single neuron神經元.
89
247000
3000
但這只是一個神經元
04:25
In order訂購 to find connectomesconnectomes,
90
250000
2000
如果我們要找尋聯結體
04:27
we have to see all the neurons神經元 at the same相同 time.
91
252000
3000
我們需要同時看見所有神經元
04:30
So let's meet遇到 Bobby鮑比 Kasthuri卡斯特里,
92
255000
2000
讓我們認識這位Bobby Kasthuri
04:32
who works作品 in the laboratory實驗室 of Jeff傑夫 Lichtman利希特曼
93
257000
2000
他在哈佛 Jeff Lichtman 實驗室
04:34
at Harvard哈佛 University大學.
94
259000
2000
裡面工作
04:36
Bobby鮑比 is holding保持 fantastically飛馳 thin slices
95
261000
2000
Bobby 手握著一片奇妙的
04:38
of a mouse老鼠 brain.
96
263000
2000
老鼠腦。
04:40
And we're zooming縮放 in by a factor因子 of 100,000 times
97
265000
3000
讓我們放大十萬倍
04:44
to obtain獲得 the resolution解析度,
98
269000
2000
得到一個更清晰的分辨率
04:46
so that we can see the branches分支機構 of neurons神經元 all at the same相同 time.
99
271000
3000
讓我們一次看見神經元的所有分支
04:50
Except, you still may可能 not really recognize認識 them,
100
275000
3000
除了﹐你可能認不出它來
04:53
and that's because we have to work in three dimensions尺寸.
101
278000
3000
因為我們需要在三維度裡看
04:56
If we take many許多 images圖片 of many許多 slices of the brain
102
281000
2000
讓我們把一片片的腦部圖片
04:58
and stack them up,
103
283000
2000
堆起來
05:00
we get a three-dimensional三維 image圖片.
104
285000
2000
我們就得到了這個3D圖像
05:02
And still, you may可能 not see the branches分支機構.
105
287000
2000
但是你還是看不到這些分支
05:04
So we start開始 at the top最佳,
106
289000
2000
於是我們從上面
05:06
and we color顏色 in the cross-section橫截面 of one branch in red,
107
291000
3000
把橫斷面裡的分支塗成紅色
05:09
and we do that for the next下一個 slice
108
294000
2000
然後再塗下一片
05:11
and for the next下一個 slice.
109
296000
2000
再下一片
05:13
And we keep on doing that,
110
298000
2000
我們一直這樣做
05:15
slice after slice.
111
300000
3000
一片又一片
05:18
If we continue繼續 through通過 the entire整個 stack,
112
303000
2000
直到我們把整堆都塗完
05:20
we can reconstruct重建 the three-dimensional三維 shape形狀
113
305000
3000
我們就可以在3D形狀裡重現
05:23
of a small fragment分段 of a branch of a neuron神經元.
114
308000
3000
神經元分支的一小部份
05:26
And we can do that for another另一個 neuron神經元 in green綠色.
115
311000
2000
我們可以把另一個神經元涂成綠色
05:28
And you can see that the green綠色 neuron神經元 touches觸摸 the red neuron神經元
116
313000
2000
我們可以看到綠色和紅色神經元
05:30
at two locations地點,
117
315000
2000
在兩個地方接觸
05:32
and these are what are called synapses突觸.
118
317000
2000
這就是所謂的突觸(synapses)
05:34
Let's zoom放大 in on one synapse突觸,
119
319000
2000
讓我們放大這個突觸
05:36
and keep your eyes眼睛 on the interior室內 of the green綠色 neuron神經元.
120
321000
3000
繼續看著綠色神經元的內部
05:39
You should see small circles --
121
324000
2000
你會看到小小的圈
05:41
these are called vesicles囊泡.
122
326000
3000
這些就是突觸小泡(囊泡)
05:44
They contain包含 a molecule分子 know as a neurotransmitter神經遞質.
123
329000
3000
裡面有叫神經傳遞素的分子
05:47
And so when the green綠色 neuron神經元 wants to communicate通信,
124
332000
2000
當綠色神經元想和紅色神經元溝通
05:49
it wants to send發送 a message信息 to the red neuron神經元,
125
334000
2000
傳遞訊息的時候
05:51
it spits吐奶 out neurotransmitter神經遞質.
126
336000
3000
它就從突觸吐出神經傳遞素
05:54
At the synapse突觸, the two neurons神經元
127
339000
2000
兩個神經元
05:56
are said to be connected連接的
128
341000
2000
就這樣聯繫
05:58
like two friends朋友 talking on the telephone電話.
129
343000
3000
像兩個朋友講電話一樣
06:02
So you see how to find a synapse突觸.
130
347000
2000
這就是找到突觸的方法
06:04
How can we find an entire整個 connectome連接組?
131
349000
3000
但要怎麼找到整個聯結體呢
06:07
Well, we take this three-dimensional三維 stack of images圖片
132
352000
3000
我們把這堆層層疊起的3D畫面
06:10
and treat對待 it as a gigantic巨大 three-dimensional三維 coloring染色 book.
133
355000
3000
把它變成一個超大型的3D塗色簿
06:13
We color顏色 every一切 neuron神經元 in, in a different不同 color顏色,
134
358000
3000
把所有神經元涂成不同顏色
06:16
and then we look through通過 all of the images圖片,
135
361000
2000
看過所有切片圖
06:18
find the synapses突觸
136
363000
2000
找到突觸
06:20
and note注意 the colors顏色 of the two neurons神經元 involved參與 in each synapse突觸.
137
365000
3000
然後在記錄參與突觸的兩個神經元
06:23
If we can do that throughout始終 all the images圖片,
138
368000
3000
如果我們可以這樣處理所有圖片
06:26
we could find a connectome連接組.
139
371000
2000
就可以找到一個聯結體
06:29
Now, at this point,
140
374000
2000
目前為止
06:31
you've learned學到了 the basics基本 of neurons神經元 and synapses突觸.
141
376000
2000
你們已經學到了神經元和突觸的基礎
06:33
And so I think we're ready準備 to tackle滑車
142
378000
2000
我想我們已經可以處理
06:35
one of the most important重要 questions問題 in neuroscience神經科學:
143
380000
3000
神經科學裡最重要的問題之一﹕
06:39
how are the brains大腦 of men男人 and women婦女 different不同?
144
384000
3000
男人和女人的大腦有什麼不同﹖
06:42
(Laughter笑聲)
145
387000
2000
(笑聲)
06:44
According根據 to this self-help自救 book,
146
389000
2000
從這本勵志書裡看來
06:46
guys brains大腦 are like waffles威化餅;
147
391000
2000
男人的大腦像鬆餅
06:48
they keep their lives生活 compartmentalized條塊 in boxes盒子.
148
393000
3000
把生活分門別類放在小格子裡
06:51
Girls'少女時代 brains大腦 are like spaghetti意大利面;
149
396000
3000
女人的大腦則像意大利麵
06:54
everything in their life is connected連接的 to everything else其他.
150
399000
3000
人生裡的每件事都習習相關
06:57
(Laughter笑聲)
151
402000
2000
(笑聲)
06:59
You guys are laughing,
152
404000
2000
你們現在在笑
07:01
but you know, this book changed my life.
153
406000
2000
但這本書改變了我的生命
07:03
(Laughter笑聲)
154
408000
2000
(笑聲)
07:07
But seriously認真地, what's wrong錯誤 with this?
155
412000
3000
但說真的﹐它錯在哪裡﹖
07:10
You already已經 know enough足夠 to tell me -- what's wrong錯誤 with this statement聲明?
156
415000
3000
你們已經有能力可以告訴我﹐這句話錯在哪裡
07:20
It doesn't matter whether是否 you're a guy or girl女孩,
157
425000
3000
無論你是男人還是女人
07:23
everyone's大家的 brains大腦 are like spaghetti意大利面.
158
428000
3000
每個人的大腦都是意大利麵
07:26
Or maybe really, really fine capellini卡佩利尼 with branches分支機構.
159
431000
3000
或是﹐加上分支的超細天使髮麵
07:30
Just as one strand of spaghetti意大利面
160
435000
2000
就像在你盤子裡的一條意大利麵
07:32
contacts往來 many許多 other strands on your plate盤子,
161
437000
3000
碰觸其它意大利麵一樣
07:35
one neuron神經元 touches觸摸 many許多 other neurons神經元
162
440000
2000
一個神經元會用它們糾纏的分支
07:37
through通過 their entangled糾纏 branches分支機構.
163
442000
2000
觸碰許多其它神經元
07:39
One neuron神經元 can be connected連接的 to so many許多 other neurons神經元,
164
444000
3000
一個神經元可以和許多其它神經元連接
07:42
because there can be synapses突觸
165
447000
2000
因為在這些聯結點
07:44
at these points of contact聯繫.
166
449000
3000
可以有許多突觸
07:49
By now, you might威力 have sort分類 of lost丟失 perspective透視
167
454000
3000
現在你可能已經忘記
07:52
on how large this cube立方體 of brain tissue組織 actually其實 is.
168
457000
3000
這塊腦尺度究竟有多小
07:55
And so let's do a series系列 of comparisons對比 to show顯示 you.
169
460000
3000
我們來比較一下
07:58
I assure保證 you, this is very tiny. It's just six microns微米 on a side.
170
463000
3000
這真的很小﹐只有六微米(百萬分之一米)
08:03
So, here's這裡的 how it stacks up against反對 an entire整個 neuron神經元.
171
468000
3000
面對整條神經元又是怎樣呢
08:06
And you can tell that, really, only the smallest最少 fragments片段 of branches分支機構
172
471000
3000
你可以發現﹐這個方塊
08:09
are contained inside this cube立方體.
173
474000
3000
只是這個分支其中的一小小塊
08:12
And a neuron神經元, well, that's smaller than brain.
174
477000
3000
而神經元﹐當然比腦還小
08:17
And that's just a mouse老鼠 brain --
175
482000
2000
而這甚至只是老鼠的腦
08:21
it's a lot smaller than a human人的 brain.
176
486000
3000
比人腦還小很多
08:25
So when show顯示 my friends朋友 this,
177
490000
2000
於是當我給朋友看這些的時候
08:27
sometimes有時 they've他們已經 told me,
178
492000
2000
他們會告訴我
08:29
"You know, Sebastian塞巴斯蒂安, you should just give up.
179
494000
3000
“Sebastian﹐我想你放棄好了。
08:32
Neuroscience神經科學 is hopeless絕望."
180
497000
2000
神經科學簡直無可救藥。”
08:34
Because if you look at a brain with your naked eye,
181
499000
2000
因為當你用肉眼看大腦時
08:36
you don't really see how complex複雜 it is,
182
501000
2000
你不知道它到底有多麼複雜
08:38
but when you use a microscope顯微鏡,
183
503000
2000
但當你把它放在顯微鏡下
08:40
finally最後 the hidden complexity複雜 is revealed透露.
184
505000
3000
這些隱藏的複雜性就顯露出來了。
08:45
In the 17th century世紀,
185
510000
2000
在十七世紀
08:47
the mathematician數學家 and philosopher哲學家, Blaise布萊斯 Pascal帕斯卡爾,
186
512000
2000
法國哲學家和數學家巴斯卡
08:49
wrote of his dread恐懼 of the infinite無窮,
187
514000
3000
寫下他對無限的恐懼
08:52
his feeling感覺 of insignificance渺小
188
517000
2000
當他思考外太空時
08:54
at contemplating考慮 the vast廣大 reaches到達 of outer space空間.
189
519000
3000
感到自己是如何的微不足道
08:59
And, as a scientist科學家,
190
524000
2000
身為科學家
09:01
I'm not supposed應該 to talk about my feelings情懷 --
191
526000
3000
我不應該談論我的感受
09:04
too much information信息, professor教授.
192
529000
2000
教授﹐我真的不想知道
09:06
(Laughter笑聲)
193
531000
2000
(笑聲)
09:08
But may可能 I?
194
533000
2000
我...... 可以嗎﹖
09:10
(Laughter笑聲)
195
535000
2000
(笑聲)
09:12
(Applause掌聲)
196
537000
2000
(掌聲)
09:14
I feel curiosity好奇心,
197
539000
2000
我感到好奇
09:16
and I feel wonder奇蹟,
198
541000
2000
我感到驚嘆
09:18
but at times I have also felt despair絕望.
199
543000
3000
但有時我也感到絕望
09:22
Why did I choose選擇 to study研究
200
547000
2000
為什麼我要選擇學習
09:24
this organ器官 that is so awesome真棒 in its complexity複雜
201
549000
3000
這樣一個複雜到不可思議
09:27
that it might威力 well be infinite無窮?
202
552000
2000
有可能接近無限的器官﹖
09:29
It's absurd荒誕.
203
554000
2000
這太荒謬了
09:31
How could we even dare to think
204
556000
2000
我們怎麼敢妄想
09:33
that we might威力 ever understand理解 this?
205
558000
3000
我們有可能可以理解它呢﹖
09:38
And yet然而, I persist堅持 in this quixotic不切實際 endeavor努力.
206
563000
3000
但﹐我仍然想踏上這唐吉訶德式的旅程
09:41
And indeed確實, these days I harbor港口 new hopes希望.
207
566000
3000
最近﹐我有了新的希望
09:45
Someday日後,
208
570000
2000
或許某天
09:47
a fleet艦隊 of microscopes顯微鏡 will capture捕獲
209
572000
2000
某種顯微鏡能捕捉
09:49
every一切 neuron神經元 and every一切 synapse突觸
210
574000
2000
巨大圖片資料庫裡的
09:51
in a vast廣大 database數據庫 of images圖片.
211
576000
3000
每一個神經元和突觸
09:54
And some day, artificially人為 intelligent智能 supercomputers超級計算機
212
579000
3000
有這麼一天﹐一個人工智慧的超級電腦
09:57
will analyze分析 the images圖片 without human人的 assistance幫助
213
582000
3000
能在無人操作的狀況下分析這些圖像
10:00
to summarize總結 them in a connectome連接組.
214
585000
3000
並把它們總結成一個聯結體
10:04
I do not know, but I hope希望 that I will live生活 to see that day,
215
589000
3000
我不知道能不能﹐但我希望我能看到這一天
10:08
because finding發現 an entire整個 human人的 connectome連接組
216
593000
2000
因為找出人類的所有聯結體
10:10
is one of the greatest最大 technological技術性 challenges挑戰 of all time.
217
595000
3000
是科技史上最大的挑戰之一
10:13
It will take the work of generations to succeed成功.
218
598000
3000
可能要幾個世代才能成功
10:17
At the present當下 time, my collaborators合作者 and I,
219
602000
3000
目前﹐我和我的夥伴
10:20
what we're aiming瞄準 for is much more modest謙虛 --
220
605000
2000
我們的目標較為微小
10:22
just to find partial局部 connectomesconnectomes
221
607000
2000
不過是在鼠腦和人腦間
10:24
of tiny chunks of mouse老鼠 and human人的 brain.
222
609000
3000
找到部份的聯結體
10:27
But even that will be enough足夠 for the first tests測試 of this hypothesis假設
223
612000
3000
但就算只是這樣﹐也足以實驗“我就是我的聯結體“
10:30
that I am my connectome連接組.
224
615000
3000
這個假設
10:35
For now, let me try to convince說服 you of the plausibility合理性 of this hypothesis假設,
225
620000
3000
現在我只是想說服你們﹐這個假設有它的可信度
10:38
that it's actually其實 worth價值 taking服用 seriously認真地.
226
623000
3000
它是值得被認真對待的
10:42
As you grow增長 during childhood童年
227
627000
2000
在你的成長過程
10:44
and age年齡 during adulthood成年,
228
629000
3000
不同的經歷
10:47
your personal個人 identity身分 changes變化 slowly慢慢地.
229
632000
3000
你對自己的身份認同也逐漸改變
10:50
Likewise同樣, every一切 connectome連接組
230
635000
2000
同樣地﹐每個聯結體
10:52
changes變化 over time.
231
637000
2000
也隨著時間改變
10:55
What kinds of changes變化 happen發生?
232
640000
2000
怎樣的改變呢﹖
10:57
Well, neurons神經元, like trees樹木,
233
642000
2000
神經元﹐就像樹一樣
10:59
can grow增長 new branches分支機構,
234
644000
2000
可以長出新的枝幹
11:01
and they can lose失去 old ones那些.
235
646000
3000
也可以失去一些老枝幹
11:04
Synapses突觸 can be created創建,
236
649000
3000
突觸可以被創造
11:07
and they can be eliminated淘汰.
237
652000
3000
也可以被淘汰
11:10
And synapses突觸 can grow增長 larger,
238
655000
2000
突觸可以長大
11:12
and they can grow增長 smaller.
239
657000
3000
也可以變小
11:15
Second第二 question:
240
660000
2000
第二個問題是﹕
11:17
what causes原因 these changes變化?
241
662000
3000
這些改變是從哪裡來的﹖
11:20
Well, it's true真正.
242
665000
2000
沒錯
11:22
To some extent程度, they are programmed程序 by your genes基因.
243
667000
3000
某種程度而言﹐它們寫在你的基因裡
11:25
But that's not the whole整個 story故事,
244
670000
2000
但那不是全部
11:27
because there are signals信號, electrical電動 signals信號,
245
672000
2000
因為有訊號﹐電位訊號
11:29
that travel旅行 along沿 the branches分支機構 of neurons神經元
246
674000
2000
在神經元枝幹裡運行
11:31
and chemical化學 signals信號
247
676000
2000
還有化學訊號
11:33
that jump across橫過 from branch to branch.
248
678000
2000
從枝幹跳往枝幹
11:35
These signals信號 are called neural神經 activity活動.
249
680000
3000
這些訊號就叫神經活動
11:38
And there's a lot of evidence證據
250
683000
2000
有很多證據
11:40
that neural神經 activity活動
251
685000
3000
神經活動
11:43
is encoding編碼 our thoughts思念, feelings情懷 and perceptions看法,
252
688000
3000
寫就了我們的想法﹐感覺和感知
11:46
our mental心理 experiences經驗.
253
691000
2000
我們所有的思考經驗
11:48
And there's a lot of evidence證據 that neural神經 activity活動
254
693000
3000
有許多的證據證明神經活動
11:51
can cause原因 your connections連接 to change更改.
255
696000
3000
可以改變這些連結
11:54
And if you put those two facts事實 together一起,
256
699000
3000
如果你綜合這兩個事實
11:57
it means手段 that your experiences經驗
257
702000
2000
這代表著你的經驗
11:59
can change更改 your connectome連接組.
258
704000
3000
能改變你的聯結體
12:02
And that's why every一切 connectome連接組 is unique獨特,
259
707000
2000
這就是為什麼每個聯結體都獨一無二
12:04
even those of genetically基因 identical相同 twins雙胞胎.
260
709000
3000
就算是基因完全相同的雙胞胎也一樣
12:08
The connectome連接組 is where nature性質 meets符合 nurture培育.
261
713000
3000
聯結體便是先天加上後天的綜合體
12:12
And it might威力 true真正
262
717000
2000
有可能
12:14
that just the mere act法案 of thinking思維
263
719000
2000
就算只是想想而已
12:16
can change更改 your connectome連接組 --
264
721000
2000
也能改變你的聯結體
12:18
an idea理念 that you may可能 find empowering授權.
265
723000
3000
你也許能從這想法中得到力量。
12:24
What's in this picture圖片?
266
729000
2000
這是什麼﹖
12:28
A cool and refreshing清爽 stream of water, you say.
267
733000
3000
有人會說,一條清涼的小河
12:32
What else其他 is in this picture圖片?
268
737000
2000
還有呢﹖
12:37
Do not forget忘記 that groove in the Earth地球
269
742000
2000
別忘了下面那條刻在地球上的
12:39
called the stream bed.
270
744000
3000
河床
12:42
Without沒有 it, the water would not know in which哪一個 direction方向 to flow.
271
747000
3000
沒有它﹐河水不知往哪裡流
12:45
And with the stream,
272
750000
2000
我想以這條河流
12:47
I would like to propose提出 a metaphor隱喻
273
752000
2000
作為說明神經活動和連接
12:49
for the relationship關係 between之間 neural神經 activity活動
274
754000
2000
兩者關係
12:51
and connectivity連接.
275
756000
2000
的隱喻。
12:54
Neural神經 activity活動 is constantly經常 changing改變.
276
759000
3000
神經活動是一直在改變的
12:57
It's like the water of the stream; it never sits坐鎮 still.
277
762000
3000
就像河水﹐從不停息
13:00
The connections連接
278
765000
2000
而連接
13:02
of the brain's大腦的 neural神經 network網絡
279
767000
2000
大腦的神經組織
13:04
determines確定 the pathways途徑
280
769000
2000
則決定了
13:06
along沿 which哪一個 neural神經 activity活動 flows流動.
281
771000
2000
這些神經活動的方向
13:08
And so the connectome連接組 is like bed of the stream;
282
773000
3000
聯結體就像河床
13:13
but the metaphor隱喻 is richer更豐富 than that,
283
778000
3000
這個比喻還不止這樣
13:16
because it's true真正 that the stream bed
284
781000
3000
因為雖然是河床帶領著
13:19
guides導遊 the flow of the water,
285
784000
2000
河水
13:21
but over long timescales時間表,
286
786000
2000
在悠長的時間裡
13:23
the water also reshapes重塑 the bed of the stream.
287
788000
3000
河水也改變了河床的方向。
13:26
And as I told you just now,
288
791000
2000
就像我剛告訴你的
13:28
neural神經 activity活動 can change更改 the connectome連接組.
289
793000
3000
神經活動可以改變聯結體
13:33
And if you'll你會 allow允許 me to ascend
290
798000
2000
如果你允許我提昇到
13:35
to metaphorical隱喻 heights高度,
291
800000
3000
一種文學的層次
13:38
I will remind提醒 you that neural神經 activity活動
292
803000
3000
我再次提醒各位,神經活動是
13:41
is the physical物理 basis基礎 -- or so neuroscientists神經學家 think --
293
806000
2000
人類思想、感覺、感知的生物基礎﹐
13:43
of thoughts思念, feelings情懷 and perceptions看法.
294
808000
3000
至少神經科學家是這麼認為的。
13:46
And so we might威力 even speak說話 of
295
811000
2000
我們甚至可以說它是
13:48
the stream of consciousness意識.
296
813000
2000
意識流。
13:50
Neural神經 activity活動 is its water,
297
815000
3000
神經活動是河水
13:53
and the connectome連接組 is its bed.
298
818000
3000
聯結體是河床
13:57
So let's return返回 from the heights高度 of metaphor隱喻
299
822000
2000
讓我們從文學的高度回到
13:59
and return返回 to science科學.
300
824000
2000
科學上
14:01
Suppose假設 our technologies技術 for finding發現 connectomesconnectomes
301
826000
2000
假設我們的科技真的可以找出
14:03
actually其實 work.
302
828000
2000
所有的聯結體
14:05
How will we go about testing測試 the hypothesis假設
303
830000
2000
我們該如何測試這個”我就是我的聯結體“
14:07
"I am my connectome連接組?"
304
832000
3000
的假設﹖
14:10
Well, I propose提出 a direct直接 test測試.
305
835000
3000
讓我提議一個直接的測試法
14:13
Let us attempt嘗試
306
838000
2000
讓我們嘗試
14:15
to read out memories回憶 from connectomesconnectomes.
307
840000
3000
從聯結體中解讀出我們的記憶
14:18
Consider考慮 the memory記憶
308
843000
2000
想像記憶
14:20
of long temporal sequences序列 of movements運動,
309
845000
3000
是一連串綿長的短樂章
14:23
like a pianist鋼琴家 playing播放 a Beethoven貝多芬 sonata奏鳴曲.
310
848000
3000
就像一個彈奏貝多芬夜曲的鋼琴家。
14:26
According根據 to a theory理論 that dates日期 back to the 19th century世紀,
311
851000
3000
從一個十九世紀的學說看來
14:29
such這樣 memories回憶 are stored存儲
312
854000
2000
這些回憶就儲存在
14:31
as chains of synaptic突觸 connections連接 inside your brain.
313
856000
3000
你腦子裡的那串突觸聯結
14:35
Because, if the first neurons神經元 in the chain are activated活性,
314
860000
3000
如果這串聯結的第一個神經元被啟動了
14:38
through通過 their synapses突觸 they send發送 messages消息 to the second第二 neurons神經元, which哪一個 are activated活性,
315
863000
3000
開始對第二個神經元傳送訊息
14:41
and so on down the line,
316
866000
2000
一路延續下去
14:43
like a chain of falling落下 dominoes骨牌.
317
868000
2000
就像一整條骨牌
14:45
And this sequence序列 of neural神經 activation激活
318
870000
2000
這一系列的神經活動
14:47
is hypothesized假設 to be the neural神經 basis基礎
319
872000
3000
便是假設中的連續動態
14:50
of those sequence序列 of movements運動.
320
875000
2000
的神經基礎
14:52
So one way of trying to test測試 the theory理論
321
877000
2000
測試這個學說的辦法之一
14:54
is to look for such這樣 chains
322
879000
2000
便是找到聯結體中的
14:56
inside connectomesconnectomes.
323
881000
2000
這串連接
14:58
But it won't慣於 be easy簡單, because they're not going to look like this.
324
883000
3000
但這並不容易﹐因為它們看起來不會像這樣
15:01
They're going to be scrambled up.
325
886000
2000
它們會全部纏在一起
15:03
So we'll have to use our computers電腦
326
888000
2000
我們需要用電腦
15:05
to try to unscramble解讀 the chain.
327
890000
3000
嘗試把它們解開
15:08
And if we can do that,
328
893000
2000
如果我們鬆開
15:10
the sequence序列 of the neurons神經元 we recover恢復 from that unscrambling解讀
329
895000
3000
這些纏在一起的神經元序列
15:13
will be a prediction預測 of the pattern模式 of neural神經 activity活動
330
898000
3000
就能夠預測回憶時
15:16
that is replayed重播 in the brain during memory記憶 recall召回.
331
901000
3000
所重放的神經活動模式
15:19
And if that were successful成功,
332
904000
2000
如果這也成功了
15:21
that would be the first example of reading a memory記憶 from a connectome連接組.
333
906000
3000
那就會成為從聯結體中讀出記憶的第一例
15:28
(Laughter笑聲)
334
913000
2000
(笑聲)
15:30
What a mess食堂 --
335
915000
2000
真是一團亂
15:33
have you ever tried試著 to wire up a system系統
336
918000
2000
你試過為這麼複雜的系統
15:35
as complex複雜 as this?
337
920000
2000
接上線嗎﹖
15:37
I hope希望 not.
338
922000
2000
希望你沒有這種經驗
15:39
But if you have, you know it's very easy簡單 to make a mistake錯誤.
339
924000
3000
但如果你有這種經驗﹐你便知道犯錯是難免的
15:45
The branches分支機構 of neurons神經元 are like the wires電線 of the brain.
340
930000
2000
神經元的樹突就像腦子裡的電線
15:47
Can anyone任何人 guess猜測: what's the total length長度 of wires電線 in your brain?
341
932000
4000
誰能猜出﹐腦中所有電線的長度﹖
15:54
I'll give you a hint暗示. It's a big number.
342
939000
2000
給你一點提示﹐很長。
15:56
(Laughter笑聲)
343
941000
2000
(笑聲)
15:59
I estimate估計, millions百萬 of miles英里,
344
944000
3000
我測量過﹐幾百萬公里
16:02
all packed打包 in your skull頭骨.
345
947000
3000
全擠在你的腦殼裡
16:05
And if you appreciate欣賞 that number,
346
950000
2000
如果你想像這個數字
16:07
you can easily容易 see
347
952000
2000
可以很容易發現
16:09
there is huge巨大 potential潛在 for mis-wiring誤接線 of the brain.
348
954000
2000
頭腦裡很有可能會接錯線
16:11
And indeed確實, the popular流行 press loves headlines新聞頭條 like,
349
956000
3000
於是﹐許多媒體喜歡用這種頭條﹐
16:14
"Anorexic厭食症 brains大腦 are wired有線 differently不同,"
350
959000
2000
"厭食症患者腦神經異於常人“
16:16
or "Autistic自閉症 brains大腦 are wired有線 differently不同."
351
961000
2000
或"自閉症患者腦神經異於常人“
16:18
These are plausible似是而非 claims索賠,
352
963000
2000
這些聽上去都很可信
16:20
but in truth真相,
353
965000
2000
但事實上
16:22
we can't see the brain's大腦的 wiring接線 clearly明確地 enough足夠
354
967000
2000
我們根本沒法看清楚這些腦連接
16:24
to tell if these are really true真正.
355
969000
2000
更不可能知道這樣說是否真實
16:26
And so the technologies技術 for seeing眼看 connectomesconnectomes
356
971000
3000
若我們有了能看見聯結體的科技
16:29
will allow允許 us to finally最後
357
974000
2000
我們就能從大腦
16:31
read mis-wiring誤接線 of the brain,
358
976000
2000
接錯的線路中
16:33
to see mental心理 disorders障礙 in connectomesconnectomes.
359
978000
3000
從聯結體來辨識神經疾病。
16:40
Sometimes有時 the best最好 way to test測試 a hypothesis假設
360
985000
3000
有時候﹐證明假設最好的方法
16:43
is to consider考慮 its most extreme極端 implication意義.
361
988000
3000
便是把可能性推到極限。
16:46
Philosophers哲學家 know this game遊戲 very well.
362
991000
3000
哲學家很擅長這個游戲
16:50
If you believe that I am my connectome連接組,
363
995000
3000
如果你相信我就是我的聯結體
16:53
I think you must必須 also accept接受 the idea理念
364
998000
3000
你也必須接受
16:56
that death死亡 is the destruction毀壞
365
1001000
2000
死亡便是破壞聯結體
16:58
of your connectome連接組.
366
1003000
3000
的想法
17:02
I mention提到 this because there are prophets先知 today今天
367
1007000
3000
我這麼說是因為今日有許多先知
17:05
who claim要求 that technology技術
368
1010000
3000
聲明科技將會
17:08
will fundamentally從根本上 alter改變 the human人的 condition條件
369
1013000
3000
徹底的改變人類的生存狀態
17:11
and perhaps也許 even transform轉變 the human人的 species種類.
370
1016000
3000
甚至改變人類這個物種
17:14
One of their most cherished珍愛的 dreams
371
1019000
3000
其中一個最迷人的夢想
17:17
is to cheat作弊 death死亡
372
1022000
2000
就是克服死亡
17:19
by that practice實踐 known已知 as cryonics人體冷凍.
373
1024000
2000
用人體冷藏法
17:21
If you pay工資 100,000 dollars美元,
374
1026000
2000
你付上十萬美金的代價
17:23
you can arrange安排 to have your body身體 frozen凍結的 after death死亡
375
1028000
3000
就可以把死後的身體急速冷藏
17:26
and stored存儲 in liquid液體 nitrogen
376
1031000
2000
儲存在某個充滿液氮的鐵罐裡
17:28
in one of these tanks坦克 in an Arizona亞利桑那 warehouse倉庫,
377
1033000
2000
放在亞歷桑納州的某個倉庫
17:30
awaiting等待 a future未來 civilization文明
378
1035000
2000
等待未來某個先進的文明
17:32
that is advanced高級 to resurrect復活 you.
379
1037000
3000
來讓你復活
17:36
Should we ridicule嘲笑 the modern現代 seekers求職者 of immortality不朽,
380
1041000
2000
我們應該開這些追求長生不老者
17:38
calling調用 them fools傻瓜?
381
1043000
2000
的玩笑嗎﹖
17:40
Or will they someday日後 chuckle暗笑
382
1045000
2000
還是某天他們會在我們的墳前
17:42
over our graves墳墓?
383
1047000
2000
呵呵地笑﹖
17:45
I don't know --
384
1050000
2000
我不知道。
17:47
I prefer比較喜歡 to test測試 their beliefs信仰, scientifically科學.
385
1052000
3000
但我可以使用科學方法把他們的信念拿來實驗
17:50
I propose提出 that we attempt嘗試 to find a connectome連接組
386
1055000
2000
假設我們在某個冷凍的大腦裡找到
17:52
of a frozen凍結的 brain.
387
1057000
2000
一個聯結體
17:54
We know that damage損傷 to the brain
388
1059000
2000
我們知道死後急凍將會
17:56
occurs發生 after death死亡 and during freezing冷凍.
389
1061000
2000
破壞大腦組織
17:58
The question is: has that damage損傷 erased擦除 the connectome連接組?
390
1063000
3000
於是問題是﹕聯結體被破壞了嗎﹖
18:01
If it has, there is no way that any future未來 civilization文明
391
1066000
3000
如果答案是肯定的﹐未來的文明
18:04
will be able能夠 to recover恢復 the memories回憶 of these frozen凍結的 brains大腦.
392
1069000
3000
決不可能恢復這些冷凍大腦裡的記憶
18:07
Resurrection復活 might威力 succeed成功 for the body身體,
393
1072000
2000
身體或許可以復活
18:09
but not for the mind心神.
394
1074000
2000
但思想卻沒有。
18:11
On the other hand, if the connectome連接組 is still intact完整,
395
1076000
3000
另一種可能是﹐如果聯結體都還在
18:14
we cannot不能 ridicule嘲笑 the claims索賠 of cryonics人體冷凍 so easily容易.
396
1079000
3000
我們便不能隨便嘲笑這些死後冷凍的想法
18:20
I've described描述 a quest尋求
397
1085000
2000
我描繪了一個旅程
18:22
that begins開始 in the world世界 of the very small,
398
1087000
3000
從一個很小的世界開始
18:25
and propels推動 us to the world世界 of the far future未來.
399
1090000
3000
一直到很遠的未來
18:28
ConnectomesConnectomes will mark標記 a turning車削 point in human人的 history歷史.
400
1093000
3000
連接體將會成為人類史上的轉捩點
18:32
As we evolved進化 from our ape-like類人猿 ancestors祖先
401
1097000
2000
在進化的過程中﹐我們較大的腦
18:34
on the African非洲人 savanna稀樹草原,
402
1099000
2000
是唯一讓我們和非洲老祖宗猩猩
18:36
what distinguished傑出的 us was our larger brains大腦.
403
1101000
3000
唯一不同之處。
18:40
We have used our brains大腦 to fashion時尚
404
1105000
2000
我們用我們的大腦創造了
18:42
ever more amazing驚人 technologies技術.
405
1107000
3000
更多令人驚異的科技
18:45
Eventually終於, these technologies技術 will become成為 so powerful強大
406
1110000
3000
總有一天﹐這些科技會強大到
18:48
that we will use them to know ourselves我們自己
407
1113000
3000
讓我們可以使用它們來了解自己
18:51
by deconstructing解構 and reconstructing重建
408
1116000
3000
用解構﹐再重新建構
18:54
our own擁有 brains大腦.
409
1119000
3000
我們大腦的方法。
18:57
I believe that this voyage航程 of self-discovery自我發現
410
1122000
3000
我相信這個自我尋找的旅程
19:00
is not just for scientists科學家們,
411
1125000
3000
不只屬於科學家
19:03
but for all of us.
412
1128000
2000
也是屬於我們所有人的。
19:05
And I'm grateful感激 for the opportunity機會 to share分享 this voyage航程 with you today今天.
413
1130000
3000
我很榮幸今天有這個機會,能和你分享這個旅程
19:08
Thank you.
414
1133000
2000
謝謝各位
19:10
(Applause掌聲)
415
1135000
8000
(掌聲)
Translated by Coco Shen
Reviewed by Adrienne Lin

▲Back to top

ABOUT THE SPEAKER
Sebastian Seung - Computational neuroscientist
Sebastian Seung is a leader in the new field of connectomics, currently the hottest space in neuroscience, which studies, in once-impossible detail, the wiring of the brain.

Why you should listen

In the brain, neurons are connected into a complex network. Sebastian Seung and his lab at MIT are inventing technologies for identifying and describing the connectome, the totality of connections between the brain's neurons -- think of it as the wiring diagram of the brain. We possess our entire genome at birth, but things like memories are not "stored" in the genome; they are acquired through life and accumulated in the brain. Seung's hypothesis is that "we are our connectome," that the connections among neurons is where memories and experiences get stored.

Seung and his collaborators, including Winfried Denk at the Max Planck Institute and Jeff Lichtman at Harvard University, are working on a plan to thin-slice a brain (probably starting with a mouse brain) and trace, from slice to slice, each neural pathway, exposing the wiring diagram of the brain and creating a powerful new way to visualize the workings of the mind. They're not the first to attempt something like this -- Sydney Brenner won a Nobel for mapping all the 7,000 connections in the nervous system of a tiny worm, C. elegans. But that took his team a dozen years, and the worm only had 302 nerve cells. One of Seung's breakthroughs is in using advanced imagining and AI to handle the crushing amount of data that a mouse brain will yield and turn it into richly visual maps that show the passageways of thought and sensation.

More profile about the speaker
Sebastian Seung | Speaker | TED.com