What AI is -- and isn't | Sebastian Thrun and Chris Anderson

259,143 views ใƒป 2017-12-21

TED


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

๋ฒˆ์—ญ: Yoonkeun Ji ๊ฒ€ํ† : Jeeyong Park
00:12
Chris Anderson: Help us understand what machine learning is,
0
12904
2886
ํฌ๋ฆฌ์Šค ์•ค๋”์Šจ: ๋จธ์‹ ๋Ÿฌ๋‹์„ ์ดํ•ดํ•˜๋Š” ๋ฐ ๋„์›€์„ ์ฃผ์‹ค ๋ถ„์„ ๋ชจ์…จ์Šต๋‹ˆ๋‹ค.
00:15
because that seems to be the key driver
1
15814
2054
๋จธ์‹  ๋Ÿฌ๋‹์€ ์ธ๊ณต์ง€๋Šฅ์„ ๋‘˜๋Ÿฌ์‹ผ
00:17
of so much of the excitement and also of the concern
2
17892
2737
์„ธ๊ฐ„์˜ ๊ฑฑ์ •๊ณผ ๊ธฐ๋Œ€์˜
00:20
around artificial intelligence.
3
20653
1494
์ค‘์‹ฌ์— ์žˆ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
00:22
How does machine learning work?
4
22171
1643
๋จธ์‹  ๋Ÿฌ๋‹์€ ์–ด๋–ป๊ฒŒ ์ž‘๋™ํ•˜๋‚˜์š”?
00:23
Sebastian Thrun: So, artificial intelligence and machine learning
5
23838
3896
์„ธ๋ฐ”์Šค์ฐฌ: ์ธ๊ณต์ง€๋Šฅ๊ณผ ๋จธ์‹  ๋Ÿฌ๋‹์€ ์‚ฌ์‹ค
00:27
is about 60 years old
6
27758
2002
60๋…„ ์ •๋„ ๋œ ์—ฐ๊ตฌ๋ถ„์•ผ์ธ๋ฐ
00:29
and has not had a great day in its past until recently.
7
29784
4269
์ตœ๊ทผ๊นŒ์ง€๋Š” ํฐ ๊ด€์‹ฌ์„ ๋ฐ›์ง€ ๋ชปํ–ˆ์Šต๋‹ˆ๋‹ค.
00:34
And the reason is that today,
8
34077
2924
์ตœ๊ทผ์— ๋“ค์–ด์„œ์•ผ
00:37
we have reached a scale of computing and datasets
9
37025
3973
๊ธฐ๊ณ„๋ฅผ ๋˜‘๋˜‘ํ•˜๊ฒŒ ๋งŒ๋“ค ์ˆ˜ ์žˆ์„ ๋งŒํผ
00:41
that was necessary to make machines smart.
10
41022
2637
์ปดํ“จํŒ…, ๋ฐ์ดํ„ฐ ๊ธฐ์ˆ ์˜ ๋ฐœ์ „์ด ์ด๋ฃจ์–ด์กŒ๊ธฐ ๋•Œ๋ฌธ์ด์ฃ .
00:43
So here's how it works.
11
43683
1751
์ž‘๋™ ์›๋ฆฌ๋ฅผ ์‚ดํŽด๋ณด์ฃ .
00:45
If you program a computer today, say, your phone,
12
45458
3497
์Šค๋งˆํŠธํฐ์„ ์œ„ํ•œ ํ”„๋กœ๊ทธ๋žจ์„ ๋งŒ๋“ ๋‹ค๊ณ  ๊ฐ€์ •ํ•ด๋ด…์‹œ๋‹ค.
00:48
then you hire software engineers
13
48979
2335
๊ทธ๋Ÿฌ๋ฉด ์†Œํ”„ํŠธ์›จ์–ด ๊ฐœ๋ฐœ์ž๋ฅผ ๊ณ ์šฉํ•ด์„œ
00:51
that write a very, very long kitchen recipe,
14
51338
3854
์•„์ฃผ, ์•„์ฃผ ๊ธด ์Œ์‹ ์กฐ๋ฆฌ๋ฒ•์„ ์ ๊ฒŒ ํ•ฉ๋‹ˆ๋‹ค.
00:55
like, "If the water is too hot, turn down the temperature.
15
55216
3132
"๋งŒ์•ฝ ๋ฌผ์ด ๋„ˆ๋ฌด ๋œจ๊ฑฐ์šฐ๋ฉด, ์˜จ๋„๋ฅผ ๋‚ด๋ ค๋ผ.
00:58
If it's too cold, turn up the temperature."
16
58372
2279
๋„ˆ๋ฌด ์ฐจ๊ฐ€์šฐ๋ฉด, ์˜จ๋„๋ฅผ ์˜ฌ๋ ค๋ผ." ํ•˜๋Š” ์‹์œผ๋กœ ๋ง์ด์ฃ .
01:00
The recipes are not just 10 lines long.
17
60675
2849
์กฐ๋ฆฌ๋ฒ•์€ ์—ด ์ค„ ์งœ๋ฆฌ๊ฐ€ ์•„๋‹ˆ์ฃ .
01:03
They are millions of lines long.
18
63548
2603
์ˆ˜ ๋ฐฑ๋งŒ ์ค„ ์ •๋„๋Š” ๋˜์–ด์•ผ ํ•  ๊ฒ๋‹ˆ๋‹ค.
01:06
A modern cell phone has 12 million lines of code.
19
66175
4084
์ตœ์‹  ์Šค๋งˆํŠธํฐ์€ 1,200๋งŒ ์ค„์˜ ์†Œ์Šค ์ฝ”๋“œ๋กœ ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.
01:10
A browser has five million lines of code.
20
70283
2646
ํ•˜๋‚˜์˜ ๋ธŒ๋ผ์šฐ์ €๋Š” 5๋ฐฑ๋งŒ ์ค„ ์ •๋„๋กœ ๊ตฌ์„ฑ๋ฉ๋‹ˆ๋‹ค.
01:12
And each bug in this recipe can cause your computer to crash.
21
72953
4969
์กฐ๋ฆฌ๋ฒ•์— ํฌํ•จ๋œ ์˜ค๋ฅ˜ ํ•˜๋‚˜ ํ•˜๋‚˜๊ฐ€ ์ปดํ“จํ„ฐ๋ฅผ ๊ณ ์žฅ๋‚˜๊ฒŒ ํ•  ์ˆ˜ ์žˆ์ฃ .
01:17
That's why a software engineer makes so much money.
22
77946
3075
์ด๋Ÿฌํ•œ ์ด์œ ๋กœ ๊ฐœ๋ฐœ์ž๊ฐ€ ๋ˆ์„ ๋งŽ์ด ๋ฒŒ ์ˆ˜ ๋ฐ–์— ์—†์ฃ .
01:21
The new thing now is that computers can find their own rules.
23
81953
3660
๊ทธ๋Ÿฐ๋ฐ ์š”์ฆ˜์€ ์ปดํ“จํ„ฐ๊ฐ€ ์Šค์Šค๋กœ ๊ทœ์น™์„ ์ฐพ์•„๋ƒ…๋‹ˆ๋‹ค.
01:25
So instead of an expert deciphering, step by step,
24
85637
3606
๊ธฐ์กด์—๋Š” ๋ชจ๋“  ๋งŒ์ผ์˜ ์‚ฌํƒœ๋ฅผ ๋Œ€๋น„ํ•ด
01:29
a rule for every contingency,
25
89267
2148
์ „๋ฌธ๊ฐ€๊ฐ€ ๊ทœ์น™์„ ์ผ์ผ์ด ํ•ด๋…ํ•ด์•ผ ํ–ˆ์Šต๋‹ˆ๋‹ค.
01:31
what you do now is you give the computer examples
26
91439
3074
๊ทธ๋Ÿฌ๋‚˜ ์ด์ œ๋Š” ์ปดํ“จํ„ฐ์—๊ฒŒ ์˜ˆ์‹œ๋งŒ ๋˜์ ธ์ฃผ๊ณ 
01:34
and have it infer its own rules.
27
94537
1581
์Šค์Šค๋กœ ๊ทœ์น™์„ ์ถ”๋ก ํ•˜๊ฒŒ ํ•ฉ๋‹ˆ๋‹ค.
01:36
A really good example is AlphaGo, which recently was won by Google.
28
96142
4306
์ตœ๊ทผ ์Šน๋ฆฌ๋ฅผ ๊ฑฐ๋‘” ๊ตฌ๊ธ€์˜ ์•ŒํŒŒ๊ณ ๊ฐ€ ์ข‹์€ ์˜ˆ์‹œ๊ฐ€ ๋˜๊ฒ ๋„ค์š”.
01:40
Normally, in game playing, you would really write down all the rules,
29
100472
3687
๋ณดํ†ต ๊ฒŒ์ž„์„ ํ•˜๋ ค๋ฉด ๋ชจ๋“  ๊ทœ์น™์„ ์ ์–ด์ฃผ์–ด์•ผ ํ•˜๊ฒ ์ฃ .
01:44
but in AlphaGo's case,
30
104183
1785
๊ทธ๋Ÿฌ๋‚˜ ์•ŒํŒŒ๊ณ ์˜ ๊ฒฝ์šฐ๋Š” ๋‹ค๋ฆ…๋‹ˆ๋‹ค.
01:45
the system looked over a million games
31
105992
2066
์‹œ์Šคํ…œ์ด ๋ฐฑ๋งŒ ๊ฐœ๋„ ๋„˜๋Š” ๊ธฐ๋ณด๋ฅผ ๊ณต๋ถ€ํ•ด์„œ
01:48
and was able to infer its own rules
32
108082
2192
์Šค์Šค๋กœ ๊ทœ์น™์„ ์ถ”๋ก ํ–ˆ๊ณ 
01:50
and then beat the world's residing Go champion.
33
110298
2738
๊ทธ๋ ‡๊ฒŒ ์„ธ๊ณ„ ์ตœ๊ณ ์˜ ๋ฐ”๋‘‘ ๊ธฐ์‚ฌ๋ฅผ ์ด๊ฒผ์ฃ .
01:53
That is exciting, because it relieves the software engineer
34
113853
3509
์ด ๊ธฐ์ˆ ์˜ ์ข‹์€ ์ ์€ ์†Œํ”„ํŠธ์›จ์–ด๋ฅผ ๊ฐœ๋ฐœํ•˜๋Š” ์‚ฌ๋žŒ์ด
01:57
of the need of being super smart,
35
117386
1819
์•„์ฃผ ๋˜‘๋˜‘ํ•ด์•ผ ํ•œ๋‹ค๊ฑฐ๋‚˜
01:59
and pushes the burden towards the data.
36
119229
2325
๋ฐ์ดํ„ฐ๋ฅผ ์ž˜ ๋‹ค๋ฃจ์–ด์•ผ ํ•œ๋‹ค๋Š” ๋ถ€๋‹ด์„ ๋œ์–ด์ค๋‹ˆ๋‹ค.
02:01
As I said, the inflection point where this has become really possible --
37
121578
4534
๋ง์”€๋“œ๋ ธ๋“ฏ์ด, ์ด๊ฒƒ์ด ๊ฐ€๋Šฅํ•˜๊ฒŒ ๋œ ๊ณ„๊ธฐ๋Š” --
02:06
very embarrassing, my thesis was about machine learning.
38
126136
2746
๋ถ€๋„๋Ÿฝ์ง€๋งŒ, ์กธ์—… ๋…ผ๋ฌธ์„ ๊ธฐ๊ณ„ํ•™์Šต์— ๊ด€ํ•ด ์ผ์Šต๋‹ˆ๋‹ค.
02:08
It was completely insignificant, don't read it,
39
128906
2205
์ •๋ง ์‹œ์‹œํ•œ ๋…ผ๋ฌธ์ด๋‹ˆ ์ฝ์ง€๋Š” ๋ง์•„์ฃผ์„ธ์š”.
02:11
because it was 20 years ago
40
131135
1350
20๋…„ ์ „์ด์—ˆ๊ธฐ ๋•Œ๋ฌธ์—
02:12
and back then, the computers were as big as a cockroach brain.
41
132509
2907
๊ทธ ๋•Œ๋Š” ์ปดํ“จํ„ฐ๊ฐ€ ๋งค์šฐ ์ปธ์ง€์š”.
02:15
Now they are powerful enough to really emulate
42
135440
2331
์ง€๊ธˆ์€ ํŠน์ • ์‚ฌ๊ณ ๋ฅผ ์‚ฌ๋žŒ๊ณผ ๊ฑฐ์˜ ๋น„์Šทํ•˜๊ฒŒ
02:17
kind of specialized human thinking.
43
137795
2076
ํ•ด๋‚ผ ์ˆ˜ ์žˆ์„๋งŒํผ ๋˜‘๋˜‘ํ•˜๊ณ  ๊ฐ•๋ ฅํ•ด์กŒ์ฃ .
02:19
And then the computers take advantage of the fact
44
139895
2313
๊ฒŒ๋‹ค๊ฐ€ ์ปดํ“จํ„ฐ๋Š” ์‚ฌ๋žŒ๋ณด๋‹ค ๋” ๋งŽ์€ ๋ฐ์ดํ„ฐ๋ฅผ
02:22
that they can look at much more data than people can.
45
142232
2500
์ฒ˜๋ฆฌํ• ์ˆ˜ ์žˆ๋‹ค๋Š” ์ ์—์„œ ์šฐ์œ„์— ์žˆ์Šต๋‹ˆ๋‹ค.
02:24
So I'd say AlphaGo looked at more than a million games.
46
144756
3080
๊ทธ๋ž˜์„œ ์•ŒํŒŒ๊ณ ๊ฐ€ ๋ฐฑ๋งŒ ๊ฐœ ์ด์ƒ์˜ ๊ธฐ๋ณด๋ฅผ ํ•™์Šตํ•  ์ˆ˜ ์žˆ์—ˆ๋˜ ๊ฑฐ์ฃ .
02:27
No human expert can ever study a million games.
47
147860
2839
๋ฐฑ๋งŒ ๊ฐœ์˜ ๊ธฐ๋ณด๋ฅผ ๊ณต๋ถ€ํ•˜๋Š” ๊ฑด ์‚ฌ๋žŒ์—๊ฒ ๋ถˆ๊ฐ€๋Šฅํ•œ ์ผ์ด์ง€์š”.
02:30
Google has looked at over a hundred billion web pages.
48
150723
3182
๊ตฌ๊ธ€์€ 1์ฒœ์–ต ๊ฐœ๊ฐ€ ๋„˜๋Š” ์›น ํŽ˜์ด์ง€๋ฅผ ์‚ดํŽด ๋ด…๋‹ˆ๋‹ค.
02:33
No person can ever study a hundred billion web pages.
49
153929
2650
์ด ์—ญ์‹œ ์‚ฌ๋žŒ์—๊ฒŒ๋Š” ๋ฌด๋ฆฌ์ง€์š”.
02:36
So as a result, the computer can find rules
50
156603
2714
๊ฒฐ๊ณผ์ ์œผ๋กœ ์‚ฌ๋žŒ์ด ์ฐพ์„ ์ˆ˜ ์—†๋˜ ๊ทœ์น™๋“ค์„
02:39
that even people can't find.
51
159341
1755
์ปดํ“จํ„ฐ๋Š” ์ฐพ์•„๋‚ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
02:41
CA: So instead of looking ahead to, "If he does that, I will do that,"
52
161120
4312
ํฌ๋ฆฌ์Šค: "์ƒ๋Œ€ํŽธ์ด ์ด๋ ‡๊ฒŒ ํ•˜๋ฉด, ๋‚˜๋Š” ์ €๋ ‡๊ฒŒ ํ•ด์•ผ์ง€" ๋ผ๊ณ 
02:45
it's more saying, "Here is what looks like a winning pattern,
53
165456
3072
๋‚ด๋‹ค๋ณด๊ณ  ์ƒ๊ฐ ํ•˜๋Š” ๊ฒŒ ์•„๋‹ˆ๋ผ "์ด๋Ÿฐ ์‹์œผ๋กœ ํ•˜๋ฉด ์ด๊ธฐ๋Š”๊ตฌ๋‚˜."
02:48
here is what looks like a winning pattern."
54
168552
2079
"์ด๋Ÿฐ ์‹์œผ๋กœ ํ•˜๋ฉด ์ด๊ธฐ๋Š”๊ตฌ๋‚˜." ๋ผ๊ณ  ์ƒ๊ฐํ•˜๋Š” ๊ฑฐ๊ตฐ์š”.
02:50
ST: Yeah. I mean, think about how you raise children.
55
170655
2517
์„ธ๋ฐ”์Šค์ฐฌ: ๊ทธ๋ ‡์ฃ . ์•„์ด๋ฅผ ํ‚ค์šธ ๋•Œ๋ฅผ ์ƒ๊ฐํ•ด๋ณด์„ธ์š”.
02:53
You don't spend the first 18 years giving kids a rule for every contingency
56
173196
3644
18์‚ด์ด ๋  ๋•Œ๊นŒ์ง€ ๋ชจ๋“  ๋Œ๋ฐœ ์ƒํ™ฉ์— ์ผ์ผ์ด ์ง€์‹œ๋ฅผ ๋‚ด๋ฆฌ์ง€๋Š” ์•Š์ฃ .
02:56
and set them free and they have this big program.
57
176864
2347
๊ทธ์ € ์ž์—ฐ์Šค๋Ÿฝ๊ฒŒ ๋‘๋ฉด ํฐ ํ”„๋กœ๊ทธ๋žจ์„ ๊ฐ–๊ฒŒ ๋˜์ฃ .
02:59
They stumble, fall, get up, they get slapped or spanked,
58
179235
2719
ํ”๋“ค๋ฆฌ๊ณ , ๋„˜์–ด์ง€๊ณ , ๋‹ค์‹œ ์ผ์–ด์„œ๊ณ , ๋˜ ํ˜ผ๋‚˜๊ธฐ๋„ ํ•ฉ๋‹ˆ๋‹ค.
03:01
and they have a positive experience, a good grade in school,
59
181978
2884
ํ•™๊ต์—์„œ ์ข‹์€ ์„ฑ์ ์„ ๋ฐ›๋Š” ๋“ฑ ๊ธ์ •์ ์ธ ๊ฒฝํ—˜๋„ ํ•  ๊ฒ๋‹ˆ๋‹ค.
03:04
and they figure it out on their own.
60
184886
1834
๋‚˜๋ฆ„๋Œ€๋กœ ์‚ด์•„๊ฐ€๋Š” ๋ฒ•์„ ์•Œ์•„๋‚ด์ฃ .
03:06
That's happening with computers now,
61
186744
1737
์ง€๊ธˆ ์ปดํ“จํ„ฐ๊ฐ€ ์ด๋ ‡๊ฒŒ ํ•˜๊ณ  ์žˆ์–ด์š”.
03:08
which makes computer programming so much easier all of a sudden.
62
188505
3029
๋•๋ถ„์— ์ปดํ“จํ„ฐ ํ”„๋กœ๊ทธ๋ž˜๋ฐ์ด ๊ฐ‘์ž๊ธฐ ๋„ˆ๋ฌด ์‰ฌ์›Œ์กŒ์ฃ .
03:11
Now we don't have to think anymore. We just give them lots of data.
63
191558
3175
๊ทธ์ € ์ปดํ“จํ„ฐ์—๊ฒŒ ๋งŽ์€ ์–‘์˜ ์ •๋ณด๋ฅผ ๋„˜๊ฒจ์ฃผ๋ฉด ๋ฉ๋‹ˆ๋‹ค.
03:14
CA: And so, this has been key to the spectacular improvement
64
194757
3422
ํฌ๋ฆฌ์Šค: ์ž์œจ์ฃผํ–‰ ์ž๋™์ฐจ๊ฐ€ ํ˜์‹ ์ ์œผ๋กœ ๋ฐœ์ „ํ•˜๊ฒŒ ๋œ ๋ฐ์—
03:18
in power of self-driving cars.
65
198203
3064
๊ทธ๋Ÿฐ ๋ฐฐ๊ฒฝ์ด ์žˆ์—ˆ๊ตฐ์š”.
03:21
I think you gave me an example.
66
201291
1739
์˜ˆ์‹œ๋ฅผ ๊ฐ€์ง€๊ณ  ์˜ค์…จ๋Š”๋ฐ
03:23
Can you explain what's happening here?
67
203054
2685
์˜์ƒ์œผ๋กœ ์„ค๋ช… ํ•ด ์ฃผ์‹œ๊ฒ ์–ด์š”?
03:25
ST: This is a drive of a self-driving car
68
205763
3564
์„ธ๋ฐ”์Šค์ฐฌ: ์œ ๋‹ค์‹œํ‹ฐ์˜ ์ž์œจ์ฃผํ–‰ ์ž๋™์ฐจ๊ฐ€
03:29
that we happened to have at Udacity
69
209351
1957
์ฃผํ–‰ ์ค‘์ธ ๋ชจ์Šต์ž…๋‹ˆ๋‹ค.
03:31
and recently made into a spin-off called Voyage.
70
211332
2398
์ตœ๊ทผ์—๋Š” ๋ณด์•ผ์ง€(Voyage)๋ผ๋Š” ์ƒˆ ๋ฒ„์ „์„ ๋งŒ๋“ค์—ˆ์ฃ .
03:33
We have used this thing called deep learning
71
213754
2574
๋”ฅ๋Ÿฌ๋‹ ์ด๋ผ๋Š” ๊ธฐ์ˆ ์„ ์ด์šฉํ•ด์„œ
03:36
to train a car to drive itself,
72
216352
1623
์ฐจ๊ฐ€ ์Šค์Šค๋กœ ์ฃผํ–‰ํ•˜๋„๋ก ํ›ˆ๋ จ์‹œ์ผฐ์Šต๋‹ˆ๋‹ค.
03:37
and this is driving from Mountain View, California,
73
217999
2387
์บ˜๋ฆฌํฌ๋‹ˆ์•„์˜ ๋งˆ์šดํ‹ด๋ทฐ์—์„œ ์ถœ๋ฐœํ•ด์„œ
03:40
to San Francisco
74
220410
1168
์ƒŒํ”„๋ž€์‹œ์Šค์ฝ”๊นŒ์ง€
03:41
on El Camino Real on a rainy day,
75
221602
2259
๋น„์˜ค๋Š” ๋‚  '์—˜ ์นด๋ฏธ๋…ธ ๋ฆฌ์–ผ'์„ ๋”ฐ๋ผ ์ฃผํ–‰ํ•˜๋ฉฐ
03:43
with bicyclists and pedestrians and 133 traffic lights.
76
223885
3524
์ž์ „๊ฑฐ ํƒ€๋Š” ์‚ฌ๋žŒ๋“ค, ๋ณดํ–‰์ž๋“ค, 133๊ฐœ์˜ ์‹ ํ˜ธ๋“ฑ์„ ์ง€๋‚˜์ณค์ฃ .
03:47
And the novel thing here is,
77
227433
2636
์—ฌ๊ธฐ์„œ ๋Œ€๋‹จํ•œ ์ ์€
03:50
many, many moons ago, I started the Google self-driving car team.
78
230093
3120
์•„์ฃผ ์˜ค๋ž˜์ „ ์ œ๊ฐ€ ๊ตฌ๊ธ€์˜ ์ž์œจ์ฃผํ–‰ ์ž๋™์ฐจ ํŒ€์—์„œ ์ผํ–ˆ์„ ๋•Œ
03:53
And back in the day, I hired the world's best software engineers
79
233237
3181
์ €๋Š” ์„ธ๊ณ„ ์ตœ๊ณ ์˜ ๊ฐœ๋ฐœ์ž๋“ค์„ ๊ณ ์šฉํ•ด์„œ
03:56
to find the world's best rules.
80
236442
1607
์„ธ๊ณ„ ์ตœ๊ณ ์˜ ๊ทœ์น™์„ ์ฐพ์œผ๋ ค ํ–ˆ์ฃ .
03:58
This is just trained.
81
238073
1754
๊ทธ๋Ÿฐ๋ฐ ์ด ์ž๋™์ฐจ๋Š” ๊ทธ๋ƒฅ ํ›ˆ๋ จ์„ ๋ฐ›์•˜์Šต๋‹ˆ๋‹ค.
03:59
We drive this road 20 times,
82
239851
3336
๋„๋กœ ์ฃผํ–‰์„ 20๋ฒˆ ํ•œ ํ›„
04:03
we put all this data into the computer brain,
83
243211
2447
๊ทธ ์ •๋ณด๋ฅผ ์ปดํ“จํ„ฐ์˜ ๋‡Œ์— ๊ฑด๋„ค์ฃผ๊ณ 
04:05
and after a few hours of processing,
84
245682
2082
๋ช‡ ์‹œ๊ฐ„๋™์•ˆ ์ฒ˜๋ฆฌํ•˜๋„๋ก ํ–ˆ๋”๋‹ˆ
04:07
it comes up with behavior that often surpasses human agility.
85
247788
3926
์‹ฌ์ง€์–ด ์‚ฌ๋žŒ๋ณด๋‹ค ๋” ๋Šฅ์ˆ™ํ•˜๊ฒŒ ์šด์ „ํ•˜๊ธฐ๋„ ํ•˜๋Š” ๊ฒ๋‹ˆ๋‹ค.
04:11
So it's become really easy to program it.
86
251738
2017
๊ทธ๋Ÿฌ๋‹ˆ๊นŒ ํ”„๋กœ๊ทธ๋ž˜๋ฐํ•˜๊ธฐ๊ฐ€ ์‰ฌ์›Œ์กŒ์ฃ .
04:13
This is 100 percent autonomous, about 33 miles, an hour and a half.
87
253779
3803
์‹คํ—˜์—์„œ๋Š” 53km๋ฅผ ํ•œ ์‹œ๊ฐ„ ๋ฐ˜ ๋งŒ์— ์‚ฌ๋žŒ์˜ ๋„์›€์—†์ด ์ฃผํ–‰ํ–ˆ์Šต๋‹ˆ๋‹ค.
04:17
CA: So, explain it -- on the big part of this program on the left,
88
257606
3630
ํฌ๋ฆฌ์Šค: ์ด์ œ ์˜์ƒ์„ ์„ค๋ช… ํ•ด ์ฃผ์‹œ์ฃ . ์™ผ์ชฝ์—๋Š” ์ปดํ“จํ„ฐ๊ฐ€ ์ง€๊ฐํ•˜๋Š” ๋ชจ์Šต์ด์ฃ ?
04:21
you're seeing basically what the computer sees as trucks and cars
89
261260
3257
ํŠธ๋Ÿญ์ด๋‚˜ ์ฐจ๋กœ ๋ณด์ด๋Š” ์ ๋“ค์ด
04:24
and those dots overtaking it and so forth.
90
264541
2886
์ž์œจ์ฃผํ–‰์ฐจ๋ฅผ ์ถ”์›” ํ•˜๊ณ  ์žˆ๋Š” ๊ฒƒ ๊ฐ™๋„ค์š”.
04:27
ST: On the right side, you see the camera image, which is the main input here,
91
267451
3762
์„ธ๋ฐ”์Šค์ฐฌ: ์˜ค๋ฅธํŽธ์— ์นด๋ฉ”๋ผ๋กœ ์ดฌ์˜๋œ ์˜์ƒ์ด ์ปดํ“จํ„ฐ์˜ ๋ฉ”์ธ ์ž…๋ ฅ์ด์ฃ .
04:31
and it's used to find lanes, other cars, traffic lights.
92
271237
2676
์ด ์˜์ƒ์„ ํ†ตํ•ด ์ฐจ์„ ์ด๋‚˜ ๋‹ค๋ฅธ ์ฐจ, ์‹ ํ˜ธ๋“ฑ์„ ์ธ์‹ํ•ฉ๋‹ˆ๋‹ค.
04:33
The vehicle has a radar to do distance estimation.
93
273937
2489
์ฐจ๋Ÿ‰์—๋Š” ๊ฑฐ๋ฆฌ๋ฅผ ์ถ”์ •ํ•˜๋Š” ๋ ˆ์ด๋”๋„ ๋‹ฌ๋ ค์žˆ์Šต๋‹ˆ๋‹ค.
04:36
This is very commonly used in these kind of systems.
94
276450
2621
์ด๋Ÿฐ ์ž์œจ์ฃผํ–‰์ฐจ์— ํ”ํžˆ ์“ฐ์ด๋Š” ์žฅ๋น„์ž…๋‹ˆ๋‹ค.
04:39
On the left side you see a laser diagram,
95
279095
1992
์™ผํŽธ์— ๋ณด์ด๋Š” ๊ฒƒ์€ ๋ ˆ์ด์ € ๋„ํ‘œ์ž…๋‹ˆ๋‹ค.
04:41
where you see obstacles like trees and so on depicted by the laser.
96
281111
3200
๋‚˜๋ฌด ๋“ฑ์˜ ์žฅ์• ๋ฌผ์„ ๋ ˆ์ด์ €๋กœ ๊ทธ๋ฆฐ ๊ฒƒ์„ ๋ณด์‹ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
04:44
But almost all the interesting work is centering on the camera image now.
97
284335
3436
ํ•˜์ง€๋งŒ ์ •๋ง ์žฌ๋ฐŒ๋Š” ๊ฑด ์นด๋ฉ”๋ผ ์˜์ƒ ์ชฝ ์ž…๋‹ˆ๋‹ค.
04:47
We're really shifting over from precision sensors like radars and lasers
98
287795
3476
๋ ˆ์ด๋”๋‚˜ ๋ ˆ์ด์ € ๊ฐ™์€ ์ •๋ฐ€ ๊ฐ์ง€๊ธฐ๋ฅผ
04:51
into very cheap, commoditized sensors.
99
291295
1842
๊ฐ’์‹ธ๊ณ  ์ƒ์šฉํ™”๋œ ๊ฐ์ง€๊ธฐ๋กœ ๋ฐ”๊พธ๋Š” ์ค‘์ž…๋‹ˆ๋‹ค.
04:53
A camera costs less than eight dollars.
100
293161
1987
์ด์ œ 8๋‹ฌ๋Ÿฌ๋„ ์•ˆ๋˜๋Š” ์นด๋ฉ”๋ผ๋ฅผ ์”๋‹ˆ๋‹ค.
04:55
CA: And that green dot on the left thing, what is that?
101
295172
2793
ํฌ๋ฆฌ์Šค: ๊ทธ๋Ÿฐ๋ฐ ์™ผ์ชฝ์— ์ดˆ๋ก์ ์€ ๋ญ”๊ฐ€์š”?
04:57
Is that anything meaningful?
102
297989
1371
๋ญ”๊ฐ€ ์˜๋ฏธ์žˆ๋Š” ๊ฑด๊ฐ€์š”?
04:59
ST: This is a look-ahead point for your adaptive cruise control,
103
299384
3668
์„ธ๋ฐ”์Šค์ฐฌ: ์•ž ์ฐจ์™€์˜ ๊ฑฐ๋ฆฌ์— ๋”ฐ๋ผ ์†๋„๋ฅผ ์กฐ์ ˆํ•˜๋Š” ์ ์‘ํ˜• ์ˆœํ•ญ ์žฅ์น˜๊ฐ€
05:03
so it helps us understand how to regulate velocity
104
303076
2477
์–ด๋””๋ฅผ ์•ž ์ฐจ๋ผ๊ณ  ์ธ์‹ํ•˜๊ณ  ์žˆ๋Š”์ง€ ํ‘œ์‹œํ•œ ๊ฒ๋‹ˆ๋‹ค.
05:05
based on how far the cars in front of you are.
105
305577
2634
์šฐ๋ฆฌ๊ฐ€ ๋ณด๊ณ  ์†๋„ ์กฐ์ ˆ ๊ธฐ์žฌ๋ฅผ ์ดํ•ดํ•˜๋„๋ก ๋ง์ด์ฃ .
05:08
CA: And so, you've also got an example, I think,
106
308235
2716
ํฌ๋ฆฌ์Šค: ์ œ๊ฐ€ ๋“ฃ๊ธฐ๋กœ, ๊ธฐ๊ณ„๊ฐ€ ์–ด๋–ป๊ฒŒ ํ•™์Šต์„ ํ•˜๋Š”์ง€์— ๋Œ€ํ•œ
05:10
of how the actual learning part takes place.
107
310975
2381
์˜ˆ์‹œ๋„ ์˜ค๋Š˜ ์ค€๋น„ํ•˜์…จ์ฃ ?
05:13
Maybe we can see that. Talk about this.
108
313380
2458
๊ฑฐ๊ธฐ์— ๋Œ€ํ•ด ๋“ฃ๊ณ  ์‹ถ๊ตฐ์š”.
05:15
ST: This is an example where we posed a challenge to Udacity students
109
315862
3643
์„ธ๋ฐ”์Šค์ฐฌ: ์œ ๋‹ค์‹œํ‹ฐ์— ์ž์œจ์ฃผํ–‰ ์ž๋™์ฐจ '๋‚˜๋…ธ๋””๊ทธ๋ฆฌ'๋ผ๊ณ  ๋ถ€๋ฅด๋Š”
05:19
to take what we call a self-driving car Nanodegree.
110
319529
3131
๊ต์œก๊ณผ์ •์„ ๋“ฃ๋Š” ํ•™์ƒ๋“ค์ด ์žˆ๋Š”๋ฐ์š”.
05:22
We gave them this dataset
111
322684
1495
์ค€๋น„๋œ ๋ฐ์ดํ„ฐ์…‹์„ ์ฃผ๊ณ 
05:24
and said "Hey, can you guys figure out how to steer this car?"
112
324203
3054
"์ด ์ฐจ๊ฐ€ ์–ด๋–ป๊ฒŒ ๋ฐฉํ–ฅ์„ ์กฐ์ ˆํ•˜๋Š”์ง€ ์•Œ์•„๋‚ผ ์ˆ˜ ์žˆ๊ฒ ๋‹ˆ?" ๋ผ๊ณ  ๋ฌผ์—ˆ์Šต๋‹ˆ๋‹ค.
05:27
And if you look at the images,
113
327281
1624
์˜์ƒ์„ ๋ณด์‹œ๋ฉด ์•„์‹œ๊ฒ ์ง€๋งŒ
05:28
it's, even for humans, quite impossible to get the steering right.
114
328929
4073
์‚ฌ์‹ค ์‚ฌ๋žŒ๋„ ์ œ๋Œ€๋กœ ๋ฐฉํ–ฅ์„ ๋งž์ถ”๊ธฐ ํž˜๋“ญ๋‹ˆ๋‹ค.
05:33
And we ran a competition and said, "It's a deep learning competition,
115
333026
3591
์ž‘์€ ๋Œ€ํšŒ๋ฅผ ์—ด๊ณ , "์ด๊ฑด ๋”ฅ๋Ÿฌ๋‹ ๋Œ€ํšŒ์ด๊ณ , AI ๋Œ€ํšŒ์•ผ."
05:36
AI competition,"
116
336641
1173
๋ผ๊ณ  ์–˜๊ธฐํ•˜๊ณ 
05:37
and we gave the students 48 hours.
117
337838
1887
ํ•™์ƒ๋“ค์—๊ฒŒ 48์‹œ๊ฐ„์„ ์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค.
05:39
So if you are a software house like Google or Facebook,
118
339749
4172
๊ตฌ๊ธ€์ด๋‚˜ ํŽ˜์ด์Šค๋ถ๊ฐ™์€ ๊ธฐ์—…์—์„œ ๊ฐ™์€ ์ž‘์—…์„ ํ–ˆ๋‹ค๋ฉด
05:43
something like this costs you at least six months of work.
119
343945
2717
์ตœ์†Œํ•œ ์—ฌ์„ฏ ๋‹ฌ ์ •๋„๋Š” ๊ฑธ๋ ธ์„ ๊ฒ๋‹ˆ๋‹ค.
05:46
So we figured 48 hours is great.
120
346686
2202
48์‹œ๊ฐ„์ด๋ฉด ๋Œ€๋‹จํ•œ๊ฑฐ์ฃ .
05:48
And within 48 hours, we got about 100 submissions from students,
121
348912
3467
๊ฒฐ๊ตญ 48์‹œ๊ฐ„ ์•ˆ์— 100๋ช…์˜ ํ•™์ƒ๋“ค์ด ๋‹ต์•ˆ์„ ์ œ์ถœํ•˜์˜€์Šต๋‹ˆ๋‹ค.
05:52
and the top four got it perfectly right.
122
352403
3370
๊ทธ์ค‘ ๊ฐ€์žฅ ์šฐ์ˆ˜ํ•œ 4๋ช…์˜ ๋‹ต์•ˆ์€ ๊ฑฐ์˜ ์™„๋ฒฝ์— ๊ฐ€๊นŒ์› ์Šต๋‹ˆ๋‹ค.
05:55
It drives better than I could drive on this imagery,
123
355797
2640
์ œ๊ฐ€ ์šด์ „ํ•œ๋‹ค๊ณ  ํ•ด๋„
05:58
using deep learning.
124
358461
1189
๋”ฅ๋Ÿฌ๋‹ ๋ณด๋‹ค ๋ชปํ•  ๊ฒ๋‹ˆ๋‹ค.
05:59
And again, it's the same methodology.
125
359674
1799
๋‹ค์‹œ ํ•œ ๋ฒˆ ๋ง์”€๋“œ๋ฆฌ์ง€๋งŒ ๊ฐ™์€ ๋ฐฉ๋ฒ•๋ก ์ž…๋‹ˆ๋‹ค.
06:01
It's this magical thing.
126
361497
1164
์ •๋ง ๋งˆ๋ฒ•๊ฐ™์•„์š”.
06:02
When you give enough data to a computer now,
127
362685
2085
์ด์ œ ์ปดํ“จํ„ฐ์—๊ฒŒ ์ถฉ๋ถ„ํ•œ ๋ฐ์ดํ„ฐ๋ฅผ ์ฃผ๊ณ 
06:04
and give enough time to comprehend the data,
128
364794
2140
๋ฐ์ดํ„ฐ๋ฅผ ํŒŒ์•…ํ•  ์‹œ๊ฐ„์„ ์ถฉ๋ถ„ํžˆ ์ฃผ๋ฉด
06:06
it finds its own rules.
129
366958
1445
์ž์‹ ๋งŒ์˜ ๊ทœ์น™์„ ์ฐพ์•„๋ƒ…๋‹ˆ๋‹ค.
06:09
CA: And so that has led to the development of powerful applications
130
369339
4845
ํฌ๋ฆฌ์Šค: ์ด ๊ธฐ์ˆ ์ด ์ข‹์€ ์„œ๋น„์Šค๋“ค์ด ๋งŽ์ด ๊ฐœ๋ฐœ๋˜๋Š” ๋ฐ์—
06:14
in all sorts of areas.
131
374208
1525
๋™๋ ฅ์›์ด ๋˜๊ณ  ์žˆ๊ตฐ์š”.
06:15
You were talking to me the other day about cancer.
132
375757
2668
์•”์— ๊ด€๋ จํ•ด์„œ ์–˜๊ธฐํ•˜์‹ค ๊ฒŒ ์žˆ๋‹ค๊ณ  ํ•˜์…จ์ฃ .
06:18
Can I show this video?
133
378449
1189
์˜์ƒ์„ ๊ฐ™์ด ๋ณผ๊นŒ์š”?
06:19
ST: Yeah, absolutely, please. CA: This is cool.
134
379662
2354
์„ธ๋ฐ”์Šค์ฐฌ: ๋„ค, ๊ทธ๋Ÿฌ์ฃ . ํฌ๋ฆฌ์Šค: ๋Œ€๋‹จํ•œ ์˜์ƒ์ด์—์š”.
06:22
ST: This is kind of an insight into what's happening
135
382040
3534
์„ธ๋ฐ”์Šค์ฐฌ: ์™„์ „ํžˆ ๋‹ค๋ฅธ ๋ถ„์•ผ์—์„œ๋Š” ์–ด๋–ป๊ฒŒ ์ธ๊ณต์ง€๋Šฅ์ด ํ™œ์šฉ ๋˜๋Š”์ง€
06:25
in a completely different domain.
136
385598
2429
์ดํ•ด๋ฅผ ๋•๋Š” ์˜์ƒ์ž…๋‹ˆ๋‹ค.
06:28
This is augmenting, or competing --
137
388051
3752
์ฆ๊ฐ€ํ•˜๊ณ  ๊ฒฝํ•ฉํ•˜๋Š” ๊ฒƒ๋“ค์ด ๋ณด์ž…๋‹ˆ๋‹ค.
06:31
it's in the eye of the beholder --
138
391827
1749
๋ณด๋Š” ์‚ฌ๋žŒ์˜ ๋ˆˆ์—๋Š”
06:33
with people who are being paid 400,000 dollars a year,
139
393600
3454
40๋งŒ๋‹ฌ๋Ÿฌ์”ฉ ์—ฐ๋ด‰์„ ๋ฐ›๋Š”
06:37
dermatologists,
140
397078
1237
ํ”ผ๋ถ€๊ณผ ์ „๋ฌธ์˜์˜ ์ž‘์—…์ฒ˜๋Ÿผ ๋ณด์ด์ฃ .
06:38
highly trained specialists.
141
398339
1983
๊ณ ๋„๋กœ ํ›ˆ๋ จ๋ฐ›์€ ์ „๋ฌธ๊ฐ€๋“ค ๋ง์ž…๋‹ˆ๋‹ค.
06:40
It takes more than a decade of training to be a good dermatologist.
142
400346
3561
์ œ๋Œ€๋กœ ๋œ ํ”ผ๋ถ€๊ณผ์ „๋ฌธ์˜๊ฐ€ ๋˜๋ ค๋ฉด 10๋…„์ด ๋„˜๊ฒŒ ๊ฑธ๋ฆฝ๋‹ˆ๋‹ค.
06:43
What you see here is the machine learning version of it.
143
403931
3196
์ง€๊ธˆ ๋ณด์‹œ๋Š” ๊ฒƒ์€ ์‚ฌ์‹ค ๋จธ์‹ ๋Ÿฌ๋‹์˜ ๊ฒฐ๊ณผ๋ฌผ ์ž…๋‹ˆ๋‹ค.
06:47
It's called a neural network.
144
407151
1841
์ „๋ฌธ์šฉ์–ด๋กœ ์ธ๊ณต์‹ ๊ฒฝ๋ง์ด๋ผ๊ณ  ํ•˜์ฃ .
06:49
"Neural networks" is the technical term for these machine learning algorithms.
145
409016
3742
์ด๋Ÿฐ ๊ธฐ๊ณ„ํ•™์Šต ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ๋ถ€๋ฅด๋Š” ๋ง์ž…๋‹ˆ๋‹ค.
06:52
They've been around since the 1980s.
146
412782
1789
์ด ๊ฐœ๋…์€ 1980๋…„๋Œ€๋ถ€ํ„ฐ ์žˆ์—ˆ์ฃ .
06:54
This one was invented in 1988 by a Facebook Fellow called Yann LeCun,
147
414595
4640
์ด ํ”„๋กœ๊ทธ๋žจ์€ 1988๋…„๋„ ์–€ ๋ฅด์ฟค ์ด๋ผ๋Š” ํŽ˜์ด์Šค๋ถ ๊ฐœ๋ฐœ์ž์˜ ์ž‘ํ’ˆ์ž…๋‹ˆ๋‹ค.
06:59
and it propagates data stages
148
419259
3558
์—ฌ๋Ÿฌ ๋‹จ๊ณ„์— ๊ฑธ์ณ ๋ฐ์ดํ„ฐ๋ฅผ ์ „๋‹ฌํ•ฉ๋‹ˆ๋‹ค.
07:02
through what you could think of as the human brain.
149
422841
2578
๋งˆ์น˜ ์‚ฌ๋žŒ์˜ ๋‡Œ๊ฐ€ ํ•˜๋“ฏ์ด ๋ง์ด์ฃ .
07:05
It's not quite the same thing, but it emulates the same thing.
150
425443
2966
์—„๋ฐ€ํ•˜๊ฒŒ ๋งํ•˜๋ฉด ๋‹ค๋ฅด์ง€๋งŒ, ๋น„์Šทํ•˜๊ฒŒ ๋ชจ๋ฐฉํ•œ ๊ฒƒ์ด์ฃ .
07:08
It goes stage after stage.
151
428433
1302
์—ฌ๋Ÿฌ ๋‹จ๊ณ„๋ฅผ ๊ฑฐ์นฉ๋‹ˆ๋‹ค.
07:09
In the very first stage, it takes the visual input and extracts edges
152
429759
3637
๋งจ ์ฒ˜์Œ ๋‹จ๊ณ„์—์„œ๋Š” ์‹œ๊ฐ ์ž๋ฃŒ๋ฅผ ์ž…๋ ฅ๋ฐ›์€ ํ›„ ๊ทธ ์•ˆ์—์„œ
07:13
and rods and dots.
153
433420
2612
๋ชจ์„œ๋ฆฌ, ์„ , ์  ๋“ฑ์„ ๊ตฌ๋ถ„ํ•ฉ๋‹ˆ๋‹ค.
07:16
And the next one becomes more complicated edges
154
436056
3037
๋‹ค์Œ ๋‹จ๊ณ„์—์„œ๋Š” ๋” ๋ณต์žกํ•œ ๋ชจ์„œ๋ฆฌ๋‚˜
07:19
and shapes like little half-moons.
155
439117
3191
์ž‘์€ ๋ฐ˜๋‹ฌ ๊ฐ™์€ ๋ชจ์–‘์ด ๋˜์ฃ .
07:22
And eventually, it's able to build really complicated concepts.
156
442332
4443
๊ทธ๋ฆฌ๊ณ  ๋งˆ์นจ๋‚ด ์ •๋ง ๋ณต์žกํ•œ ๊ฐœ๋…๋“ค์„ ๋‹ค๋ฃฐ ์ˆ˜ ์žˆ์–ด์š”.
07:26
Andrew Ng has been able to show
157
446799
2048
์•ค๋“œ๋ฅ˜ ์‘์€ ์ธ๊ณต์‹ ๊ฒฝ๋ง์„ ์ด์šฉํ•ด
07:28
that it's able to find cat faces and dog faces
158
448871
3480
๋งŽ์€ ์‚ฌ์ง„ ์†์—์„œ ๊ณ ์–‘์ด์™€ ๊ฐœ์˜ ์–ผ๊ตด์„
07:32
in vast amounts of images.
159
452375
1661
๊ตฌ๋ณ„ํ•ด ๋‚ผ ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์„ ๋ณด์—ฌ์คฌ์ฃ .
07:34
What my student team at Stanford has shown is that
160
454060
2724
์Šคํƒ ํผ๋“œ ๋Œ€ํ•™๊ต์— ์ œ๊ฐ€ ๋ฐ๋ฆฌ๊ณ ์žˆ๋Š” ํ•™์ƒ๋“ค์€
07:36
if you train it on 129,000 images of skin conditions,
161
456808
6073
ํ‘์ƒ‰์ข…๊ณผ ํ”ผ๋ถ€์•”์„ ํฌํ•จํ•œ 12๋งŒ 9์ฒœ ์žฅ์˜ ํ”ผ๋ถ€ ์‚ฌ์ง„์„ ๊ฐ€์ง€๊ณ 
07:42
including melanoma and carcinomas,
162
462905
2565
์ธ๊ณต์ง€๋Šฅ์„ ํ›ˆ๋ จ์‹œํ‚ค๋ฉด
07:45
you can do as good a job
163
465494
3301
ํ˜„์กดํ•˜๋Š” ์ตœ๊ณ ์˜ ์‹ค๋ ฅ์„ ๊ฐ€์ง„ ํ”ผ๋ถ€๊ณผ ์ „๋ฌธ์˜ ๋งŒํผ์˜
07:48
as the best human dermatologists.
164
468819
2197
์‹ค๋ ฅ์„ ๊ฐ–์ถ”๊ฒŒ ํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์„ ๋ณด์—ฌ์คฌ์Šต๋‹ˆ๋‹ค.
07:51
And to convince ourselves that this is the case,
165
471040
2549
์ด ํ”„๋กœ๊ทธ๋žจ ๋˜ํ•œ ๊ทธ๋งŒํผ ์‹ค๋ ฅ์ด ์žˆ๋‹ค๋Š” ๊ฒƒ์„ ์ฆ๋ช…ํ•˜๊ธฐ ์œ„ํ•ด
07:53
we captured an independent dataset that we presented to our network
166
473613
3990
์ธ๊ณต์ง€๋Šฅ ํ”„๋กœ๊ทธ๋žจ์—์„œ ๋ฝ‘์•„๋‚ธ ๋…๋ฆฝ์ ์ธ ๋ฐ์ดํ„ฐ์…‹๊ณผ
07:57
and to 25 board-certified Stanford-level dermatologists,
167
477627
4342
์Šคํƒ ํฌ๋“œ ์ˆ˜์ค€์œผ๋กœ ์‹ค๋ ฅ์ด ์ฆ๋ช…๋œ ํ”ผ๋ถ€๊ณผ ์ „๋ฌธ์˜๋“ค์˜ ๋ฐ์ดํƒ€์…‹์„ ๋ชจ์•„์„œ
08:01
and compared those.
168
481993
1672
๊ฒฐ๊ณผ๋ฅผ ๋น„๊ต ํ–ˆ์Šต๋‹ˆ๋‹ค.
08:03
And in most cases,
169
483689
1504
๋Œ€๋ถ€๋ถ„์˜ ๊ฒฝ์šฐ์—
08:05
they were either on par or above the performance classification accuracy
170
485217
3875
์ธ๊ณต์ง€๋Šฅ์˜ ๊ฒฐ๊ณผ๊ฐ€ ํ”ผ๋ถ€๊ณผ ์ „๋ฌธ์˜๊ฐ€ ๋‚ด๋ฆฐ ๊ฒฐ๊ณผ์™€ ๋น„์Šทํ•˜๊ฑฐ๋‚˜ ๋” ๋›ฐ์–ด๋‚œ
08:09
of human dermatologists.
171
489116
1467
๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋ณด์˜€์Šต๋‹ˆ๋‹ค.
08:10
CA: You were telling me an anecdote.
172
490607
1746
ํฌ๋ฆฌ์Šค: ํ•  ์ด์•ผ๊ธฐ๊ฐ€ ์žˆ์œผ์‹œ์ฃ .
08:12
I think about this image right here.
173
492377
1957
์—ฌ๊ธฐ ์žˆ๋Š” ์‚ฌ์ง„์— ๊ด€ํ•œ๊ฑด๋ฐ
08:14
What happened here?
174
494358
1484
์–ด๋–ค ์–˜๊ธฐ ์˜€์ฃ ?
08:15
ST: This was last Thursday. That's a moving piece.
175
495866
4008
์„ธ๋ฐ”์Šค์ฐฌ: ์ง€๋‚œ ๋ชฉ์š”์ผ์ด์—ˆ์–ด์š”.
08:19
What we've shown before and we published in "Nature" earlier this year
176
499898
3600
์ด์ „์— ๊ณต๊ฐœํ–ˆ๊ณ  ์˜ฌํ•ด ์ดˆ ๋„ค์ด์ณ ์ง€์—๋„ ์‹ค๋ฆฐ ๋‚ด์šฉ์ธ๋ฐ์š”.
08:23
was this idea that we show dermatologists images
177
503522
2484
ํ”ผ๋ถ€๊ณผ ์˜์‚ฌ์™€ ์ปดํ“จํ„ฐ์—๊ฒŒ
08:26
and our computer program images,
178
506030
1539
์ด๋ฏธ์ง€๋ฅผ ๋ณด์—ฌ์ค€ ๋’ค
08:27
and count how often they're right.
179
507593
1627
๋‘˜ ์ค‘ ๋ˆ„๊ฐ€ ๋” ์ •ํ™•๋„๊ฐ€ ๋†’์•˜๋Š”์ง€ ์ธก์ •ํ•ด๋ณด์•˜์ฃ .
08:29
But all these images are past images.
180
509244
1778
์ด ์‚ฌ์ง„๋“ค์€ ๋ชจ๋‘ ๋ถ„๋ฅ˜๊ฐ€ ๋๋‚œ ๊ฒƒ๋“ค์ž…๋‹ˆ๋‹ค.
08:31
They've all been biopsied to make sure we had the correct classification.
181
511046
3460
์ •ํ™•ํ•œ ๋ถ„๋ฅ˜๋ฅผ ์œ„ํ•ด ์ด๋ฏธ ์ƒ์ฒด ๊ฒ€์‚ฌ๊ฐ€ ๋๋‚œ ๊ฒƒ๋“ค๋งŒ ๋ชจ์•˜์ฃ .
08:34
This one wasn't.
182
514530
1172
์ด ํ•œ ์žฅ๋งŒ ๋นผ๊ณ ์š”.
08:35
This one was actually done at Stanford by one of our collaborators.
183
515726
3179
์ด ์‚ฌ์ง„์€ ์šฐ๋ฆฌ ์ค‘ ํ•œ ๋ช…์ด ์Šคํƒ ํฌ๋“œ์—์„œ ์ฐ์€ ์‚ฌ์ง„์ด์—์š”.
08:38
The story goes that our collaborator,
184
518929
2314
์–ด๋–ป๊ฒŒ ๋œ๊ฑฐ๋ƒ๋ฉด ์šฐ๋ฆฌ ํŒ€์› ์ค‘ ํ•œ ๋ช…์ด ํ”ผ๋ถ€๊ณผ ์˜์‚ฌ์ธ๋ฐ์š”.
08:41
who is a world-famous dermatologist, one of the three best, apparently,
185
521267
3391
์ „ ์„ธ๊ณ„์—์„œ ์œ ๋ช…ํ•œ, ์„ธ ์†๊ฐ€๋ฝ ์ค‘์— ํ•˜๋‚˜์— ๋“œ๋Š” ์‚ฌ๋žŒ์ด์ฃ .
08:44
looked at this mole and said, "This is not skin cancer."
186
524682
2935
์ด ๋ฐ˜์ ์„ ๋ณด๊ณ ๋Š” "ํ”ผ๋ถ€์•”์ด ์•„๋‹ˆ๋‹ค"๊ณ  ํ–ˆ์–ด์š”.
08:47
And then he had a second moment, where he said,
187
527641
2476
ํ•˜์ง€๋งŒ ์ž ์‹œ ๋’ค์— ๋‹ค์‹œ ์ด๋ ‡๊ฒŒ ๋งํ–ˆ์–ด์š”.
08:50
"Well, let me just check with the app."
188
530141
1866
"์ž ๊น, ์–ดํ”Œ๋กœ ํ™•์ธํ•ด๋ณด๊ณ  ์‹ถ์€๋ฐ"
08:52
So he took out his iPhone and ran our piece of software,
189
532031
2699
๊ทธ๋ฆฌ๊ณ ๋Š” ์ž๊ธฐ ์•„์ดํฐ์„ ๊บผ๋‚ด์„œ ์šฐ๋ฆฌ๊ฐ€ ๋งŒ๋“  ์†Œํ”„ํŠธ์›จ์–ด
08:54
our "pocket dermatologist," so to speak,
190
534754
2121
๋งํ•˜์ž๋ฉด "์ฃผ๋จธ๋‹ˆ ์† ํ”ผ๋ถ€ ์ „๋ฌธ์˜"์ฃ .
08:56
and the iPhone said: cancer.
191
536899
2994
๊ทธ๊ฒƒ์„ ๊บผ๋‚ด์„œ ํ™•์ธํ•ด๋ณด๋‹ˆ ์•”์ด๋ผ๊ณ  ์ง„๋‹จํ–ˆ์–ด์š”.
08:59
It said melanoma.
192
539917
1306
ํ‘์ƒ‰์ข…์ด๋ผ๊ณ  ํ–ˆ์ฃ .
09:01
And then he was confused.
193
541849
1233
์นœ๊ตฌ๋Š” ๋ง์„ค์ด๋‹ค๊ฐ€ ์ด๋ ‡๊ฒŒ ๋งํ–ˆ์ฃ .
09:03
And he decided, "OK, maybe I trust the iPhone a little bit more than myself,"
194
543106
4551
"์ข‹์•„, ์•„์ดํฐ์„ ํ•œ ๋ฒˆ ๋ฏฟ์–ด๋ด์•ผ๊ฒ ์–ด."
09:07
and he sent it out to the lab to get it biopsied.
195
547681
2735
๊ทธ๋ฆฌ๊ณ  ์กฐ์ง๊ฒ€์‚ฌ๋ฅผ ํ•˜๊ธฐ ์œ„ํ•ด ์‚ฌ์ง„์„ ์—ฐ๊ตฌ์†Œ๋กœ ๋ณด๋ƒˆ์Šต๋‹ˆ๋‹ค.
09:10
And it came up as an aggressive melanoma.
196
550440
2469
๊ทธ ๊ฒฐ๊ณผ ์•…์„ฑ ํ‘์ƒ‰์ข…์œผ๋กœ ํŒ๋ช…๋˜์—ˆ์ฃ .
09:13
So I think this might be the first time that we actually found,
197
553545
3067
์ด๊ฒƒ์ด ์šฐ๋ฆฌ๊ฐ€ ๋”ฅ๋Ÿฌ๋‹ ๊ธฐ์ˆ ์„ ์ ์šฉํ•ด
09:16
in the practice of using deep learning,
198
556636
2487
ํ‘์ƒ‰์ข…์ด ๋ฐœ์ƒํ–ˆ์ง€๋งŒ ๋†“์น  ๋ป”ํ–ˆ๋˜ ํ™˜์ž๋ฅผ
09:19
an actual person whose melanoma would have gone unclassified,
199
559147
3372
๋”ฅ๋Ÿฌ๋‹์„ ํ†ตํ•ด ์ฐพ์•„๋‚ธ
09:22
had it not been for deep learning.
200
562543
2115
์ฒซ ๋ฒˆ์งธ ์‚ฌ๋ก€์ž…๋‹ˆ๋‹ค.
09:24
CA: I mean, that's incredible.
201
564682
1560
ํฌ๋ฆฌ์Šค: ์ •๋ง ๋Œ€๋‹จํ•˜๋„ค์š”.
09:26
(Applause)
202
566266
1769
(๋ฐ•์ˆ˜)
09:28
It feels like there'd be an instant demand for an app like this right now,
203
568059
3600
์ด ์–ดํ”Œ์„ ์ถœ์‹œํ•˜์ž๋งˆ์ž ์–ด๋งˆ์–ด๋งˆํ•œ ์ˆซ์ž์˜ ์‚ฌ๋žŒ๋“ค์ด
09:31
that you might freak out a lot of people.
204
571683
1966
๊ทธ ์„ฑ๋Šฅ์— ๊นœ์ง ๋†€๋ผ ์ด์šฉํ•˜๊ณ  ์‹ถ์–ดํ•  ๊ฒƒ ๊ฐ™์€๋ฐ์š”.
09:33
Are you thinking of doing this, making an app that allows self-checking?
205
573673
3527
์ž๊ฐ€ ์ง„๋‹จ์ด ๊ฐ€๋Šฅํ•œ ์–ดํ”Œ์„ ๋งŒ๋“ค ๊ณ„ํš์„ ๊ฐ–๊ณ  ์žˆ๋‚˜์š”?
09:37
ST: So my in-box is flooded about cancer apps,
206
577224
4973
์„ธ๋ฐ”์Šค์ฐฌ: ์ง€๊ธˆ ์ œ ์ด๋ฉ”์ผ์€ ๊ฐ€์Šด ์•„ํ”ˆ ์‚ฌ์—ฐ์„ ๊ฐ€์ง„ ์‚ฌ๋žŒ๋“ค์ด ๋ณด๋‚ธ
09:42
with heartbreaking stories of people.
207
582221
2303
์–ดํ”Œ ๊ด€๋ จ ๋ฉ”์ผ๋กœ ํญ์ฃผํ•˜๋Š” ์ƒํ™ฉ์ธ๋ฐ์š”.
09:44
I mean, some people have had 10, 15, 20 melanomas removed,
208
584548
3204
์–ด๋–ค ์‚ฌ๋žŒ๋“ค์€ 10๊ฐœ, 15๊ฐœ, 20๊ฐœ์˜ ํ‘์ƒ‰์ข…์„ ์ œ๊ฑฐํ•˜๊ณ ๋„
09:47
and are scared that one might be overlooked, like this one,
209
587776
3952
์ง€๊ธˆ์ฒ˜๋Ÿผ ๋†“์นœ ํ‘์ƒ‰์ข…์ด ํ•˜๋‚˜๋ผ๋„ ์žˆ์„๊นŒ ๋‘๋ ค์›Œํ•ฉ๋‹ˆ๋‹ค.
09:51
and also, about, I don't know,
210
591752
1741
๋˜ ์ €๋„ ์ž˜ ๋ชจ๋ฅด์ง€๋งŒ
09:53
flying cars and speaker inquiries these days, I guess.
211
593517
2732
๋‚ ์•„๋‹ค๋‹ˆ๋Š” ์ž๋™์ฐจ๋ฅผ ๋น„๋กฏํ•ด์„œ
09:56
My take is, we need more testing.
212
596273
2738
์•„์ง ๋” ๋งŽ์€ ์‹œํ–‰์ฐฉ์˜ค๊ฐ€ ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค.
09:59
I want to be very careful.
213
599449
1778
์‹ ์ค‘ํ•˜๊ฒŒ ๊ฐ€๊ณ  ์‹ถ์–ด์š”.
10:01
It's very easy to give a flashy result and impress a TED audience.
214
601251
3666
๋†€๋ผ์šด ๊ฒฐ๊ณผ๋กœ TED์˜ ์ฒญ์ค‘๋“ค์—๊ฒŒ ๊นŠ์€ ์ธ์ƒ์„ ์‹ฌ์–ด์ฃผ๋Š” ์ผ์€ ์‰ฝ์ง€๋งŒ
10:04
It's much harder to put something out that's ethical.
215
604941
2627
์œค๋ฆฌ์ ์ธ ๋ฌธ์ œ๊ฐ€ ๊ฒฐ๋ถ€๋˜์–ด ์žˆ๋Š”๋งŒํผ ์„ฃ๋ถˆ๋ฆฌ ์›€์ง์ผ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค.
10:07
And if people were to use the app
216
607592
2394
๋งŒ์•ฝ ์‚ฌ๋žŒ๋“ค์ด ์˜์‚ฌ์™€ ์ƒ๋‹ดํ•˜๋Š” ๋Œ€์‹ 
10:10
and choose not to consult the assistance of a doctor
217
610010
2797
์šฐ๋ฆฌ ์–ดํ”Œ์„ ์ด์šฉํ•˜๊ณ 
10:12
because we get it wrong,
218
612831
1583
๊ทธ ๋•Œ๋ฌธ์— ๋‚˜์œ ๊ฒฐ๊ณผ๊ฐ€ ๋ฐœ์ƒํ•œ๋‹ค๋ฉด
10:14
I would feel really bad about it.
219
614438
1653
์ €๋Š” ๊ต‰์žฅํžˆ ๋ถˆํŽธํ•ด์งˆ ๊ฒƒ ๊ฐ™์•„์š”.
10:16
So we're currently doing clinical tests,
220
616115
1925
ํ˜„์žฌ ์šฐ๋ฆฌ๋Š” ์ž„์ƒ ์‹คํ—˜์„ ์ง„ํ–‰์ค‘์ž…๋‹ˆ๋‹ค.
10:18
and if these clinical tests commence and our data holds up,
221
618064
2798
์ž„์ƒ ์‹คํ—˜์ด ์‹œ์ž‘๋˜๊ณ 
10:20
we might be able at some point to take this kind of technology
222
620886
2990
์–ธ์  ๊ฐ€๋Š” ์ด ๊ธฐ์ˆ ์„
10:23
and take it out of the Stanford clinic
223
623900
1892
์Šคํƒ ํฌ๋“œ ๋ณ‘์› ๋ฟ๋งŒ ์•„๋‹ˆ๋ผ
10:25
and bring it to the entire world,
224
625816
1658
์Šคํƒ ํฌ๋“œ์˜ ์˜์‚ฌ๋“ค์ด ์ฐพ์•„๊ฐˆ ์ˆ˜ ์—†๋Š”
10:27
places where Stanford doctors never, ever set foot.
225
627498
2456
์ „ ์„ธ๊ณ„์—์„œ ํ™œ์šฉํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋˜๊ฒ ์ฃ .
10:30
CA: And do I hear this right,
226
630617
2580
ํฌ๋ฆฌ์Šค: ์ œ๊ฐ€ ์ œ๋Œ€๋กœ ๋“ค์€๊ฒŒ ๋งž๋‹ค๋ฉด
10:33
that it seemed like what you were saying,
227
633221
1966
๋‹น์‹  ๋ง์ธ์ฆ‰์Šจ,
10:35
because you are working with this army of Udacity students,
228
635211
4254
์œ ๋‹ค์‹œํ‹ฐ์˜ ํ•™์ƒ๋“ค๊ณผ ํ•จ๊ป˜ ์ž‘์—…ํ•˜๊ธฐ ๋•Œ๋ฌธ์—
10:39
that in a way, you're applying a different form of machine learning
229
639489
3221
๋‹น์‹ ์€ ๊ธฐ์—…์ด๋‚˜ ๊ฐ€๋Šฅํ•  ๋ฒ•ํ•œ, ๋จธ์‹  ๋Ÿฌ๋‹ ๊ธฐ์ˆ ๊ณผ ์ง‘๋‹จ ์ง€์„ฑ์„
10:42
than might take place in a company,
230
642734
1735
๊ฒฐํ•ฉํ•ด์„œ ์ƒˆ๋กœ์šด ์ข…๋ฅ˜์˜ ๋จธ์‹  ๋Ÿฌ๋‹์„
10:44
which is you're combining machine learning with a form of crowd wisdom.
231
644493
3484
ํ˜„์‹ค ๋ถ„์•ผ์— ์ ์šฉํ•˜๋ ค๊ณ  ํ•œ๋‹ค๋Š” ๋ง์ธ๊ฐ€์š”?
10:48
Are you saying that sometimes you think that could actually outperform
232
648001
3384
๋‹น์‹  ๋ง๋Œ€๋กœ๋ผ๋ฉด ์–ธ์  ๊ฐ€ ๋จธ์‹  ๋Ÿฌ๋‹์ด
์ผ๊ฐœ ํšŒ์‚ฌ, ์‹ฌ์ง€์–ด๋Š” ๊ฑฐ๋Œ€ ํšŒ์‚ฌ์˜ ๋Šฅ๋ ฅ์„ ์•ž์„ค ์ˆ˜ ์žˆ๋‹ค๊ณ  ๋ณด์‹ญ๋‹ˆ๊นŒ?
10:51
what a company can do, even a vast company?
233
651409
2050
10:53
ST: I believe there's now instances that blow my mind,
234
653483
2940
์„ธ๋ฐ”์Šค์ฐฌ: ์ € ์Šค์Šค๋กœ๋„ ๊ฐํƒ„ํ•˜๋Š” ์ˆœ๊ฐ„๋“ค์ด ์žˆ์ฃ .
10:56
and I'm still trying to understand.
235
656447
1758
์•„์ง๋„ ์ดํ•ดํ•˜๋ ค๊ณ  ๋…ธ๋ ฅํ•ด์š”.
10:58
What Chris is referring to is these competitions that we run.
236
658229
3937
ํฌ๋ฆฌ์Šค๋Š” ์šฐ๋ฆฌ๊ฐ€ ์ง„ํ–‰ํ•˜๋Š” ๊ฒฝ์—ฐ๋Œ€ํšŒ๋ฅผ ๋งํ•˜๋Š” ๊ฑฐ์—์š”.
11:02
We turn them around in 48 hours,
237
662190
2268
์ œํ•œ ์‹œ๊ฐ„์ด 48์‹œ๊ฐ„์ด๊ตฌ์š”.
11:04
and we've been able to build a self-driving car
238
664482
2252
๊ทธ ์‹œ๊ฐ„ ๋‚ด์— ์ž์œจ ์ฃผํ–‰ ์ž๋™์ฐจ๋ฅผ ๋งŒ๋“ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
11:06
that can drive from Mountain View to San Francisco on surface streets.
239
666758
3387
์ด ์ฐจ๋Š” ํ‰๋ฉด ๊ฐ€๋กœ์—์„œ ๋งˆ์šดํ‹ด ๋ทฐ๋ถ€ํ„ฐ ์ƒŒํ”„๋ž€์‹œ์Šค์ฝ”๊นŒ์ง€ ์ฃผํ–‰ ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.
11:10
It's not quite on par with Google after seven years of Google work,
240
670169
3584
7๋…„์ด๋‚˜ ์—ฐ๊ตฌํ•œ ๊ตฌ๊ธ€๊ณผ ๋น„๊ตํ•  ๋ฐ”๋Š” ์•„๋‹ˆ์ง€๋งŒ
11:13
but it's getting there.
241
673777
2528
๊ฝค ์„ฑ๊ณต์ ์ด์—์š”.
11:16
And it took us only two engineers and three months to do this.
242
676329
3084
์—ฐ๊ตฌ์› ๋‘ ๋ช…์ด ๊ณ ์ž‘ ์„ธ ๋‹ฌ๋งŒ์— ์™„์„ฑํ–ˆ์ฃ .
11:19
And the reason is, we have an army of students
243
679437
2856
๊ทธ ์ด์œ ์ธ์ฆ‰์Šจ ์šฐ๋ฆฌ์—๊ฒŒ๋Š” ๋Œ€ํšŒ์— ์ฐธ๊ฐ€ํ•œ
11:22
who participate in competitions.
244
682317
1850
์ˆ˜ ๋งŽ์€ ํ•™์ƒ๋“ค์ด ์žˆ๊ฑฐ๋“ ์š”.
11:24
We're not the only ones who use crowdsourcing.
245
684191
2220
ํฌ๋ผ์šฐ๋“œ์†Œ์‹ฑ์„ ํ™œ์šฉํ•˜๋Š” ๊ฒƒ์€ ์šฐ๋ฆฌ๋ฟ๋งŒ์ด ์•„๋‹™๋‹ˆ๋‹ค.
11:26
Uber and Didi use crowdsource for driving.
246
686435
2223
์šฐ๋ฒ„์™€ ๋””๋””๋„ ์ฃผํ–‰์— ํฌ๋ผ์šฐ๋“œ์†Œ์‹ฑ์„ ํ™œ์šฉํ•˜์ฃ .
11:28
Airbnb uses crowdsourcing for hotels.
247
688682
2759
์—์–ด๋น„์•ค๋น„๋„ ํฌ๋ผ์šฐ๋“œ์†Œ์‹ฑ์„ ํ˜ธํ…”์— ์ ์šฉํ–ˆ์Šต๋‹ˆ๋‹ค.
11:31
There's now many examples where people do bug-finding crowdsourcing
248
691465
4007
๋ฒ„๊ทธ๋ฅผ ์žก๊ธฐ ์œ„ํ•ด ํฌ๋ผ์šฐ๋“œ์†Œ์‹ฑ์„ ํ™œ์šฉํ•˜๋Š” ์˜ˆ๋Š” ์–ผ๋งˆ๋“ ์ง€ ์žˆ์Šต๋‹ˆ๋‹ค.
11:35
or protein folding, of all things, in crowdsourcing.
249
695496
2804
๋‹จ๋ฐฑ์งˆ ์ค‘์ฒฉ์ด๋‚˜ ๋ชจ๋“  ๋ถ„์•ผ์—์„œ ๋ง์ด์ฃ .
11:38
But we've been able to build this car in three months,
250
698324
2915
ํ•˜์ง€๋งŒ ์šฐ๋ฆฌ๋Š” 3๋‹ฌ๋งŒ์— ์ฐจ๋ฅผ ์™„์„ฑํ–ˆ๊ณ 
11:41
so I am actually rethinking
251
701263
3655
์ €๋Š” ํšŒ์‚ฌ๋ฅผ ์–ด๋–ป๊ฒŒ ๊ตฌ์„ฑํ•ด์•ผ ํ• ์ง€
11:44
how we organize corporations.
252
704942
2238
๋‹ค์‹œ ๊ตฌ์ƒ์ค‘์ž…๋‹ˆ๋‹ค.
11:47
We have a staff of 9,000 people who are never hired,
253
707204
4696
์ง€๊ธˆ ์šฐ๋ฆฌ์—๊ฒŒ๋Š” ๊ณ ์šฉํ•œ ์ ๋„ ์—†๊ณ  ํ•ด๊ณ ํ•˜์ง€๋„ ์•Š๋Š” ์ง์›์ด
11:51
that I never fire.
254
711924
1308
9์ฒœ๋ช… ์žˆ์Šต๋‹ˆ๋‹ค.
11:53
They show up to work and I don't even know.
255
713256
2362
์ €๋„ ๋ชจ๋ฅด๋Š” ์ƒˆ์— ์‚ฌ๋žŒ๋“ค์€ ์ผ์„ ํ•˜๊ณ 
11:55
Then they submit to me maybe 9,000 answers.
256
715642
3058
9์ฒœ ๊ฐœ์˜ ์‘๋‹ต์ด ์ €์—๊ฒŒ ์ „๋‹ฌ๋ฉ๋‹ˆ๋‹ค.
11:58
I'm not obliged to use any of those.
257
718724
2176
๊ทธ ์ค‘ ์–ด๋Š ๊ฒƒ์„ ์‚ฌ์šฉํ•ด์•ผ๋งŒ ํ•œ๋‹ค๋Š” ๊ทœ์ •๋„ ์—†์Šต๋‹ˆ๋‹ค.
12:00
I end up -- I pay only the winners,
258
720924
1991
๋‹ต๋ณ€ ์ค‘ ๊ฐ€์žฅ ์ข‹์€ ๊ฒƒ์„ ๊ณจ๋ผ ๋Œ“๊ฐ€๋ฅผ ์ง€๋ถˆํ•˜์ฃ .
12:02
so I'm actually very cheapskate here, which is maybe not the best thing to do.
259
722939
3718
์ด ์ ์—์„œ๋Š” ๊ฝค๋‚˜ ๊ตฌ๋‘์‡ ์ฒ˜๋Ÿผ ํ–‰๋™ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ตœ์„ ์€ ์•„๋‹ˆ์ฃ .
12:06
But they consider it part of their education, too, which is nice.
260
726681
3185
ํ•˜์ง€๋งŒ ์‘๋‹ต์ž๋“ค์—๊ฒŒ๋„ ๊ฒฝํ—˜์ด ๋  ๊ฑฐ๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
12:09
But these students have been able to produce amazing deep learning results.
261
729890
4201
ํ•™์ƒ๋“ค๋„ ๋”ฅ๋Ÿฌ๋‹์„ ์ด์šฉํ•ด ๋†€๋ผ์šด ๊ฒฐ๊ณผ๋ฌผ์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ์ฃ .
12:14
So yeah, the synthesis of great people and great machine learning is amazing.
262
734115
3861
ํ›Œ๋ฅญํ•œ ์‚ฌ๋žŒ๋“ค๊ณผ ํ›Œ๋ฅญํ•œ ๋จธ์‹  ๋Ÿฌ๋‹์ด ๋งŒ๋‚œ ๊ฒฐ๊ณผ๋Š” ์ •๋ง ๋†€๋ž์Šต๋‹ˆ๋‹ค.
12:18
CA: I mean, Gary Kasparov said on the first day [of TED2017]
263
738000
2814
ํฌ๋ฆฌ์Šค: ๊ฐœ๋ฆฌ ์นด์ŠคํŒŒ๋กœํ”„๊ฐ€ TED2017์˜ ์ฒซ์งธ๋‚  ํ•œ ๋ง์ธ๋ฐ์š”.
12:20
that the winners of chess, surprisingly, turned out to be two amateur chess players
264
740848
5412
์ฒด์Šค ๋Œ€ํšŒ์˜ ์šฐ์Šน์ž๊ฐ€ ์‚ฌ์‹ค์€ ๋‘ ๋ช…์˜ ์•„๋งˆ์ถ”์–ด ์ฒด์Šค ๊ธฐ์‚ฌ
12:26
with three mediocre-ish, mediocre-to-good, computer programs,
265
746284
5371
์„ธ ๊ฐ€์ง€์˜ ์ค‘๊ฐ„, ํ˜น์€ ์ค‘์ƒ๊ธ‰์˜ ์ปดํ“จํ„ฐ ํ”„๋กœ๊ทธ๋žจ์„ ์ด์šฉํ•˜์—ฌ
12:31
that could outperform one grand master with one great chess player,
266
751679
3163
ํ•œ ๋ช…์˜ ๊ทธ๋žœ๋“œ ๋งˆ์Šคํ„ฐ, ํ•œ ๋ช…์˜ ๋›ฐ์–ด๋‚œ ์ฒด์Šค ๊ธฐ์‚ฌ๋ฅผ ์ด๊ฒผ๋‹ค๋Š” ๊ฑฐ์š”.
12:34
like it was all part of the process.
267
754866
1743
์ด๊ฒƒ๋„ ์—ญ์‹œ ๊ทธ ๊ณผ์ •์˜ ์ผ๋ถ€์˜€๋˜๊ฑฐ์ฃ .
12:36
And it almost seems like you're talking about a much richer version
268
756633
3335
๋‹น์‹ ์ด ๋งํ•˜๋Š” ๋ฒ„์ „์€
12:39
of that same idea.
269
759992
1200
์ด๋ณด๋‹ค ๋” ์ƒ์œ„ํ˜ธํ™˜ ๋ฒ„์ „ ๊ฐ™์€๋ฐ์š”.
12:41
ST: Yeah, I mean, as you followed the fantastic panels yesterday morning,
270
761216
3857
์„ธ๋ฐ”์Šค์ฐฌ: ์–ด์ œ ์•„์นจ์— ๋“ค์œผ์…จ๋˜ ๊ฒƒ์ฒ˜๋Ÿผ
12:45
two sessions about AI,
271
765097
1994
๋‘ ๋ฒˆ์˜ ์„ธ์…˜์„ ํ†ตํ•ด ์ธ๊ณต์ง€๋Šฅ
12:47
robotic overlords and the human response,
272
767115
2167
๋กœ๋ด‡์— ์˜ํ•œ ์ธ๊ฐ„ ์ง€๋ฐฐ๊ณผ ์‚ฌ๋žŒ๋“ค์˜ ๋ฐ˜์‘ ๋“ฑ
12:49
many, many great things were said.
273
769306
1982
์•„์ฃผ ํ›Œ๋ฅญํ•œ ์ฃผ์ œ๋“ค์— ๊ด€ํ•ด ์ด์•ผ๊ธฐํ–ˆ์ฃ .
12:51
But one of the concerns is that we sometimes confuse
274
771312
2687
๋‹ค๋งŒ ์‚ฌ๋žŒ๋“ค์ด ์ข…์ข… ์šฐ๋ ค๋ฅผ ํ‘œํ•˜๋Š” ๋ถ€๋ถ„์€
12:54
what's actually been done with AI with this kind of overlord threat,
275
774023
4062
์šฐ๋ฆฌ๊ฐ€ ๋งŒ๋“œ๋Š” AI๊ฐ€ ๊ฐ–๋Š” ์ž์˜์‹์ด ๋กœ๋ด‡์˜ ์ธ๊ฐ„ ์ง€๋ฐฐ๋ผ๋Š”
12:58
where your AI develops consciousness, right?
276
778109
3424
๊ด€์ ์—์„œ ๋น„์ถฐ๋ณผ ๋•Œ ์–ด๋Š ์ •๋„ ์ˆ˜์ค€๊นŒ์ง€ ๋„๋‹ฌํ–ˆ๋ƒํ•˜๋Š” ๊ฑฐ์ฃ .
13:01
The last thing I want is for my AI to have consciousness.
277
781557
2971
์ œ ๋งˆ์ง€๋ง‰ ๋ชฉํ‘œ๋Š” AI์— ์˜์‹์„ ์‹ฌ์–ด์ฃผ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
13:04
I don't want to come into my kitchen
278
784552
1716
์ €๋Š” ๋ถ€์—Œ์— ๋“ค์–ด๊ฐ”์„ ๋•Œ
13:06
and have the refrigerator fall in love with the dishwasher
279
786292
4193
์‹๊ธฐ ์„ธ์ฒ™๊ธฐ์™€ ์‚ฌ๋ž‘์— ๋น ์ง„ ๋ƒ‰์žฅ๊ณ ๊ฐ€
13:10
and tell me, because I wasn't nice enough,
280
790509
2124
์ œ๊ฐ€ ์นœ์ ˆํ•˜๊ฒŒ ๊ตด์ง€ ์•Š์•„์„œ
13:12
my food is now warm.
281
792657
1837
์Œ์‹๋“ค์ด ๋…น๊ฒŒ ๋‚ด๋ฒ„๋ ค๋’€๋‹ค๊ณ  ๋งํ•˜๋Š” ์ƒํ™ฉ์„ ์›ํ•˜์ง€ ์•Š์•„์š”.
13:14
I wouldn't buy these products, and I don't want them.
282
794518
2891
์ด๋Ÿฐ ์ œํ’ˆ๋“ค์€ ์‚ฌ๊ณ  ์‹ถ์ง€๋„ ์•Š๊ณ  ์›ํ•˜์ง€๋„ ์•Š์•„์š”.
13:17
But the truth is, for me,
283
797825
1802
ํ•˜์ง€๋งŒ ์ œ ์ƒ๊ฐ์— ํ•œ๊ฐ€์ง€ ํ™•์‹คํ•œ ๊ฒƒ์€
13:19
AI has always been an augmentation of people.
284
799651
2720
AI๋Š” ์–ธ์ œ๋‚˜ ์‚ฌ๋žŒ๋“ค์„ ๋•๊ธฐ ์œ„ํ•œ ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
13:22
It's been an augmentation of us,
285
802893
1676
์šฐ๋ฆฌ๋ฅผ ๋•๊ธฐ ์œ„ํ•œ ๊ฒƒ
13:24
to make us stronger.
286
804593
1457
๋” ๋‚˜์•„์ง€๋„๋ก ๋„์›€์„ ์ฃผ๋Š” ๊ฒƒ์ด์ฃ .
13:26
And I think Kasparov was exactly correct.
287
806074
2831
์ €๋Š” ์นด์ŠคํŒŒ๋กœํ”„๊ฐ€ ํ•œ ๋ง์ด ์ •ํ™•ํ–ˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ด์š”.
13:28
It's been the combination of human smarts and machine smarts
288
808929
3849
์ธ๊ฐ„์˜ ์ง€๋Šฅ๊ณผ ๊ธฐ๊ณ„์˜ ์ง€๋Šฅ์ด ๋งŒ๋‚˜์„œ
13:32
that make us stronger.
289
812802
1464
๋” ํฐ ํž˜์ด ๋˜๋Š” ๊ฒƒ์ด์ฃ .
13:34
The theme of machines making us stronger is as old as machines are.
290
814290
4587
๊ธฐ๊ณ„๊ฐ€ ํƒ„์ƒํ•œ ์ดํ›„๋กœ ๊ทธ ๋ชฉ์ ์€ ์ค„๊ณง ์ธ๋ฅ˜๋ฅผ ์œ„ํ•œ ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
13:39
The agricultural revolution took place because it made steam engines
291
819567
3758
๋†์—… ํ˜๋ช…์€ ์ฆ๊ธฐ๊ธฐ๊ด€๊ณผ
13:43
and farming equipment that couldn't farm by itself,
292
823349
2666
๋†์ž‘ ๊ธฐ๊ณ„๋“ค์„ ๋งŒ๋“ค์–ด ๋ƒˆ์ง€๋งŒ ์ด๊ฒƒ๋“ค์€ ์ธ๊ฐ„์„ ๋Œ€์ฒดํ•˜์ง€ ์•Š์•˜๊ณ 
13:46
that never replaced us; it made us stronger.
293
826039
2122
์˜คํžˆ๋ ค ์šฐ๋ฆฌ์˜ ๋†์—… ํ™œ๋™์„ ์ข€ ๋” ์šฉ์ดํ•˜๊ฒŒ ํ•ด์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค.
13:48
And I believe this new wave of AI will make us much, much stronger
294
828185
3738
๋”ฐ๋ผ์„œ ์ €๋Š” ์ง€๊ธˆ์˜ ์ƒˆ๋กœ์šด ์ธ๊ณต์ง€๋Šฅ ๋ฐ”๋žŒ๋„ ์ธ๋ฅ˜๋ฅผ ๋”์šฑ
13:51
as a human race.
295
831947
1183
ํŽธ๋ฆฌํ•˜๊ฒŒ ํ•ด์ค„ ๊ฒƒ์ด๋ผ ๋ฏฟ์Šต๋‹ˆ๋‹ค.
13:53
CA: We'll come on to that a bit more,
296
833765
1813
ํฌ๋ฆฌ์Šค: ์ด ์ ์— ๋Œ€ํ•ด ์ข€ ๋” ์ด์•ผ๊ธฐํ•  ๊ฒƒ์ด ์žˆ๋Š”๋ฐ์š”.
13:55
but just to continue with the scary part of this for some people,
297
835602
3671
์ด ์ ์— ๋‘๋ ค์›€์„ ๊ฐ–๊ณ  ๊ณ„์‹  ๋ถ„๋“ค์„ ์œ„ํ•ด์„œ ๋ง์ด์ฃ .
13:59
like, what feels like it gets scary for people is when you have
298
839297
3558
์Šค์Šค๋กœ ์ฝ”๋“œ๋ฅผ ๋‹ค์‹œ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋Š” ์ปดํ“จํ„ฐ๋ฅผ ๋งŒ๋“ค๊ณ 
14:02
a computer that can, one, rewrite its own code,
299
842879
4618
์ด ์ปดํ“จํ„ฐ๊ฐ€ ์ž์‹ ๊ณผ ๋™์ผํ•œ ์ปดํ“จํ„ฐ๋ฅผ ์žฌ์ƒ์‚ฐํ•  ์ˆ˜ ์žˆ๋Š” ๋Šฅ๋ ฅ์€ ๋ฌผ๋ก 
14:07
so, it can create multiple copies of itself,
300
847521
3584
๋‹ค๋ฅธ ๋ฒ„์ „์˜ ์ฝ”๋“œ๋ฅผ ๊ฐ€์ง„
14:11
try a bunch of different code versions,
301
851129
1897
์ปดํ“จํ„ฐ๋ฅผ ์—ฌ๋Ÿฌ ๋Œ€ ๋งŒ๋“ค์–ด ๋ณด๋Š” ๊ฑฐ์ฃ .
14:13
possibly even at random,
302
853050
1775
์‹ฌ์ง€์–ด ๋ฌด์ž‘์œ„๋กœ์š”.
14:14
and then check them out and see if a goal is achieved and improved.
303
854849
3632
๊ทธ๋ฆฌ๊ณ  ๋ชฉํ‘œ๊ฐ€ ๋‹ฌ์„ฑ๋˜๊ณ  ๋” ํ–ฅ์ƒ๋˜์—ˆ๋Š”์ง€ ํ™•์ธํ•˜๋Š” ๊ฒ๋‹ˆ๋‹ค.
14:18
So, say the goal is to do better on an intelligence test.
304
858505
3641
๊ทธ ๋ชฉํ‘œ๋ž€ ์ง€๋Šฅ ์‹œํ—˜์—์„œ ๋” ์ข‹์€ ์„ฑ์  ์„ ๋‚ด๋Š” ๊ฒƒ์ด๋ผ๊ณ  ์น˜์ฃ .
14:22
You know, a computer that's moderately good at that,
305
862170
3894
์ปดํ“จํ„ฐ๋Š” ๊ทธ๋Ÿฐ ๋ถ„์•ผ์—์„œ๋Š” ๊ฝค ํ›Œ๋ฅญํ•˜์ž–์•„์š”.
14:26
you could try a million versions of that.
306
866088
2509
๋ฐฑ๋งŒ ๊ฐœ์˜ ๋‹ค๋ฅธ ๋ฒ„์ „์„ ์‹œํ—˜ํ•ด๋ณผ ์ˆ˜ ์žˆ๊ณ 
14:28
You might find one that was better,
307
868621
2090
์ข‹์€ ๊ฒƒ์„ ์ฐพ์•„๋‚ด์„œ
14:30
and then, you know, repeat.
308
870735
2004
๊ทธ๊ฒƒ๋“ค์„ ๋ชจ์•„์„œ ๋‹ค์‹œ ์œ„ ๊ณผ์ •์„ ๋ฐ˜๋ณตํ•˜๋Š” ๊ฑฐ์ฃ .
14:32
And so the concern is that you get some sort of runaway effect
309
872763
3040
์‚ฌ๋žŒ๋“ค์˜ ์šฐ๋ ค๋ž€ ์ปดํ“จํ„ฐ๋ฅผ ํ†ต์ œํ•˜์ง€ ๋ชปํ•˜๋Š” ์ƒํ™ฉ๊ณผ ๋งˆ์ฃผ์น˜์ง€ ์•Š์„๊นŒ ํ•˜๋Š”
14:35
where everything is fine on Thursday evening,
310
875827
3008
๊ฑด๋ฐ์š”. ๋ชฉ์š”์ผ ๋ฐค๊นŒ์ง€ ๋ชจ๋“  ๊ฒŒ ์ •์ƒ์ด์—ˆ๋Š”๋ฐ
14:38
and you come back into the lab on Friday morning,
311
878859
2336
๊ธˆ์š”์ผ ์•„์นจ์— ์—ฐ๊ตฌ์†Œ์— ๋“ค์–ด์˜ค์ž
14:41
and because of the speed of computers and so forth,
312
881219
2449
์ปดํ“จํ„ฐ๊ฐ€ ํญ์ฃผํ•˜๋ฉด์„œ ๋ฌธ์ œ๋ฅผ ์ผ์œผํ‚ค๊ณ 
14:43
things have gone crazy, and suddenly --
313
883692
1903
์ƒํ™ฉ์ด ํ†ต์ œ๋ถˆ๋Šฅ์œผ๋กœ ๋ณ€ํ•ด๋ฒ„๋ฆฌ๋Š”
14:45
ST: I would say this is a possibility,
314
885619
2020
์„ธ๋ฐ”์Šค์ฐฌ: ๊ฐ€๋Šฅ์„ฑ์— ๋ถˆ๊ณผํ•ฉ๋‹ˆ๋‹ค.
14:47
but it's a very remote possibility.
315
887663
1916
ํ•˜์ง€๋งŒ ์•„์ฃผ ๋จผ ๋ฏธ๋ž˜์— ์ผ์–ด๋‚  ๊ฐ€๋Šฅ์„ฑ์ด์ฃ .
14:49
So let me just translate what I heard you say.
316
889603
3337
์ œ ์‹๋Œ€๋กœ ํ•œ ๋ฒˆ ๋งํ•ด๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.
14:52
In the AlphaGo case, we had exactly this thing:
317
892964
2704
์•ŒํŒŒ๊ณ ๋ฅผ ํ†ตํ•ด ์šฐ๋ฆฌ๋Š” ๋‹ค์Œ ์‚ฌ์‹ค์„ ๋ชฉ๊ฒฉํ–ˆ์Šต๋‹ˆ๋‹ค.
14:55
the computer would play the game against itself
318
895692
2315
์ปดํ“จํ„ฐ๊ฐ€ ์ž๊ธฐ ์ž์‹ ์„ ์ƒ๋Œ€๋กœ ๋ฐ”๋‘‘์„ ๋‘˜ ์ˆ˜ ์žˆ๊ณ 
14:58
and then learn new rules.
319
898031
1250
์ƒˆ๋กœ์šด ๊ทœ์น™์„ ๋ฐฐ์šธ ์ˆ˜ ์žˆ์ฃ .
14:59
And what machine learning is is a rewriting of the rules.
320
899305
3235
ํ•˜์ง€๋งŒ ๋จธ์‹  ๋Ÿฌ๋‹์€ ์ด๋ฏธ ์กด์žฌํ•˜๋Š” ๊ทœ์น™์„ ๋‹ค์‹œ ์‚ฌ์šฉํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
15:02
It's the rewriting of code.
321
902564
1769
์›๋ž˜ ์žˆ๋˜ ์ฝ”๋“œ๋ฅผ ๋‹ค์‹œ ์‚ฌ์šฉํ•˜๋Š” ๊ฑฐ๊ตฌ์š”.
15:04
But I think there was absolutely no concern
322
904357
2845
์ €๋Š” ์•ŒํŒŒ๊ณ ๊ฐ€ ์ง€๊ตฌ๋ฅผ ์ง€๋ฐฐํ•  ๊ฑฐ๋ผ๋Š” ๊ฑฑ์ •์€
15:07
that AlphaGo would take over the world.
323
907226
2426
์™„์ „ํžˆ ๊ธฐ์šฐ๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
15:09
It can't even play chess.
324
909676
1464
์•ŒํŒŒ๊ณ ๋Š” ์ฒด์Šค๋„ ๋‘์ง€ ๋ชปํ•ด์š”.
15:11
CA: No, no, no, but now, these are all very single-domain things.
325
911164
5147
ํฌ๋ฆฌ์Šค: ํ•˜์ง€๋งŒ ์šฐ๋ฆฌ๋Š” ์ง€๊ธˆ ํ•œ ๋ถ„์•ผ์— ๋Œ€ํ•ด์„œ๋งŒ ์ด์•ผ๊ธฐํ•˜๊ณ  ์žˆ์–ด์š”.
15:16
But it's possible to imagine.
326
916335
2879
๊ฐ€๋Šฅํ•  ๋ฒ•ํ•œ ์ƒํ™ฉ์ด๋ผ๊ณ  ๋ณด๋Š”๋ฐ์š”.
15:19
I mean, we just saw a computer that seemed nearly capable
327
919238
3089
์šฐ๋ฆฌ๋Š” ๋Œ€ํ•™ ์ž…ํ•™ ์‹œํ—˜์„ ํ†ต๊ณผํ•  ์ˆ˜ ์žˆ์„ ์ •๋„์˜
15:22
of passing a university entrance test,
328
922351
2655
์ปดํ“จํ„ฐ๋ฅผ ๋ชฉ๊ฒฉํ–ˆ์Šต๋‹ˆ๋‹ค.
15:25
that can kind of -- it can't read and understand in the sense that we can,
329
925030
3688
์‚ฌ๋žŒ์ฒ˜๋Ÿผ ๊ธ€์„ ์ฝ๊ณ  ์ดํ•ดํ•  ์ˆ˜๋Š” ์—†์ง€๋งŒ
15:28
but it can certainly absorb all the text
330
928742
1987
๋Œ€์‹  ๋ฌธ์žฅ ์ž์ฒด๋ฅผ ํ†ต์งธ๋กœ ํก์ˆ˜ํ•ด๋ฒ„๋ฆฌ์ฃ .
15:30
and maybe see increased patterns of meaning.
331
930753
2899
๊ทธ๋ฆฌ๊ณ  ์ดํ•ด์˜ ํญ์„ ๋„“ํž์ง€๋„ ๋ชจ๋ฆ…๋‹ˆ๋‹ค.
15:33
Isn't there a chance that, as this broadens out,
332
933676
3694
๊ทธ๋ ‡๋‹ค๋ฉด ๊ทธ ์˜์—ญ์„ ์ ์ฐจ ๋„“ํ˜€์„œ
15:37
there could be a different kind of runaway effect?
333
937394
2466
์–ธ์  ๊ฐ€ ํ†ต์ œ๊ฐ€ ๋ถˆ๊ฐ€๋Šฅํ•œ ์ƒํ™ฉ์ด ์ƒ๊ธธ ์ˆ˜ ์žˆ์ง€ ์•Š์„๊นŒ์š”?
15:39
ST: That's where I draw the line, honestly.
334
939884
2078
์„ธ๋ฐ”์Šค์ฐฌ: ๊ทธ๋ ‡์ง€ ์•Š๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
15:41
And the chance exists -- I don't want to downplay it --
335
941986
2643
๊ฐ€๋Šฅ์„ฑ์€ ์žˆ๊ฒ ์ฃ . ๊ทธ๊ฒƒ๊นŒ์ง€ ๋ถ€์ •ํ•˜๊ณ  ์‹ถ์ง„ ์•Š์•„์š”.
15:44
but I think it's remote, and it's not the thing that's on my mind these days,
336
944653
3672
ํ•˜์ง€๋งŒ ํ˜„์žฌ๋กœ์„  ์•„์ฃผ ๋จผ ์ด์•ผ๊ธฐ์ด๊ณ  ์ง€๊ธˆ ์—ผ๋‘์— ๋‘๊ณ  ์žˆ๋Š” ๋ฌธ์ œ๋Š” ์•„๋‹™๋‹ˆ๋‹ค.
15:48
because I think the big revolution is something else.
337
948349
2512
์ €๋Š” ์ง„์งœ ํ˜๋ช…์€ ๋‹ค๋ฅธ ๊ณณ์—์„œ ์ผ์•„๋‚˜๊ณ  ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ด์š”.
15:50
Everything successful in AI to the present date
338
950885
2922
์ง€๊ธˆ๊นŒ์ง€ ๋ชจ๋“  ์„ฑ๊ณต์ ์ธ AI๋Š”
15:53
has been extremely specialized,
339
953831
2214
๊ทน๋„๋กœ ํŠน์ˆ˜ํ™”, ์ „๋ฌธํ™”๋œ ๋ถ„์•ผ์— ํ•œ์ •๋˜์–ด ์žˆ๊ตฌ์š”.
15:56
and it's been thriving on a single idea,
340
956069
2489
์—„์ฒญ๋‚œ ์–‘์˜ ๋ฐ์ดํ„ฐ๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ํ•œ๋‹ค๋Š” ๋ฐœ์ƒ์— ๊ธฐ๋Œ€์–ด
15:58
which is massive amounts of data.
341
958582
2739
์ง€๊ธˆ๊นŒ์ง€ ์™”์Šต๋‹ˆ๋‹ค.
16:01
The reason AlphaGo works so well is because of massive numbers of Go plays,
342
961345
4147
์•ŒํŒŒ๊ณ ๊ฐ€ ๋ฐ”๋‘‘์„ ์ž˜ ๋‘๋Š” ์ด์œ ๋Š” ์—„์ฒญ๋‚œ ํšŸ์ˆ˜์˜ ๋Œ€๊ตญ์„ ๋‘๊ธฐ ๋•Œ๋ฌธ์ด์ฃ .
16:05
and AlphaGo can't drive a car or fly a plane.
343
965516
3255
์•ŒํŒŒ๊ณ ๋Š” ์ž๋™์ฐจ๋ฅผ ์šด์ „ํ•˜๊ฑฐ๋‚˜ ๋น„ํ–‰๊ธฐ๋ฅผ ๋ชฐ ์ˆ˜ ์—†์–ด์š”.
16:08
The Google self-driving car or the Udacity self-driving car
344
968795
3031
๊ตฌ๊ธ€์ด๋‚˜ ์œ ๋‹ค์‹œํ‹ฐ์˜ ์ž์œจ์ฃผํ–‰์ฐจ๋Š”
16:11
thrives on massive amounts of data, and it can't do anything else.
345
971850
3240
๋งŽ์€ ์–‘์˜ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•  ๋ฟ ๊ทธ ์ด์ƒ์˜ ๊ฒƒ์€ ๋ถˆ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.
16:15
It can't even control a motorcycle.
346
975114
1727
์˜คํ† ๋ฐ”์ด๋„ ์šด์ „ํ•  ์ˆ˜ ์—†์–ด์š”.
16:16
It's a very specific, domain-specific function,
347
976865
2762
์ •๋ง ํŠน์ • ๋ถ„์•ผ์— ํ•œ์ •๋˜์–ด ํ™œ์šฉ ๊ฐ€๋Šฅํ•˜๊ณ 
16:19
and the same is true for our cancer app.
348
979651
1907
์šฐ๋ฆฌ๊ฐ€ ๋งŒ๋“  ์•” ์ง„๋‹จ ์•ฑ๋„ ๋งˆ์ฐฌ๊ฐ€์ง€์ž…๋‹ˆ๋‹ค.
16:21
There has been almost no progress on this thing called "general AI,"
349
981582
3236
์ธ๊ณต์ง€๋Šฅ์—๊ฒŒ "์ƒ๋Œ€์„ฑ ์ด๋ก ์ด๋‚˜ ๋ˆ ์ด๋ก ์„ ๋งŒ๋“ค์–ด ๋ด"๋ผ๊ณ 
16:24
where you go to an AI and say, "Hey, invent for me special relativity
350
984842
4000
๋ช…๋ นํ•  ์ˆ˜ ์žˆ๋Š”, ํ†ต์นญ "์ผ๋ฐ˜ AI" ๋ผ๊ณ  ํ•˜๋Š” ์˜์—ญ์—์„œ๋Š”
16:28
or string theory."
351
988866
1666
๊ฑฐ์˜ ์ง„์ „์ด ์—†๋Š” ์ƒํƒœ์ž…๋‹ˆ๋‹ค.
16:30
It's totally in the infancy.
352
990556
1931
์•„์ง ์œ ์•„๊ธฐ ๋‹จ๊ณ„์ฃ .
16:32
The reason I want to emphasize this,
353
992511
2127
์ œ๊ฐ€ ์ด ์ ์„ ๊ฐ•์กฐํ•˜๋Š” ์ด์œ ๋Š”
16:34
I see the concerns, and I want to acknowledge them.
354
994662
3838
์‚ฌ๋žŒ๋“ค์˜ ์šฐ๋ ค๊ฐ€ ๋ญ”์ง€ ์•Œ๊ธฐ์— ์ œ๋Œ€๋กœ ํ™•์ธ์‹œ์ผœ ์ฃผ๊ณ  ์‹ถ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
16:38
But if I were to think about one thing,
355
998524
2886
์ €๋Š” ์Šค์Šค๋กœ์—๊ฒŒ ์ด๋Ÿฐ ์งˆ๋ฌธ์„ ๋˜์ ธ๋ด…๋‹ˆ๋‹ค.
16:41
I would ask myself the question, "What if we can take anything repetitive
356
1001434
5563
"๋งŒ์•ฝ ์šฐ๋ฆฌ๊ฐ€ ์–ด๋–ค ๋ฐ˜๋ณต ์ž‘์—…์ด๋“ ์ง€
16:47
and make ourselves 100 times as efficient?"
357
1007021
3473
์ง€๊ธˆ๋ณด๋‹ค 100๋ฐฐ ๋” ํšจ์œจ์ ์œผ๋กœ ์ผํ•  ์ˆ˜ ์žˆ๋‹ค๋ฉด ์–ด๋–ป๊ฒŒ ๋ ๊นŒ?"
16:51
It so turns out, 300 years ago, we all worked in agriculture
358
1011170
4249
300๋…„ ์ „ ์ธ๋ฅ˜๋Š” ๋ชจ๋‘ ๋†์‚ฌ๋ฅผ ์ง€์œผ๋ฉฐ
16:55
and did farming and did repetitive things.
359
1015443
2051
๋ฐ˜๋ณต์ ์ธ ์ž‘์—…์„ ํ–ˆ์ฃ .
16:57
Today, 75 percent of us work in offices
360
1017518
2556
์˜ค๋Š˜๋‚  ์šฐ๋ฆฌ ์ค‘ 75%๊ฐ€ ์‚ฌ๋ฌด์‹ค์—์„œ ์ผํ•˜๊ณ 
17:00
and do repetitive things.
361
1020098
2124
์—ญ์‹œ ๋ฐ˜๋ณต ์ž‘์—…์„ ํ•ฉ๋‹ˆ๋‹ค.
17:02
We've become spreadsheet monkeys.
362
1022246
2183
์Šคํ”„๋ ˆ๋“œ์‹œํŠธ๋งŒ ๋งŒ๋“œ๋Š” ์›์ˆญ์ด์ฒ˜๋Ÿผ ๋˜๋ฒ„๋ ธ์–ด์š”.
17:04
And not just low-end labor.
363
1024453
2054
๋‹จ์ˆœํžˆ ์ €์†Œ๋“ ์ง์žฅ ๋ฟ๋งŒ์ด ์•„๋‹ˆ์—์š”.
17:06
We've become dermatologists doing repetitive things,
364
1026531
2754
ํ”ผ๋ถ€๊ณผ ์ „๋ฌธ์˜๋„ ๊ฐ™์€ ์ผ์„ ๋ฐ˜๋ณตํ•˜๊ณ 
17:09
lawyers doing repetitive things.
365
1029309
1749
๋ณ€ํ˜ธ์‚ฌ๋„ ๊ฐ™์€ ์ผ์„ ๋ฐ˜๋ณตํ•˜์ฃ .
17:11
I think we are at the brink of being able to take an AI,
366
1031082
3823
์ €๋Š” ์šฐ๋ฆฌ๊ฐ€ AI๋ฅผ ์ด์šฉํ•ด
17:14
look over our shoulders,
367
1034929
1718
์ง€๊ธˆ๊นŒ์ง€์˜ ์ž์‹ ์„ ๋„˜์–ด
17:16
and they make us maybe 10 or 50 times as effective in these repetitive things.
368
1036671
4058
ํ˜„์žฌ์˜ ๋ฐ˜๋ณต ์ž‘์—…์—์„œ 10๋ฐฐ, 50๋ฐฐ์˜ ํšจ์œจ์„ ๋‚ผ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
17:20
That's what is on my mind.
369
1040753
1275
๊ทธ๊ฒŒ ์š”์ฆ˜ ์ œ๊ฐ€ ํ•˜๋Š” ์ƒ๊ฐ์ž…๋‹ˆ๋‹ค.
17:22
CA: That sounds super exciting.
370
1042052
2450
ํฌ๋ฆฌ์Šค: ์ •๋ง ๋ฉ‹์ง„ ์ด์•ผ๊ธฐ๋„ค์š”.
17:24
The process of getting there seems a little terrifying to some people,
371
1044526
3530
๊ทธ ์ˆ˜์ค€๊นŒ์ง€ ๊ฐ€๋Š” ์ผ์ด ์–ด๋–ค ์ด๋“ค์—๊ฒŒ๋Š” ๋‘๋ ค์›€์œผ๋กœ ๋‹ค๊ฐ€์˜ฌ ๊ฒƒ ๊ฐ™์•„์š”.
17:28
because once a computer can do this repetitive thing
372
1048080
3180
์ปดํ“จํ„ฐ๊ฐ€ ํ”ผ๋ถ€๊ณผ ์˜์‚ฌ๋ณด๋‹ค
17:31
much better than the dermatologist
373
1051284
3434
๋ฐ˜๋ณต ์ž‘์—…์„ ๋” ์ž˜ ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋œ๋‹ค๋ฉด ๋ง์ด์ฃ .
17:34
or than the driver, especially, is the thing that's talked about
374
1054742
3230
์•„๋‹ˆ๋ฉด ์šด์ „ ๊ธฐ์‚ฌ๋“ค๋„ ๊ทธ๋ ‡์ฃ . ์ž์œจ์ฃผํ–‰ ์ž๋™์ฐจ ๋“ฑ์œผ๋กœ
17:37
so much now,
375
1057996
1290
๋งŽ์ด ์–ธ๊ธ‰ํ–ˆ๋Š”๋ฐ์š”.
17:39
suddenly millions of jobs go,
376
1059310
1958
๊ฐ‘์ž๊ธฐ ์ˆ˜๋ฐฑ๋งŒ ๊ฐœ์˜ ์ผ์ž๋ฆฌ๊ฐ€ ์‚ฌ๋ผ์ง€์ง€ ์•Š์„๊นŒ์š”?
17:41
and, you know, the country's in revolution
377
1061292
2695
์šฐ๋ฆฌ๊ฐ€ AI๋ฅผ ํ†ตํ•ด ๋งŒ๋‚  ์ˆ˜ ์žˆ๋Š” ํฌ๋ง์ฐฌ ๋ฏธ๋ž˜์— ๋„๋‹ฌํ•˜๊ธฐ ์ „์—
17:44
before we ever get to the more glorious aspects of what's possible.
378
1064011
4329
์ด๋Ÿฐ ์ผ์ด ์ผ์–ด๋‚  ๊ฑฐ ๊ฐ™์€๋ฐ์š”.
17:48
ST: Yeah, and that's an issue, and it's a big issue,
379
1068364
2517
์„ธ๋ฐ”์Šค์ฐฌ: ๋„ค ๋ฌธ์ œ์ฃ . ์ •๋ง ํฐ ๋ฌธ์ œ์ž…๋‹ˆ๋‹ค.
17:50
and it was pointed out yesterday morning by several guest speakers.
380
1070905
4196
์–ด์ œ ์•„์นจ์—๋„ ์—ฌ๋Ÿฌ ์ดˆ์ฒญ ์—ฐ์‚ฌ๋“ค์ด ๊ทธ ์ ์„ ์ง€์ ํ–ˆ์–ด์š”.
17:55
Now, prior to me showing up onstage,
381
1075125
2754
์ œ ์˜๊ฒฌ์„ ๋งํ•˜๊ธฐ์— ์•ž์„œ
17:57
I confessed I'm a positive, optimistic person,
382
1077903
3739
์ €๋Š” ๊ธ์ •์ ์ด๊ณ  ๋‚™๊ด€์ ์ธ ์‚ฌ๋žŒ์ž„์„ ๊ณ ๋ฐฑํ•ฉ๋‹ˆ๋‹ค.
18:01
so let me give you an optimistic pitch,
383
1081666
2389
์ž ๊ทธ๋Ÿผ ๋‚™๊ด€์ ์ธ ์ด์•ผ๊ธฐ๋ฅผ ํ•˜๊ธฐ ์œ„ํ•ด
18:04
which is, think of yourself back 300 years ago.
384
1084079
4795
300๋…„ ์ „์œผ๋กœ ๋Œ์•„๊ฐ€ ๋ณด์ฃ .
18:08
Europe just survived 140 years of continuous war,
385
1088898
3996
์œ ๋Ÿฝ์€ 140๋…„๊ฐ„ ์ง€์†์ ์œผ๋กœ ์ „์Ÿ์— ์‹œ๋‹ฌ๋ ธ๊ณ 
18:12
none of you could read or write,
386
1092918
1711
๋ˆ„๊ตฌ๋„ ์ฝ๊ฑฐ๋‚˜ ์“ธ ์ˆ˜ ์—†์—ˆ์œผ๋ฉฐ
18:14
there were no jobs that you hold today,
387
1094653
2945
ํ˜„์žฌ ์—ฌ๋Ÿฌ๋ถ„์ด ๊ฐ–๊ณ  ์žˆ๋Š” ์ง์—… ์ค‘ ์•„๋ฌด๊ฒƒ๋„ ์กด์žฌํ•˜์ง€ ์•Š์•˜์Šต๋‹ˆ๋‹ค.
18:17
like investment banker or software engineer or TV anchor.
388
1097622
4096
์€ํ–‰ ํˆฌ์ž๊ฐ€๋‚˜ ์†Œํ”„ํŠธ์›จ์–ด ์—”์ง€๋‹ˆ์–ด, TV ์•ต์ปค ๊ฐ™์€ ์ง์—…์ด์š”.
18:21
We would all be in the fields and farming.
389
1101742
2414
๋ชจ๋‘ ๋…ผ๊ณผ ๋ฐญ์— ๋‚˜๊ฐ€์„œ ์ผ์„ ํ–ˆ์—ˆ์ฃ .
18:24
Now here comes little Sebastian with a little steam engine in his pocket,
390
1104180
3573
๊ทธ ๋•Œ ์–ด๋ฆฐ ์„ธ๋ฐ”์Šค์ฐฌ์ด ์ฃผ๋จธ๋‹ˆ ์†์— ์žˆ๋Š” ์ž‘์€ ์ฆ๊ธฐ ๊ธฐ๊ด€์„ ๊บผ๋‚ด ๋ณด์—ฌ์ฃผ๋ฉฐ
18:27
saying, "Hey guys, look at this.
391
1107777
1548
"์–˜๋“ค์•„ ์ด๊ฒƒ ์ข€ ๋ด. ์ด ์ฆ๊ธฐ๊ธฐ๊ด€์„ ์‚ฌ์šฉํ•˜๋ฉด
18:29
It's going to make you 100 times as strong, so you can do something else."
392
1109349
3595
100๋ฐฐ๋‚˜ ํšจ์œจ์ ์œผ๋กœ ์ผํ•ด์„œ ๋‚จ์€ ์‹œ๊ฐ„์„ ํ™œ์šฉํ•  ์ˆ˜ ์žˆ์–ด."
18:32
And then back in the day, there was no real stage,
393
1112968
2470
๋‹ค์‹œ ์˜›๋‚ ๋กœ ๋Œ์•„๊ฐ€์„œ
18:35
but Chris and I hang out with the cows in the stable,
394
1115462
2526
ํฌ๋ฆฌ์Šค์™€ ์ œ๊ฐ€ ์™ธ์–‘๊ฐ„์˜ ์†Œ๋“ค์„ ๋‘˜๋Ÿฌ๋ณด๋Š” ๋™์•ˆ
18:38
and he says, "I'm really concerned about it,
395
1118012
2100
ํฌ๋ฆฌ์Šค๊ฐ€ ๋งํ•˜๊ฒ ์ฃ . "์š”์ฆ˜ ์ƒ๊ฐํ•˜๊ณ  ์žˆ๋Š” ๊ฒŒ ํ•˜๋‚˜ ์žˆ๋Š”๋ฐ
18:40
because I milk my cow every day, and what if the machine does this for me?"
396
1120136
3652
๋งŒ์•ฝ ๋งค์ผ ์šฐ์œ  ์งœ๋Š” ์ผ์„ ๊ธฐ๊ณ„๊ฐ€ ๋Œ€์‹ ํ•ด์ฃผ๋ฉด ์–ด๋–จ๊นŒ?"
18:43
The reason why I mention this is,
397
1123812
1702
์ œ๊ฐ€ ์ด ๋ง์„ ํ•˜๋Š” ์ด์œ ๋Š”
18:46
we're always good in acknowledging past progress and the benefit of it,
398
1126360
3603
์šฐ๋ฆฌ๋Š” ํ•ญ์ƒ ๊ณผ๊ฑฐ๋ฅผ ๋Œ์•„๋ณด๋ฉฐ ์ข‹์€ ์ ์„ ์ทจํ•  ์ค„ ์•Œ์•˜์–ด์š”.
18:49
like our iPhones or our planes or electricity or medical supply.
399
1129987
3354
์•„์ดํฐ์ด๋‚˜ ๋น„ํ–‰๊ธฐ, ์ „๊ธฐ, ์˜๋ฃŒ๊ธฐ๊ธฐ ๊ฐ™์€ ๊ฒƒ๋“ค ๋ง์ด์—์š”.
18:53
We all love to live to 80, which was impossible 300 years ago.
400
1133365
4245
์šฐ๋ฆฌ๋Š” ๋ชจ๋‘ 80์„ธ๊นŒ์ง€ ์‚ด ์ˆ˜ ์žˆ์ฃ . 300๋…„ ์ „์—๋Š” ๋ถˆ๊ฐ€๋Šฅํ–ˆ๋˜ ์ผ์ž…๋‹ˆ๋‹ค.
18:57
But we kind of don't apply the same rules to the future.
401
1137634
4156
ํ•˜์ง€๋งŒ ๋ฏธ๋ž˜์—๋„ ์ง€๊ธˆ๊ณผ ๋˜‘๊ฐ™์œผ๋ฆฌ๋ผ ๊ฐ€์ •ํ•  ์ˆ˜๋Š” ์—†๊ฒ ์ฃ .
19:02
So if I look at my own job as a CEO,
402
1142621
3207
CEO๋กœ์„œ ์ œ ์ผ์„ ๋ณด๋ฉด
19:05
I would say 90 percent of my work is repetitive,
403
1145852
3140
์ผ์˜ 90%๊ฐ€ ๋ฐ˜๋ณต์ž‘์—…์ด๊ณ 
19:09
I don't enjoy it,
404
1149016
1351
๋ณ„๋กœ ์ข‹์•„ํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
19:10
I spend about four hours per day on stupid, repetitive email.
405
1150391
3978
ํ•˜๋ฃจ์— 4์‹œ๊ฐ„์„ ๋ฐ”๋ณด๊ฐ™์€, ๋˜‘๊ฐ™์€ ์ด๋ฉ”์ผ ์ž‘์„ฑ์— ํ—ˆ๋น„ํ•ด์š”.
19:14
And I'm burning to have something that helps me get rid of this.
406
1154393
3641
๊ทธ๋ฆฌ๊ณ  ๋‚ญ๋น„ํ•˜๋Š” ์‹œ๊ฐ„์„ ์ค„์ผ ์ˆ˜ ์žˆ๋‹ค๋ฉด ์ •๋ง ์ข‹๊ฒ ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
19:18
Why?
407
1158058
1158
์™œ๋ƒ๊ตฌ์š”?
19:19
Because I believe all of us are insanely creative;
408
1159240
3003
์ €๋Š” ์šฐ๋ฆฌ๊ฐ€ ๋ชจ๋‘๊ฐ€ ๋†€๋ผ์šธ ์ •๋„๋กœ ์ฐฝ์˜์ ์ด๋ผ๊ณ  ๋ฏฟ์œผ๋‹ˆ๊นŒ์š”.
19:22
I think the TED community more than anybody else.
409
1162731
3194
ํŠนํžˆ TED ์ปค๋ฎค๋‹ˆํ‹ฐ์˜ ์—ฌ๋Ÿฌ๋ถ„์€ ๊ทธ๋Ÿด๊ฑฐ๋ผ ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
19:25
But even blue-collar workers; I think you can go to your hotel maid
410
1165949
3559
์œก์ฒด ๋…ธ๋™์ž๋“ค๋„ ๋งˆ์ฐฌ๊ฐ€์ง€์—์š”. ํ˜ธํ…”์˜ ์ข…์—…์›๊ณผ ์–ด์šธ๋ฆฌ๋ฉฐ
19:29
and have a drink with him or her,
411
1169532
2402
ํ•œ ์ž” ํ•  ์ˆ˜๋„ ์žˆ๊ฒ ์ฃ .
19:31
and an hour later, you find a creative idea.
412
1171958
2717
๊ทธ๋ฆฌ๊ณ  1์‹œ๊ฐ„ ๋’ค์—, ์ฐฝ์˜์ ์ธ ์ƒ๊ฐ์ด ๋– ์˜ค๋ฅด๋Š” ๊ฒ๋‹ˆ๋‹ค.
19:34
What this will empower is to turn this creativity into action.
413
1174699
4140
AI์˜ ๋„์›€์œผ๋กœ ์šฐ๋ฆฌ๋Š” ์ฐฝ์˜๋ ฅ์„ ์‹ค์ œ ํ–‰๋™์œผ๋กœ ์˜ฎ๊ธธ ์ˆ˜ ์žˆ๊ฒŒ ๋˜๋Š” ๊ฒ๋‹ˆ๋‹ค.
19:39
Like, what if you could build Google in a day?
414
1179265
3442
๊ตฌ๊ธ€๊ฐ™์€ ํšŒ์‚ฌ๋ฅผ ํ•˜๋ฃจ๋งŒ์— ๋งŒ๋“ค ์ˆ˜ ์žˆ๋‹ค๋ฉด ์–ด๋–จ๊นŒ์š”?
19:43
What if you could sit over beer and invent the next Snapchat,
415
1183221
3316
๋งฅ์ฃผ ํ•œ ์ž”์„ ๋งˆ์‹œ๋ฉฐ ๋˜ ๋‹ค๋ฅธ '์Šค๋ƒ…์ฑ—'์„ ๋ฐœ๋ช…ํ•˜๊ณ 
19:46
whatever it is,
416
1186561
1165
๊ทธ๊ฒŒ ๋ฌด์—‡์ด๋“  ๊ฐ„์—
19:47
and tomorrow morning it's up and running?
417
1187750
2187
๋‚ด์ผ ์•„์นจ์€ ๋˜ ๋‹ค๊ฐ€์˜ค๊ฒ ์ฃ .
19:49
And that is not science fiction.
418
1189961
1773
๊ทธ๋Ÿฐ ์ผ์€ ๊ณต์ƒ๊ณผํ•™์ด ์•„๋‹™๋‹ˆ๋‹ค.
19:51
What's going to happen is,
419
1191758
1254
์•ž์œผ๋กœ ์ผ์–ด๋‚  ์ผ์€
19:53
we are already in history.
420
1193036
1867
์šฐ๋ฆฌ๊ฐ€ ์ด๋ฏธ ์—ญ์‚ฌ์˜ ์ผ๋ถ€๊ฐ€ ๋  ๊ฒƒ์ด๋ผ๋Š” ์ ์ž…๋‹ˆ๋‹ค.
19:54
We've unleashed this amazing creativity
421
1194927
3228
์šฐ๋ฆฌ๋Š” ๋†์—…์—์„œ ํ•ด๋ฐฉ๋˜๋ฉด์„œ
19:58
by de-slaving us from farming
422
1198179
1611
๊ทธ๋ฆฌ๊ณ  ํ›„์—๋Š” ๊ณต์žฅ ์ž‘์—…์—์„œ ํ•ด๋ฐฉ๋˜๋ฉด์„œ
19:59
and later, of course, from factory work
423
1199814
3363
์šฐ๋ฆฌ๊ฐ€ ๊ฐ€์ง„ ๋†€๋ผ์šด ์ฐฝ์˜๋ ฅ์„ ํ•ด๋ฐฉ์‹œํ‚ฌ ์ˆ˜ ์žˆ์—ˆ๊ณ 
20:03
and have invented so many things.
424
1203201
3162
๋งŽ์€ ๊ฒƒ๋“ค์„ ๋ฐœ๋ช…ํ–ˆ์Šต๋‹ˆ๋‹ค.
20:06
It's going to be even better, in my opinion.
425
1206387
2178
์ €๋Š” ์•ž์œผ๋กœ ๋”์šฑ ๋‚˜์•„์งˆ๊ฑฐ๋ผ ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
20:08
And there's going to be great side effects.
426
1208589
2072
๋ฌผ๋ก  ๋ถ€์ž‘์šฉ๋„ ๋งŒ๋งŒ์น˜ ์•Š๊ฒ ์ฃ .
20:10
One of the side effects will be
427
1210685
1489
๊ทธ ์ค‘ ํ•˜๋‚˜๋Š”
20:12
that things like food and medical supply and education and shelter
428
1212198
4795
์Œ์‹๊ณผ ์˜๋ฃŒ ์„œ๋น„์Šค, ๊ต์œก, ์ฃผ๊ฑฐ, ์šด์†ก์ˆ˜๋‹จ๋“ฑ์ด
20:17
and transportation
429
1217017
1177
๋ถ€์ž๋“ค๋ฟ๋งŒ ์•„๋‹ˆ๋ผ
20:18
will all become much more affordable to all of us,
430
1218218
2441
์šฐ๋ฆฌ ๋ชจ๋‘์—๊ฒŒ ์ข€ ๋”
20:20
not just the rich people.
431
1220683
1322
๊ณตํ‰ํ•˜๊ฒŒ ์ฃผ์–ด์งˆ ๊ฒƒ์ด๋ผ๋Š” ์ ์ž…๋‹ˆ๋‹ค.
20:22
CA: Hmm.
432
1222029
1182
ํฌ๋ฆฌ์Šค: ์œผ์Œ...
20:23
So when Martin Ford argued, you know, that this time it's different
433
1223235
4341
๋งˆํ‹ด ํฌ๋“œ๊ฐ€ ์ด๋ฒˆ์—” ์ƒํ™ฉ์ด ์กฐ๊ธˆ ๋‹ค๋ฅด๋‹ค๊ณ  ํ–ˆ๋˜ ๊ฒƒ ๊ธฐ์–ตํ•˜์‹œ๋‚˜์š”?
20:27
because the intelligence that we've used in the past
434
1227600
3453
์ง€๊ธˆ๊นŒ์ง€ ์šฐ๋ฆฌ๊ฐ€ ์ƒˆ๋กœ์šด ๊ธธ์„
20:31
to find new ways to be
435
1231077
2483
์ฐพ๊ธฐ ์œ„ํ•ด ์‚ฌ์šฉํ•œ ์ธ๊ณต์ง€๋Šฅ์€
20:33
will be matched at the same pace
436
1233584
2279
๊ฐ™์€ ์†๋„๋กœ ์ปดํ“จํ„ฐ์— ์˜ํ•ด
20:35
by computers taking over those things,
437
1235887
2291
๊ทธ ์˜์—ญ์„ ์žƒ๊ฒŒ ๋  ๊ฑฐ๋ผ๋Š” ๊ฑฐ์˜€์ฃ .
20:38
what I hear you saying is that, not completely,
438
1238202
3078
๋‹น์‹ ์€ ์ธ๊ฐ„์ด ๊ฐ€์ง„ ์ฐฝ์กฐ์„ฑ ๋•Œ๋ฌธ์—
20:41
because of human creativity.
439
1241304
2951
์™„์ „ํžˆ ๋นผ์•—๊ธฐ์ง€๋Š” ์•Š์„๊ฑฐ๋ผ๋Š” ๊ฑฐ์ฃ ?
20:44
Do you think that that's fundamentally different from the kind of creativity
440
1244279
3785
๋‹น์‹ ์€ ์ธ๊ฐ„์˜ ์ฐฝ์˜์„ฑ์ด ์ปดํ“จํ„ฐ์˜ ๊ทธ๊ฒƒ๊ณผ
20:48
that computers can do?
441
1248088
2696
๊ทผ๋ณธ์ ์œผ๋กœ ๋‹ค๋ฅด๋‹ค๊ณ  ๋ณด์‹ญ๋‹ˆ๊นŒ?
20:50
ST: So, that's my firm belief as an AI person --
442
1250808
4434
์„ธ๋ฐ”์Šค์ฐฌ: ๋„ค ๊ทธ๊ฒƒ์ด ์ธ๊ณต์ง€๋Šฅ์„ ๊ฐœ๋ฐœ ํ•˜๋Š” ์‚ฌ๋žŒ์œผ๋กœ์„œ ์ œ๊ฐ€ ๊ฐ€์ง„ ์‹ ๋…์ž…๋‹ˆ๋‹ค.
20:55
that I haven't seen any real progress on creativity
443
1255266
3803
์ฐฝ์˜๋ ฅ๊ณผ ํ‹€์„ ๋ฒ—์–ด๋‚œ ๋…์ฐฝ์„ฑ์—์„œ
20:59
and out-of-the-box thinking.
444
1259949
1407
AI๋Š” ์•„๋ฌด๋Ÿฐ ์ง„๋ณด๋„ ์ด๋ค„๋‚ด์ง€ ๋ชปํ–ˆ์–ด์š”.
21:01
What I see right now -- and this is really important for people to realize,
445
1261380
3623
์ธ๊ณต ์ง€๋Šฅ์ด๋ผ๋Š” ๋‹จ์–ด๊ฐ€ ์ฃผ๋Š” ์œ„ํ˜‘์ ์ธ ๋Š๋‚Œ๊ณผ
21:05
because the word "artificial intelligence" is so threatening,
446
1265027
2903
์Šคํ‹ฐ๋ธ ์Šคํ•„๋ฒ„๊ทธ๊ฐ€ ๋งŒ๋“  ์˜ํ™” ์†์—์„œ
21:07
and then we have Steve Spielberg tossing a movie in,
447
1267954
2523
๊ฐ‘์ž๊ธฐ ๊ธฐ๊ณ„๊ฐ€ ์„ธ๊ณ„๋ฅผ ์ง€๋ฐฐํ•˜๋Š” ๋“ฑ์˜ ์„ค์ • ๋•Œ๋ฌธ์—
21:10
where all of a sudden the computer is our overlord,
448
1270501
2413
์‚ฌ๋žŒ๋“ค์ด ๋ถˆ์•ˆํ•ด ํ•˜๋Š” ๊ฒƒ์„ ์•Œ์ง€๋งŒ
21:12
but it's really a technology.
449
1272938
1452
์ง€๊ธˆ ์ œ๊ฐ€ ๋งŒ๋“œ๋Š” ๊ฑด ๋‹จ์ˆœํ•œ '๊ธฐ์ˆ '์ž…๋‹ˆ๋‹ค.
21:14
It's a technology that helps us do repetitive things.
450
1274414
2982
๋ฐ˜๋ณต ์ž‘์—…์— ๋“œ๋Š” ์ˆ˜๊ณ ๋ฅผ ๋œ์–ด์ฃผ๋Š” ๊ธฐ์ˆ ์  ์ง„๋ณด์ฃ .
21:17
And the progress has been entirely on the repetitive end.
451
1277420
2913
AI์˜ ์ง„๋ณด๋Š” ๋ฐ˜๋ณต ์ž‘์—…๊ณผ ์—ฐ๊ด€๋œ ๋ถ„์•ผ์—์„œ๋งŒ ์ด๋ค„์กŒ์Šต๋‹ˆ๋‹ค.
21:20
It's been in legal document discovery.
452
1280357
2228
๋ฒ•๋ฅ  ๋ฌธ์„œ์˜ ๋ฐœ๊ฒฌ์ด๋‚˜
21:22
It's been contract drafting.
453
1282609
1680
๊ณ„์•ฝ์„œ ์ž‘์„ฑ
21:24
It's been screening X-rays of your chest.
454
1284313
4223
ํ‰๋ถ€ X-ray์ดฌ์˜ ๊ฐ™์€ ๋ถ„์•ผ์š”.
21:28
And these things are so specialized,
455
1288560
1773
๋งค์šฐ ์ „๋ฌธํ™”๋œ ๋ถ„์•ผ์— ํ•œ์ •๋˜์–ด ์žˆ๊ธฐ ๋•Œ๋ฌธ์—
21:30
I don't see the big threat of humanity.
456
1290357
2391
์ธ๋ฅ˜์—๊ฒŒ ํฐ ์œ„ํ˜‘์„ ์ค„ ์ •๋„๋Š” ์•„๋‹ ๊ฒƒ ๊ฐ™์Šต๋‹ˆ๋‹ค.
21:32
In fact, we as people --
457
1292772
1794
์‚ฌ์‹ค์ƒ ์šฐ๋ฆฌ๋“ค์€
21:34
I mean, let's face it: we've become superhuman.
458
1294590
2385
์†”์งํžˆ ๋งˆ์ฃผํ•˜์ฃ . ์šฐ๋ฆฌ๋Š” ์ด๋ฏธ ์ดˆ์ธ๊ฐ„์ž…๋‹ˆ๋‹ค.
21:36
We've made us superhuman.
459
1296999
1764
์šฐ๋ฆฌ ์Šค์Šค๋กœ ๊ทธ ๋ฐœ์ „์„ ์ด๋ค„๋ƒˆ์ฃ .
21:38
We can swim across the Atlantic in 11 hours.
460
1298787
2632
๋Œ€์„œ์–‘์„ 11์‹œ๊ฐ„๋งŒ์— ๊ฐ€๋กœ์ง€๋ฅผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
21:41
We can take a device out of our pocket
461
1301443
2074
์ฃผ๋จธ๋‹ˆ์—์„œ ๊บผ๋‚ธ ๊ธฐ๊ธฐ๋กœ
21:43
and shout all the way to Australia,
462
1303541
2147
ํ˜ธ์ฃผ์— ์‚ฌ๋Š” ๋ˆ„๊ตฐ๊ฐ€์™€ ์—ฐ๋ฝํ•  ์ˆ˜ ์žˆ๊ณ 
21:45
and in real time, have that person shouting back to us.
463
1305712
2600
๊ทธ ์‚ฌ๋žŒ์—๊ฒŒ์„œ ์‹ค์‹œ๊ฐ„์œผ๋กœ ์—ฐ๋ฝ์„ ๋ฐ›๋Š” ๊ฒƒ ์—ญ์‹œ ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.
21:48
That's physically not possible. We're breaking the rules of physics.
464
1308336
3624
๋ฌผ๋ฆฌ์ ์œผ๋กœ๋Š” ๋ถˆ๊ฐ€๋Šฅํ•œ ์ผ์ด์ง€๋งŒ ์šฐ๋ฆฌ๋Š” ๋ฌผ๋ฆฌ๋ฒ•์น™์„ ๋ฌด๋„ˆ๋œจ๋ฆฌ๊ณ  ์žˆ์–ด์š”.
21:51
When this is said and done, we're going to remember everything
465
1311984
2943
์ œ๊ฐ€ ๋งํ•œ ๊ฒƒ๋“ค์ด ์ด๋ค„์งˆ ๋•Œ ์—ฌ๋Ÿฌ๋ถ„์€ ๋ณด๊ณ  ๋“ค์€
21:54
we've ever said and seen,
466
1314951
1213
๋ชจ๋“  ๊ฒƒ์„ ๊ธฐ์–ตํ•˜๊ฒŒ ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
21:56
you'll remember every person,
467
1316188
1496
๋งŒ๋‚ฌ๋˜ ๋ชจ๋“  ์‚ฌ๋žŒ์„ ๊ธฐ์–ตํ•˜๊ฒŒ ๋  ๊ฒƒ์ด๊ตฌ์š”.
21:57
which is good for me in my early stages of Alzheimer's.
468
1317708
2626
์•Œ์ธ ํ•˜์ด๋จธ ์ดˆ๊ธฐ์ธ ์ €์—๊ฒŒ๋Š” ์ข‹์€ ์ผ์ด์ฃ .
22:00
Sorry, what was I saying? I forgot.
469
1320358
1677
์ œ๊ฐ€ ๋ญ๋ผ๊ณ  ๋งํ–ˆ์—ˆ์ฃ ? ๊ธฐ์–ต์ด ์•ˆ ๋‚˜๋„ค์š”.
22:02
CA: (Laughs)
470
1322059
1578
ํฌ๋ฆฌ์Šค: (์›ƒ์Œ)
22:03
ST: We will probably have an IQ of 1,000 or more.
471
1323661
3077
์„ธ๋ฐ”์Šค์ฐฌ: ์šฐ๋ฆฌ์˜ ์•„์ดํ๋Š” 1000์ด๋‚˜ ๊ทธ ์ด์ƒ์ด ๋  ๊ฒ๋‹ˆ๋‹ค.
22:06
There will be no more spelling classes for our kids,
472
1326762
3425
์•„์ด๋“ค์€ ๋” ์ด์ƒ ์ฒ ์ž ์ˆ˜์—…์„ ๋“ฃ์ง€ ์•Š์•„๋„ ๋ฉ๋‹ˆ๋‹ค.
22:10
because there's no spelling issue anymore.
473
1330211
2086
์•ž์œผ๋กœ ์ฒ ์ž ๋•Œ๋ฌธ์— ๋ฌธ์ œ ์ƒ๊ธธ ์ผ์ด ์—†์–ด์งˆ ๊ฑฐ๋‹ˆ๊นŒ์š”.
22:12
There's no math issue anymore.
474
1332321
1832
์ˆ˜ํ•™ ์—ญ์‹œ ๋งˆ์ฐฌ๊ฐ€์ง€์ž…๋‹ˆ๋‹ค.
22:14
And I think what really will happen is that we can be super creative.
475
1334177
3510
๋Œ€์‹  ์šฐ๋ฆฌ๋Š” ์—„์ฒญ๋‚˜๊ฒŒ ์ฐฝ์˜์ ์ธ ์ธ๊ฐ„์ด ๋  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
22:17
And we are. We are creative.
476
1337711
1857
์šฐ๋ฆฌ๋Š” ์ด๋ฏธ ์ฐฝ์˜์ ์ด์ฃ .
22:19
That's our secret weapon.
477
1339592
1552
๊ทธ๊ฒƒ์ด ์ธ๊ฐ„์˜ ๋ฌด๊ธฐ์ž…๋‹ˆ๋‹ค.
22:21
CA: So the jobs that are getting lost,
478
1341168
2153
ํฌ๋ฆฌ์Šค: ์ผ์ž๋ฆฌ์˜ ์ƒ์‹ค์€ ์–ด๋Š ์ •๋„ ๋ฐœ์ƒํ•  ๊ฒƒ์ด๊ณ 
22:23
in a way, even though it's going to be painful,
479
1343345
2494
๊ทธ๋กœ์ธํ•œ ํƒ€๊ฒฉ์ด ๊ฝค ์žˆ๊ฒ ์ง€๋งŒ
22:25
humans are capable of more than those jobs.
480
1345863
2047
์šฐ๋ฆฌ๋Š” ์ผ์ž๋ฆฌ์˜ ์ƒ์‹ค๋ณด๋‹ค ๋” ๋งŽ์€ ๊ฒƒ์„ ์–ป๊ฒŒ ๋  ๊ฒƒ์ด๋‹ค.
22:27
This is the dream.
481
1347934
1218
๊ฟˆ๊ณผ ๊ฐ™์€ ์ด์•ผ๊ธฐ๋„ค์š”.
22:29
The dream is that humans can rise to just a new level of empowerment
482
1349176
4247
์ธ๊ฐ„์—๊ฒŒ ์ƒˆ๋กœ์šด ๊ธฐํšŒ์™€ ๋ฐœ๊ฒฌ์„ ์ œ๊ณตํ•  ์ˆ˜ ์žˆ๋Š”
22:33
and discovery.
483
1353447
1657
๊ฟˆ์ž…๋‹ˆ๋‹ค.
22:35
That's the dream.
484
1355128
1452
๊ฟˆ์ด์š”.
22:36
ST: And think about this:
485
1356604
1643
์„ธ๋ฐ”์Šค์ฐฌ: ํ•œ ๋ฒˆ ์ƒ๊ฐํ•ด ๋ณด์„ธ์š”.
22:38
if you look at the history of humanity,
486
1358271
2021
์ธ๋ฅ˜์˜ ์—ญ์‚ฌ๋ฅผ ๋Œ์•„๋ณด๋ฉด
22:40
that might be whatever -- 60-100,000 years old, give or take --
487
1360316
3328
6๋งŒ๋…„์—์„œ 10๋งŒ๋…„๊นŒ์ง€ ๊ฑฐ์Šฌ๋Ÿฌ ์˜ฌ๋ผ๊ฐ€์ง€๋งŒ
22:43
almost everything that you cherish in terms of invention,
488
1363668
3726
์šฐ๋ฆฌ์—๊ฒŒ ๋„์›€์ด ๋˜๋Š” ๋ฐœ๋ช…ํ’ˆ๋“ค
22:47
of technology, of things we've built,
489
1367418
2151
๊ธฐ์ˆ ์  ์ง„๋ณด์ด๋“  ๊ฑด์ถ•์ด๋“ ๊ฐ„์—
22:49
has been invented in the last 150 years.
490
1369593
3099
๊ฑฐ์˜ ๋Œ€๋ถ€๋ถ„์ด ์ง€๋‚œ 150๋…„ ์‚ฌ์ด์— ๋‚˜์˜จ ๊ฒƒ๋“ค์ž…๋‹ˆ๋‹ค.
22:53
If you toss in the book and the wheel, it's a little bit older.
491
1373756
3048
์ฑ…์ด๋‚˜ ๋ฐ”ํ€ด, ๋„๋ผ ๊ฐ™์€ ๊ฒƒ๋“ค์€
22:56
Or the axe.
492
1376828
1169
์ข€ ์˜ค๋ž˜๋œ ๋ฐœ๋ช…ํ’ˆ๋“ค์ด์ฃ .
22:58
But your phone, your sneakers,
493
1378021
2790
ํ•˜์ง€๋งŒ ์ „ํ™”๋‚˜ ์Šค๋‹ˆ์ปค์ฆˆ ๊ฐ™์€ ์‹ ๋ฐœ
23:00
these chairs, modern manufacturing, penicillin --
494
1380835
3551
์˜์ž, ๊ณต์‚ฐํ’ˆ, ํŽ˜๋‹ˆ์‹ค๋ฆฐ
23:04
the things we cherish.
495
1384410
1714
์šฐ๋ฆฌ์—๊ฒŒ ๊ผญ ํ•„์š”ํ•œ ๊ฒƒ๋“ค์€ ์ „๋ถ€ ์ตœ๊ทผ์— ํƒ„์ƒํ•œ ๊ฒƒ๋“ค์ž…๋‹ˆ๋‹ค.
23:06
Now, that to me means
496
1386148
3658
์ด๋Š” ๊ณง ์ €์—๊ฒŒ
23:09
the next 150 years will find more things.
497
1389830
3041
๋‹ค์Œ 150๋…„๊ฐ„ ์ธ๋ฅ˜๋Š” ๋” ๋งŽ์€ ๋ฐœ๋ช…์„ ํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ์˜๋ฏธ๋กœ ์—ฌ๊ฒจ์ง‘๋‹ˆ๋‹ค.
23:12
In fact, the pace of invention has gone up, not gone down, in my opinion.
498
1392895
4154
์ €๋Š” ๋ฐœ๋ช…์˜ ์†๋„๊ฐ€ ๊ณ„์† ๋นจ๋ผ์ง€๊ณ  ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
23:17
I believe only one percent of interesting things have been invented yet. Right?
499
1397073
4905
ํ›Œ๋ฅญํ•œ ๋ฌผ๊ฑด์˜ ๋ฐœ๋ช…์€ ์•„์ง 1%๋ฐ–์— ์ด๋ค„์ง€์ง€ ์•Š์•˜๋‹ค๊ณ  ์ƒ๊ฐํ•ด์š”.
23:22
We haven't cured cancer.
500
1402002
1988
์šฐ๋ฆฌ๋Š” ์•„์ง ์•”์„ ์น˜๋ฃŒํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค.
23:24
We don't have flying cars -- yet. Hopefully, I'll change this.
501
1404014
3718
ํ•˜๋Š˜์„ ๋‚˜๋Š” ์ž๋™์ฐจ๋„ ๋ฉ€์—ˆ์ฃ . ํ•˜์ง€๋งŒ ์ €๋Š” ๊ณ„์† ๋„์ „ํ•  ๊ฒ๋‹ˆ๋‹ค.
23:27
That used to be an example people laughed about. (Laughs)
502
1407756
3257
์‚ฌ๋žŒ๋“ค์ด ๋น„์›ƒ๋˜ ๋ฐœ๋ช…์˜ ์ „ํ˜•์ ์ธ ์˜ˆ์ฃ . (์›ƒ์Œ)
23:31
It's funny, isn't it? Working secretly on flying cars.
503
1411037
2992
์žฌ๋ฐŒ๋Š” ๋ฐœ์ƒ ์•„๋‹Œ๊ฐ€์š”? ๋‚จ ๋ชฐ๋ž˜ ํ•˜๋Š˜์„ ๋‚˜๋Š” ์ž๋™์ฐจ๋ฅผ ๊ฐœ๋ฐœํ•œ๋‹ค๋Š” ๊ฑฐ์š”.
23:34
We don't live twice as long yet. OK?
504
1414053
2683
์šฐ๋ฆฌ๋Š” ์•„์ง ํ‰๊ท  ์ˆ˜๋ช…์˜ ๋‘ ๋ฐฐ๋งŒํผ ์‚ด์ง€๋„ ๋ชปํ•ฉ๋‹ˆ๋‹ค.
23:36
We don't have this magic implant in our brain
505
1416760
2785
์šฐ๋ฆฌ๊ฐ€ ์›ํ•˜๋Š” ์ •๋ณด๋ฅผ ๋ฐ”๋กœ ์ทจ๋“ํ•  ์ˆ˜ ์žˆ๋„๋ก
23:39
that gives us the information we want.
506
1419569
1832
๋‡Œ ์†์— ์นฉ์„ ์ด์‹ํ•˜๋Š” ์ผ๋„ ์•„์ง ๋ฉ€์—ˆ์ฃ .
23:41
And you might be appalled by it,
507
1421425
1526
๋‘๋ ค์›€์„ ๊ฐ€์งˆ ์ˆ˜๋„ ์žˆ์ง€์š”.
23:42
but I promise you, once you have it, you'll love it.
508
1422975
2444
ํ•œ ๊ฐ€์ง€ ์•ฝ์†ํ•  ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์€ ์—ฌ๋Ÿฌ๋ถ„์˜ ๋งˆ์Œ์— ์™ ๋“ค๊ฒŒ ๋ ๊ฑฐ๋ผ๋Š” ๊ฒ๋‹ˆ๋‹ค.
23:45
I hope you will.
509
1425443
1166
๊ทธ๋Ÿฌ๊ธธ ๋ฐ”๋ž๋‹ˆ๋‹ค.
23:46
It's a bit scary, I know.
510
1426633
1909
์กฐ๊ธˆ ๋ฌด์„œ์šด ์ผ์ผ์ˆ˜๋„ ์žˆ์ง€๋งŒ์š”.
23:48
There are so many things we haven't invented yet
511
1428566
2254
์ œ ์ƒ๊ฐ์— ์ธ๋ฅ˜๊ฐ€ ๋งŒ๋“ค์–ด๋‚ผ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•˜๋Š” ๊ฒƒ๋“ค ์ค‘์—
23:50
that I think we'll invent.
512
1430844
1268
์ˆ˜๋งŽ์€ ๊ฒƒ๋“ค์ด ์•„์ง ๋ฐœ๋ช…๋˜์ง€ ์•Š์•˜์–ด์š”.
23:52
We have no gravity shields.
513
1432136
1306
์˜ˆ๋ฅผ๋“ค์–ด ์ค‘๋ ฅ ์‰ด๋“œ ๋ผ๋˜์ง€,
23:53
We can't beam ourselves from one location to another.
514
1433466
2553
๋ ˆ์ด์ €๋กœ ๋จผ ๊ฑฐ๋ฆฌ๋ฅผ ์ˆœ๊ฐ„์ด๋™ ํ•˜๋Š” ์žฅ์น˜ ๋ง์ž…๋‹ˆ๋‹ค.
23:56
That sounds ridiculous,
515
1436043
1151
์ง€๊ธˆ์œผ๋กœ์„œ๋Š” ํ„ฐ๋ฌด๋‹ˆ์—†๋Š” ์†Œ๋ฆฌ์ฃ .
23:57
but about 200 years ago,
516
1437218
1288
๊ทธ๋Ÿฌ๋‚˜ 200์—ฌ๋…„ ์ „์— ์ „๋ฌธ๊ฐ€๋“ค์€
23:58
experts were of the opinion that flight wouldn't exist,
517
1438530
2667
์‚ฌ๋žŒ์„ ํƒœ์šฐ๋Š” ๋น„ํ–‰๊ธฐ๋Š” ์กด์žฌํ•˜์ง€ ๋ชปํ• ๊ฑฐ๋ผ ์ƒ๊ฐํ–ˆ๊ณ 
24:01
even 120 years ago,
518
1441221
1324
120๋…„ ์ „ ๋งŒํ•ด๋„ ์šฐ๋ฆฌ๋Š”
24:02
and if you moved faster than you could run,
519
1442569
2582
์‚ฌ๋žŒ์ด ๋‹ฌ๋ฆฌ๋Š” ๊ฒƒ ๋ณด๋‹ค ๋” ๋น ๋ฅด๊ฒŒ ์›€์ง์ด๋ฉด
24:05
you would instantly die.
520
1445175
1520
์ฆ‰์‚ฌํ•  ๊ฒƒ์ด๋ผ๊ณ  ๋ฏฟ์—ˆ์ฃ .
24:06
So who says we are correct today that you can't beam a person
521
1446719
3569
๊ทธ๋ž˜์„œ ์ง€๊ตฌ์—์„œ ํ™”์„ฑ๊นŒ์ง€ ์ˆœ๊ฐ„์ด๋™์„ ํ•  ์ˆ˜ ์—†๋‹ค ๋ผ๋Š”๊ฒŒ
24:10
from here to Mars?
522
1450312
2249
์‚ฌ์‹ค์ด๋ผ๊ณ  ๋ˆ„๊ฐ€ ๋‹จ์ •ํ•  ์ˆ˜ ์žˆ์ฃ ?
24:12
CA: Sebastian, thank you so much
523
1452585
1569
ํฌ๋ฆฌ์Šค: ์˜ค๋Š˜ ์˜๊ฐ์„ ์ฃผ๋Š” ํ˜œ์•ˆ๊ณผ ์žฌ๋Šฅ์„ ๊ณต์œ ํ•ด ์ฃผ์…”์„œ
24:14
for your incredibly inspiring vision and your brilliance.
524
1454178
2682
์ •๋ง ๊ฐ์‚ฌ๋“œ๋ฆฝ๋‹ˆ๋‹ค.
24:16
Thank you, Sebastian Thrun.
525
1456884
1323
์„ธ๋ฐ”์Šค์ฐฌ ์“ฐ๋Ÿฐ๋‹˜, ๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.
24:18
That was fantastic. (Applause)
526
1458231
1895
์ •๋ง ์ข‹์€ ๋Œ€๋‹ด์ด์—ˆ์Šต๋‹ˆ๋‹ค. (๋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