Are Insect Brains the Secret to Great AI? | Frances S. Chance | TED

70,399 views ใƒป 2023-01-02

TED


์•„๋ž˜ ์˜๋ฌธ์ž๋ง‰์„ ๋”๋ธ”ํด๋ฆญํ•˜์‹œ๋ฉด ์˜์ƒ์ด ์žฌ์ƒ๋ฉ๋‹ˆ๋‹ค.

๋ฒˆ์—ญ: Hyeryung Kim ๊ฒ€ํ† : DK Kim
00:05
Creating intelligence on a computer.
0
5210
2377
์ปดํ“จํ„ฐ์— ์ง€๋Šฅ์„ ์ฐฝ์กฐํ•˜๋Š” ๊ฒƒ.
00:08
This has been the Holy Grail for artificial intelligence
1
8129
2837
์ด๊ฒƒ์€ ์˜ค๋žœ ์‹œ๊ฐ„ ๋™์•ˆ ์ธ๊ณต ์ง€๋Šฅ์˜ ๊ฟˆ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
00:11
for quite some time.
2
11007
1377
00:12
But how do we get there?
3
12968
1668
๊ทธ๋Ÿฐ๋ฐ ์ด ๊ฟˆ์„ ์–ด๋–ป๊ฒŒ ์ด๋ฃฐ๊นŒ์š”?
00:15
So we view ourselves as highly intelligent beings.
4
15679
3128
์ธ๊ฐ„์€ ์ž์‹ ์„ ๊ณ ๋„์˜ ์ง€์  ์กด์žฌ๋กœ ์—ฌ๊น๋‹ˆ๋‹ค.
00:18
So it's logical to study our own brains,
5
18848
2795
๊ทธ๋ž˜์„œ ๋‹น์—ฐํžˆ ์ธ์‹์˜ ๋ฐ”ํƒ•์ธ ์ธ๊ฐ„์˜ ๋‡Œ๋ฅผ ์—ฐ๊ตฌํ•ด์„œ
00:21
the substrate of our cognition, for creating artificial intelligence.
6
21685
4212
์ธ๊ณต ์ง€๋Šฅ์„ ๋งŒ๋“ค๋ ค ํ•˜์ฃ .
00:27
Imagine if we could replicate how our own brains work on a computer.
7
27148
4130
์šฐ๋ฆฌ ๋‡Œ๊ฐ€ ์ผํ•˜๋Š” ๋ฐฉ์‹์„ ์ปดํ“จํ„ฐ ์•ˆ์— ๊ทธ๋Œ€๋กœ ๋ณต์ œํ•œ๋‹ค๊ณ  ์ƒ๊ฐํ•ด ๋ณด์„ธ์š”.
00:32
But now consider the journey that would be required.
8
32612
2503
๊ทธ๋Ÿผ ๋ญ๊ฐ€ ํ•„์š”ํ• ์ง€ ํ•œ๋ฒˆ ์ƒ๊ฐํ•ด ๋ด…์‹œ๋‹ค.
00:37
The human brain contains 86 billion neurons.
9
37367
3712
์ธ๊ฐ„์˜ ๋‡Œ์—๋Š” ์‹ ๊ฒฝ ์„ธํฌ๊ฐ€ 860์–ต ๊ฐœ ์žˆ์Šต๋‹ˆ๋‹ค.
00:42
Each is constantly communicating with thousands of others,
10
42289
3628
๊ฐ๊ฐ์˜ ์‹ ๊ฒฝ ์„ธํฌ๋Š” ์ง€์†์ ์œผ๋กœ ๋‹ค๋ฅธ ์‹ ๊ฒฝ ์„ธํฌ ์ˆ˜์ฒœ ๊ฐœ์™€ ์†Œํ†ตํ•˜๋ฉฐ
00:45
and each has individual characteristics of its own.
11
45959
3045
๊ฐ ์‹ ๊ฒฝ ์„ธํฌ๋งˆ๋‹ค ๋…ํŠนํ•œ ํŠน์ง•์ด ์žˆ์Šต๋‹ˆ๋‹ค.
00:49
Capturing the human brain on a computer
12
49879
2378
์ธ๊ฐ„์˜ ๋‡Œ๋ฅผ ํ†ต์งธ๋กœ ์ปดํ“จํ„ฐ ์•ˆ์— ๋„ฃ๋Š” ์ผ์€
00:52
may simply be too big and too complex a problem
13
52299
3670
ํ˜„์žฌ ์šฐ๋ฆฌ๊ฐ€ ๊ฐ€์ง„ ๊ธฐ์ˆ ๊ณผ ์ง€์‹๋งŒ์œผ๋กœ ํ•ด๋‚ด๊ธฐ์—”
๋„ˆ๋ฌด ๋ณต์žกํ•˜๊ณ  ๋จผ ์ด์•ผ๊ธฐ์ฒ˜๋Ÿผ ๋“ค๋ฆด์ง€๋„ ๋ชจ๋ฆ…๋‹ˆ๋‹ค.
00:56
to tackle with the technology and the knowledge that we have today.
14
56011
3795
01:01
I believe that we can capture a brain on a computer,
15
61266
2878
์ €๋Š” ๋‡Œ๋ฅผ ์ปดํ“จํ„ฐ ์•ˆ์— ๋„ฃ์„ ์ˆ˜ ์žˆ๋‹ค๊ณ  ๋ฏฟ์ง€๋งŒ
01:04
but we have to start smaller.
16
64144
2461
์ž‘์€ ๊ฒƒ์—์„œ๋ถ€ํ„ฐ ์‹œ์ž‘ํ•ด์•ผ ํ•˜์ฃ .
01:07
Much smaller.
17
67230
1168
์•„์ฃผ ์ž‘์€ ๊ฒƒ์—์„œ๋ถ€ํ„ฐ์š”.
01:10
These insects have three of the most fascinating brains in the world to me.
18
70734
4713
์—ฌ๊ธฐ ์ด ์„ธ ๊ณค์ถฉ์˜ ๋‡Œ๋Š” ๊ทธ์•ผ๋ง๋กœ ์„ธ์ƒ์—์„œ ๊ฐ€์žฅ ๊ฒฝ์ด๋กญ์Šต๋‹ˆ๋‹ค.
01:16
While they do not possess human-level intelligence,
19
76448
2836
์ด ๊ณค์ถฉ๋“ค์˜ ์ง€๋Šฅ์ด ์ธ๊ฐ„ ์ˆ˜์ค€์€ ์•„๋‹ˆ์ง€๋งŒ
01:19
each is remarkable at a particular task.
20
79326
3169
ํŠน์ •ํ•œ ์ž‘์—…์—๋Š” ํ›Œ๋ฅญํ•ฉ๋‹ˆ๋‹ค.
01:22
Think of them as highly trained specialists.
21
82954
2544
๊ณ ๋„๋กœ ํ›ˆ๋ จ๋œ ์ „๋ฌธ๊ฐ€์™€ ๊ฐ™์ฃ .
01:26
African dung beetles are really good at rolling large balls in straight lines.
22
86750
5088
์•„ํ”„๋ฆฌ์นด ์‡ ๋˜ฅ๊ตฌ๋ฆฌ๋Š” โ€˜๋˜‘๋ฐ”๋กœ ๊ณต ๊ตด๋ฆฌ๊ธฐโ€™ ์ „๋ฌธ๊ฐ€์ž…๋‹ˆ๋‹ค.
01:31
(Laughter)
23
91880
1710
(์›ƒ์Œ)
01:33
Now, if you've ever made a snowman,
24
93632
1710
๋ˆˆ์‚ฌ๋žŒ์„ ๋งŒ๋“ค์–ด๋ณด์…จ๋‹ค๋ฉด ์•„์‹œ๊ฒ ์ง€๋งŒ
01:35
you know that rolling a large ball is not easy.
25
95342
2460
ํฐ ๊ณต์„ ๊ตด๋ฆฌ๋Š” ์ผ์€ ์‰ฝ์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
01:39
Now picture trying to make that snowman
26
99095
2211
๋ˆˆ์‚ฌ๋žŒ์„ ๋งŒ๋“ค ๋•Œ
01:41
when the ball of snow is as big as you are
27
101348
2377
์ž์‹ ๋งŒํผ ํฐ ๋ˆˆ๋ฉ์ด๋ฅผ ๊ตด๋ฆฐ๋‹ค๊ณ  ์ƒ๊ฐํ•ด ๋ณด์„ธ์š”.
01:43
and you're standing on your head.
28
103725
1835
๊ทธ ์™€์ค‘์— ๋ฌผ๊ตฌ๋‚˜๋ฌด๊นŒ์ง€ ์„ ๋‹ค๋ฉด์š”?
01:45
(Laughter)
29
105602
1168
(์›ƒ์Œ)
01:47
Sahara desert ants are navigation specialists.
30
107228
3796
์‚ฌํ•˜๋ผ ์‚ฌ๋ง‰ ๊ฐœ๋ฏธ๋Š” โ€˜๊ธธ ์ฐพ๊ธฐโ€™ ์ „๋ฌธ๊ฐ€์ž…๋‹ˆ๋‹ค.
01:51
They might have to wander a considerable distance to forage for food.
31
111775
3753
๊ฐœ๋ฏธ๋“ค์€ ๋จน์ด๋ฅผ ์ฐพ์œผ๋Ÿฌ ์ƒ๋‹นํžˆ ๋ฉ€๋ฆฌ๊นŒ์ง€๋„ ๋Œ์•„๋‹ค๋…€์•ผ ํ•˜์ฃ .
01:55
But once they do find sustenance,
32
115862
1752
ํ•˜์ง€๋งŒ ์ผ๋‹จ ๋จน์ด๋ฅผ ์ฐพ์œผ๋ฉด ์ง‘๊นŒ์ง€ ์ตœ๋‹จ ๊ฑฐ๋ฆฌ๋ฅผ ์••๋‹ˆ๋‹ค.
01:57
they know how to calculate the straightest path home.
33
117656
2585
02:01
And the dragonfly is a hunting specialist.
34
121910
2919
๊ทธ๋ฆฌ๊ณ  ์ž ์ž๋ฆฌ๋Š” โ€˜์‚ฌ๋ƒฅโ€™ ์ „๋ฌธ๊ฐ€์ž…๋‹ˆ๋‹ค.
02:05
In the wild, dragonflies capture approximately 95 percent
35
125205
3587
์ž์—ฐ์—์„œ ์ž ์ž๋ฆฌ์˜ ์‚ฌ๋ƒฅ ์„ฑ๊ณต๋ฅ ์€ ์•ฝ 95ํผ์„ผํŠธ์— ์ด๋ฅธ๋‹ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค.
02:08
of the prey they choose to go after.
36
128833
1752
02:11
These insects are so good at their specialties
37
131336
3003
์ด ๊ณค์ถฉ๋“ค์€ ๊ฐ์ž์˜ ๋ถ„์•ผ์—์„œ ์ตœ๊ณ  ์ „๋ฌธ๊ฐ€๋ผ์„œ
02:14
that neuroscientists such as myself study them as model systems
38
134381
4337
์ € ๊ฐ™์€ ์‹ ๊ฒฝ ๊ณผํ•™์ž๋“ค์€ ์ด ๊ณค์ถฉ๋“ค์„ ๋ชจ๋ธ ์‚ผ์•„ ์—ฐ๊ตฌํ•ฉ๋‹ˆ๋‹ค.
02:18
to understand how animal nervous systems solve particular problems.
39
138760
4046
๋™๋ฌผ ์‹ ๊ฒฝ๊ณ„๊ฐ€ ํŠน์ •ํ•œ ๊ณผ์ œ๋ฅผ ํ•ด๋‚ด๋Š” ๋น„๊ฒฐ์„ ์—ฌ๊ธฐ์„œ ์ฐพ์„ ์ˆ˜ ์žˆ์ฃ .
02:23
And in my own research, I study brains to bring these solutions,
40
143973
3879
์ €๋Š” ๋‡Œ๋ฅผ ์—ฐ๊ตฌํ•ด์„œ ์ด๋Ÿฐ ๋น„๊ฒฐ๋“ค์„ ์ปดํ“จํ„ฐ์— ์˜ฎ๊น๋‹ˆ๋‹ค.
02:27
the best that biology has to offer, to computers.
41
147894
3337
์ƒ๋ฌผํ•™์ด ์ปดํ“จํ„ฐ์— ์ค„ ์ˆ˜ ์žˆ๋Š” ์ตœ๊ณ ์˜ ์„ ๋ฌผ์ด์ฃ .
02:31
So consider the dragonfly brain.
42
151272
1877
์ž ์ž๋ฆฌ ๋‡Œ๋ฅผ ์ƒ๊ฐํ•ด ๋ด…์‹œ๋‹ค.
02:33
It has only on the order of one million neurons.
43
153900
3003
์ž ์ž๋ฆฌ ๋‡Œ ์•ˆ์˜ ์‹ ๊ฒฝ ์„ธํฌ๋Š” ๊ณ ์ž‘ 100๋งŒ ๊ฐœ ์ •๋„์ž…๋‹ˆ๋‹ค.
02:37
Now, it's still not easy to unravel a circuit of even one million neurons.
44
157570
4672
ํ•˜์ง€๋งŒ ์ด๋ ‡๊ฒŒ ์ ์€ ์‹ ๊ฒฝ ์„ธํฌ์กฐ์ฐจ๋„ ๊ทธ ํšŒ๋กœ๋ฅผ ํ’€์–ด๋‚ด๊ธฐ๋ž€ ์‰ฝ์ง€ ์•Š์ฃ .
02:42
But given the choice
45
162701
1167
ํ•˜์ง€๋งŒ ๋งŒ์•ฝ ์—ฌ๋Ÿฌ๋ถ„์—๊ฒŒ
02:43
between trying to tease apart the one-million-neuron brain
46
163910
3003
์‹ ๊ฒฝ ์„ธํฌ๊ฐ€ 100๋งŒ ๊ฐœ ์žˆ๋Š” ๋‡Œ์™€ 860์–ต ๊ฐœ ์žˆ๋Š” ๋‡Œ๋ฅผ
02:46
versus the 86-billion-neuron brain,
47
166955
2586
์กฐ๊ฐ์กฐ๊ฐ ํ•ด๋ถ€ํ•˜๋ผ๊ณ  ํ•œ๋‹ค๋ฉด
02:49
which would you choose to try first?
48
169541
2169
์–ด๋–ค ๊ฑธ ๋จผ์ € ํ•ด๋ณด์‹œ๊ฒ ์Šต๋‹ˆ๊นŒ?
02:51
(Laughter)
49
171710
1167
(์›ƒ์Œ)
02:53
When studying these smaller insect brains,
50
173837
2586
๊ณค์ถฉ์˜ ๋” ์ž‘์€ ๋‡Œ๋ฅผ ์—ฐ๊ตฌํ•˜๋Š” ๊ฒƒ์€
02:56
the immediate goal is not human intelligence.
51
176464
2461
์ธ๊ฐ„ ์ง€๋Šฅ์„ ์•„๋Š” ๊ฒŒ ์ง์ ‘ ๋ชฉํ‘œ๊ฐ€ ์•„๋‹™๋‹ˆ๋‹ค.
02:59
We study these brains for what the insects do well.
52
179551
3461
๊ณค์ถฉ์˜ ์ฃผํŠน๊ธฐ๋ฅผ ์•Œ์•„๋ณด๋ ค๋Š” ๊ฑฐ์ฃ .
03:03
And in the case of the dragonfly, that's interception.
53
183805
2878
์ž ์ž๋ฆฌ์˜ ์ฃผํŠน๊ธฐ๋Š” โ€˜๊ฐ€๋กœ์ฑ„๊ธฐโ€™์ž…๋‹ˆ๋‹ค.
03:07
So when dragonflies are hunting,
54
187559
1543
์ž ์ž๋ฆฌ๊ฐ€ ์‚ฌ๋ƒฅ์„ ํ•  ๋•Œ
03:09
they do more than just fly straight at the prey.
55
189144
2502
๋จน์ž‡๊ฐ์„ ํ–ฅํ•ด ์ผ์ง์„ ์œผ๋กœ ๋‹จ์ˆœํžˆ ๋Œ์ง„ํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹™๋‹ˆ๋‹ค.
03:12
They fly in such a way that they will intercept it.
56
192021
2503
๋จน์ž‡๊ฐ์„ ๊ฐ€๋กœ์ฑŒ ์ˆ˜ ์žˆ๋„๋ก ๋‚˜๋Š” ๊ฒ๋‹ˆ๋‹ค.
03:14
They aim for where the prey is going to be.
57
194566
2460
๋จน์ž‡๊ฐ์ด ๊ฐˆ ์ง€์ ์„ ๊ฒจ๋ƒฅํ•ด ๋‚ ์•„๊ฐ€์ฃ .
03:17
Much like a soccer player, running to intercept a pass.
58
197402
3003
ํŒจ์Šค๋ฅผ ์ฐจ๋‹จํ•˜๋Ÿฌ ๋‹ฌ๋ ค๊ฐ€๋Š” ์ถ•๊ตฌ ์„ ์ˆ˜์— ํ›จ์”ฌ ๋” ๊ฐ€๊น์ฃ .
03:21
To do this correctly,
59
201865
1459
์ด๊ฑธ ์ •ํ™•ํžˆ ํ•ด๋‚ด๊ธฐ ์œ„ํ•ด
03:23
dragonflies need to perform what is known as a coordinate transformation,
60
203366
3879
์ž ์ž๋ฆฌ๋Š” โ€˜์ขŒํ‘œ ๋ณ€ํ™˜โ€™์œผ๋กœ ์•Œ๋ ค์ง„ ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.
03:27
going from the eyeโ€™s frame of reference, or what the dragonfly sees,
61
207245
3504
์‹œ๊ฐ ๊ธฐ์ค€, ์ฆ‰, ๋ˆˆ์œผ๋กœ ๋ณด๋Š” ๊ฒƒ์„
03:30
to the body's frame of reference,
62
210790
1585
๋ชธ์ฒด ๊ธฐ์ค€, ์ฆ‰, ๊ฐ€๋กœ์ฑ„๊ธฐ๋ฅผ ํ•˜๋Ÿฌ ๋ชธ์„ ์›€์ง์ด๋Š” ๊ฒƒ์œผ๋กœ ๋ฐ”๊ฟ”์•ผ ํ•ฉ๋‹ˆ๋‹ค.
03:32
or how the dragonfly needs to turn its body to intercept.
63
212375
2836
03:36
Coordinate transformations are a basic calculation
64
216004
3044
์ด๋Ÿฐ ์ขŒํ‘œ ๋ณ€ํ™˜ ๊ธฐ์ˆ ์€
๋™๋ฌผ์ด ์„ธ์ƒ๊ณผ ์†Œํ†ตํ•˜๊ธฐ ์œ„ํ•ด ๊ผญ ํ•„์š”ํ•œ ๊ธฐ๋ณธ ๊ธฐ์ˆ ์ž…๋‹ˆ๋‹ค.
03:39
that animals need to perform to interact with the world.
65
219048
3713
03:43
We do them instinctively every time we reach for something.
66
223261
3086
์šฐ๋ฆฐ ๋ฌด์–ธ๊ฐ€๋ฅผ ์žก์„ ๋•Œ ๋ณธ๋Šฅ์ ์œผ๋กœ ์ด ๊ธฐ์ˆ ์„ ์”๋‹ˆ๋‹ค.
03:47
When I reach for an object straight in front of me,
67
227098
2878
์ œ๊ฐ€ ๋ฐ”๋กœ ์•ž์— ์žˆ๋Š” ๋ฌผ๊ฑด์„ ์žก์œผ๋ ค๊ณ  ํ•  ๋•Œ
03:50
my arm takes a very different trajectory than if I turn my head,
68
230018
3545
์ œ ํŒ”์ด ์›€์ง์ด๋Š” ๊ถค์ ์€
๊ฐ™์€ ๋ฌผ์ฒด๊ฐ€ ์˜†์œผ๋กœ ๋–จ์–ด์ ธ ์žˆ์–ด์„œ
03:53
look at that same object when it is off to one side
69
233605
2460
๋จธ๋ฆฌ๋ฅผ ๋Œ๋ฆฌ๊ณ  ์žก์œผ๋ ค๊ณ  ํ•  ๋•Œ์™€๋Š” ์ „ํ˜€ ๋‹ค๋ฆ…๋‹ˆ๋‹ค.
03:56
and reach for it there.
70
236107
1335
03:58
In both cases, my eyes see the same image of that object,
71
238109
3712
๋‘ ๊ฒฝ์šฐ ๋ชจ๋‘, ๋ˆˆ์œผ๋กœ๋Š” ๊ฐ™์€ ๋ฌผ์ฒด๋ฅผ ๋ณด์ง€๋งŒ
04:01
but my brain is sending my arm on a very different trajectory
72
241821
3712
๋‡Œ๋Š” ํŒ”์„ ์ „ํ˜€ ๋‹ค๋ฅธ ๊ถค๋„๋กœ ์›€์ง์ด๊ฒŒ ํ•˜์ฃ .
04:05
based on the position of my neck.
73
245575
1919
๋ชฉ์˜ ์œ„์น˜์— ๋”ฐ๋ผ์„œ์š”.
04:12
And dragonflies are fast.
74
252624
1960
๊ทธ๋ฆฌ๊ณ  ์ž ์ž๋ฆฌ๋Š” ๋งค์šฐ ๋น ๋ฆ…๋‹ˆ๋‹ค.
04:15
This means they calculate fast.
75
255293
2085
๊ณ„์‚ฐ์ด ๋น ๋ฅด๋‹ค๋Š” ๋œป์ด์ฃ .
04:18
The latency, or the time it takes for a dragonfly to respond
76
258046
4004
์ž ์ž๋ฆฌ์˜ ๋ฐ˜์‘ ์‹œ๊ฐ„,
์ฆ‰, ๋จน์ž‡๊ฐ์˜ ์›€์ง์ž„์„ ๋ณด๊ณ  ๋ฐ˜์‘ํ•˜๋Š” ๋ฐ ๊ฑธ๋ฆฌ๋Š” ์‹œ๊ฐ„์€
04:22
once it sees the prey turn,
77
262091
1752
04:23
is about 50 milliseconds.
78
263885
1960
๋Œ€๋žต 0.05์ดˆ ์ •๋„์ž…๋‹ˆ๋‹ค.
04:27
This latency is remarkable.
79
267180
2044
์—„์ฒญ๋‚˜๊ฒŒ ์งง์€ ์‹œ๊ฐ„์ด์ฃ .
04:30
For one thing, it's only half the time of a human eye blink.
80
270016
3378
์ธ๊ฐ„์ด ๋ˆˆ์„ ํ•œ๋ฒˆ ๊นœ๋นก์ด๋Š” ๋ฐ ๊ฑธ๋ฆฌ๋Š” ์‹œ๊ฐ„์˜ ์ ˆ๋ฐ˜๋ฐ–์— ์•ˆ ๋ฉ๋‹ˆ๋‹ค.
๊ทธ๋Ÿฐ๋ฐ ์ด๊ฑธ ๋ฐ”๊ฟ” ๋งํ•˜๋ฉด,
04:34
But for another thing,
81
274020
1668
04:35
it suggests that dragonflies capture how to intercept
82
275688
3003
์ž ์ž๋ฆฌ๊ฐ€ ๋จน์ž‡๊ฐ์˜ ๊ฒฝ๋กœ๋ฅผ ๊ฐ€๋กœ์ฑ„๋Š” ๋ฐฉ๋ฒ•์„ ๊ณ„์‚ฐํ•˜๋Š” ๋ฐ
04:38
in only relatively or surprisingly few computational steps.
83
278733
4755
์ƒ๋Œ€์ ์œผ๋กœ ๋˜๋Š” ๋†€๋ž„ ์ •๋„๋กœ ์ ์€ ๋‹จ๊ณ„๋ฅผ ๊ฑฐ์นœ๋‹ค๋Š” ๋œป์ด๊ธฐ๋„ ํ•˜์ฃ .
04:44
So in the brain,
84
284364
1376
๋‡Œ์—์„œ ๊ณ„์‚ฐ ๋‹จ๊ณ„ ํ•˜๋‚˜๋Š” ์‹ ๊ฒฝ ์„ธํฌ ํ•œ ๊ฐœ๊ฐ€ ๋‹ด๋‹นํ•˜๊ฑฐ๋‚˜
04:45
a computational step is a single neuron
85
285782
2878
04:48
or a layer of neurons working in parallel.
86
288660
2460
ํ˜น์€ ๋™์‹œ์— ์ž‘๋™ํ•˜๋Š” ์‹ ๊ฒฝ ์„ธํฌ ํ•œ ์ธต์ด ๋‹ด๋‹นํ•ฉ๋‹ˆ๋‹ค.
04:51
It takes a single neuron about 10 milliseconds
87
291996
3087
์‹ ๊ฒฝ ์„ธํฌ ํ•˜๋‚˜๊ฐ€ ๋“ค์–ด์˜ค๋Š” ๋ชจ๋“  ์ •๋ณด๋ฅผ ๋ชจ์•„์„œ
04:55
to add up all its inputs and respond.
88
295124
2336
๋ฐ˜์‘ํ•˜๊ธฐ๊นŒ์ง€๋Š” ๋Œ€๋žต 0.01์ดˆ ์ •๋„ ๊ฑธ๋ฆฝ๋‹ˆ๋‹ค.
04:58
The 50-millisecond response time
89
298169
2336
๋ฐ˜์‘ ์‹œ๊ฐ„์ด 0.05์ดˆ๋ผ๋Š” ๊ฒƒ์€
05:00
means that once the dragonfly sees its prey turn,
90
300547
3503
์ž ์ž๋ฆฌ๊ฐ€ ๋จน์ž‡๊ฐ์ด ๊ฒฝ๋กœ๋ฅผ ๋ฐ”๊พธ๋Š” ๊ฒƒ์„ ํฌ์ฐฉํ•˜๊ณ  ๋‚˜์„œ
05:04
there's only time for maybe four of these computational steps
91
304092
3336
๋‹จ์ง€ ๋„ค ๋‹จ๊ณ„ ์ •๋„๋ฅผ ๊ฑฐ์ณ์„œ,
05:07
or four layers of neurons, working in sequence, one after the other,
92
307470
3837
๋˜๋Š” ์‹ ๊ฒฝ ์„ธํฌ ๋„ค ์ธต์ด ํ•œ ์ธต์”ฉ ์ฐจ๋ก€๋กœ ๋ฐ˜์‘ํ•ด์„œ
05:11
to calculate how the dragonfly needs to turn.
93
311349
2377
์–ด๋–ป๊ฒŒ ๋ชธ์„ ๋Œ๋ ค์•ผ ํ• ์ง€ ๊ณ„์‚ฐํ•œ๋‹ค๋Š” ๋ง์ž…๋‹ˆ๋‹ค.
05:14
In other words, if I want to study
94
314811
2085
๋”ฐ๋ผ์„œ ๋งŒ์•ฝ ์ œ๊ฐ€
05:16
how the dragonfly does coordinate transformations,
95
316896
4213
์ž ์ž๋ฆฌ์˜ ์ขŒํ‘œ ๋ณ€ํ™˜์„ ์—ฐ๊ตฌํ•˜๊ณ  ์‹ถ๋‹ค๋ฉด
05:21
the neural circuit that I need to understand,
96
321109
2753
์ œ๊ฐ€ ์•Œ์•„์•ผ ํ•  ์‹ ๊ฒฝ ํšŒ๋กœ, ์—ฐ๊ตฌํ•ด์•ผ ํ•  ์‹ ๊ฒฝ ํšŒ๋กœ๋Š”
05:23
the neural circuit that I need to study,
97
323903
2252
05:26
can have at most four layers of neurons.
98
326197
2670
๋งŽ์•„์•ผ ๋„ค ์ธต์งœ๋ฆฌ ์‹ ๊ฒฝ ์„ธํฌ์ธต์ด๋ผ๋Š” ๊ฒ๋‹ˆ๋‹ค.
05:29
Each layer may have many neurons,
99
329784
2377
์ธต๋งˆ๋‹ค ์‹ ๊ฒฝ ์„ธํฌ๊ฐ€ ์—ฌ๋Ÿฌ ๊ฐœ์”ฉ ์žˆ์–ด๋„ ์•„์ฃผ ์ž‘์€ ์‹ ๊ฒฝ ํšŒ๋กœ์— ๋ถˆ๊ณผํ•˜์ฃ .
05:32
but this is a small neural circuit.
100
332203
2127
05:34
Small enough that we can identify it
101
334789
2002
ํ˜„์žฌ ์žˆ๋Š” ๋„๊ตฌ๋ฅผ ์ด์šฉํ•ด
05:36
and study it with the tools that are available today.
102
336833
2669
๊ทœ๋ช…ํ•˜๊ณ  ์—ฐ๊ตฌํ•  ์ˆ˜ ์žˆ์„ ๋งŒํผ ์ž‘์€ ํšŒ๋กœ์ž…๋‹ˆ๋‹ค.
05:40
And this is what I'm trying to do.
103
340670
1835
์ œ๊ฐ€ ํ•˜๋ ค๋Š” ์ผ์€ ์ด๋Ÿฐ ๊ฒ๋‹ˆ๋‹ค.
05:43
I have built a model of what I believe is the neural circuit
104
343298
3044
์ž ์ž๋ฆฌ๊ฐ€ ์–ด๋–ป๊ฒŒ ์›€์ง์ผ์ง€ ์ถ”์ธกํ•  ์ˆ˜ ์žˆ๋Š”
05:46
that calculates how the dragonfly should turn.
105
346384
2544
์‹ ๊ฒฝ ํšŒ๋กœ ๋ชจํ˜•์„ ๋งŒ๋“ค์—ˆ์ฃ .
05:49
And here is the cool result.
106
349596
1584
๊ฒฐ๊ณผ๋Š” ๊ฝค ์„ฑ๊ณต์ ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
05:51
In the model,
107
351222
1627
์ด ์‹คํ—˜ ๋ชจํ˜•์—์„œ,
05:52
dragonflies do coordinate transformations in only one computational step,
108
352891
4713
์ž ์ž๋ฆฌ๋Š” ํ•œ ๋‹จ๊ณ„๋งŒ์„ ๊ฑฐ์ณ ์ขŒํ‘œ ๋ณ€ํ™˜์„ ํ•ฉ๋‹ˆ๋‹ค.
05:57
one layer of neurons.
109
357645
1627
์‹ ๊ฒฝ ์„ธํฌ์ธต ํ•˜๋‚˜์ž…๋‹ˆ๋‹ค.
05:59
This is something we can test and understand.
110
359898
2877
์‹คํ—˜ํ•˜๊ณ  ์—ฐ๊ตฌํ•  ์ˆ˜ ์žˆ๋Š” ๋Œ€์ƒ์ž…๋‹ˆ๋‹ค.
06:03
In a computer simulation,
111
363651
1502
์ปดํ“จํ„ฐ ๋ชจ์˜ ์‹คํ—˜์—์„œ๋Š”
06:05
I can predict the activities of individual neurons
112
365194
3170
์ž ์ž๋ฆฌ๊ฐ€ ์‚ฌ๋ƒฅํ•  ๋•Œ ๊ฐ ์‹ ๊ฒฝ ์„ธํฌ๊ฐ€ ์–ด๋–ป๊ฒŒ ํ™œ๋™ํ•˜๋Š”์ง€ ์˜ˆ์ธกํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
06:08
while the dragonfly is hunting.
113
368364
1669
06:11
For example, here I am predicting the action potentials, or the spikes,
114
371367
4338
์˜ˆ๋ฅผ ๋“ค์–ด ์ž ์ž๋ฆฌ๊ฐ€ ๋จน์ž‡๊ฐ์ด ์›€์ง์ด๋Š” ๊ฑธ ๋ณผ ๋•Œ
06:15
that are fired by one of these neurons
115
375747
1835
ํ•œ ์‹ ๊ฒฝ ์„ธํฌ์—์„œ ํ™œ์„ฑํ™”๋˜๋Š” ํ™œ๋™ ์ „์œ„ ์ฆ‰, ์ŠคํŒŒ์ดํฌ๋ฅผ ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค.
06:17
when the dragonfly sees the prey move.
116
377582
2627
06:22
To test the model, my collaborators and I
117
382545
2169
์ €์™€ ๋™๋ฃŒ๋“ค์€ ์ด ๋ชจํ˜•์„ ๊ฒ€์ฆํ•˜๊ธฐ ์œ„ํ•ด
06:24
are now comparing these predicted neural responses
118
384756
2836
๋ชจํ˜•์—์„œ ์˜ˆ์ธก๋œ ์‹ ๊ฒฝ ์„ธํฌ ๋ฐ˜์‘์„
06:27
with responses of neurons recorded in living dragonfly brains.
119
387592
3837
์‚ด์•„์žˆ๋Š” ์ž ์ž๋ฆฌ ๋‡Œ์—์„œ ๊ธฐ๋ก๋œ ์‹ ๊ฒฝ ์„ธํฌ ๋ฐ˜์‘๊ณผ ๋น„๊ตํ•ฉ๋‹ˆ๋‹ค.
06:33
These are ongoing experiments
120
393431
1668
ํ˜„์žฌ๋„ ์ง„ํ–‰๋˜๊ณ  ์žˆ๋Š” ์‹คํ—˜์ธ๋ฐ
06:35
in which we put living dragonflies in virtual reality.
121
395099
3796
์ž ์ž๋ฆฌ์—๊ฒŒ ๊ฐ€์ƒ ํ˜„์‹ค์„ ์ฒดํ—˜ํ•˜๊ฒŒ ํ•ด์ฃผ๋Š” ๊ฑฐ์ฃ .
06:40
(Laughter)
122
400188
1918
(์›ƒ์Œ)
06:42
Now, it's not practical to put VR goggles on a dragonfly.
123
402148
3754
์ž ์ž๋ฆฌํ•œํ…Œ ๊ฐ€์ƒ ํ˜„์‹ค ์•ˆ๊ฒฝ์„ ์”Œ์›Œ ์ค„ ์ˆ˜๋Š” ์—†์Šต๋‹ˆ๋‹ค.
06:47
So instead, we show movies of moving targets to the dragonfly,
124
407070
4588
๊ทธ๋ž˜์„œ ๋Œ€์‹ ์— ์›€์ง์ด๋Š” ๋จน์ž‡๊ฐ์˜ ์˜์ƒ์„ ์ž ์ž๋ฆฌ์—๊ฒŒ ๋ณด์—ฌ ์ค๋‹ˆ๋‹ค.
06:51
while an electrode records activity patterns of individual neurons
125
411658
3670
๊ทธ๋ฆฌ๊ณ  ๊ทธ๋™์•ˆ ์ž ์ž๋ฆฌ์˜ ๋‡Œ ์•ˆ์—์„œ ์›€์ง์ด๋Š”
๊ฐ ์‹ ๊ฒฝ ์„ธํฌ์˜ ํ™œ๋™ ์–‘์ƒ์„ ์ „๊ทน์„ ํ†ตํ•ด ๊ธฐ๋กํ•˜์ฃ .
06:55
in the brain.
126
415328
1251
06:58
Yeah, he likes the movies.
127
418331
1335
๋„ค, ์ž ์ž๋ฆฌ๋Š” ์˜ํ™”๋ฅผ ์ข‹์•„ํ•ด์š”.
07:01
If the responses that we record in the brain
128
421167
2628
์ด๋ ‡๊ฒŒ ๊ธฐ๋กํ•œ ๋ฐ˜์‘๋“ค์ด ๋ชจํ˜•์—์„œ ์˜ˆ์ธกํ–ˆ๋˜ ๋ฐ˜์‘๊ณผ ์ผ์น˜ํ•˜๋ฉด
07:03
match those predicted by the model,
129
423836
2128
07:06
we will have identified which neurons are responsible
130
426005
2711
์–ด๋–ค ์‹ ๊ฒฝ ์„ธํฌ๊ฐ€ ์ขŒํ‘œ ๋ณ€ํ™˜์— ๊ด€์—ฌํ•˜๋Š”์ง€ ์•Œ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
07:08
for coordinate transformations.
131
428758
1543
07:11
The next step will be to understand the specifics
132
431010
2294
๋‹ค์Œ ๋‹จ๊ณ„๋Š” ์ด ์‹ ๊ฒฝ ์„ธํฌ๋“ค์ด ์›€์ง์ž„์„ ๊ณ„์‚ฐํ•  ๋•Œ
07:13
of how these neurons work together to do the calculation.
133
433346
3337
์–ด๋–ค ์‹์œผ๋กœ ํ˜‘๋ ฅํ•˜๋Š”์ง€ ์•Œ์•„๋ณด๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
07:16
But this is how we begin to understand how brains do basic
134
436724
3462
๊ทธ๋Ÿฐ๋ฐ ์ด๊ฑด ๋‡Œ๊ฐ€ ๊ธฐ์ดˆ์ ์ธ ๊ณ„์‚ฐ์„ ํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์ดํ•ดํ•˜๋Š” ์ฒซ๊ฑธ์Œ์ž…๋‹ˆ๋‹ค.
07:20
or primitive calculations.
135
440186
1627
07:22
Calculations that I regard as building blocks for more complex functions,
136
442188
4880
์ €๋Š” ์ด โ€˜๊ณ„์‚ฐโ€™์ด๋ผ๋Š” ๊ฒƒ์„ ๋ชฉํ‘œ๋ฌผ์„ ์žก์•„์ฑ„๋Š” ๊ฒƒ๋ฟ ์•„๋‹ˆ๋ผ
07:27
not only for interception but also for cognition.
137
447110
3086
์ธ์ง€์ฒ˜๋Ÿผ ๋” ๋ณต์žกํ•œ ๊ธฐ๋Šฅ์„ ์ˆ˜ํ–‰ํ•˜๋Š” ๊ธฐ๋ณธ ๊ตฌ์„ฑ ๋‹จ์œ„๋กœ ๋ด…๋‹ˆ๋‹ค.
07:32
The way that these neurons compute may be different from anything
138
452156
3170
์ด ์‹ ๊ฒฝ ์„ธํฌ๋“ค์ด ๊ณ„์‚ฐํ•˜๋Š” ๋ฐฉ๋ฒ•์€
ํ˜„์กดํ•˜๋Š” ์–ด๋–ค ์ปดํ“จํ„ฐ์—๋„ ์—†๋Š” ์ „ํ˜€ ์ƒˆ๋กœ์šด ๋ฐฉ๋ฒ•์ผ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
07:35
that exists on a computer today.
139
455368
1877
07:37
And the goal of this work is to do more than just write code
140
457870
3587
์ด ์ž‘์—…์˜ ๊ถ๊ทน์ ์ธ ๋ชฉํ‘œ๋Š”
์‹ ๊ฒฝ ์„ธํฌ์˜ ํ™œ๋™ ์–‘์ƒ์„ ๋ณต์ œํ•˜๋Š” ์ฝ”๋“œ๋ฅผ ์“ฐ๋Š” ๊ฒƒ ์ด์ƒ์ž…๋‹ˆ๋‹ค.
07:41
that replicates the activity patterns of neurons.
141
461457
2419
07:44
We aim to build a computer chip
142
464293
1794
์šฐ๋ฆฌ๋Š” ์ปดํ“จํ„ฐ ์นฉ์„ ๋งŒ๋“ค๊ณ ์ž ํ•ฉ๋‹ˆ๋‹ค.
07:46
that not only does the same things as biological brains
143
466129
2794
์ƒ๋ฌผํ•™์  ๋‡Œ์˜ ํ–‰๋™์„ ๊ทธ๋Œ€๋กœ ํ•  ๋ฟ ์•„๋‹ˆ๋ผ
07:48
but does them in the same way as biological brains.
144
468965
2711
์ƒ๋ฌผํ•™์  ๋‡Œ์™€ ๊ฐ™์€ ๋ฐฉ์‹์œผ๋กœ ์›€์ง์ด๋Š” ์นฉ์ด์ฃ .
07:52
This could lead to drones driven by computers
145
472885
3254
์ด๊ฒƒ์„ ์ปดํ“จํ„ฐ๋กœ ์กฐ์ ˆ๋˜๊ฒŒ ๋งŒ๋“ค๋ฉด
07:56
the same size of the dragonfly's brain
146
476139
2002
์ž ์ž๋ฆฌ ๋‡Œ์™€ ํฌ๊ธฐ๊ฐ€ ๊ฐ™๊ณ 
07:58
that captures some targets and avoid others.
147
478182
2837
๋‹ค๋ฅธ ๊ฒƒ์€ ํ”ผํ•˜๋ฉด์„œ ๋ชฉํ‘œ๋ฌผ๋งŒ ์ •ํ•™ํžˆ ์žก๋Š” ๋“œ๋ก ์ด ๋ฉ๋‹ˆ๋‹ค.
08:01
Personally, I'm hoping for a small army of these
148
481853
2586
๊ฐœ์ธ์ ์œผ๋กœ ์ด ์ž‘์€ ๋ฌด๋ฆฌ๋“ค์ด
08:04
to defend my backyard from mosquitoes in the summer.
149
484480
2461
์—ฌ๋ฆ„์— ์ €ํฌ ์ง‘ ๋’ท๋งˆ๋‹น์„ ๋ชจ๊ธฐ๋กœ๋ถ€ํ„ฐ ์ง€์ผœ์ฃผ๋ฉด ์ข‹๊ฒ ์Šต๋‹ˆ๋‹ค.
08:06
(Laughter)
150
486983
1460
(์›ƒ์Œ)
08:09
The GPS on your phone could be replaced by a new navigation device
151
489027
4170
ํœด๋Œ€์ „ํ™” GPS๊ฐ€ ์ƒˆ๋กœ์šด ๊ธธ์•ˆ๋‚ด ์žฅ์น˜๋กœ ๋ฐ”๋€” ์ˆ˜๋„ ์žˆ์„ ๊ฒ๋‹ˆ๋‹ค.
08:13
based on dung beetles or ants
152
493239
1710
์‡ ๋˜ฅ๊ตฌ๋ฆฌ๋‚˜ ๊ฐœ๋ฏธ์˜ ๋‡Œ๋ฅผ ์ฐธ๊ณ ํ•ด ๋งŒ๋“  ์žฅ์น˜๋กœ
08:14
that could guide you to the straight or the easy path home.
153
494991
3045
๋˜‘๋ฐ”๋กœ ์ง‘์œผ๋กœ ๊ฐ€๋Š” ๊ธธ, ํ˜น์€ ์ง‘์— ๊ฐ€๋Š” ์‰ฌ์šด ๊ธธ์„ ์•Œ๋ ค ์ฃผ๋Š” ๊ฑฐ์ฃ .
08:18
And what would the power requirements of these devices be like?
154
498953
3378
์ด๋Ÿฌํ•œ ์žฅ์น˜์— ํ•„์š”ํ•œ ์ „๋ ฅ๋Ÿ‰์€ ์–ผ๋งˆ๋‚˜ ๋ ๊นŒ์š”?
08:23
As small as it is --
155
503458
1501
๊ทธ ํฌ๊ธฐ๋งŒํผ ์ ๊ฒ ์ฃ , ์•„, ๊ทธ ํฌ๊ธฐ๋งŒํผ ํฌ๊ฒ ์ฃ .
08:25
Or, sorry -- as large as it is,
156
505001
1877
08:26
the human brain is estimated to have the same power requirements
157
506878
3628
์ธ๊ฐ„ ๋‘๋‡Œ์— ํ•„์š”ํ•œ ์ „๋ ฅ์€
08:30
as a 20-watt light bulb.
158
510548
2044
20์™€ํŠธ์งœ๋ฆฌ ์ „๊ตฌ ํ•˜๋‚˜๋ฅผ ์ผค ์–‘์ด๋ผ๊ณ  ํ•ฉ๋‹ˆ๋‹ค.
08:32
Imagine if all brain-inspired computers
159
512633
2211
๋‘๋‡Œ์—์„œ ๋”ฐ์˜จ ๊ธฐ๊ณ„๋“ค์ด ๋ชจ๋‘
08:34
had the same extremely low-power requirements.
160
514844
2669
์ด๋ ‡๊ฒŒ ๊ทนํžˆ ์ ์€ ์ „๋ ฅ๋งŒ์œผ๋กœ ์›€์ง์ธ๋‹ค๊ณ  ์ƒ๊ฐํ•ด ๋ณด์„ธ์š”.
08:38
Your smartphone or your smartwatch probably needs charging every day.
161
518431
3837
ํœด๋Œ€์ „ํ™”๋‚˜ ์Šค๋งˆํŠธ์›Œ์น˜๋Š” ๋งค์ผ ์ถฉ์ „ํ•ด์•ผ ํ•  ๊ฒ๋‹ˆ๋‹ค.
08:42
Your new brain-inspired device might only need charging every few months,
162
522268
3587
๋‘๋‡Œ๋ฅผ ๋ณธ๋œฌ ์ƒˆ๋กœ์šด ์žฅ์น˜๋Š”
๋ช‡ ๋‹ฌ ์‹ฌ์ง€์–ด ๋ช‡ ๋…„์— ํ•œ ๋ฒˆ๋งŒ ์ถฉ์ „ํ•˜๋ฉด ๋˜๋Š” ๊ฑฐ์˜ˆ์š”.
08:45
or maybe even every few years.
163
525897
1668
08:49
The famous physicist, Richard Feynman, once said,
164
529275
3670
์œ ๋ช…ํ•œ ๋ฌผ๋ฆฌํ•™์ž ๋ฆฌ์ฒ˜๋“œ ํŒŒ์ธ๋งŒ์€ ์ด๋Ÿฐ ๋ง์„ ํ–ˆ์Šต๋‹ˆ๋‹ค.
08:52
"What I cannot create, I do not understand."
165
532945
2878
โ€œ๋งŒ๋“ค ์ˆ˜ ์—†๋‹ค๋ฉด ์ดํ•ดํ•˜์ง€ ๋ชปํ•˜๋Š” ๊ฒƒ์ด๋‹ค.โ€
08:56
What I see in insect nervous systems
166
536783
2085
๊ณค์ถฉ ์‹ ๊ฒฝ๊ณ„์—์„œ ์ œ๊ฐ€ ๋ณด๋Š” ๊ฒƒ์€
08:58
is an opportunity to understand brains
167
538868
2169
๋‘๋‡Œ์™€ ๊ฐ™์€ ๋ฐฉ์‹์œผ๋กœ ์ผํ•˜๋Š” ์ปดํ“จํ„ฐ๋ฅผ ๋งŒ๋“ฆ์œผ๋กœ์จ
09:01
through the creation of computers that work as brains do.
168
541079
3003
๋‘๋‡Œ๋ฅผ ์ดํ•ดํ•  ์ˆ˜ ์žˆ๋Š” ๊ธฐํšŒ์ž…๋‹ˆ๋‹ค.
09:05
And creation of these computers will not just be for knowledge.
169
545083
3712
์ด๋Ÿฐ ์ปดํ“จํ„ฐ๋ฅผ ๋งŒ๋“ค์–ด์„œ ๋‹จ์ง€ ์ง€์‹๋งŒ ์–ป๋Š” ๊ฒŒ ์•„๋‹™๋‹ˆ๋‹ค.
09:08
There's potential for real impact on your devices, your vehicles,
170
548795
4254
๊ธฐ๊ณ„ ์žฅ์น˜, ๊ตํ†ต ์ˆ˜๋‹จ์— ์‹ค์งˆ์ ์œผ๋กœ ํฐ ์˜ํ–ฅ์„ ๋ฏธ์น  ์ˆ˜ ์žˆ์ฃ .
09:13
maybe even artificial intelligences.
171
553091
2210
์–ด์ฉŒ๋ฉด ์‹ฌ์ง€์–ด ์ธ๊ณต ์ง€๋Šฅ์—๊นŒ์ง€์š”.
09:16
So next time you see an insect,
172
556177
2419
๋‹ค์Œ์— ๊ณค์ถฉ์„ ๋ณด์‹œ๋ฉด
09:18
consider that these tiny brains can lead to remarkable computers.
173
558638
4045
๊ทธ ์กฐ๊ทธ๋งˆํ•œ ๋‡Œ๊ฐ€ ๋งค๋ ฅ์ ์ธ ์ปดํ“จํ„ฐ๊ฐ€ ๋  ์ˆ˜ ์žˆ๋‹ค๋Š” ์ ์„ ๊ธฐ์–ตํ•ด์ฃผ์„ธ์š”.
09:23
And think of the potential that they offer us for the future.
174
563518
3211
๋˜ํ•œ ๋ฏธ๋ž˜์˜ ์ž ์žฌ๋ ฅ๋„ ํ•œ ๋ฒˆ์”ฉ ์ƒ๊ฐํ•ด ์ฃผ์‹œ๊ธฐ๋ฅผ ๋ฐ”๋ž๋‹ˆ๋‹ค.
09:27
Thank you.
175
567271
1126
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.
09:28
(Applause)
176
568439
3379
(๋ฐ•์ˆ˜)
์ด ์›น์‚ฌ์ดํŠธ ์ •๋ณด

์ด ์‚ฌ์ดํŠธ๋Š” ์˜์–ด ํ•™์Šต์— ์œ ์šฉํ•œ YouTube ๋™์˜์ƒ์„ ์†Œ๊ฐœํ•ฉ๋‹ˆ๋‹ค. ์ „ ์„ธ๊ณ„ ์ตœ๊ณ ์˜ ์„ ์ƒ๋‹˜๋“ค์ด ๊ฐ€๋ฅด์น˜๋Š” ์˜์–ด ์ˆ˜์—…์„ ๋ณด๊ฒŒ ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ฐ ๋™์˜์ƒ ํŽ˜์ด์ง€์— ํ‘œ์‹œ๋˜๋Š” ์˜์–ด ์ž๋ง‰์„ ๋”๋ธ” ํด๋ฆญํ•˜๋ฉด ๊ทธ๊ณณ์—์„œ ๋™์˜์ƒ์ด ์žฌ์ƒ๋ฉ๋‹ˆ๋‹ค. ๋น„๋””์˜ค ์žฌ์ƒ์— ๋งž์ถฐ ์ž๋ง‰์ด ์Šคํฌ๋กค๋ฉ๋‹ˆ๋‹ค. ์˜๊ฒฌ์ด๋‚˜ ์š”์ฒญ์ด ์žˆ๋Š” ๊ฒฝ์šฐ ์ด ๋ฌธ์˜ ์–‘์‹์„ ์‚ฌ์šฉํ•˜์—ฌ ๋ฌธ์˜ํ•˜์‹ญ์‹œ์˜ค.

https://forms.gle/WvT1wiN1qDtmnspy7