Ajit Narayanan: A word game to communicate in any language

114,906 views ใƒป 2014-03-10

TED


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

๋ฒˆ์—ญ: Sunphil Ga ๊ฒ€ํ† : Jaewon Hur
00:12
I work with children with autism.
0
12721
2670
์žํ์•„๋ฅผ ๊ฐ€์ง„ ์•„์ด๋“ค๊ณผ ์ €๋Š” ์ผํ•ฉ๋‹ˆ๋‹ค.
00:15
Specifically, I make technologies
1
15391
1914
๊ตฌ์ฒด์ ์œผ๋กœ, ์•„์ด๋“ค์ด ์†Œํ†ตํ•  ์ˆ˜ ์žˆ๋„๋ก
00:17
to help them communicate.
2
17305
2171
๋„์™€์ฃผ๋Š” ๊ธฐ๊ณ„๋ฅผ ๋งŒ๋“ญ๋‹ˆ๋‹ค.
00:19
Now, many of the problems that children
3
19476
1539
์ง€๊ธˆ, ์žํ๋ฅผ ๊ฐ€์ง„ ์•„์ด๋“ค์ด ์ง๋ฉดํ•œ ๋ฌธ์ œ๋Š”,
00:21
with autism face, they have a common source,
4
21015
3763
๊ทธ๋“ค์ด ํ•˜๋‚˜์˜ ๊ณตํ†ต ๊ทผ์›์„ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
00:24
and that source is that they find it difficult
5
24778
2094
๋ฌธ์ œ๋Š” ๊ทธ๋“ค์ด ์ถ”์ƒ์ ์ด๊ณ  ์ƒ์ง•์ ์ธ ์˜๋ฏธ๋ฅผ
00:26
to understand abstraction, symbolism.
6
26872
5260
์ดํ•ดํ•˜๋Š”๋ฐ ์–ด๋ ค์›Œํ•˜๋Š” ๊ฒƒ์„ ์•Œ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
00:32
And because of this, they have a lot of difficulty with language.
7
32132
4652
์ด ๋ฌธ์ œ ๋•Œ๋ฌธ์—, ๊ทธ๋“ค์€ ์–ธ์–ด๋ฅผ ๋งค์šฐ ์–ด๋ ค์›Œ ํ•˜๋Š” ๊ฒƒ์ด์ฃ .
00:36
Let me tell you a little bit about why this is.
8
36784
3015
๊ทธ๋“ค์ด ์™œ ์–ด๋ ค์›Œํ•˜๋Š”์ง€ ๊ฐ„๋žตํ•˜๊ฒŒ ๋ง์”€๋“œ๋ฆฌ๊ฒ ์Šต๋‹ˆ๋‹ค.
00:39
You see that this is a picture of a bowl of soup.
9
39799
3934
๋ณด๋‹ค์‹œํ”ผ ์ด ์‚ฌ์ง„์€ ์Šคํ‘ธ๊ฐ€ ๋‹ด๊ธด ์‚ฌ๋ฐœ์ž…๋‹ˆ๋‹ค.
00:43
All of us can see it. All of us understand this.
10
43733
2485
๋ชจ๋“  ๋ถ„๋“ค๊ป˜์„œ ์•Œ ์ˆ˜ ์žˆ์œผ๋ฉฐ, ์ดํ•ดํ•˜์‹ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
00:46
These are two other pictures of soup,
11
46218
2312
์Šคํ‘ธ๊ฐ€ ๋‹ด๊ธด ๋‹ค๋ฅธ ๊ทธ๋ฆผ๋“ค์ž…๋‹ˆ๋‹ค,
00:48
but you can see that these are more abstract
12
48530
2067
๋ณด์‹œ๋Š” ๊ฒƒ์ฒ˜๋Ÿผ ์ข€๋” ์ถ”์ƒ์ ์ด๋ฉฐ,
00:50
These are not quite as concrete.
13
50597
1856
๊ตฌ์ฒด์ ์ด์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
00:52
And when you get to language,
14
52453
2174
์–ธ์–ด๋กœ ์ ‘๊ทผํ–ˆ์„ ๋•Œ๋Š”
00:54
you see that it becomes a word
15
54627
1868
๋ณด๊ณ  ๋“ฃ๋Š” ๋ฐฉ์‹์œผ๋กœ
00:56
whose look, the way it looks and the way it sounds,
16
56495
3261
ํ•˜๋‚˜์˜ ๋‹จ์–ด๊ฐ€ ๋˜๋Š” ๊ฒƒ์„ ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค,
00:59
has absolutely nothing to do with what it started with,
17
59756
2912
ํ•˜์ง€๋งŒ ์ด๋Š”๋ฌด์—‡๊ณผ ์‹œ์ž‘ํ•˜๊ณ  ํ˜น์€
01:02
or what it represents, which is the bowl of soup.
18
62668
2830
๋Œ€ํ‘œํ•˜๋Š”์ง€ ๊ด€๋ จ์ด ์—†์Šต๋‹ˆ๋‹ค, ๊ทธ์ € ์Šคํ‘ธ ํ•œ ์‚ฌ๋ฐœ์„ ์˜๋ฏธํ•˜์ฃ .
01:05
So it's essentially a completely abstract,
19
65498
2900
๋•Œ๋ฌธ์— ์™„์ „ํžˆ ์ถ”์ƒ์ ์ž…๋‹ˆ๋‹ค,
01:08
a completely arbitrary representation of something
20
68398
2576
์‹ค์ œ์— ์กด์žฌํ•˜๋Š” ๋Œ€์ƒ์„
01:10
which is in the real world,
21
70974
1163
์ถ”์ƒ์ ์œผ๋กœ ๋Œ€ํ‘œํ•ฉ๋‹ˆ๋‹ค,
01:12
and this is something that children with autism
22
72137
1791
์ด ๋ถ€๋ถ„์ด ์žํ์ฆ์„ ๊ฐ€์ง„ ์•„์ด๋“ค์ด
01:13
have an incredible amount of difficulty with.
23
73928
3164
์–ด๋ ค์›€์„ ๊ฐ€์ง€๋Š” ๋ถ€๋ถ„์ž…๋‹ˆ๋‹ค.
01:17
Now that's why most of the people that work with children with autism --
24
77092
2751
๋–„๋ฌธ์— ์žํ์•„ ์•„์ด๋“ค๊ณผ ํ•จ๊ป˜ ์ผ ํ•˜๋Š”
01:19
speech therapists, educators --
25
79843
1878
์–ธ์–ด์น˜๋ฃŒ์‚ฌ, ๊ต์œก์ž ๋“ฑ ๋Œ€๋ถ€๋ถ„์˜ ์‚ฌ๋žŒ๋“ค์€
01:21
what they do is, they try to help children with autism
26
81721
2633
์žํ๋ฅผ ๊ฐ€์ง„ ์•„์ด๋“ค์ด ์–ธ์–ด๊ฐ€ ์•„๋‹Œ ์ด๋ฏธ์ง€๋ฅผ ๊ฐ€์ง€๊ณ 
01:24
communicate not with words, but with pictures.
27
84354
3229
์ปค๋ฎค๋‹ˆ์ผ€์ด์…˜ ํ•˜๋„๋ก ๋„์™€ ์ฃผ๋„๋ก ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
01:27
So if a child with autism wanted to say,
28
87583
1930
์žํ๋ฅผ ๊ฐ€์ง„ ์•„์ด๊ฐ€
01:29
"I want soup," that child would pick
29
89513
2458
"์Šคํ‘ธ๋ฅผ ์›ํ•ด์š”."๋ผ๊ณ  ๋งํ•œ๋‹ค๋ฉด,
01:31
three different pictures, "I," "want," and "soup,"
30
91971
2260
์•„์ด๋Š” ์„ธ๊ฐœ์˜ ๋‹ค๋ฅธ ์ด๋ฏธ์ง€ "๋‚˜," "์›ํ•˜๋‹ค." ๊ทธ๋ฆฌ๊ณ  "์Šคํ‘ธ."๋ฅผ
01:34
and they would put these together,
31
94231
1609
์ง‘๊ณ  ํ•˜๋‚˜๋กœ ๋ชจ์๋‹ˆ๋‹ค.,
01:35
and then the therapist or the parent would
32
95840
1867
๊ทธ๋Ÿผ ์–ธ์–ด ์น˜๋ฃŒ์‚ฌ ํ˜น์€ ๋ถ€๋ชจ๋Š”
01:37
understand that this is what the kid wants to say.
33
97707
1887
์ด ํ‘œํ˜„์ด ์•„์ด๊ฐ€ ๋งํ•˜๊ณ  ์‹ถ์€ ๊ฒƒ์ด๋ผ๊ณ  ์ดํ•ดํ•ฉ๋‹ˆ๋‹ค.
01:39
And this has been incredibly effective;
34
99594
1778
์ด ๋ฐฉ๋ฒ•์€ ๋งค์šฐ ํšจ๊ณผ์ ์ž…๋‹ˆ๋‹ค;
01:41
for the last 30, 40 years
35
101372
2141
์ง€๋‚œ 30, 40๋…„ ๋™์•ˆ
01:43
people have been doing this.
36
103513
1613
์ธ๋ฅ˜๋Š” ์ง€์†์ ์œผ๋กœ ์ด ๋ฐฉ๋ฒ•์„ ํ–‰ํ•ด์™”์Šต๋‹ˆ๋‹ค.
01:45
In fact, a few years back,
37
105126
1349
์‚ฌ์‹ค, ๋ช‡ ๋…„์ „์—
01:46
I developed an app for the iPad
38
106475
2675
์ €๋Š” ์•„์ดํŒจ๋“œ์—์„œ
01:49
which does exactly this. It's called Avaz,
39
109150
2255
์ด๋ฅผ ์‹คํ˜„ํ•  ์ˆ˜ ์žˆ๋Š” Avaz๋ผ๋Š” ์•ฑ์„ ๊ฐœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค,
01:51
and the way it works is that kids select
40
111405
2279
์ด ์ž‘๋™ ๋ฐฉ๋ฒ•์€ ์•„์ด๋“ค์ด ๋‹ค๋ฅธ ์ด๋ฏธ์ง€๋ฅผ
01:53
different pictures.
41
113684
1321
์„ ํƒํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
01:55
These pictures are sequenced together to form sentences,
42
115005
2570
์ด ์‚ฌ์ง„๋“ค์€ ์ˆœ์ฐจ์ ์œผ๋กœ ๋ชจ์—ฌ ๋ฌธ์žฅ์„ ๋งŒ๋“ค๊ณ 
01:57
and these sentences are spoken out.
43
117575
1719
์†Œ๋ฆฌ๊ฐ€ ๋‚˜์˜ต๋‹ˆ๋‹ค.
01:59
So Avaz is essentially converting pictures,
44
119294
3025
Avaz๋Š” ๊ทผ๋ณธ์ ์œผ๋กœ ์ด๋ฏธ์ง€๋ฅผ ๋ณ€ํ™˜์‹œํ‚ต๋‹ˆ๋‹ค,
02:02
it's a translator, it converts pictures into speech.
45
122319
3960
์ด๋ฏธ์ง€๋ฅผ ์†Œ๋ฆฌ๋กœ ์ „ํ™”์‹œํ‚ค๋Š” ๋ฒˆ์—ญ๊ฐ€์™€ ๊ฐ™์Šต๋‹ˆ๋‹ค.
02:06
Now, this was very effective.
46
126279
1718
์ง€๊ธˆ, ์ด๋Š” ๋งค์šฐ ํšจ๊ณผ์ ์ž…๋‹ˆ๋‹ค.
02:07
There are thousands of children using this,
47
127997
1384
์•Œ๋‹ค์‹œํ”ผ ์ „์„ธ๊ณ„์— ์ด๋ฅผ ์‚ฌ์šฉํ•˜๋Š”
02:09
you know, all over the world,
48
129381
1430
์ˆ˜ ์ฒœ๋ช…์˜ ์•„์ด๋“ค์ด ์žˆ์Šต๋‹ˆ๋‹ค,
02:10
and I started thinking about
49
130811
2175
์ €๋Š” ์ด ํ”„๋กœ๊ทธ๋žจ์„ ํ†ตํ•ด ๋ฌด์—‡์ด ๊ฐ€๋Šฅํ•œ์ง€, ๋ถˆ๊ฐ€๋Šฅํ•œ์ง€
02:12
what it does and what it doesn't do.
50
132986
2654
์ƒ๊ฐํ•ด๋ณด๊ธฐ ์‹œ์ž‘ํ–ˆ๊ณ 
02:15
And I realized something interesting:
51
135640
1684
์žฌ๋ฏธ๋‚œ ๊ฒƒ์„ ๊นจ๋‹ฌ์•˜์Šต๋‹ˆ๋‹ค.
02:17
Avaz helps children with autism learn words.
52
137324
4203
Avaz๋Š” ์žํ๋ฅผ ๊ฐ€์ง„ ์•„์ด๋“ค์ด ์–ธ์–ด๋ฅผ ํ•™์Šตํ•˜๋Š”๋ฐ ๋„์›€์„ ์ค๋‹ˆ๋‹ค,
02:21
What it doesn't help them do is to learn
53
141527
2405
ํ•˜์ง€๋งŒ ๊ทธ๋“ค์ด ๋‹จ์–ด ํŒจํ„ด์„ ํ•™์Šตํ•˜๋Š”๋ฐ๋Š”
02:23
word patterns.
54
143932
2748
๋„์›€์„ ์ฃผ์ง€ ๋ชปํ•ฉ๋‹ˆ๋‹ค.
02:26
Let me explain this in a little more detail.
55
146680
2472
์ด ๋ถ€๋ถ„์„ ์ข€๋” ๊ตฌ์ฒด์ ์œผ๋กœ ์„ค๋ช…ํ•ด๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.
02:29
Take this sentence: "I want soup tonight."
56
149152
3057
์ด ๋ฌธ์žฅ์„ ๋ณด์‹œ์ฃ , "์ €๋Š” ์˜ค๋Š˜ ๋ฐค ์Šคํ‘ธ๋ฅผ ์›ํ•ด์š”."
02:32
Now it's not just the words here that convey the meaning.
57
152209
4080
์—ฌ๊ธฐ์— ์žˆ๋Š” ๊ฐœ๊ฐœ์˜ ๋‹จ์–ด๋“ค๋งŒ์œผ๋กœ๋Š” ์˜๋ฏธ๋ฅผ ์ถฉ๋ถ„ํžˆ ์ „๋‹ฌํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค.
02:36
It's also the way in which these words are arranged,
58
156289
3140
๋ฌธ์žฅ์€ ๋ฐ”๋กœ ๊ฐœ๊ฐœ์˜ ๋‹จ์–ด๋“ค์ด ์–ด๋–ป๊ฒŒ ๋‚˜์—ด๋˜์–ด ์žˆ๊ณ  ๋ณ€ํ˜•๋˜์–ด ์žˆ๋Š”์ง€์— ๋”ฐ๋ผ ๊ตฌ์„ฑ๋˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค
02:39
the way these words are modified and arranged.
59
159429
2515
02:41
And that's why a sentence like "I want soup tonight"
60
161959
2306
๋•Œ๋ฌธ์— "์ €๋Š” ์˜ค๋Š˜ ๋ฐค ์Šคํ‘ธ๋ฅผ ์›ํ•ด์š”."
02:44
is different from a sentence like
61
164265
1984
์ด ๋ฌธ์žฅ์€ "์Šคํ‘ธ๋Š” ์ €๋ฅผ ์›ํ•ด์š” ์˜ค๋Š˜๋ฐค." ์ด๋ผ๋Š” ์™„์ „ํžˆ ์˜๋ฏธ๊ฐ€ ์—†๋Š” ๋ฌธ์žฅ๊ณผ๋Š”
02:46
"Soup want I tonight," which is completely meaningless.
62
166249
3312
๋‹ค๋ฆ…๋‹ˆ๋‹ค.
02:49
So there is another hidden abstraction here
63
169561
2619
์—ฌ๊ธฐ์— ์žํ์•„๋“ค์ด ๋Œ€์ฒ˜ํ•˜๊ธฐ ์–ด๋ ค์›Œํ•˜๋Š”
02:52
which children with autism find a lot of difficulty coping with,
64
172180
3557
๋˜ ๋‹ค๋ฅธ ์ถ”์ƒ์ ์ธ ๊ฒƒ์ด ์กด์žฌํ•ฉ๋‹ˆ๋‹ค,
02:55
and that's the fact that you can modify words
65
175737
2840
์ด๋Š” ๋‹ค๋ฅธ ์˜๋ฏธ๋ฅผ ๊ฐ€์ง€๊ฑฐ๋‚˜,
02:58
and you can arrange them to have
66
178577
2101
๋‹ค๋ฅธ ์•„์ด๋””์–ด๋ฅผ ๋‚˜์—ดํ•˜๊ธฐ ์œ„ํ•ด
03:00
different meanings, to convey different ideas.
67
180678
2895
๋‹จ์–ด๋“ค์„ ์ˆ˜์ •ํ•˜๊ณ  ๋ฐฐ์—ดํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
03:03
Now, this is what we call grammar.
68
183573
3459
์ง€๊ธˆ, ์ด๊ฒƒ์„ ์šฐ๋ฆฌ๋Š” ๋ฌธ๋ฒ•์ด๋ผ๊ณ  ํ•ฉ๋‹ˆ๋‹ค.
03:07
And grammar is incredibly powerful,
69
187032
2036
๋ฌธ๋ฒ•์€ ๋†€๋ž๋„๋ก ๊ฐ•๋ ฅํ•ฉ๋‹ˆ๋‹ค,
03:09
because grammar is this one component of language
70
189068
3157
๋ฌธ๋ฒ•์€ ์šฐ๋ฆฌ๊ฐ€ ๊ฐ€์ง„ ํ•œ์ •์ ์ธ ์–ดํœ˜๋ฅผ ๊ฐ€์ง„
03:12
which takes this finite vocabulary that all of us have
71
192225
3489
์–ธ์–ด์˜ ์กฐํ•ฉ ์ค‘ ํ•˜๋‚˜์ด๋ฉฐ
03:15
and allows us to convey an infinite amount of information,
72
195714
4531
๋ฌดํ•œ์˜ ์ •๋ณด, ์•„์ด๋””์–ด๋ฅผ
03:20
an infinite amount of ideas.
73
200245
2134
์ „๋‹ฌํ•  ์ˆ˜ ์žˆ๋„๋ก ๋งŒ๋“ค์–ด์ฃผ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
03:22
It's the way in which you can put things together
74
202379
2002
์ด ๋ฐฉ๋ฒ•์€ ์›ํ•˜๋Š” ์–ด๋–ค ๊ฒƒ์ด๋“  ์ „๋‹ฌํ•˜๊ธฐ ์œ„ํ•ด
03:24
in order to convey anything you want to.
75
204381
2168
๋‹จ์–ด๋“ค์„ ํ•˜๋‚˜๋กœ ๋ชจ์œผ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
03:26
And so after I developed Avaz,
76
206549
2127
Avaz๋ฅผ ๊ฐœ๋ฐœํ•œ ์ดํ›„์—,
03:28
I worried for a very long time
77
208676
1568
์˜ค๋žœ ์‹œ๊ฐ„ ๋™์•ˆ ์ €๋Š” ๊ฑฑ์ •ํ–ˆ์Šต๋‹ˆ๋‹ค,
03:30
about how I could give grammar to children with autism.
78
210244
3910
์–ด๋–ป๊ฒŒ ๋ฌธ๋ฒ•์„ ์žํ์•„์—๊ฒŒ ์ „๋‹ฌํ•  ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์— ๊ด€ํ•˜์—ฌ ๋ง์ด์ฃ .
03:34
The solution came to me from a very interesting perspective.
79
214154
2275
๋งค์šฐ ํฅ๋ฏธ๋กœ์šด ๊ด€์ ์„ ํ†ตํ•ด ํ•ด๊ฒฐ์ ์„ ๋„์ถœํ•˜์˜€์Šต๋‹ˆ๋‹ค.
03:36
I happened to chance upon a child with autism
80
216429
3449
์ €๋Š” ์šฐ์—ฐํžˆ ์žํ๋ฅผ ๊ฐ€์ง„ ์†Œ๋…€๊ฐ€
03:39
conversing with her mom,
81
219878
2109
์—„๋งˆ์™€ ํ•จ๊ป˜ ๋Œ€ํ™”๋ฅผ ๋‚˜๋ˆˆ ๊ฒƒ์„ ๋ฐœ๊ฒฌํ–ˆ์Šต๋‹ˆ๋‹ค,
03:41
and this is what happened.
82
221987
2094
์ด๋Ÿฐ ์ƒํ™ฉ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
03:44
Completely out of the blue, very spontaneously,
83
224081
2186
์™„์ „ํžˆ ์šฐ์—ฐํžˆ์˜€์œผ๋ฉฐ, ์ž์—ฐ์Šค๋Ÿฝ๊ฒŒ
03:46
the child got up and said, "Eat."
84
226267
2463
์•„์ด๋Š” ์ผ์–ด๋‚˜ ๋งํ–ˆ์ฃ . "๋จน๋‹ค"
03:48
Now what was interesting was
85
228730
1770
ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์€
03:50
the way in which the mom was trying to tease out
86
230500
4244
์–ด๋จธ๋‹ˆ๊ฐ€ ์•„์ด๊ฐ€ ๋งํ•˜๊ณ  ์‹ถ์€ ์˜๋ฏธ๋ฅผ
03:54
the meaning of what the child wanted to say
87
234744
2213
์ •๋ฆฌํ•˜๊ณ  ์žˆ๋Š” ๋ฐฉ๋ฒ• ์ด์—ˆ์Šต๋‹ˆ๋‹ค,
03:56
by talking to her in questions.
88
236957
2260
์•„์ด์˜ ์งˆ๋ฌธ์— ๋‹ตํ•˜๋ฉด์„œ ๋ง์ด์ฃ .
03:59
So she asked, "Eat what? Do you want to eat ice cream?
89
239217
2593
์–ด๋จธ๋‹ˆ๋Š” ๋ฌป์Šต๋‹ˆ๋‹ค, "๋จน๋‹ค, ๋ญ๋ฅผ? ์•„์ด์Šคํฌ๋ฆผ ๋จน๊ณ ์‹ถ๋‹ˆ?
04:01
You want to eat? Somebody else wants to eat?
90
241810
2112
์•„๊ฐ€์•ผ ๋จน๊ณ ์‹ถ๋‹ˆ? ๋‹ค๋ฅธ ๋ˆ„๊ตฐ๊ฐ€๋„ ๋จน๊ณ  ์‹ถ์€๊ฑฐ๋‹ˆ?
04:03
You want to eat cream now? You want to eat ice cream in the evening?"
91
243922
3313
์•„์ด์Šคํฌ๋ฆผ ์ง€๊ธˆ ๋จน๊ณ ์‹ถ๋‹ˆ? ๋ฐค์— ์•„์ด์Šคํฌ๋ฆผ ๋จน๊ณ  ์‹ถ๋‹ˆ?
04:07
And then it struck me that
92
247235
1514
์ด๋Š” ์ œ๊ฒŒ ์ถฉ๊ฒฉ์„ ์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค
04:08
what the mother had done was something incredible.
93
248749
2028
์•„์ด์˜ ์–ด๋จธ๋‹ˆ๊ฐ€ ํ•ด์˜จ ๊ฒƒ์€ ๋งค์šฐ ๋†€๋ผ์šด ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
04:10
She had been able to get that child to communicate
94
250777
1994
๊ทธ๋…€๋Š” ๋ฌธ๋ฒ• ์—†์ด ์•„์ด๊ฐ€ ์ƒ๊ฐ์„ ๊ทธ๋…€์—๊ฒŒ
04:12
an idea to her without grammar.
95
252771
4138
๋Œ€ํ™”ํ•  ์ˆ˜ ์žˆ๋„๋ก ์ด๋Œ์—ˆ์Šต๋‹ˆ๋‹ค.
04:16
And it struck me that maybe this is what
96
256909
2696
์ €์—๊ฒŒ ์ถฉ๊ฒฉ์ ์ด์—ˆ์Šต๋‹ˆ๋‹ค
04:19
I was looking for.
97
259605
1385
์ œ๊ฐ€ ๊ทธ๋™์•ˆ ์ฐพ์•„์™”๋˜ ๋ฐฉ๋ฒ•์ด์—ˆ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
04:20
Instead of arranging words in an order, in sequence,
98
260990
4142
๋‹จ์–ด๋ฅผ ์ˆœ์„œ๋Œ€๋กœ ๋‚˜์—ดํ•˜๋Š” ๋Œ€์‹ ์—, ์ˆœ์ฐจ์ ์œผ๋กœ
04:25
as a sentence, you arrange them
99
265132
2172
ํ•˜๋‚˜์˜ ๋ฌธ์žฅ์ฒ˜๋Ÿผ ์ด ์ง€๋„์— ๋‹จ์–ด๋“ค์„ ๋‚˜์—ดํ•˜๋ฉด
04:27
in this map, where they're all linked together
100
267304
3811
๊ทธ๊ฒƒ๋“ค์€ ํ•˜๋‚˜๋กœ ๋ชจ๋‘ ์—ฐ๊ฒฐํ•˜๊ฒŒ ๋ฉ๋‹ˆ๋‹ค
04:31
not by placing them one after the other
101
271115
2143
ํ•˜๋‚˜ ํ•˜๋‚˜ ๋†“์•„์„œ ๋งŒ๋“œ ๊ฒƒ์ด ์•„๋‹ˆ์—ˆ์ฃ .
04:33
but in questions, in question-answer pairs.
102
273258
3284
ํ•˜์ง€๋งŒ ์งˆ๋ฌธ๋“ค์—์„œ๋Š”, ์งˆ๋ฌธ ๋‹ต๋ณ€ ํ•œ์Œ์—์„œ๋Š” ๋‹ค๋ฆ…๋‹ˆ๋‹ค.
04:36
And so if you do this, then what you're conveying
103
276542
2358
๋งŒ์•ฝ ์—ฌ๋Ÿฌ๋ถ„๊ป˜์„œ ์ด๋ฅผ ํ•œ๋‹ค๋ฉด,
04:38
is not a sentence in English,
104
278900
1986
์—ฌ๋Ÿฌ๋ถ„๋“ค์ด ์ „๋‹ฌํ•˜๋Š” ๊ฒƒ์€ ์˜์–ด ๋ฌธ์žฅ์ผ ๋ฟ๋งŒ ์•„๋‹ˆ๋ผ,
04:40
but what you're conveying is really a meaning,
105
280886
2966
์˜์–ด๋กœ ๋œ ๋ฌธ์žฅ์˜ ์˜๋ฏธ๋ฅผ
04:43
the meaning of a sentence in English.
106
283852
1511
์ „๋‹ฌํ•˜๊ธฐ๋„ ํ•ฉ๋‹ˆ๋‹ค.
04:45
Now, meaning is really the underbelly, in some sense, of language.
107
285363
2932
์ง€๊ธˆ, ์˜๋ฏธ๋Š” ์–ธ์–ด ๊ด€์ ์—์„œ ๋งค์šฐ ์ทจ์•ฝํ•œ ๋ถ€๋ถ„์ž…๋‹ˆ๋‹ค.
04:48
It's what comes after thought but before language.
108
288295
3821
์ด๋Š” ์–ธ์–ด ์ด์ „์— ์‚ฌ๊ณ ๋กœ๋ถ€ํ„ฐ ์˜ต๋‹ˆ๋‹ค.
04:52
And the idea was that this particular representation
109
292116
2503
์ด ์•„์ด๋””์–ด๋Š” ํŠน์ • ๋Œ€ํ‘œ ๋‹จ์–ด๋“ค์ด
04:54
might convey meaning in its raw form.
110
294619
3261
์žˆ๋Š” ๊ทธ๋Œ€๋กœ์˜ ํ˜•ํƒœ๋กœ ์˜๋ฏธ๋ฅผ ์ „๋‹ฌํ• ์ง€๋„ ๋ชจ๋ฅธ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
04:57
So I was very excited by this, you know,
111
297880
1771
์ €๋Š” ์ด ์•„์ด๋””์–ด์— ๋งค์šฐ ํฅ๋ถ„ํ–ˆ์Šต๋‹ˆ๋‹ค,
04:59
hopping around all over the place,
112
299651
1493
๋ชจ๋“  ๊ณณ์— ํฌ๋ง์„ ๊ฐ€์ง€๋ฉฐ,
05:01
trying to figure out if I can convert
113
301144
1771
์ œ๊ฐ€ ๋“ค์€ ๊ฐ€๋Šฅํ•œ ๋ชจ๋“  ๋ฌธ์žฅ์„ ์—ฌ๊ธฐ์—
05:02
all possible sentences that I hear into this.
114
302915
2524
์ „ํ™˜์‹œํ‚ฌ ์ˆ˜ ์žˆ๋Š”์ง€ ๊นจ๋‹ซ๊ธฐ ์œ„ํ•ด ๋…ธ๋ ฅํ–ˆ์Šต๋‹ˆ๋‹ค.
05:05
And I found that this is not enough.
115
305439
1773
๊ทธ๋ ‡์ง€๋งŒ ์ด๋Š” ์ถฉ๋ถ„์น˜ ์•Š๋‹ค๊ณ  ๊บ ๋‹ฌ์•˜์Šต๋‹ˆ๋‹ค.
05:07
Why is this not enough?
116
307212
1385
์™œ ์ถฉ๋ถ„ํ•˜์ง€ ์•Š์„๊นŒ์š”?
05:08
This is not enough because if you wanted to convey
117
308597
1711
์ด๋Š” ๋ถˆ์ถฉ๋ถ„ํ•ฉ๋‹ˆ๋‹ค ์™œ๋ƒํ•˜๋ฉด ์—ฌ๋Ÿฌ๋ถ„๊ป˜์„œ ๋ถ€์ • ํ‘œํ˜„์„
05:10
something like negation,
118
310308
2250
์ „๋‹ฌํ•˜๋ ค ํ•œ๋‹ค๋ฉด,
05:12
you want to say, "I don't want soup,"
119
312558
1736
์—ฌ๋Ÿฌ๋ถ„์€ "์ €๋Š” ์Šคํ‘ธ๋ฅผ ์›ํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค."๋ผ๊ณ 
05:14
then you can't do that by asking a question.
120
314294
2220
๋งํ•˜๋ฉด ์งˆ๋ฌธํ•  ํ•„์š”๊ฐ€ ์—†์Šต๋‹ˆ๋‹ค.
05:16
You do that by changing the word "want."
121
316514
2285
"์›ํ•ด."๋ผ๊ณ  ๋‹จ์–ด๋ฅผ ๋ฐ”๊ฟ” ๋ถ€์ •์„ ํ‘œํ˜„ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
05:18
Again, if you wanted to say,
122
318799
1637
๋‹ค์‹œ, ์—ฌ๋Ÿฌ๋ถ„๊ป˜์„œ "์–ด์ œ ์Šคํ‘ธ๋ฅผ ์›ํ–ˆ์–ด์š”."๋ผ๊ณ 
05:20
"I wanted soup yesterday,"
123
320436
1980
๋งํ•˜๊ณ  ์‹ถ๋‹ค๋ฉด,
05:22
you do that by converting the word "want" into "wanted."
124
322416
2737
"์›ํ•˜๋‹ค."๋ฅผ "์›ํ–ˆ์Šต๋‹ˆ๋‹ค."๋กœ ๋ณ€ํ™˜ํ•จ์œผ๋กœ์จ ํ‘œํ˜„ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
05:25
It's a past tense.
125
325153
1666
๊ณผ๊ฑฐํ˜•์ž…๋‹ˆ๋‹ค.
05:26
So this is a flourish which I added
126
326819
2103
์ด๋Š” ์ œ๊ฐ€ ์™„๋ฒฝํ•œ ์‹œ์Šคํ…œ์„ ๋งŒ๋“ค๊ธฐ ์œ„ํ•ด
05:28
to make the system complete.
127
328922
1576
์ถ”๊ฐ€ํ•œ ํ”Œ๋Ÿฌ๋ฆฌ์‰ฌ ์ž…๋‹ˆ๋‹ค.
05:30
This is a map of words joined together
128
330498
1977
์ด๋Š” ์งˆ๋ฌธ๊ณผ ๋‹ต๋ณ€์ฒ˜๋Ÿผ
05:32
as questions and answers,
129
332475
1656
๋‹จ์–ด๋“ค์„ ํ•˜๋‚˜๋กœ ๋ฌถ์€ ์ง€๋„์ž…๋‹ˆ๋‹ค,
05:34
and with these filters applied on top of them
130
334131
2264
์ด ์ƒ๋‹จ ๋ถ€๋ถ„์— ์ ์šฉ๋œ ํ•„ํ„ฐ๋“ค์„ ํ†ตํ•ด
05:36
in order to modify them to represent
131
336395
1817
๋‹จ์–ด๋“ค์„ ์ˆ˜์ •ํ•˜๊ฑฐ๋‚˜ ๋ถ„๋ช…ํ•œ
05:38
certain nuances.
132
338212
1709
๋‰˜์•™์Šค๋ฅผ ํ‘œํ˜„ํ•  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.
05:39
Let me show you this with a different example.
133
339921
1951
๋‹ค๋ฅธ ์‚ฌ๋ก€๋ฅผ ์—ฌ๋Ÿฌ๋ถ„๊ป˜ ๋ณด์—ฌ๋“œ๋ฆฌ๊ฒ ์Šต๋‹ˆ๋‹ค.
05:41
Let's take this sentence:
134
341872
1254
์ด ๋ฌธ์žฅ์„ ๋ณด์‹œ์ฃ ,
05:43
"I told the carpenter I could not pay him."
135
343126
1980
"๋ชฉ์ˆ˜์—๊ฒŒ ๋‚˜๋Š” ๊ทธ์—๊ฒŒ ์ง€๋ถˆ ํ•  ์ˆ˜ ์—†์—ˆ๋‹ค๊ณ  ๋งํ•ด์—ˆ๋‹ค."
05:45
It's a fairly complicated sentence.
136
345106
1792
์ด ๋ฌธ์žฅ์€ ๊ฝค ๋ณต์žกํ•œ ๋ฌธ์žฅ์ž…๋‹ˆ๋‹ค.
05:46
The way that this particular system works,
137
346898
1893
์ด ํŠน์ • ์‹œ์Šคํ…œ์ด ์ž‘๋™ํ•˜๋Š” ๋ฐฉ๋ฒ•์€,
05:48
you can start with any part of this sentence.
138
348791
2578
์ด ๋ฌธ์žฅ์˜ ์–ด๋Š ๋ถ€๋ถ„์ด๋“  ์ƒ๊ด€์—†์ด ์‹œ์ž‘ํ• ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
05:51
I'm going to start with the word "tell."
139
351369
1698
"๋งํ•˜๋‹ค."๋กœ ์‹œ์ž‘ํ•ด๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.
05:53
So this is the word "tell."
140
353067
1462
"๋งํ•˜๋‹ค." ๋‹จ์–ด์ž…๋‹ˆ๋‹ค.
05:54
Now this happened in the past,
141
354529
1600
๊ณผ๊ฑฐํ˜•์ด์—ˆ์ฃ ,
05:56
so I'm going to make that "told."
142
356129
2223
์ด ๋‹จ์–ด๋ฅผ "๋งํ–ˆ์—ˆ๋‹ค." ๊ณผ๊ฑฐํ˜•์œผ๋กœ ๋งŒ๋“ค์–ด ๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.
05:58
Now, what I'm going to do is,
143
358352
1708
์ง€๊ธˆ ์ œ๊ฐ€ ํ•˜๊ณ ์ž ํ•˜๋Š” ๊ฒƒ์€,
06:00
I'm going to ask questions.
144
360060
1756
์งˆ๋ฌธ์„ ํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
06:01
So, who told? I told.
145
361816
2364
์ด์™€ ๊ฐ™์ฃ , "๋ˆ„๊ฐ€ ๋งํ–ˆ๋‹ˆ?" ๋‚ด๊ฐ€ ๋งํ•ด์—ˆ๋‹ค.
06:04
I told whom? I told the carpenter.
146
364180
1927
๋‚ด๊ฐ€ ๋ˆ„๊ตฌ์—๊ฒŒ ๋งํ–ˆ๋‹ˆ? "๋‚˜๋Š” ๋ชฉ์ˆ˜์—๊ฒŒ ๋งํ–ˆ์—ˆ๋‹ค."
06:06
Now we start with a different part of the sentence.
147
366107
1751
์ง€๊ธˆ์€ ๋ฌธ์žฅ์˜ ๋‹ค๋ฅธ ๋ถ€๋ถ„๊ณผ ์‹œ์ž‘ํ•ด๋ณด๋„๋ก ํ•˜๊ฒ ์Šต๋‹ˆ๋‹ค.
06:07
We start with the word "pay,"
148
367858
1867
"์ง€๋ถˆํ•˜๋‹ค." ๋‹จ์–ด๋กœ ์‹œ์ž‘ํ•ด๋ณด์ฃ ,
06:09
and we add the ability filter to it to make it "can pay."
149
369725
4577
ํ•„ํ„ฐ ๋Šฅ๋ ฅ์„ ์ถ”๊ฐ€ํ•ด "์ง€๋ถˆ ํ•  ์ˆ˜ ์žˆ๋‹ค."๋ผ๋Š” ๋ฌธ์žฅ์„ ๋งŒ๋“ค ์ดํ›„์—
06:14
Then we make it "can't pay,"
150
374302
2101
"์ง€๋ถˆํ•  ์ˆ˜ ์—†์–ด."๋ผ๊ณ  ๋งŒ๋“ญ๋‹ˆ๋‹ค,
06:16
and we can make it "couldn't pay"
151
376403
1599
๊ทธ๋ฆฌ๊ณ  "์ง€๋ถˆ ํ•  ์ˆ˜ ์—†์—ˆ๋‹ค."๋ผ๊ณ  ๋งŒ๋“ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค,
06:18
by making it the past tense.
152
378002
1663
๊ณผ๊ฑฐํ˜•์œผ๋กœ ๋ณ€ํ™˜์‹œํ‚ค๋ฉด์„œ ๋ง์ด์ฃ .
06:19
So who couldn't pay? I couldn't pay.
153
379665
1923
๋‹ค์Œ์œผ๋กœ ๋ˆ„๊ฐ€ ์ง€๋ถˆํ•  ์ˆ˜ ์—†์—ˆ๋‹ˆ? "๋‚ด๊ฐ€ ์ง€๋ถˆํ•  ์ˆ˜ ์—†์—ˆ๋‹ค. "
06:21
Couldn't pay whom? I couldn't pay the carpenter.
154
381588
2676
๋ˆ„๊ตฌ์—๊ฒŒ ์ง€๋ถˆ ํ•  ์ˆ˜ ์—†์—ˆ๋‹ˆ? "๋‚˜๋Š” ๋ชฉ์ˆ˜์—๊ฒŒ ์ง€๋ถˆํ•  ์ˆ˜ ์—†์—ˆ๋‹ค."
06:24
And then you join these two together
155
384264
1731
์ดํ›„ ๋‘๊ฐœ๋ฅผ ์กฐํ•ฉํ•ฉ๋‹ˆ๋‹ค
06:25
by asking this question:
156
385995
1350
์ด ์งˆ๋ฌธ์„ ํ†ตํ•ด์„œ ๋ง์ด์ฃ ,
06:27
What did I tell the carpenter?
157
387345
1737
๋ชฉ์ˆ˜์—๊ฒŒ ๋งํ•œ ๊ฒƒ์€ ๋ฌด์—‡์ธ๊ฐ€?
06:29
I told the carpenter I could not pay him.
158
389082
4049
๋ชฉ์ˆ˜์—๊ฒŒ ๋‚˜๋Š” ๊ทธ์—๊ฒŒ ์ง€๋ถˆํ•  ์ˆ˜ ์—†์—ˆ๋‹ค๊ณ  ๋งํ–ˆ์—ˆ๋‹ค.
06:33
Now think about this. This is
159
393131
1937
์ด๊ฒƒ์— ๊ด€ํ•˜์—ฌ ์ƒ๊ฐํ•ด ๋ณด์„ธ์š”. ์ด๋Š”
06:35
โ€”(Applause)โ€”
160
395068
3542
(๋ฐ•์ˆ˜)
06:38
this is a representation of this sentence
161
398610
3672
์ด๋Š” ์–ธ์–ด๊ฐ€ ์—†๋Š” ๋Œ€ํ‘œ์ ์ธ
06:42
without language.
162
402282
2435
๋ฌธ์žฅ์ž…๋‹ˆ๋‹ค.
06:44
And there are two or three interesting things about this.
163
404717
2192
์ด๊ฒƒ์— ๊ด€ํ•˜์—ฌ ๋‘ ๊ฐ€์ง€ ํ˜น์€ ์„ธ ๊ฐ€์ง€ ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์ด ์žˆ์Šต๋‹ˆ๋‹ค.
06:46
First of all, I could have started anywhere.
164
406909
3131
์ฒซ ๋ฒˆ์งธ๋กœ, ์–ด๋””์—์„œ๋“  ์‹œ์ž‘ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
06:50
I didn't have to start with the word "tell."
165
410040
2243
"๋งํ•˜๋‹ค." ๋กœ ์‹œ์ž‘ํ•  ํ•„์š”๋Š” ์—†์Šต๋‹ˆ๋‹ค.
06:52
I could have started anywhere in the sentence,
166
412283
1416
๋ฌธ์žฅ์—์„œ ์–ด๋””์—์„œ๋“ ์ง€ ์‹œ์ž‘ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค,
06:53
and I could have made this entire thing.
167
413699
1507
๊ทธ๋ฆฌ๊ณ  ์ „์ฒด ๋ฌธ์žฅ์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
06:55
The second thing is, if I wasn't an English speaker,
168
415206
2776
๋‘ ๋ฒˆ์งธ๋กœ ๋งŒ์•ฝ ์ œ๊ฐ€ ์˜์–ด๊ถŒ์ด ์•„๋‹ˆ๊ณ ,
06:57
if I was speaking in some other language,
169
417982
2175
๋‹ค๋ฅธ ์–ธ์–ด๋ฅผ ๋งํ•œ๋‹ค๋ฉด,
07:00
this map would actually hold true in any language.
170
420157
3156
์ด ์ง€๋„๋Š” ์‹ค์ œ๋กœ ๋‹ค๋ฅธ ์–ธ์–ด์—์„œ๋„ ํšจ์šฉ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.
07:03
So long as the questions are standardized,
171
423313
1990
์งˆ๋ฌธ์ด ํ‘œ์ค€ํ™” ๋œ๋‹ค๋ฉด,
07:05
the map is actually independent of language.
172
425303
4287
์‹ค์ œ๋กœ ์ด ์ง€๋„๋Š” ์–ธ์–ด์— ๊ตฌ์• ๋ฐ›์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
07:09
So I call this FreeSpeech,
173
429590
2115
์ด๋ฅผ ์ €๋Š” "์ž์œ ์—ฐ์„ค"์ด๋ผ๊ณ  ๋ถ€๋ฆ…๋‹ˆ๋‹ค.
07:11
and I was playing with this for many, many months.
174
431705
2935
๊ทธ๋ฆฌ๊ณ  ์ €๋Š” ์ˆ˜๊ฐœ์›” ์ด๋ฅผ ๊ฐ€์ง€๊ณ  ๋†€์•˜์Šต๋‹ˆ๋‹ค.
07:14
I was trying out so many different combinations of this.
175
434640
2726
์ €๋Š” ์ด๊ฒƒ์˜ ๋‹ค์–‘ํ•œ ์กฐํ•ฉ์„ ๋งค์šฐ ๋งŽ์ด ์‹œ๋„ํ•ด๋ณด์•˜์Šต๋‹ˆ๋‹ค.
07:17
And then I noticed something very interesting about FreeSpeech.
176
437366
2289
๊ทธ๋ฆฌ๊ณ  ์ž์œ ์—ฐ์„ค์— ๊ด€ํ•œ ๋งค์šฐ ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์„ ๋ˆˆ์น˜ ์ฑ˜์Šต๋‹ˆ๋‹ค.
07:19
I was trying to convert language,
177
439655
3243
์ €๋Š” ์–ธ์–ด๋ฅผ ๋ณ€ํ™”ํ•˜๋ ค๊ณ  ์‹œ๋„ํ–ˆ์Šต๋‹ˆ๋‹ค,
07:22
convert sentences in English into sentences in FreeSpeech,
178
442898
2384
์˜์–ด๋กœ ๋œ ๋ฌธ์žฅ์„ ๋ฐ˜๋Œ€๋กœ ์•ž ๋’ค๋กœ
07:25
and vice versa, and back and forth.
179
445282
1752
์ž์œ ์—ฐ์„ค๋กœ ๋ณ€ํ™˜์‹œ์ผฐ์Šต๋‹ˆ๋‹ค.
07:27
And I realized that this particular configuration,
180
447034
2255
์ œ๊ฐ€ ๊นจ๋‹ฌ์€ ๊ฒƒ์€ ํŠน์ •๋ฐฐ์—ด,
07:29
this particular way of representing language,
181
449289
2026
์–ธ์–ด๋ฅผ ๋Œ€ํ‘œํ•˜๋Š” ์ด ํŠน์ • ๋ฐฉ๋ฒ•์ด,
07:31
it allowed me to actually create very concise rules
182
451315
4395
์‹ค์ œ๋กœ ๋งค์šฐ ๊ฐ„๊ฒฐํ•œ ๋ฒ•์น™๋“ค์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ๊ฒŒ ํ•ด์ค€๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค,
07:35
that go between FreeSpeech on one side
183
455710
2734
์ด ๋ฒ•์น™์€ ์ž์œ ์—ฐ์„ค์˜ ํ•œ ์ธก๋ฉด๊ณผ
07:38
and English on the other.
184
458444
1488
์˜์–ด์˜ ๋‹ค๋ฅธ ์ธก๋ฉด ์‚ฌ์ด์— ์žˆ์Šต๋‹ˆ๋‹ค.
07:39
So I could actually write this set of rules
185
459932
2180
์‹ค์ œ๋กœ ์ €๋Š” ์ด ๊ทœ์น™ ์ฒด๊ณ„๋ฅผ ์ ์„ ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค,
07:42
that translates from this particular representation into English.
186
462112
3395
์ด ๊ทœ์น™๋“ค์€ ํŠน์ • ๋Œ€ํ‘œ ๋‹จ์–ด๋ฅผ ์˜์–ด๋กœ ๋ณ€ํ™” ์‹œํ‚ต๋‹ˆ๋‹ค.
07:45
And so I developed this thing.
187
465507
1831
๊ทธ๋ฆฌ๊ณ  ์ด๊ฒƒ๋“ค์„ ๊ฐœ๋ฐœํ•˜์˜€์Šต๋‹ˆ๋‹ค.
07:47
I developed this thing called the FreeSpeech Engine
188
467338
2232
์ž์œ ์—ฐ์„ค ์—”์ง„์ด๋ผ ๋ถˆ๋ฆฌ๋Š” ๊ฒƒ์ด์—ˆ์ฃ .
07:49
which takes any FreeSpeech sentence as the input
189
469570
2561
์ด๋Š” ์ž์œ ์—ฐ์„ค ๋ฌธ์žฅ์„ ํŠน์ • ์ธํ’‹์œผ๋กœ ๋ฐ›์•„๋“ค์—ฌ
07:52
and gives out perfectly grammatical English text.
190
472131
3930
์™„๋ฒฝํ•œ ์˜์–ด ๋ฌธ๋ฒ• ๋ฌธ์žฅ์œผ๋กœ ๋ณ€ํ™”์‹œ์ผœ์ค๋‹ˆ๋‹ค.
07:56
And by putting these two pieces together,
191
476061
1605
๋Œ€ํ‘œ ๋‹จ์–ด ๊ทธ๋ฆฌ๊ณ  ์—”์ง„ ๋“ฑ
07:57
the representation and the engine,
192
477666
1881
์ด ๋‘๊ฐœ ์กฐ๊ฐ๋“ค์„ ํ•˜๋‚˜๋กœ ๋ฌถ์Œ์œผ๋กœ์จ,
07:59
I was able to create an app, a technology for children with autism,
193
479547
3796
์žํ์•„ ์•„์ด๋“ค์„ ์œ„ํ•œ ํ•˜๋‚˜์˜ ์•ฑ์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค,
08:03
that not only gives them words
194
483343
2499
๊ทธ๋“ค์—๊ฒŒ ๋‹จ์–ด๋“ค์„ ์ค„๋ฟ๋งŒ ์•„๋‹ˆ๋ผ
08:05
but also gives them grammar.
195
485842
3941
๋ฌธ๋ฒ•๊นŒ์ง€ ์ค„ ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.
08:09
So I tried this out with kids with autism,
196
489783
2360
๊ทธ๋ž˜์„œ ์žํ์•„ ์•„์ด๋“ค์—๊ฒŒ ์ด๋ฅผ ์‹œํ—˜ํ•ด ๋ณด์•˜๊ณ ,
08:12
and I found that there was an incredible amount of identification.
197
492143
5013
๊ทธ๋“ค์˜ ๋งค์šฐ ๋†€๋ผ์šด ์‹๋ณ„ ๋Šฅ๋ ฅ์„ ์ฐพ์„ ์ฐพ์„์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.
08:17
They were able to create sentences in FreeSpeech
198
497156
2720
๊ทธ๋“ค์€ ์ž์œ ์—ฐ์„ค์˜ ๋ฌธ์žฅ๋“ค์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.
08:19
which were much more complicated but much more effective
199
499876
2558
์ด๋Š” ๋”์šฑ ๋” ๋ณต์žกํ•˜๊ณ  ํšจ๊ณผ์ ์ด์—ˆ์Šต๋‹ˆ๋‹ค,
08:22
than equivalent sentences in English,
200
502434
2899
์˜์–ด์˜ ๋ณดํ†ต ๋ฌธ์žฅ๋ณด๋‹ค ๋ง์ด์ฃ ,
08:25
and I started thinking about
201
505333
1682
์ €๋Š” ์™œ ๊ทธ๊ฒƒ์ด ๊ฐ€๋Šฅํ•œ์ง€์— ๋Œ€ํ•ด์„œ
08:27
why that might be the case.
202
507015
1969
์ƒ๊ฐํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์—ˆ์Šต๋‹ˆ๋‹ค.
08:28
And I had an idea, and I want to talk to you about this idea next.
203
508984
4287
์ €๋Š” ์ด์— ๊ด€ํ•ด ํ•˜๋‚˜์˜ ๊ฒฌํ•ด๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค. ๋‹ค์Œ์œผ๋กœ ์—ฌ๋Ÿฌ๋ถ„๋“ค๊ป˜ ์ด ๊ฒฌํ•ด์— ๋Œ€ํ•ด ๋งํ•˜๊ณ  ์‹ถ๋„ค์š”.
08:33
In about 1997, about 15 years back,
204
513271
3142
15๋…„ ์ „์œผ๋กœ ๋Œ์•„๊ฐ€ 1997 ๋…„,
08:36
there were a group of scientists that were trying
205
516413
2011
๋‡Œ๊ฐ€ ์–ธ์–ด๋ฅผ ๋งŒ๋“ค์–ด๋‚ด๋Š” ๋ฐฉ๋ฒ•์„ ์ดํ•ดํ•˜๋ ค
08:38
to understand how the brain processes language,
206
518424
2389
๋…ธ๋ ฅํ•˜๋Š” ํ•œ ๊ณผํ•™์ž ์ง‘๋‹จ์ด ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค,
08:40
and they found something very interesting.
207
520813
1779
๊ทธ๋“ค์€ ๋งค์šฐ ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์„ ๋ฐœ๊ฒฌํ–ˆ์Šต๋‹ˆ๋‹ค.
08:42
They found that when you learn a language
208
522592
1872
๊ทธ๋“ค์€ 2์‚ด ๋œ ์•„์ด๊ฐ€ ์–ธ์–ด๋ฅผ ๋ฐฐ์šธ ๋•Œ
08:44
as a child, as a two-year-old,
209
524464
2912
๋‡Œ์˜ ํŠน์ • ๋ถ€๋ถ„์—์„œ
08:47
you learn it with a certain part of your brain,
210
527376
2366
ํ•™์Šตํ•˜๊ณ  ์žˆ๋‹ค๋Š” ๊ฒƒ๊ณผ
08:49
and when you learn a language as an adult --
211
529742
1600
์„ฑ์ธ์ด ๋˜์–ด์„œ ์–ธ์–ด๋ฅผ ํ•™์Šตํ•  ๋•Œ
08:51
for example, if I wanted to learn Japanese right now โ€”
212
531342
3911
์˜ˆ๋ฅผ๋“ค์–ด, ์ œ๊ฐ€ ๋งŒ์•ฝ ์ผ๋ณธ์–ด๋ฅผ ์ง€๊ธˆ๋‹น์žฅ ๋ฐฐ์šฐ๊ณ  ์‹ถ๋‹ค๋ฉด
08:55
a completely different part of my brain is used.
213
535253
2707
๋‡Œ์˜ ์™„๋ฒฝํ•˜๊ฒŒ ๋‹ค๋ฅธ ๋ถ€๋ถ„์ด ์‚ฌ์šฉ๋˜๊ณ  ์žˆ๋‹ค๋Š” ์‚ฌ์‹ค์„ ์ฐพ์•˜์Šต๋‹ˆ๋‹ค.
08:57
Now I don't know why that's the case,
214
537960
1831
์ง€๊ธˆ ์ €๋Š” ์™œ ๊ทธ๋Ÿฐ์ง€ ๋ชจ๋ฆ…๋‹ˆ๋‹ค,
08:59
but my guess is that that's because
215
539791
1991
ํ•˜์ง€๋งŒ ์ œ ์ถ”์ธก์—
09:01
when you learn a language as an adult,
216
541782
2437
์„ฑ์ธ์ผ ๋•Œ ์–ธ์–ด๋ฅผ ๋ฐฐ์šฐ๋ฉด
09:04
you almost invariably learn it
217
544219
1616
๊ฑฐ์˜ ์–ธ์ œ๋‚˜
09:05
through your native language, or through your first language.
218
545835
4266
๋ชจ๊ตญ์–ด๋ฅผ ํ†ตํ•ด ํ•™์Šตํ•˜๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
09:10
So what's interesting about FreeSpeech
219
550101
3252
์ž์œ ์—ฐ์„ค์— ๊ด€ํ•˜์—ฌ ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์€
09:13
is that when you create a sentence
220
553353
1802
๋ฌธ์žฅ์„ ๋งŒ๋“ค ๋•Œ ํ˜น์€
09:15
or when you create language,
221
555155
1695
์–ธ์–ด๋ฅผ ๋งŒ๋“ค ๋•Œ
09:16
a child with autism creates language with FreeSpeech,
222
556850
3070
์žํ๋ฅผ ๊ฐ€์ง„ ์•„์ด๊ฐ€ ์ž์œ ์—ฐ์„ค๋กœ ์–ธ์–ด๋ฅผ ๋งŒ๋“ ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค,
09:19
they're not using this support language,
223
559920
1833
๊ทธ๋“ค์€ ์ด ์ง€์› ์–ธ์–ด๋ฅผ ์‚ฌ์šฉํ•˜๊ณ  ์žˆ์ง€ ์•Š์Šต๋‹ˆ๋‹ค,
09:21
they're not using this bridge language.
224
561753
2211
๊ทธ๋“ค์€ ์ด ์—ฐ๊ฒฐ ์–ธ์–ด๋ฅผ ์‚ฌ์šฉํ•˜๊ณ  ์žˆ์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
09:23
They're directly constructing the sentence.
225
563964
2657
๊ทธ๋“ค์€ ๋ถ„๋ช…ํ•˜๊ฒŒ ๋ฌธ์žฅ์„ ๋งŒ๋“ค๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
09:26
And so this gave me this idea.
226
566621
2193
๊ทธ๋ฆฌ๊ณ  ์ด ๋ฐฉ๋ฒ•์€ ์ €์—๊ฒŒ ์ด ์•„์ด๋””์–ด๋ฅผ ์ฃผ์—ˆ์Šต๋‹ˆ๋‹ค.
09:28
Is it possible to use FreeSpeech
227
568814
2024
์ž์œ ์—ฐ์„ค์„ ์‚ฌ์šฉํ•˜๋Š”๊ฒŒ ๊ฐ€๋Šฅํ• ๊นŒ์š”,
09:30
not for children with autism
228
570838
2510
์žํ๋ฅผ ๊ฐ€์ง„ ์•„์ด๋“ค๋ฟ๋งŒ ์•„๋‹ˆ๋ผ,
09:33
but to teach language to people without disabilities?
229
573348
6262
์žฅ์• ๊ฐ€ ์—†๋Š” ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ์–ธ์–ด๋ฅผ ๊ฐ€๋ฅด์น˜๋Š”๋ฐ์—๋„ ๋ง์ด์ฃ ?
09:39
And so I tried a number of experiments.
230
579610
1978
๋•Œ๋ฌธ์— ์ €๋Š” ๋งŽ์€ ์‹คํ—˜์„ ํ–ˆ์Šต๋‹ˆ๋‹ค.
09:41
The first thing I did was I built a jigsaw puzzle
231
581588
2948
์ฒซ ๋ฒˆ์งธ๋กœ ์กฐ๊ฐ๊ทธ๋ฆผ ํผ์ฆ์„ ๋งŒ๋“œ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค
09:44
in which these questions and answers
232
584536
1970
์—ฌ๊ธฐ์—์„œ ์ด ์งˆ๋ฌธ๋“ค๊ณผ ๋‹ต๋ณ€๋“ค์€
09:46
are coded in the form of shapes,
233
586506
1835
ํŠน์ • ํ˜•ํƒœ์˜ ๋ชจ์–‘
09:48
in the form of colors,
234
588341
1138
์ƒ‰๊น”๋กœ ์•”ํ˜ธํ™” ๋˜์–ด์žˆ๊ธฐ ๋•Œ๋ฌธ์—,
09:49
and you have people putting these together
235
589479
1849
์‚ฌ๋žŒ๋“ค์—๊ฒŒ ์ด๊ฒƒ๋“ค์„ ํ•˜๋‚˜๋กœ ๋ชจ์•„
09:51
and trying to understand how this works.
236
591328
1773
์–ด๋–ป๊ฒŒ ์ž‘๋™ํ•˜๋Š”์ง€ ์ดํ•ด์‹œํ‚ฌ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
09:53
And I built an app out of it, a game out of it,
237
593101
2376
์ €๋Š” ์ด๋ฅผ ์•ฑ์œผ๋กœ ๊ฒŒ์ž„์œผ๋กœ ๋งŒ๋“ค์—ˆ์Šต๋‹ˆ๋‹ค,
์—ฌ๊ธฐ์„œ ์•„์ด๋“ค์€ ์—ฌ๋Ÿฌ ๋‹จ์–ด๋“ค์„ ๊ฐ€์ง€๊ณ  ๋†€๋ฉฐ
09:55
in which children can play with words
238
595477
2661
09:58
and with a reinforcement,
239
598138
1704
๋ฐ˜๋ณต์ ์œผ๋กœ ๊ฐ•ํ™”ํ•™์Šตํ•˜๊ณ 
09:59
a sound reinforcement of visual structures,
240
599842
2585
๋˜ ์‹œ๊ฐ์ ์ธ ๊ตฌ์กฐ์— ๋Œ€ํ•ด ์†Œ๋ฆฌ๋ฅผ ํ†ตํ•ด ๋‹ค์‹œ ์žฌํ•™์Šตํ•˜๋ฉฐ ๋†€ ์ˆ˜ ์žˆ๊ธฐ ๋•Œ๋ฌธ์—
10:02
they're able to learn language.
241
602427
2013
๊ทธ๋“ค์€ ์–ธ์–ด๋ฅผ ํ•™์Šตํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
10:04
And this, this has a lot of potential, a lot of promise,
242
604440
2736
์ด๋Š” ๋งŽ์€๊ฐ€๋Šฅ์„ฑ์„ ๊ฐ€์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค,
10:07
and the government of India recently
243
607176
1975
์ตœ๊ทผ์— ์ธ๋„ ์ •๋ถ€๋Š”
10:09
licensed this technology from us,
244
609151
1404
์šฐ๋ฆฌ๋กœ๋ถ€ํ„ฐ ์ด ๊ธฐ์ˆ ์„ ๋ผ์ด์„ผ์Šค ๋ฐ›์•˜์Šต๋‹ˆ๋‹ค,
10:10
and they're going to try it out with millions of different children
245
610555
2074
๊ทธ๋“ค์€ ์ด๋ฅผ ์˜์–ด๋ฅผ ๋ฐฐ์šฐ๋Š” ์ˆ˜ ๋ฐฑ๋งŒ๋ช…์˜ ๋‹ค๋ฅธ ์•„์ด๋“ค์—
10:12
trying to teach them English.
246
612629
2605
์‹œํ—˜ํ•ด๋ณด๋ ค ํ•ฉ๋‹ˆ๋‹ค.
10:15
And the dream, the hope, the vision, really,
247
615234
2614
์ด๊ฒƒ์— ๋Œ€ํ•œ ์ €์˜ ๊ฟˆ, ํฌ๋ง, ๋น„์ ผ์€
10:17
is that when they learn English this way,
248
617848
3082
๊ทธ๋“ค์ด ์ด๋Ÿฐ๋ฐฉ์‹์œผ๋กœ ์˜์–ด๋ฅผ ๋ฐฐ์šธ ๋•Œ
10:20
they learn it with the same proficiency
249
620930
2643
๊ทธ๋“ค์€ ๋ชจ๊ตญ์–ด๋ฅผ ํ•™์Šตํ•˜๋Š” ๊ฒƒ์ฒ˜๋Ÿผ
10:23
as their mother tongue.
250
623573
3718
ํšจ์œจ์ ์œผ๋กœ ํ•™์Šตํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
10:27
All right, let's talk about something else.
251
627291
3816
์ข‹์Šต๋‹ˆ๋‹ค, ๋‹ค๋ฅธ ์ด์•ผ๊ธฐ๋ฅผ ํ•ด๋ณด์ฃ .
10:31
Let's talk about speech.
252
631107
1997
์—ฐ์„ค์— ๊ด€ํ•˜์—ฌ ๋งํ•ด๋ณด์ฃ .
10:33
This is speech.
253
633104
1271
์ด๋Š” ์—ฐ์„ค์ž…๋‹ˆ๋‹ค.
10:34
So speech is the primary mode of communication
254
634375
1962
๊ทธ๋ ‡๋‹ค๋ฉด ์—ฐ์„ค์€ ์šฐ๋ฆฌ ๋ชจ๋‘ ์‚ฌ์ด์—์„œ ์ผ์–ด๋‚˜๋Š”
10:36
delivered between all of us.
255
636337
1613
์ปค๋ฎค๋‹ˆ์ผ€์ด์…˜์˜ ๊ธฐ์ดˆ ๋ชจ๋ธ์ž…๋‹ˆ๋‹ค.
10:37
Now what's interesting about speech is that
256
637950
1855
์—ฐ์„ค์— ๊ด€ํ•œ ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์€
10:39
speech is one-dimensional.
257
639805
1245
์—ฐ์„ค์€ 1 ์ฐจ์›์ ์ด๋ผ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
10:41
Why is it one-dimensional?
258
641050
1359
์™œ 1 ์ฐจ์›์ผ๊นŒ์š”?
10:42
It's one-dimensional because it's sound.
259
642409
1568
์ด์œ ๋Š” ์†Œ๋ฆฌ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
10:43
It's also one-dimensional because
260
643977
1539
1 ์ฐจ์›์ธ ๋˜ ๋‹ค๋ฅธ ์ด์œ ๋Š”
10:45
our mouths are built that way.
261
645516
1205
์šฐ๋ฆฌ์˜ ์ž…์ด ์ด๋Ÿฐ ๋ฐฉ์‹์œผ๋กœ ๋งŒ๋“ค์–ด์กŒ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
10:46
Our mouths are built to create one-dimensional sound.
262
646721
3512
์šฐ๋ฆฌ์˜ ์ž…์€ 1์ฐจ์›์  ์†Œ๋ฆฌ๋ฅผ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋„๋ก ๋งŒ๋“ค์–ด์กŒ์Šต๋‹ˆ๋‹ค.
10:50
But if you think about the brain,
263
650233
2866
ํ•˜์ง€๋งŒ ๋‡Œ๋ฅผ ์ƒ๊ฐํ•œ๋‹ค๋ฉด,
10:53
the thoughts that we have in our heads
264
653099
1764
์šฐ๋ฆฌ์˜ ๋จธ๋ฆฌ์— ์žˆ๋Š” ์ƒ๊ฐ์€
10:54
are not one-dimensional.
265
654863
2102
1 ์ฐจ์›์ ์ด์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
10:56
I mean, we have these rich,
266
656965
1459
์ œ ๋ง์€, ์šฐ๋ฆฌ๋Š” ์ด ํ’๋ถ€ํ•˜๊ณ ,
10:58
complicated, multi-dimensional ideas.
267
658424
3028
๋ณต์žกํ•˜๋ฉฐ, ๋‹ค๋ฉด์ ์ธ ์•„์ด๋””์–ด๋ฅผ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
11:01
Now, it seems to me that language
268
661452
1690
์ง€๊ธˆ, ์ €์—๊ฒŒ ์–ธ์–ด๋Š” ์ด ํ’๋ถ€ํ•˜๊ณ ,
11:03
is really the brain's invention
269
663142
2332
๋‹ค๋ฉด์  ์ƒ๊ฐ์˜ ํ•œ ์ธก๋ฉด์„
11:05
to convert this rich, multi-dimensional thought
270
665474
3096
๋‹ค๋ฅธ ํ•œ ์ธก๋ฉด์ธ ์—ฐ์„ค๋กœ ์ „ํ™˜์‹œํ‚ค๊ธฐ์œ„ํ•œ
11:08
on one hand
271
668570
1587
๋‡Œ์˜ ๋ฐœ๋ช…ํ’ˆ์ฒ˜๋Ÿผ
11:10
into speech on the other hand.
272
670157
1923
๋Š๊ปด์ง‘๋‹ˆ๋‹ค.
11:12
Now what's interesting is that
273
672080
1762
์ง€๊ธˆ ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์€
11:13
we do a lot of work in information nowadays,
274
673842
2568
ํ˜„์žฌ ์šฐ๋ฆฌ๊ฐ€ ์ •๋ณด ๋ถ„์•ผ์—์„œ ๋งŽ์€ ์ผ์„ ํ•˜๋ฉฐ,
11:16
and almost all of that is done in the language domain.
275
676410
3079
๋Œ€๋ถ€๋ถ„์˜ ๋ชจ๋“  ๊ฒƒ์ด ์–ธ์–ด ๋„๋ฉ”์ธ์—์˜ํ•ด ์ด๋ฃจ์–ด์ง‘๋‹ˆ๋‹ค.
11:19
Take Google, for example.
276
679489
1939
๊ตฌ๊ธ€์˜ ์‚ฌ๋ก€๋ฅผ ๋ณด์ฃ .
11:21
Google trawls all these countless billions of websites,
277
681428
2677
๊ตฌ๊ธ€์€ ์ˆ˜ ์‹ญ์–ต๊ฐœ์˜ ์›น์‚ฌ์ดํŠธ๋ฅผ ์ƒ…์ƒ…์ด ์กฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.
11:24
all of which are in English, and when you want to use Google,
278
684105
2725
๋ชจ๋“  ๊ฒƒ์€ ์˜์–ด๋กœ ๋˜์–ด์žˆ์Šต๋‹ˆ๋‹ค, ๊ตฌ๊ธ€์„ ์‚ฌ์šฉํ•˜๊ณ  ์‹ถ์„ ๋•Œ,
11:26
you go into Google search, and you type in English,
279
686830
2450
์—ฌ๋Ÿฌ๋ถ„์€ ๊ตฌ๊ธ€ ๊ฒ€์ƒ‰์œผ๋กœ ๊ฐ€์„œ ์˜์–ด๋กœ ํƒ€์ดํ•‘ํ•˜๋ฉด,
11:29
and it matches the English with the English.
280
689280
4163
๊ฒ€์ƒ‰์—”์ง„์€ ์˜๋‹จ์–ด๋ฅผ ๋งค์นญ์‹œํ‚ต๋‹ˆ๋‹ค.
11:33
What if we could do this in FreeSpeech instead?
281
693443
3583
๋งŒ์•ฝ ์šฐ๋ฆฌ๊ฐ€ ์ด๊ฒƒ์„ ์ž์œ ์—ฐ์„ค๋กœ ๋Œ€์ฒ˜ํ•  ์ˆ˜ ์žˆ๋‹ค๋ฉด ์–ด๋–จ๊นŒ์š”?
11:37
I have a suspicion that if we did this,
282
697026
2301
์ €๋Š” ์ด๊ฒŒ ๊ฐ€๋Šฅํ•œ์ง€ ๋งค์šฐ ์˜์‹ฌ์Šค๋Ÿฝ์Šต๋‹ˆ๋‹ค,
11:39
we'd find that algorithms like searching,
283
699327
2068
์šฐ๋ฆฌ๋Š” ๊ฒ€์ƒ‰๊ณผ ๊ฐ™์€ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ฐพ๊ฒŒ ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค,
11:41
like retrieval, all of these things,
284
701395
2325
ํšŒ์ˆ˜ ํ•˜๋Š” ๊ฒƒ์ฒ˜๋Ÿผ, ์ด ๋ชจ๋“  ๊ฒƒ๋“ค์€,
11:43
are much simpler and also more effective,
285
703720
3075
๋ณด๋‹ค ๊ฐ„๋‹จํ•˜๋ฉด์„œ๋„ ํšจ๊ณผ์ ์ž…๋‹ˆ๋‹ค,
11:46
because they don't process the data structure of speech.
286
706795
4417
๊ทธ ์ด์œ ๋Š” ๊ทธ๊ฒƒ๋“ค์€ ์—ฐ์„ค ๋ฐ์ดํ„ฐ ๊ตฌ์กฐ๋ฅผ ์‹คํ–‰ํ•˜์ง€ ๋ชปํ•˜๋Š” ๋Œ€์‹ ์—
11:51
Instead they're processing the data structure of thought.
287
711212
5976
์ƒ๊ฐ์˜ ๋ฐ์ดํ„ฐ ๊ตฌ์กฐ๋ฅผ ์‹คํ–‰ํ•˜๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
11:57
The data structure of thought.
288
717188
2808
์ƒ๊ฐ์˜ ๋ฐ์ดํ„ฐ ๊ตฌ์กฐ๋ฅผ ๋ง์ด์ฃ .
11:59
That's a provocative idea.
289
719996
2076
๋งค์šฐ ์ง„๋ณด์ ์ธ ์•„์ด๋””์–ด์ž…๋‹ˆ๋‹ค.
12:02
But let's look at this in a little more detail.
290
722072
2142
์ด๊ฒƒ์„ ์ข€๋” ์„ธ๋ฐ€ํ•˜๊ฒŒ ๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.
12:04
So this is the FreeSpeech ecosystem.
291
724214
2366
์ด๊ฒƒ๋“ค ์ž์œ ์—ฐ์„ค ์ƒํƒœ๊ณ„์ž…๋‹ˆ๋‹ค.
12:06
We have the Free Speech representation on one side,
292
726580
2884
์šฐ๋ฆฌ๋Š” ํ•œ ์ธก๋ฉด์˜ ๋Œ€ํ‘œ ์ž์œ ์—ฐ์„ค์„ ๊ฐ€์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
12:09
and we have the FreeSpeech Engine, which generates English.
293
729464
2228
์˜์–ด๋กœ ์‹คํ–‰๊ฐ€๋Šฅํ•œ ์ž์œ ์—ฐ์„ค์—”์ง„์ž…๋‹ˆ๋‹ค.
12:11
Now if you think about it,
294
731694
1725
์—ฌ๋Ÿฌ๋ถ„๊ป˜์„œ ์ด๊ฒƒ์— ๊ด€ํ•ด ์ƒ๊ฐํ•ด๋ณด์‹ ๋‹ค๋ฉด
12:13
FreeSpeech, I told you, is completely language-independent.
295
733419
2544
์ œ๊ฐ€ ์–ธ๊ธ‰ํ•œ ์ž์œ ์—ฐ์„ค์€ ์™„์ „ํžˆ ์–ธ์–ด์— ๊ตฌ์• ๋ฐ›์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
12:15
It doesn't have any specific information in it
296
735963
2087
์˜์–ด์— ๊ด€ํ•œ ํŠน์ • ์ •๋ณด๋ฅผ
12:18
which is about English.
297
738050
1228
๊ฐ€์ง€๊ณ  ์žˆ์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
12:19
So everything that this system knows about English
298
739278
2800
์ด ์‹œ์Šคํ…œ์ด ์˜์–ด์— ๊ด€ํ•˜์—ฌ ์•Œ๊ณ  ์žˆ๋Š” ๋ชจ๋“  ๊ฒƒ์€
12:22
is actually encoded into the engine.
299
742078
4620
์‹ค์ œ๋กœ ํŠน์ • ์—”์ง„์œผ๋กœ ์•”ํ˜ธํ™” ๋˜์–ด์žˆ์Šต๋‹ˆ๋‹ค.
12:26
That's a pretty interesting concept in itself.
300
746698
2237
๊ทธ๊ฒƒ์€ ๊ฝค ํฅ๋ฏธ๋กœ์šด ๊ฐœ๋…์ž…๋‹ˆ๋‹ค.
12:28
You've encoded an entire human language
301
748935
3604
์—ฌ๋Ÿฌ๋ถ„์€ ์ „์ฒด ์ธ๊ฐ„์–ธ์–ด๋ฅผ
12:32
into a software program.
302
752539
2645
์†Œํ”„ํŠธ์›จ์–ด ํ”„๋กœ๊ทธ๋žจ์— ์•”ํ˜ธํ™” ํ–ˆ์Šต๋‹ˆ๋‹ค.
12:35
But if you look at what's inside the engine,
303
755184
2531
ํ•˜์ง€๋งŒ ์—ฌ๋Ÿฌ๋ถ„๊ป˜์„œ ์—”์ง„ ์•ˆ์— ๋ฌด์—‡์ด ์žˆ๋Š”์ง€ ๊ด€์ฐฐํ•œ๋‹ค๋ฉด,
12:37
it's actually not very complicated.
304
757715
2358
์ด๋Š” ์‹ค์ œ๋กœ ๋ณต์žกํ•˜์ง€๋Š” ์•Š์Šต๋‹ˆ๋‹ค,
12:40
It's not very complicated code.
305
760073
2105
๋งค์šฐ ๋ณต์žกํ•˜์ง€ ์•Š์€ ์ฝ”๋“œ์ด์ฃ .
12:42
And what's more interesting is the fact that
306
762178
2672
๋”์šฑ๋” ํฅ๋ฏธ๋กœ์šด ์‚ฌ์‹ค์€
12:44
the vast majority of the code in that engine
307
764850
2203
์—”์ง„์˜ ๋Œ€๋ถ€๋ถ„์˜ ๋งŽ์€ ์ฝ”๋“œ๋Š”
12:47
is not really English-specific.
308
767053
2412
๋”ฑํžˆ ์˜์–ด๋กœ ์ƒ์„ธํ™” ๋˜์ง€ ์•Š์•˜๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
12:49
And that gives this interesting idea.
309
769465
1895
๊ทธ๊ฒƒ์€ ๋งค์šฐ ํฅ๋ฏธ๋กœ์šด ์•„์ด๋””์–ด๋ฅผ ์ œ์‹œํ•ด์ค๋‹ˆ๋‹ค.
12:51
It might be very easy for us to actually
310
771360
2038
์‹ค์ œ๋กœ ์šฐ๋ฆฌ๊ฐ€ ํžŒ๋‘์–ด, ๋ถˆ์–ด, ๋…์–ด, ์Šค์™€ํž๋ฆฌ์–ด์™€ ๊ฐ™์ด
12:53
create these engines in many, many different languages,
311
773398
3826
๋งŽ์€ ๋‹ค๋ฅธ ์–ธ์–ด์˜ ์ด๋Ÿฐ ์—”์ง„๋“ค์„
12:57
in Hindi, in French, in German, in Swahili.
312
777224
6354
์‰ฝ๊ฒŒ ๋งŒ๋“ค์ง€ ๋ชจ๋ฆ…๋‹ˆ๋‹ค,
13:03
And that gives another interesting idea.
313
783578
2799
๊ทธ๋ฆฌ๊ณ  ๋˜ ๋‹ค๋ฅธ ํฅ๋ฏธ๋กœ์šด ์•„์ด๋””์–ด๋ฅผ ์ œ์‹œํ•ด์ค๋‹ˆ๋‹ค.
13:06
For example, supposing I was a writer,
314
786377
2654
์˜ˆ๋ฅผ๋“ค์–ด, ์ œ๊ฐ€ ๋‰ด์Šค๋‚˜, ๋งค๊ฑฐ์ง„
13:09
say, for a newspaper or for a magazine.
315
789031
2122
์ž‘๊ฐ€๋ผ๊ณ  ๊ฐ€์ •ํ–ˆ์„ ๋•Œ,
13:11
I could create content in one language, FreeSpeech,
316
791153
5011
์ €๋Š” ํ•˜๋‚˜์˜ ์–ธ์–ด, ์ž์œ ์—ฐ์„ค์˜ ์ปจํ…์ธ ๋ฅผ ๋งŒ๋“ค์–ด ๋‚ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค, ๊ทธ๋ฆฌ๊ณ 
13:16
and the person who's consuming that content,
317
796164
2056
์ปจํ…Œ์ธ ๋ฅผ ์†Œ๋น„ํ•˜๊ฑฐ๋‚˜
13:18
the person who's reading that particular information
318
798220
3061
ํŠน์ • ์ •๋ณด๋ฅผ ์ฝ๋Š” ์‚ฌ๋žŒ๋“ค์€
13:21
could choose any engine,
319
801281
2495
์–ด๋– ํ•œ ์—”์ง„์ด๋“  ์„ ํƒํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ,
13:23
and they could read it in their own mother tongue,
320
803776
2736
๊ทธ๋“ค์€ ๋ชจ๊ตญ์–ด๋กœ ์ฝ์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
13:26
in their native language.
321
806512
3939
์ž๊ตญ์–ด๋กœ ๋ง์ด์ฃ .
13:30
I mean, this is an incredibly attractive idea,
322
810451
2722
์ œ ๋ง์€, ์ด ์•„์ด๋””์–ด๋Š” ๋งค์šฐ ๋งค๋ ฅ์ ์ž…๋‹ˆ๋‹ค,
13:33
especially for India.
323
813173
1999
ํŠนํžˆ ์ธ๋„์—์„œ ๋ง์ด์ฃ .
13:35
We have so many different languages.
324
815172
1690
์ธ๋„๋Š” ๊ต‰์žฅํžˆ ๋‹ค์–‘ํ•œ ์–ธ์–ด๋ฅผ ์‚ฌ์šฉํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
13:36
There's a song about India, and there's a description
325
816862
2142
์ธ๋„์— ๊ด€ํ•œ ํ•˜๋‚˜์˜ ๋…ธ๋ž˜ ์„ค๋ช…์ด ์žˆ์Šต๋‹ˆ๋‹ค,
13:39
of the country as, it says,
326
819004
2344
๊ทธ๊ฒƒ์€,
13:41
(in Sanskrit).
327
821348
2360
(์‚ฐ์Šคํฌ๋ฆฌํŠธ์–ด)
13:43
That means "ever-smiling speaker
328
823708
2773
๊ทธ ์˜๋ฏธ๋Š” "๋ฏธ์†Œ๋Š” ์•„๋ฆ„๋‹ค์šด ์–ธ์–ด."
13:46
of beautiful languages."
329
826481
4519
์ž…๋‹ˆ๋‹ค.
13:51
Language is beautiful.
330
831000
1964
์–ธ์–ด๋Š” ์•„๋ฆ„๋‹ต์Šต๋‹ˆ๋‹ค.
13:52
I think it's the most beautiful of human creations.
331
832964
2454
์ €๋Š” ์–ธ์–ด๋Š” ์ธ๊ฐ„์ด ๋งŒ๋“ค์–ด ๋‚ธ ๊ฐ€์žฅ ์•„๋ฆ„๋‹ค์šด ๊ฒƒ์ด๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
13:55
I think it's the loveliest thing that our brains have invented.
332
835418
3978
์šฐ๋ฆฌ์˜ ๋‡Œ๊ฐ€ ๋งŒ๋“ค์–ด๋‚ธ ๊ฐ€์žฅ ์‚ฌ๋ž‘์Šค๋Ÿฌ์šด ๊ฒƒ์ด๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
13:59
It entertains, it educates, it enlightens,
333
839396
3584
๊ธฐ์จ์„ ์ฃผ๊ณ , ๊ต์œก ์‹œํ‚ค๋ฉฐ, ๊นจ๋‹ฌ์Œ์„ ์ค๋‹ˆ๋‹ค,
14:02
but what I like the most about language
334
842980
2044
ํ•˜์ง€๋งŒ ์–ธ์–ด์— ๊ด€ํ•˜์—ฌ ์ œ๊ฐ€ ๊ฐ€์žฅ ์ข‹์•„ํ•˜๋Š” ๋ถ€๋ถ„์€
14:05
is that it empowers.
335
845024
1500
์ž์œจ๊ถŒ์„ ์ค€๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
14:06
I want to leave you with this.
336
846524
1838
์ €๋Š” ์ด ๋ถ€๋ถ„์„ ์—ฌ๋Ÿฌ๋ถ„๊ป˜์„œ ๊ธฐ์–ตํ•˜์‹œ๊ธฐ๋ฅผ ๋ฐ”๋ž๋‹ˆ๋‹ค.
14:08
This is a photograph of my collaborators,
337
848362
2385
์ด ์‚ฌ์ง„์€ ์ €์˜ ๋™๋ฃŒ๋“ค์ž…๋‹ˆ๋‹ค,
14:10
my earliest collaborators
338
850747
997
์ €์˜ ์ดˆ์ฐฝ๊ธฐ ๋™๋ฃŒ๋“ค์ž…๋‹ˆ๋‹ค,
14:11
when I started working on language
339
851744
1462
์ œ๊ฐ€ ์–ธ์–ด์™€ ์žํ ๊ทธ๋ฆฌ๊ณ  ๋งŽ์€ ๊ฒƒ๋“ค์— ๊ด€ํ•˜์—ฌ
14:13
and autism and various other things.
340
853206
1502
์—ฐ๊ตฌํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์„ ๋•Œ์ฃ .
14:14
The girl's name is Pavna,
341
854708
1417
์ € ์†Œ๋…€์˜ ์ด๋ฆ„์€ ํŒŒ๋ธŒ๋‚˜์ž…๋‹ˆ๋‹ค,
14:16
and that's her mother, Kalpana.
342
856125
1902
๊ทธ๋…€์˜ ์–ด๋จธ๋‹ˆ๋Š” ์นผํ”„๋‚˜์ž…๋‹ˆ๋‹ค.
14:18
And Pavna's an entrepreneur,
343
858027
2138
ํŒŒ๋ธŒ๋‚˜๋Š” ์‚ฌ์—…๊ฐ€์ž…๋‹ˆ๋‹ค,
14:20
but her story is much more remarkable than mine,
344
860165
2371
ํ•˜์ง€๋งŒ ๊ทธ๋…€์˜ ์ด์•ผ๊ธฐ๋Š” ์ œ๊ฐ€ ๊ฐ€์ง„ ์ด์•ผ๊ธฐ ๋ณด๋‹ค ๋ˆˆ์˜ ๋„์ž…๋‹ˆ๋‹ค,
14:22
because Pavna is about 23.
345
862536
2400
๊ทธ๋…€๋Š” 23 ์‚ด์ด๋ฉฐ,
14:24
She has quadriplegic cerebral palsy,
346
864936
2552
๋‡Œ์„ฑ๋งˆ๋น„๋ฅผ ๊ฐ€์ง€๊ณ  ๊ฒช๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค,
14:27
so ever since she was born,
347
867488
1640
๊ทธ๋…€๊ฐ€ ํƒœ์–ด๋‚˜๋ฉด์„œ ๋ถ€ํ„ฐ ๋ง์ด์ฃ .
14:29
she could neither move nor talk.
348
869128
3600
๊ทธ๋…€๋Š” ์›€์ง์ผ ์ˆ˜๋„ ๋ง์„ ํ•  ์ˆ˜๋„ ์—†์Šต๋‹ˆ๋‹ค.
14:32
And everything that she's accomplished so far,
349
872728
2403
์ง€๊ธˆ๊นŒ์ง€ ๊ทธ๋…€๊ฐ€ ์ด๋ฃฌ ๋ชจ๋“  ๊ฒƒ์€
14:35
finishing school, going to college,
350
875131
2227
์ •๊ทœ ๊ณผ์ •์„ ๋งˆ์น˜๊ณ , ๋Œ€ํ•™์— ๊ฐ€
14:37
starting a company,
351
877358
1416
ํšŒ์‚ฌ๋ฅผ ์šด์˜ํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ๊ณ ,
14:38
collaborating with me to develop Avaz,
352
878774
2140
Avaz๋ฅผ ๊ฐœ๋ฐœํ•˜๊ธฐ ์œ„ํ•ด ์ €์™€ ํ˜‘๋ ฅํ•˜์˜€์Šต๋‹ˆ๋‹ค,
14:40
all of these things she's done
353
880914
1892
๊ทธ๋…€๊ฐ€ ์ด๋ฃฌ ๋ชจ๋“  ๊ฒƒ๋“ค์€
14:42
with nothing more than moving her eyes.
354
882806
5523
๊ทธ๋…€์˜ ๋ˆˆ์„ ์›€์ง์ธ ๊ฒƒ ๊ทธ ์ด์ƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
14:48
Daniel Webster said this:
355
888329
2689
๋‹ค๋‹ˆ์—˜ ์›น์Šคํ„ฐ๊ฐ€ ๋งํ•˜๊ธฐ๋ฅผ
14:51
He said, "If all of my possessions were taken
356
891018
2940
"์ œ๊ฐ€ ๊ฐ€์ง„ ๋ชจ๋“  ๊ฒƒ ์ค‘
14:53
from me with one exception,
357
893958
2988
๋‹จ ํ•˜๋‚˜๋งŒ ์ทจํ•œ๋‹ค๋ฉด
14:56
I would choose to keep the power of communication,
358
896946
2981
์ €๋Š” ์ปค๋ฎค๋‹ˆ์ผ€์ด์…˜ ๋Šฅ๋ ฅ์„ ์„ ํƒํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค,
14:59
for with it, I would regain all the rest."
359
899927
3903
์ปค๋ฎค๋‹ˆ์ผ€์ด์…˜ ๋Šฅ๋ ฅ์œผ๋กœ ๋‚˜๋จธ์ง€ ๊ฒƒ๋“ค์„ ๋˜์ฐพ์„ ์ˆ˜ ์žˆ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
15:03
And that's why, of all of these incredible applications of FreeSpeech,
360
903830
5116
์ด๋Ÿฌํ•œ ์ด์œ ๋กœ, ์ž์œ ์—ฐ์„ค์— ๊ด€ํ•œ ์ด ๋†€๋ผ์šด ์–ดํ”Œ๋ฆฌ์ผ€์ด์…˜์€
15:08
the one that's closest to my heart
361
908946
2080
์ €์˜ ๊ฐ€์Šด ๊นŠ์ด ๋†“์—ฌ์žˆ์Šต๋‹ˆ๋‹ค
15:11
still remains the ability for this
362
911026
2068
์ด๋Š” ์—ฌ์ „ํžˆ ์žฅ์• ๋ฅผ ๊ฐ€์ง„ ์•„์ด์—๊ฒŒ
15:13
to empower children with disabilities
363
913094
2380
์ž์œจ๊ถŒ์„ ์ฃผ๋ฉฐ, ์˜์‚ฌ์†Œํ†ตํ•  ์ˆ˜ ์žˆ๊ฒŒ
15:15
to be able to communicate,
364
915474
1773
ํ•ด์ค๋‹ˆ๋‹ค,
15:17
the power of communication,
365
917247
1789
์ปค๋ฎค๋‹ˆ์ผ€์ด์…˜ ๋Šฅ๋ ฅ์€
15:19
to get back all the rest.
366
919036
2240
๋‚˜๋จธ์ง€ ๋ชจ๋“  ๊ฒƒ์„ ๋Œ๋ ค ๋†“์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
15:21
Thank you.
367
921276
1397
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.
15:22
(Applause)
368
922673
1332
(๋ฐ•์ˆ˜)
15:24
Thank you. (Applause)
369
924005
4199
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค. (๋ฐ•์ˆ˜)
15:28
Thank you. Thank you. Thank you. (Applause)
370
928204
5323
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค. ๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค. ๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.(๋ฐ•์ˆ˜)
15:33
Thank you. Thank you. Thank you. (Applause)
371
933527
4000
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค. ๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค. ๊ฐ์‚ฌํ•ฉ๋‹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