Jennifer Golbeck: The curly fry conundrum: Why social media "likes" say more than you might think

376,125 views

2014-04-03 ใƒป TED


New videos

Jennifer Golbeck: The curly fry conundrum: Why social media "likes" say more than you might think

376,125 views ใƒป 2014-04-03

TED


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

๋ฒˆ์—ญ: Jihyeon J. Kim ๊ฒ€ํ† : Catherine YOO
00:12
If you remember that first decade of the web,
0
12738
1997
์ธํ„ฐ๋„ท์˜ ์ดˆ๊ธฐ 10๋…„์„ ์ƒ๊ฐํ•˜๋ฉด
00:14
it was really a static place.
1
14735
2255
๊ฑฐ๊ธด ์ •๋ง ์ •์ ์ธ ๊ณณ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
00:16
You could go online, you could look at pages,
2
16990
2245
์˜จ๋ผ์ธ์œผ๋กœ ํŽ˜์ด์ง€๋“ค์„ ๋ณผ ์ˆ˜ ์žˆ์—ˆ์ฃ 
00:19
and they were put up either by organizations
3
19235
2513
๊ทธ๊ฒƒ์€ ํŒ€์œผ๋กœ ์ด๋ฃจ์–ด์ง„ ๋‹จ์ฒด์—์„œ
00:21
who had teams to do it
4
21748
1521
์˜ฌ๋ฆฐ ๊ฒƒ์ด๊ฑฐ๋‚˜
00:23
or by individuals who were really tech-savvy
5
23269
2229
๊ทธ ๋‹น์‹œ ์ปดํ“จํ„ฐ ๊ธฐ์ˆ ์— ๋Šฅ์ˆ™ํ•œ
00:25
for the time.
6
25498
1737
๊ฐœ์ธ๋“ค์ด ์˜ฌ๋ฆฐ ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
00:27
And with the rise of social media
7
27235
1575
2000๋…„๋Œ€ ์ดˆ๊ธฐ์— ๋ถ€์ƒํ•˜๊ธฐ ์‹œ์ž‘ํ•œ
00:28
and social networks in the early 2000s,
8
28810
2399
์†Œ์…œ๋ฏธ๋””์–ด์™€ ๋„คํŠธ์›Œํฌ๋กœ ์ธํ•ด
00:31
the web was completely changed
9
31209
2149
์ธํ„ฐ๋„ท์€ ์™„์ „ํžˆ ๋ฐ”๋€Œ์—ˆ์Šต๋‹ˆ๋‹ค.
00:33
to a place where now the vast majority of content
10
33358
3608
์šฐ๋ฆฌ๊ฐ€ ๋ณด๋Š” ์–ด๋งˆ์–ด๋งˆํ•œ ์ฝ˜ํ…์ธ ๋“ค์„
00:36
we interact with is put up by average users,
11
36966
3312
๋ณดํ†ต์˜ ์‚ฌ์šฉ์ž๋“ค์ด ์˜ฌ๋ฆฌ๋Š” ์žฅ์†Œ๋กœ ๋ง์ž…๋‹ˆ๋‹ค.
00:40
either in YouTube videos or blog posts
12
40278
2697
์œ ํˆฌ๋ธŒ ๋™์˜์ƒ์ด๋ผ๋“ ์ง€ ๋ธ”๋กœ๊ทธ,
00:42
or product reviews or social media postings.
13
42975
3315
๋˜๋Š” ์ƒํ’ˆํ›„๊ธฐ๋‚˜ ์†Œ์…œ๋ฏธ๋””์–ด ๊ธ€๋“ค์ด์š”.
00:46
And it's also become a much more interactive place,
14
46290
2347
์‚ฌ๋žŒ๋“ค์ด ์„œ๋กœ ์†Œํ†ตํ•˜๋Š”
00:48
where people are interacting with others,
15
48637
2637
ํ›จ์”ฌ ์ƒํ˜ธ์ ์ธ ์žฅ์†Œ๊ฐ€ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
00:51
they're commenting, they're sharing,
16
51274
1696
๋‹ต๊ธ€์„ ๋‚จ๊ธฐ๊ณ , ๊ณต์œ ํ•ฉ๋‹ˆ๋‹ค.
00:52
they're not just reading.
17
52970
1614
๊ทธ์ € ์ฝ๊ธฐ๋งŒ ํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
00:54
So Facebook is not the only place you can do this,
18
54584
1866
ํŽ˜์ด์Šค๋ถ์€ ์ด๋Ÿฐ ๊ฒƒ์„ ํ•  ์ˆ˜ ์žˆ๋Š” ์œ ์ผํ•œ ๊ณณ์ผ ๋ฟ๋งŒ ์•„๋‹ˆ๋ผ
00:56
but it's the biggest,
19
56450
1098
๊ฐ€์žฅ ํฐ ๊ณณ์ด๊ธฐ๋„ ํ•ฉ๋‹ˆ๋‹ค.
00:57
and it serves to illustrate the numbers.
20
57548
1784
์ˆซ์ž๊ฐ€ ๋ฐ”๋กœ ์„ค๋ช…ํ•ด์ฃผ์ฃ .
00:59
Facebook has 1.2 billion users per month.
21
59332
3477
ํŽ˜์ด์Šค๋ถ์€ ํ•œ ๋‹ฌ๊ธฐ์ค€์œผ๋กœ 12์–ต๋ช…์ด ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค.
01:02
So half the Earth's Internet population
22
62809
1930
๊ทธ๋Ÿฌ๋‹ˆ๊นŒ ์ง€๊ตฌ์ƒ์˜ ์ธํ„ฐ๋„ท ์‚ฌ์šฉ์ž ์ค‘
01:04
is using Facebook.
23
64739
1653
์ ˆ๋ฐ˜์ด ํŽ˜์ด์Šค๋ถ์„ ์“ฐ๊ณ  ์žˆ์–ด์š”.
01:06
They are a site, along with others,
24
66392
1932
๋‹ค๋ฅธ ์‚ฌ์ดํŠธ๋“ค๊ณผ ๊ฐ™์ด
01:08
that has allowed people to create an online persona
25
68324
3219
๋ณ„๋‹ค๋ฅธ ๊ธฐ์ˆ ์ด ์—†์–ด๋„ ์˜จ๋ผ์ธ ์ƒ์˜ ์ธ๋ฌผ์„
01:11
with very little technical skill,
26
71543
1782
๋งŒ๋“ค์ˆ˜ ์žˆ๋„๋ก ํ•ด์ค๋‹ˆ๋‹ค.
01:13
and people responded by putting huge amounts
27
73325
2476
์‚ฌ๋žŒ๋“ค์€ ์—„์ฒญ๋‚œ ์–‘์˜ ๊ฐœ์ธ์ ์ธ ๋‚ด์šฉ์„
01:15
of personal data online.
28
75801
1983
์˜ฌ๋ฆฌ๋Š” ๊ฒƒ์œผ๋กœ ๊ทธ๊ฒƒ์„ ์‚ฌ์šฉํ•˜์ฃ .
01:17
So the result is that we have behavioral,
29
77784
2543
๊ทธ๊ฒƒ์œผ๋กœ ์ธํ•œ ๊ฒฐ๊ณผ๋Š” ์—ญ์‚ฌ์— ์„ ๋ก€์—†๋Š”
01:20
preference, demographic data
30
80327
1986
์ˆ˜ ์‹ญ์–ต ๋ช…์˜ ์‚ฌ๋žŒ๋“ค์— ๊ด€ํ•œ
01:22
for hundreds of millions of people,
31
82313
2101
ํ–‰๋™ํŒจํ„ด, ์„ ํ˜ธ๋„, ์ธ๊ตฌ์  ๋ฐ์ดํƒ€๋ฅผ
01:24
which is unprecedented in history.
32
84414
2026
์–ป๊ฒŒ ๋œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
01:26
And as a computer scientist, what this means is that
33
86440
2560
์ปดํ“จํ„ฐ ๊ณผํ•™์ž๋กœ์„œ, ์ด๊ฒƒ์˜ ์˜๋ฏธ๋Š”
01:29
I've been able to build models
34
89000
1664
์—ฌ๋Ÿฌ๋ถ„๋“ค์ด ๊ณต์œ ํ•˜๊ณ  ์žˆ๋Š” ์ค„๋„ ๋ชจ๋ฅด๋Š” ์ •๋ณด๋“ค๋กœ
01:30
that can predict all sorts of hidden attributes
35
90664
2322
๊ทธ ์•ˆ์— ์ˆจ๊ฒจ์ง„ ์—ฌ๋Ÿฌ๋ถ„์˜ ์˜๋„๋ฅผ
01:32
for all of you that you don't even know
36
92986
2284
์˜ˆ์ธกํ•ด ๋‚ผ ์ˆ˜ ์žˆ๋Š” ๋ชจํ˜•์„
01:35
you're sharing information about.
37
95270
2202
๋งŒ๋“ค์–ด ๋‚ผ ์ˆ˜ ์žˆ์—ˆ๋‹ค๋Š” ๊ฒ๋‹ˆ๋‹ค.
01:37
As scientists, we use that to help
38
97472
2382
๊ณผํ•™์ž๋กœ์„œ, ์šฐ๋ฆฌ๋Š” ์‚ฌ๋žŒ๋“ค์ด ์˜จ๋ผ์ธ์ƒ์—์„œ
01:39
the way people interact online,
39
99854
2114
์ƒํ˜ธ์ž‘์šฉํ•˜๋Š” ๋ฐฉ์‹์— ๋„์›€์„ ์ฃผ๋Š”๋ฐ ์”๋‹ˆ๋‹ค.
01:41
but there's less altruistic applications,
40
101968
2499
๊ทธ๋Ÿฌ๋‚˜ ๋„์›€์ด ๋˜๋Š” ์ชฝ์€ ์ข€ ์ ์Šต๋‹ˆ๋‹ค.
01:44
and there's a problem in that users don't really
41
104467
2381
์‚ฌ์šฉ์ž๋“ค์€ ์ด ๊ธฐ์ˆ ์ด๋‚˜ ๊ณผ์ •์ด ์–ด๋–ค์ง€
01:46
understand these techniques and how they work,
42
106848
2470
์‹ค์ œ๋กœ ์ดํ•ดํ•˜์ง€ ๋ชปํ•œ๋‹ค๋Š” ๊ฒŒ ๋ฌธ์ œ์ž…๋‹ˆ๋‹ค.
01:49
and even if they did, they don't have a lot of control over it.
43
109318
3128
์•ˆ๋‹ค๊ณ  ํ•ด๋„, ๊ทธ๊ฒƒ์„ ํ†ต์ œํ•  ์ˆ˜๊ฐ€ ์—†์ฃ .
01:52
So what I want to talk to you about today
44
112446
1490
๊ทธ๋ž˜์„œ ์˜ค๋Š˜ ๋ง์”€๋“œ๋ฆฌ๋ ค๊ณ  ํ•˜๋Š” ๊ฒƒ์€
01:53
is some of these things that we're able to do,
45
113936
2702
์ด์— ๋Œ€ํ•ด ์šฐ๋ฆฌ๊ฐ€ ํ•  ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์— ๋Œ€ํ•œ ๊ฒƒ๊ณผ
01:56
and then give us some ideas of how we might go forward
46
116638
2763
์–ด๋–ป๊ฒŒ ํ•˜๋ฉด ์‚ฌ์šฉ์ž๋“ค์—๊ฒŒ ๊ทธ ํ†ต์ œ๊ถŒ์„ ๋Œ๋ ค ์ฃผ๋„๋ก
01:59
to move some control back into the hands of users.
47
119401
2769
ํ•  ์ˆ˜ ์žˆ๋Š”๊ฐ€์— ๊ด€ํ•œ ์ƒ๊ฐ๋“ค์ž…๋‹ˆ๋‹ค.
02:02
So this is Target, the company.
48
122170
1586
์ด๊ฑด ํƒ€๊ฒŸ์ด๋ผ๋Š” ํšŒ์‚ฌ์ž…๋‹ˆ๋‹ค.
02:03
I didn't just put that logo
49
123756
1324
์—ฌ๊ธฐ ์ž„์‹ ํ•œ ์—ฌ์„ฑ์˜ ๋ฐฐ์— ์žˆ๋Š” ๋กœ๊ณ ๋Š”
02:05
on this poor, pregnant woman's belly.
50
125080
2170
์ œ๊ฐ€ ๋„ฃ์€ ๊ฒŒ ์•„๋‹™๋‹ˆ๋‹ค.
02:07
You may have seen this anecdote that was printed
51
127250
1840
์ด ์ด์•ผ๊ธฐ๋ฅผ ๋“ค์–ด ๋ณด์‹  ์ ์ด ์žˆ์œผ์‹ค ๊ฑฐ์—์š”.
02:09
in Forbes magazine where Target
52
129090
2061
ํฌ๋ธŒ์Šค ์žก์ง€์— ์‹ค๋ฆฐ ๊ฒƒ์ธ๋ฐ
02:11
sent a flyer to this 15-year-old girl
53
131151
2361
ํƒ€๊ฒŸ ํšŒ์‚ฌ์—์„œ 15์‚ด์งœ๋ฆฌ ์†Œ๋…€์—๊ฒŒ
02:13
with advertisements and coupons
54
133512
1710
์ –๋ณ‘, ๊ธฐ์ €๊ท€, ์š”๋žŒ๋ฅผ ์œ„ํ•œ ๊ด‘๊ณ ์™€ ์ฟ ํฐ ์ „๋‹จ์ง€๋ฅผ
02:15
for baby bottles and diapers and cribs
55
135222
2554
๊ทธ๋…€๊ฐ€ ์ž์‹ ์˜ ๋ถ€๋ชจ๋‹˜์—๊ฒŒ
02:17
two weeks before she told her parents
56
137776
1684
์ž„์‹ ํ–ˆ๋‹ค๋Š” ์‚ฌ์‹ค์„ ๋งํ•˜๊ธฐ 2์ฃผ ์ „์—
02:19
that she was pregnant.
57
139460
1864
๋ณด๋ƒˆ๋‹ค๋Š” ๊ฒ๋‹ˆ๋‹ค.
02:21
Yeah, the dad was really upset.
58
141324
2704
๋„ค, ๊ทธ๋ž˜์š”. ์•„๋ฒ„์ง€๊ฐ€ ๋ฌด์ฒ™ ํ™”๊ฐ€ ๋‚˜์…จ์ฃ .
02:24
He said, "How did Target figure out
59
144028
1716
๊ทธ๊ฐ€, "์–ด๋–ป๊ฒŒ ๊ณ ๋“ฑํ•™์ƒ์ด ์ž„์‹ ํ–ˆ๋‹ค๋Š” ๊ฒƒ์„
02:25
that this high school girl was pregnant
60
145744
1824
๋ถ€๋ชจ๊ฐ€ ์•Œ๊ธฐ๋„ ์ „์—
02:27
before she told her parents?"
61
147568
1960
ํƒ€๊ฒŸ์ด ๋จผ์ € ์•Œ ์ˆ˜๊ฐ€ ์žˆ์ฃ ?" ๋ผ๊ณ  ๋งํ–ˆ์–ด์š”.
02:29
It turns out that they have the purchase history
62
149528
2621
์•Œ๊ณ  ๋ณด๋‹ˆ ๊ทธ๋“ค์€ ์ˆ˜ ์‹ญ๋งŒ๋ช…์˜ ๊ณ ๊ฐ๋“ค์— ๋Œ€ํ•œ
02:32
for hundreds of thousands of customers
63
152149
2301
๊ตฌ๋งค์ด๋ ฅ์„ ๊ฐ€์ง€๊ณ  ์žˆ์—ˆ๊ณ 
02:34
and they compute what they call a pregnancy score,
64
154450
2730
๊ทธ๋“ค๋งŒ์˜ ์ž„์‹ ์ฒ™๋„๋ผ๊ณ  ํ•˜๋Š” ๊ฒƒ์„ ๊ณ„์‚ฐํ–ˆ์–ด์š”.
02:37
which is not just whether or not a woman's pregnant,
65
157180
2332
๋‹จ์ง€ ์ž„์‹ ์„ ํ–ˆ๋Š”์ง€ ์•ˆํ–ˆ๋Š”์ง€๋งŒ ๋ณด๋Š”๊ฒŒ ์•„๋‹ˆ๋ผ
02:39
but what her due date is.
66
159512
1730
์˜ˆ์ •์ผ์ด ์–ธ์ œ์ธ์ง€ ๊นŒ์ง€์š”.
02:41
And they compute that
67
161242
1304
๊ทธ๋“ค์ด ๊ณ„์‚ฐ ํ•ด๋‚ธ ๊ฒƒ์ด
02:42
not by looking at the obvious things,
68
162546
1768
์•„๊ธฐ์š”๋žŒ์ด๋‚˜ ์˜ท์„ ์‚ฌ๋Š” ๊ฒƒ๊ณผ ๊ฐ™์€
02:44
like, she's buying a crib or baby clothes,
69
164314
2512
๋ป”ํ•œ ๊ฒƒ์„ ํ†ตํ•ด์„œ ์•Œ์•„ ๋‚ธ ๊ฒƒ์ด ์•„๋‹ˆ๊ณ 
02:46
but things like, she bought more vitamins
70
166826
2943
ํ‰์†Œ์— ์ƒ€๋˜ ๊ฒƒ๋ณด๋‹ค
02:49
than she normally had,
71
169769
1717
๋” ๋งŽ์€ ๋น„ํƒ€๋ฏผ์„ ์‚ฐ๋‹ค๋“ ์ง€
02:51
or she bought a handbag
72
171486
1464
์•„๋‹ˆ๋ฉด ๊ธฐ์ €๊ท€๋ฅผ ๋„ฃ์„ ์ˆ˜ ์žˆ์„ ๋งŒํผ
02:52
that's big enough to hold diapers.
73
172950
1711
ํฐ ๊ฐ€๋ฐฉ์„ ์‚ฐ๋‹ค๋“ ์ง€ ํ•˜๋Š” ๊ฒƒ์œผ๋กœ ์•„๋Š” ๊ฒ๋‹ˆ๋‹ค.
02:54
And by themselves, those purchases don't seem
74
174661
1910
๊ทธ ๊ตฌ๋งค์ž์ฒด๋งŒ ๋ด์„œ๋Š”
02:56
like they might reveal a lot,
75
176571
2469
๋ณ„๋กœ ๋ณด์—ฌ์ฃผ๋Š” ๊ฒŒ ์—†์Šต๋‹ˆ๋‹ค.
02:59
but it's a pattern of behavior that,
76
179040
1978
๊ทธ๊ฒƒ์€ ํ–‰๋™ ํŒจํ„ด์œผ๋กœ
03:01
when you take it in the context of thousands of other people,
77
181018
3117
์ˆ˜ ์ฒœ๋ช…์˜ ๋‹ค๋ฅธ ์‚ฌ๋žŒ๋“ค ์†์—์„œ ๋ณด๋ฉด
03:04
starts to actually reveal some insights.
78
184135
2757
์ดํ•ดํ•  ๋งŒํ•œ ๊ฒƒ์ด ๋ณด์ž…๋‹ˆ๋‹ค.
03:06
So that's the kind of thing that we do
79
186892
1793
๊ทธ๊ฒŒ ์šฐ๋ฆฌ๊ฐ€ ํ•˜๋Š” ์ผ์ธ๋ฐ์š”
03:08
when we're predicting stuff about you on social media.
80
188685
2567
์†Œ์…œ๋ฏธ๋””์–ด์—์„œ ์—ฌ๋Ÿฌ๋ถ„์— ๋Œ€ํ•œ ๊ฒƒ์„ ์˜ˆ์ธกํ•˜๋Š” ๊ฒ๋‹ˆ๋‹ค.
03:11
We're looking for little patterns of behavior that,
81
191252
2796
์—ฌ๋Ÿฌ๋ถ„์ด ์ˆ˜ ๋ฐฑ๋งŒ๋ช…์˜ ์‚ฌ๋žŒ๋“ค ์†์—์„œ ๋ฌด์–ธ๊ฐ€ ๋ฐœ๊ฒฌํ•˜๋ฉด
03:14
when you detect them among millions of people,
82
194048
2682
์—ฌ๋Ÿฌ๊ฐ€์ง€ ๊ฒƒ๋“ค์„ ์•Œ ์ˆ˜ ์žˆ๊ฒŒ ํ•ด์ฃผ๋Š”
03:16
lets us find out all kinds of things.
83
196730
2706
์–ด๋–ค ํ–‰๋™ ํŒจํ„ด๋“ค์„ ์ฐพ์Šต๋‹ˆ๋‹ค.
03:19
So in my lab and with colleagues,
84
199436
1747
์—ฐ๊ตฌ์‹ค์—์„œ ๋™๋ฃŒ๋“ค๊ณผ
03:21
we've developed mechanisms where we can
85
201183
1777
๊ฝค ์ •ํ™•ํ•˜๊ฒŒ ์˜ˆ์ธกํ•  ์ˆ˜ ์žˆ๋Š”
03:22
quite accurately predict things
86
202960
1560
๋ฐฉ๋ฒ•์„ ๊ฐœ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค.
03:24
like your political preference,
87
204520
1725
์—ฌ๋Ÿฌ๋ถ„์˜ ์ •์น˜์  ์„ฑํ–ฅ,
03:26
your personality score, gender, sexual orientation,
88
206245
3752
์„ฑ๊ฒฉ ์ ์ˆ˜, ์„ฑ๋ณ„, ์„ฑ์  ์„ฑํ–ฅ,
03:29
religion, age, intelligence,
89
209997
2873
์ข…๊ต, ๋‚˜์ด, ์ง€๋Šฅ,
03:32
along with things like
90
212870
1394
๊ฑฐ๊ธฐ์—๋‹ค๊ฐ€
03:34
how much you trust the people you know
91
214264
1937
์‚ฌ๋žŒ๋“ค์„ ์–ผ๋งˆ๋‚˜ ์‹ ๋ขฐํ•˜๊ณ  ์žˆ๊ณ 
03:36
and how strong those relationships are.
92
216201
1804
๊ด€๊ณ„๊ฐ€ ์–ผ๋งˆ๋‚˜ ๋‘ํ„ฐ์šด์ง€๋„์š”.
03:38
We can do all of this really well.
93
218005
1785
์ €ํฌ๋Š” ์ด๋Ÿฐ ๊ฑธ ์ •๋ง ์ž˜ํ•ฉ๋‹ˆ๋‹ค.
03:39
And again, it doesn't come from what you might
94
219790
2197
๋‹ค์‹œ ๋ง์”€๋“œ๋ฆฌ์ง€๋งŒ, ์—ฌ๋Ÿฌ๋ถ„์ด ์ƒ๊ฐํ•˜์‹ค ๋งŒํ•œ ๋ป”ํ•œ ์ •๋ณด๋กœ
03:41
think of as obvious information.
95
221987
2102
์ด๋Ÿฐ ๊ฒƒ๋“ค์„ ์•Œ์•„๋‚ด๋Š” ๊ฒŒ ์•„๋‹™๋‹ˆ๋‹ค.
03:44
So my favorite example is from this study
96
224089
2281
์˜ฌํ•ด ๋ฏธ๊ตญ ๊ตญ๋ฆฝ๊ณผํ•™์› ํšŒ๋ณด์— ์‹ค๋ฆฐ
03:46
that was published this year
97
226370
1240
์ด ์—ฐ๊ตฌ์—์„œ
03:47
in the Proceedings of the National Academies.
98
227610
1795
์ œ๊ฐ€ ์ œ์ผ ์ข‹์•„ํ•˜๋Š” ๋ถ€๋ถ„์ž…๋‹ˆ๋‹ค.
03:49
If you Google this, you'll find it.
99
229405
1285
๊ตฌ๊ธ€ ๊ฒ€์ƒ‰์„ ํ•ด๋ณด์‹œ๋ฉด ๋‚˜์˜ต๋‹ˆ๋‹ค.
03:50
It's four pages, easy to read.
100
230690
1872
์ฝ๊ธฐ ์‰ฌ์šด 4ํŽ˜์ด์ง€ ๋ถ„๋Ÿ‰์ด์—์š”.
03:52
And they looked at just people's Facebook likes,
101
232562
3003
์—ฌ๋Ÿฌ๋ถ„๋“ค์ด ์ข‹์•„ํ•˜์‹œ๋Š” ๊ฒƒ๋“ค์— ํ‘œ์‹œํ•˜๋Š”
03:55
so just the things you like on Facebook,
102
235565
1920
ํŽ˜์ด์Šค๋ถ์˜ "์ข‹์•„์š”"๋ฅผ ๋ณด๊ณ ์„œ
03:57
and used that to predict all these attributes,
103
237485
2138
๋‹ค๋ฅธ ๊ฒƒ๋“ค๊ณผ ํ•จ๊ป˜
03:59
along with some other ones.
104
239623
1645
์–ด๋–ค ์š”์ธ์„ ์˜ˆ์ธกํ•˜๋Š”๋ฐ ์“ฐ๋Š” ๊ฒ๋‹ˆ๋‹ค.
04:01
And in their paper they listed the five likes
105
241268
2961
๋ณด๊ณ ์„œ์—๋Š” 5๊ฐœ์˜ "์ข‹์•„์š”" ๋ฆฌ์ŠคํŠธ๊ฐ€ ์žˆ๋Š”๋ฐ
04:04
that were most indicative of high intelligence.
106
244229
2787
๋†’์€ ์ง€๋Šฅ์„ ๋‚˜ํƒ€๋‚ด๋Š” ๊ฐ•๋ ฅํ•œ ์ง€ํ‘œ์˜€์Šต๋‹ˆ๋‹ค.
04:07
And among those was liking a page
107
247016
2324
๊ทธ ๋‹ค์„ฏ ๊ฐ€์ง€ ์ค‘ ํ•˜๋‚˜๊ฐ€
๊ผฌ๋ถ€๋ž‘ ๊ฐ์ž์— ๋Œ€ํ•œ ํŽ˜์ด์ง€๋ฅผ "์ข‹์•„์š”"ํ•˜๋Š” ๊ฒƒ์ด์—ˆ์–ด์š”. (์›ƒ์Œ)
04:09
for curly fries. (Laughter)
108
249340
1905
04:11
Curly fries are delicious,
109
251245
2093
๊ผฌ๋ถ€๋ž‘ ๊ฐ์ž๋Š” ๋ง›์žˆ์ง€๋งŒ,
04:13
but liking them does not necessarily mean
110
253338
2530
๊ทธ๊ฒƒ์„ ์ข‹๋‹ค๊ณ  ํ•˜๋Š” ๊ฒƒ์ด
04:15
that you're smarter than the average person.
111
255868
2080
๊ผญ ํ‰๊ท ๋ณด๋‹ค ๋˜‘๋˜‘ํ•˜๋‹ค๋Š” ๊ฒƒ์„ ์˜๋ฏธํ•˜์ง„ ์•Š์Šต๋‹ˆ๋‹ค.
04:17
So how is it that one of the strongest indicators
112
257948
3207
๊ทธ๋Ÿผ ์–ด๋–ป๊ฒŒ ์—ฌ๋Ÿฌ๋ถ„์˜ ์ง€๋Šฅ์— ๊ด€ํ•œ
04:21
of your intelligence
113
261155
1570
ํ™•์‹คํ•œ ์ง€ํ‘œ์ค‘ ํ•˜๋‚˜๊ฐ€
04:22
is liking this page
114
262725
1447
์ด ํŽ˜์ด์ง€๋ฅผ ์ข‹๋‹ค๊ณ  ํ•˜๋Š” ๊ฑธ๊นŒ์š”?
04:24
when the content is totally irrelevant
115
264172
2252
์ด ๋‚ด์šฉ์ด ์ง€๋Šฅ์ด๋ผ๋Š” ์š”์ธ๊ณผ๋Š”
04:26
to the attribute that's being predicted?
116
266424
2527
์•„๋ฌด๋Ÿฐ ๊ด€๋ จ์„ฑ์ด ์—†๋Š”๋ฐ๋„ ๋ง์ด์—์š”.
04:28
And it turns out that we have to look at
117
268951
1584
์™œ ๊ทธ๋Ÿฐ์ง€ ๋ณด๋ ค๋ฉด
04:30
a whole bunch of underlying theories
118
270535
1618
๊ทธ ์ด๋ฉด์— ์žˆ๋Š” ์—ฌ๋Ÿฌ ๊ฐ€์ง€ ์ด๋ก ๋“ค์„
04:32
to see why we're able to do this.
119
272153
2569
์•Œ์•„ ๋ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.
04:34
One of them is a sociological theory called homophily,
120
274722
2913
๊ทธ ์ค‘์˜ ํ•˜๋‚˜๊ฐ€ ํ˜ธ๋ชจํ•„๋ฆฌ๋ผ๊ณ  ํ•˜๋Š” ์‚ฌํšŒ์ด๋ก ์ž…๋‹ˆ๋‹ค.
04:37
which basically says people are friends with people like them.
121
277635
3092
๋น„์Šทํ•œ ์‚ฌ๋žŒ๋“ค๋ผ๋ฆฌ ์นœ๊ตฌ๊ฐ€ ๋œ๋‹ค๋Š” ์ด๋ก ์ž…๋‹ˆ๋‹ค.
04:40
So if you're smart, you tend to be friends with smart people,
122
280727
2014
๋˜‘๋˜‘ํ•œ ์‚ฌ๋žŒ์€ ๋˜‘๋˜‘ํ•œ ์‚ฌ๋žŒ๋ผ๋ฆฌ
04:42
and if you're young, you tend to be friends with young people,
123
282741
2630
์ Š์€ ์‚ฌ๋žŒ์€ ์ Š์€ ์‚ฌ๋žŒ๋ผ๋ฆฌ ์นœ๊ตฌ๊ฐ€ ๋˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ๊ณ 
04:45
and this is well established
124
285371
1627
์ด๊ฒƒ์€ ์ˆ˜ ๋ฐฑ๋…„ ๋™์•ˆ
04:46
for hundreds of years.
125
286998
1745
๊ทธ๋ ‡๊ฒŒ ๊ตณ์–ด์กŒ์Šต๋‹ˆ๋‹ค.
04:48
We also know a lot
126
288743
1232
์šฐ๋ฆฌ๋Š” ์ž˜ ์••๋‹ˆ๋‹ค.
04:49
about how information spreads through networks.
127
289975
2550
์ •๋ณด๊ฐ€ ์–ด๋–ป๊ฒŒ ์ธ๋งฅ์„ ํ†ตํ•ด ํผ์ง€๋Š”์ง€์š”.
04:52
It turns out things like viral videos
128
292525
1754
ํžˆํŠธ ๋™์˜์ƒ์ด๋‚˜
04:54
or Facebook likes or other information
129
294279
2406
ํŽ˜์ด์Šค๋ถ์˜ "์ข‹์•„์š”", ๋˜๋Š” ๋‹ค๋ฅธ ์ •๋ณด๋“ค์ด
04:56
spreads in exactly the same way
130
296685
1888
์‚ฌํšŒ์  ๋ง์„ ํ†ตํ•ด์„œ
04:58
that diseases spread through social networks.
131
298573
2454
์งˆ๋ณ‘์ด ํผ์ง€๋Š” ๋ฐฉ๋ฒ•๊ณผ ๋˜‘๊ฐ™๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
05:01
So this is something we've studied for a long time.
132
301027
1791
์ด๊ฒƒ์ด ์ €ํฌ๊ฐ€ ์˜ค๋žซ๋™์•ˆ ์—ฐ๊ตฌํ•œ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
05:02
We have good models of it.
133
302818
1576
์ž˜ ์„ค๋ช…ํ•ด ์ค„ ์ˆ˜ ์žˆ๋Š” ๋ชจ๋ธ์ด ์žˆ์ฃ .
05:04
And so you can put those things together
134
304394
2157
๊ทธ๋Ÿฌ๋‹ˆ๊นŒ ์ด ๋ชจ๋“  ์ •๋ณด๋“ค์„ ๋ชจ์•„์„œ
05:06
and start seeing why things like this happen.
135
306551
3088
์™œ ์ด๋Ÿฐ ์ผ์ด ์ƒ๊ธฐ๋Š”์ง€ ์•Œ ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์ด์ฃ .
05:09
So if I were to give you a hypothesis,
136
309639
1814
๊ฐ€๋Šฅํ•œ ๊ฐ€์„ค์„ ๋ง์”€๋“œ๋ฆฌ์ž๋ฉด,
05:11
it would be that a smart guy started this page,
137
311453
3227
์–ด๋–ค ๋˜‘๋˜‘ํ•œ ์‚ฌ๋žŒ์ด ์ด ํŽ˜์ด์ง€๋ฅผ ์‹œ์ž‘ํ–ˆ๊ฑฐ๋‚˜
05:14
or maybe one of the first people who liked it
138
314680
1939
"์ข‹์•„์š”" ๋ผ๊ณ  ์‹œ์ž‘ํ•œ ์‚ฌ๋žŒ์ด
05:16
would have scored high on that test.
139
316619
1736
์ง€๋Šฅํ…Œ์ŠคํŠธ์—์„œ ๋†’์€ ์ ์ˆ˜๋ฅผ ๋ฐ›์•˜๋˜ ๊ฑฐ์ฃ .
05:18
And they liked it, and their friends saw it,
140
318355
2288
"์ข‹์•„์š”" ํ•œ ํ›„์— ๊ทธ ์นœ๊ตฌ๋“ค์ด ๋ดค๊ณ ,
05:20
and by homophily, we know that he probably had smart friends,
141
320643
3122
ํ˜ธ๋ชจํ•„๋ฆฌ ์ด๋ก ์ƒ, ๊ทธ์—๊ฒ ๋˜‘๋˜‘ํ•œ ์นœ๊ตฌ๋“ค์ด ์žˆ์„๊ฒ๋‹ˆ๋‹ค.
05:23
and so it spread to them, and some of them liked it,
142
323765
3056
๊ทธ๊ฒŒ ๊ทธ๋“ค์—๊ฒŒ ํผ์ง„ ๊ฒƒ์ด๊ณ , ๋ช‡๋ช‡์€ "์ข‹์•„์š”"๋ฅผ ๋ˆŒ๋ €์„ ๊ฑฐ๊ตฌ์š”.
05:26
and they had smart friends,
143
326821
1189
๊ทธ๋“ค์—๊ฒŒ ์žˆ๋Š” ๋˜‘๋˜‘ํ•œ ์นœ๊ตฌ๋“ค์—๊ฒŒ
05:28
and so it spread to them,
144
328010
807
05:28
and so it propagated through the network
145
328817
1973
๋‹ค์‹œ ํผ์ง€๋Š” ๊ฒ๋‹ˆ๋‹ค.
๊ทธ๋ž˜์„œ ๊ทธ ๊ด€๊ณ„๋ฅผ ํ†ตํ•ด ์„ ์ „์ด ๋˜๋Š” ๊ฒƒ์ด์ฃ .
05:30
to a host of smart people,
146
330790
2569
๋˜‘๋˜‘ํ•œ ์‚ฌ๋žŒ๋“ค ์ง‘๋‹จ์—๊ฒŒ์š”.
05:33
so that by the end, the action
147
333359
2056
๊ทธ๋ž˜์„œ ๋งˆ์นจ๋‚ด ๊ผฌ๋ถ€๋ž‘ ๊ฐ์ž๋ฅผ
05:35
of liking the curly fries page
148
335415
2544
"์ข‹์•„์š”" ํ•œ ํ–‰๋™์ด
05:37
is indicative of high intelligence,
149
337959
1615
๋†’์€ ์ง€๋Šฅ์˜ ํ‘œ์‹œ๊ฐ€ ๋˜๋Š” ๊ฒ๋‹ˆ๋‹ค.
05:39
not because of the content,
150
339574
1803
๊ทธ๊ฒƒ์ด ๋‚ด์šฉ ๋•Œ๋ฌธ์ด ์•„๋‹ˆ๋ผ
05:41
but because the actual action of liking
151
341377
2522
"์ข‹์•„์š”"๋ฅผ ๋ˆ„๋ฅด๋Š”
05:43
reflects back the common attributes
152
343899
1900
ํ–‰๋™์„ ํ•œ ์‚ฌ๋žŒ๋“ค์˜
05:45
of other people who have done it.
153
345799
2468
๊ณตํ†ต์ ์ธ ์š”์ธ ๋•Œ๋ฌธ์ธ ๊ฒƒ์ด์ฃ .
05:48
So this is pretty complicated stuff, right?
154
348267
2897
๊ฝค ๋ณต์žกํ•˜๊ฒŒ ๋“ค๋ฆฌ์‹œ์ฃ ?
05:51
It's a hard thing to sit down and explain
155
351164
2199
์ด๊ฑธ ๋ณดํ†ต ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ์„ค๋ช…ํ•˜๊ธฐ๊ฐ€
05:53
to an average user, and even if you do,
156
353363
2848
์–ด๋ ต์Šต๋‹ˆ๋‹ค. ํ•œ๋‹ค๊ณ  ํ•ด๋„
05:56
what can the average user do about it?
157
356211
2188
๊ทธ๋“ค์ด ์ด๊ฑธ ๊ฐ–๊ณ  ๋ญ˜ ํ• ๊นŒ์š”?
05:58
How do you know that you've liked something
158
358399
2048
์—ฌ๋Ÿฌ๋ถ„์ด ์–ด๋–ค ๊ฒƒ์„ ์ข‹๋‹ค๊ณ  ํ•˜๋Š” ๊ฒƒ์ด
06:00
that indicates a trait for you
159
360447
1492
์ข‹์•„ํ•˜๋Š” ๋‚ด์šฉ๊ณผ๋Š” ์ „ํ˜€ ๊ด€๋ จ์ด ์—†๋Š”
06:01
that's totally irrelevant to the content of what you've liked?
160
361939
3545
์–ด๋–ค ํŠน์ง•์„ ๋ณด์—ฌ์ฃผ๋Š” ๊ฒƒ์„ ์–ด๋–ป๊ฒŒ ์•Œ๊ฒ ์–ด์š”?
06:05
There's a lot of power that users don't have
161
365484
2546
์ด๋Ÿฐ ๋ฐ์ดํƒ€๊ฐ€ ์‚ฌ์šฉ๋˜๋Š” ๋ฐฉ๋ฒ•์— ๋Œ€ํ•ด
06:08
to control how this data is used.
162
368030
2230
์‚ฌ์šฉ์ž๋“ค์€ ํ†ต์ œ๋ ฅ์ด ๊ฑฐ์˜ ์—†์Šต๋‹ˆ๋‹ค.
06:10
And I see that as a real problem going forward.
163
370260
3112
์ด๊ฒƒ์ด ์•ž์œผ๋กœ ๋‚˜์•„๊ฐ€๋Š”๋ฐ ํฐ ๋ฌธ์ œ๋ผ๊ณ  ๋ด…๋‹ˆ๋‹ค.
06:13
So I think there's a couple paths
164
373372
1977
์šฐ๋ฆฌ๊ฐ€ ์‚ดํŽด๋ด์•ผ ํ• 
06:15
that we want to look at
165
375349
1001
๋ช‡ ๊ฐ€์ง€ ๊ฒฝ์šฐ๊ฐ€ ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
06:16
if we want to give users some control
166
376350
1910
๋ฐ์ดํƒ€๊ฐ€ ์‚ฌ์šฉ๋˜๋Š” ๊ฒƒ์— ๋Œ€ํ•œ ํ†ต์ œ๊ถŒ์„
06:18
over how this data is used,
167
378260
1740
์‚ฌ์šฉ์ž์—๊ฒŒ ์ฃผ๊ณ  ์‹ถ์€์ง€์— ๋Œ€ํ•ด์„œ์š”.
06:20
because it's not always going to be used
168
380000
1940
์™œ๋ƒํ•˜๋ฉด ์ด๊ฒƒ์ด ์–ธ์ œ๋‚˜ ์‚ฌ์šฉ์ž์—๊ฒŒ
06:21
for their benefit.
169
381940
1381
์ข‹์€ ์ชฝ์œผ๋กœ๋งŒ ์‚ฌ์šฉ๋˜๋Š” ๊ฒƒ์€ ์•„๋‹ˆ๋‹ˆ๊นŒ์š”.
06:23
An example I often give is that,
170
383321
1422
์ œ๊ฐ€ ์ž์ฃผ ๋“œ๋Š” ์˜ˆ๊ฐ€ ์žˆ๋Š”๋ฐ์š”,
06:24
if I ever get bored being a professor,
171
384743
1646
์ œ๊ฐ€ ๋งŒ์•ฝ ๊ต์ˆ˜์ง์„ ํ•˜๋Š” ๊ฒƒ์ด ์‹ซ์–ด์ง„๋‹ค๋ฉด
06:26
I'm going to go start a company
172
386389
1653
์ด๋Ÿฐ ๋ชจ๋“  ์š”์ธ๋“ค์„ ์˜ˆ์ธกํ•˜๋Š”
06:28
that predicts all of these attributes
173
388042
1454
ํšŒ์‚ฌ๋ฅผ ์ฐจ๋ฆด ์ˆ˜ ์žˆ์„ ๊ฒ๋‹ˆ๋‹ค.
06:29
and things like how well you work in teams
174
389496
1602
์—ฌ๋Ÿฌ๋ถ„์ด ํŒ€์—์„œ ์ผ์„ ์ž˜ ํ•  ์ง€,
06:31
and if you're a drug user, if you're an alcoholic.
175
391098
2671
๋งˆ์•ฝ๋ณต์šฉ์ž์ธ์ง€, ์•Œ์ฝœ ์ค‘๋…์ž์ธ์ง€ ์•Œ๋ ค์ฃผ๋Š” ํšŒ์‚ฌ์š”.
06:33
We know how to predict all that.
176
393769
1440
์˜ˆ์ธกํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์•Œ๊ณ  ์žˆ๊ฑฐ๋“ ์š”.
06:35
And I'm going to sell reports
177
395209
1761
์ €๋Š” ๊ทธ ๋ณด๊ณ ์„œ๋ฅผ ํŒ๋งคํ•˜๋Š” ๊ฑฐ์ฃ .
06:36
to H.R. companies and big businesses
178
396970
2100
์—ฌ๋Ÿฌ๋ถ„์„ ๊ณ ์šฉํ•˜๋ ค๋Š” ๋Œ€๊ธฐ์—…์ด๋‚˜
06:39
that want to hire you.
179
399070
2273
์ธ๋ ฅํšŒ์‚ฌ์— ๋ง์ด์—์š”.
06:41
We totally can do that now.
180
401343
1177
์ง€๊ธˆ ๊ทธ ์ผ๋“ค์„ ํ•  ์ˆ˜๊ฐ€ ์žˆ์–ด์š”.
06:42
I could start that business tomorrow,
181
402520
1788
๋‚ด์ผ ๊ทธ ์‚ฌ์—…์„ ์‹œ์ž‘ํ•  ์ˆ˜๋„ ์žˆ์ฃ .
06:44
and you would have absolutely no control
182
404308
2052
์ œ๊ฐ€ ๊ทธ๋Ÿฐ ๋ฐ์ดํƒ€๋ฅผ ์“ฐ๋Š” ๊ฒƒ์— ๋Œ€ํ•ด
06:46
over me using your data like that.
183
406360
2138
์—ฌ๋Ÿฌ๋ถ„์€ ํ†ต์ œ๊ถŒ์ด ์•„์˜ˆ ์—†์Šต๋‹ˆ๋‹ค.
06:48
That seems to me to be a problem.
184
408498
2292
๊ทธ๊ฒŒ ์ €์—๊ฒŒ๋Š” ๋ฌธ์ œ๋กœ ๋Š๊ปด์ง‘๋‹ˆ๋‹ค.
06:50
So one of the paths we can go down
185
410790
1910
์šฐ๋ฆฌ๊ฐ€ ํƒํ•  ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ• ์ค‘ ํ•˜๋‚˜๋Š”
06:52
is the policy and law path.
186
412700
2032
์ •์ฑ…๊ณผ ๋ฒ•๋ฅ ์ž…๋‹ˆ๋‹ค.
06:54
And in some respects, I think that that would be most effective,
187
414732
3046
์–ด๋–ค ๋ฉด์—์„œ, ์ œ ์ƒ๊ฐ์—” ๊ฐ€์žฅ ํšจ๊ณผ์ ์ผ ๊ฒƒ ๊ฐ™์Šต๋‹ˆ๋‹ค.
06:57
but the problem is we'd actually have to do it.
188
417778
2756
๋ฌธ์ œ๋Š” ์‹ค์ œ๋กœ ๊ทธ๋ ‡๊ฒŒ ํ•ด์•ผ ํ•œ๋‹ค๋Š” ๊ฒ๋‹ˆ๋‹ค.
07:00
Observing our political process in action
189
420534
2780
์ •์น˜์ ์ธ ๊ณผ์ •์ด ๋Œ์•„๊ฐ€๋Š” ๊ฒƒ์„ ๋ณด๋ฉด
07:03
makes me think it's highly unlikely
190
423314
2379
๊ทธ๋ ‡๊ฒŒ ๋  ๊ฒƒ ๊ฐ™์ง€๊ฐ€ ์•Š์Šต๋‹ˆ๋‹ค.
07:05
that we're going to get a bunch of representatives
191
425693
1597
์ˆ˜ ๋งŽ์€ ์˜์›๋‹˜๋“ค์„ ๋ฐ๋ฆฌ๊ณ 
07:07
to sit down, learn about this,
192
427290
1986
์•‰ํ˜€ ๋†“๊ณ , ์ƒํ™ฉ์„ ์•Œ๋ ค ์ค€ ๋‹ค์Œ์—
07:09
and then enact sweeping changes
193
429276
2106
๋ฏธ๊ตญ์˜ ์ง€์ ์žฌ์‚ฐ๊ถŒ๋ฒ•์—
07:11
to intellectual property law in the U.S.
194
431382
2157
์—„์ฒญ๋‚œ ๋ณ€ํ™”๋ฅผ ์ค„ ๋ฒ•์„ ์ œ์ •ํ•ด์„œ
07:13
so users control their data.
195
433539
2461
์‚ฌ์šฉ์ž๋“ค์ด ๋ฐ์ดํƒ€๋ฅผ ํ†ต์ œํ•˜๊ฒŒ ํ•˜๋Š” ๊ฒƒ์ด์ฃ .
07:16
We could go the policy route,
196
436000
1304
์†Œ์…œ๋ฏธ๋””์–ด ํšŒ์‚ฌ๋“ค์ด ๋งํ•˜๋Š”
07:17
where social media companies say,
197
437304
1479
์ •์ฑ…์ ์ธ ๋ฐฉ๋ฒ•์„ ์„ ํƒํ•  ์ˆ˜ ๋„ ์žˆ์–ด์š”.
07:18
you know what? You own your data.
198
438783
1402
์•„์„ธ์š”? ๋ฐ์ดํƒ€๋Š” ์—ฌ๋Ÿฌ๋ถ„ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
07:20
You have total control over how it's used.
199
440185
2489
๋ฐ์ดํƒ€ ์‚ฌ์šฉ์— ๋Œ€ํ•œ ํ†ต์ œ๊ถŒ์ด ์—ฌ๋Ÿฌ๋ถ„์—๊ฒŒ ์žˆ์–ด์š”.
07:22
The problem is that the revenue models
200
442674
1848
๋ฌธ์ œ๋Š” ๋Œ€๋ถ€๋ถ„์˜ ์†Œ์…œ๋ฏธ๋””์–ด ๊ธฐ์—…๋“ค์ด
07:24
for most social media companies
201
444522
1724
์–ด๋–ค ๋ฐฉ์‹์œผ๋กœ๋“  ์‚ฌ์šฉ์ž๋“ค์˜ ๋ฐ์ดํƒ€๋ฅผ
07:26
rely on sharing or exploiting users' data in some way.
202
446246
4031
๊ณต์œ ํ•˜๊ณ  ์ด์šฉํ•˜๋Š” ๊ฒƒ์œผ๋กœ ๋งค์ถœ์„ ์–ป๊ณ  ์žˆ๋‹ค๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
07:30
It's sometimes said of Facebook that the users
203
450277
1833
ํŽ˜์ด์Šค๋ถ์—๋Š” ์‚ฌ์šฉ์ž๊ฐ€ ์žˆ๋Š”๊ฒŒ ์•„๋‹ˆ๋ผ
07:32
aren't the customer, they're the product.
204
452110
2528
์ƒํ’ˆ๋“ค์ด ์žˆ๋Š” ๊ฑฐ๋ผ๊ณ  ์–˜๊ธฐํ•˜๊ธฐ๋„ ํ•ฉ๋‹ˆ๋‹ค.
07:34
And so how do you get a company
205
454638
2714
์–ด๋–ป๊ฒŒ ํ•ด์•ผ ํšŒ์‚ฌ๋“ค์ด
07:37
to cede control of their main asset
206
457352
2558
์ฃผ์š” ์ž์‚ฐ์˜ ํ†ต์ œ๊ถŒ์„
07:39
back to the users?
207
459910
1249
์‚ฌ์šฉ์ž๋“ค์—๊ฒŒ ๋Œ๋ ค ์ค„๊นŒ์š”?
07:41
It's possible, but I don't think it's something
208
461159
1701
๊ฐ€๋Šฅํ•œ ์ผ์ด์ง€๋งŒ, ์ œ ์ƒ๊ฐ์—”
07:42
that we're going to see change quickly.
209
462860
2320
๋‹น์žฅ ๋ฒŒ์–ด์ง€์ง„ ์•Š์„ ๊ฒ๋‹ˆ๋‹ค.
07:45
So I think the other path
210
465180
1500
๊ทธ๋ž˜์„œ ๋‹ค๋ฅธ ๋ฐฉ๋ฒ•์œผ๋กœ
07:46
that we can go down that's going to be more effective
211
466680
2288
ํ›จ์”ฌ ํšจ๊ณผ์ ์œผ๋กœ ๋Œ€์‘ํ•  ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์ด
07:48
is one of more science.
212
468968
1508
๊ณผํ•™์ ์ธ ๋ฐฉ๋ฒ•์„ ์“ฐ๋Š” ๊ฒ๋‹ˆ๋‹ค.
07:50
It's doing science that allowed us to develop
213
470476
2510
์• ์ดˆ์— ๊ฐœ์ธ์ ์ธ ๋ฐ์ดํƒ€๋ฅผ
07:52
all these mechanisms for computing
214
472986
1750
๊ณ„์‚ฐํ•ด๋‚ด๋Š” ๋ฉ”์นด๋‹ˆ์ฆ˜์„
07:54
this personal data in the first place.
215
474736
2052
๋งŒ๋“ค์–ด ๋‚ด๋Š” ๊ฒƒ์ด ๊ณผํ•™์ž…๋‹ˆ๋‹ค.
07:56
And it's actually very similar research
216
476788
2106
์‚ฌ์‹ค ์ด๊ฑด ์šฐ๋ฆฌ๊ฐ€ ํ•ด์•ผ ํ•˜๋Š” ์—ฐ๊ตฌ์™€
07:58
that we'd have to do
217
478894
1438
์•„์ฃผ ๋น„์Šทํ•ฉ๋‹ˆ๋‹ค.
08:00
if we want to develop mechanisms
218
480332
2386
์‚ฌ์šฉ์ž์—๊ฒŒ ์ด๋Ÿฐ ๊ฒฝ๊ณ ๋ฅผ ํ•ด์ฃผ๋Š”
08:02
that can say to a user,
219
482718
1421
ํ”„๋กœ๊ทธ๋žจ์„ ๋งŒ๋“ค๋ ค๋ฉด์š”.
08:04
"Here's the risk of that action you just took."
220
484139
2229
"๋ฐฉ๊ธˆ ํ•˜์‹  ์ผ์—๋Š” ์œ„ํ—˜์ด ๋”ฐ๋ฆ…๋‹ˆ๋‹ค."
08:06
By liking that Facebook page,
221
486368
2080
ํŽ˜์ด์Šค๋ถ์˜ "์ข‹์•„์š”"๋ฅผ ๋ˆ„๋ฅด์‹œ๊ณ 
08:08
or by sharing this piece of personal information,
222
488448
2535
๊ฐœ์ธ ์ •๋ณด๋ฅผ ๊ณต์œ ํ•˜์‹ฌ์œผ๋กœ์จ
08:10
you've now improved my ability
223
490983
1502
์ œ๊ฐ€ ์˜ˆ์ธกํ•˜๋Š” ๋Šฅ๋ ฅ์„ ํ‚ค์›Œ ์ฃผ์‹œ๋Š” ๊ฑฐ์—์š”.
08:12
to predict whether or not you're using drugs
224
492485
2086
์—ฌ๋Ÿฌ๋ถ„์ด ๋งˆ์•ฝ์„ ํ•˜์‹œ๋Š”์ง€
08:14
or whether or not you get along well in the workplace.
225
494571
2862
ํšŒ์‚ฌ์—์„œ ์ž˜ ์ง€๋‚ด์‹œ๋Š”์ง€์š”.
08:17
And that, I think, can affect whether or not
226
497433
1848
์ œ ์ƒ๊ฐ์—”, ๊ทธ๊ฒŒ ์˜ํ–ฅ์„ ์ค„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
08:19
people want to share something,
227
499281
1510
์–ด๋–ค ๊ฒƒ์„ ๊ณต์œ ํ•  ์ง€
08:20
keep it private, or just keep it offline altogether.
228
500791
3239
๋น„๋ฐ€๋กœ ํ• ์ง€, ์•„์˜ˆ ์˜คํ”„๋ผ์ธ์œผ๋กœ๋งŒ ๊ฐ–๊ณ  ์žˆ์„์ง€์š”.
08:24
We can also look at things like
229
504030
1563
์ด๋Ÿฐ ๊ฒƒ๋„ ์ƒ๊ฐ ํ•ด ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
08:25
allowing people to encrypt data that they upload,
230
505593
2728
์—…๋กœ๋“œํ•˜๋Š” ๋ฐ์ดํƒ€๋ฅผ ์•”ํ˜ธํ™”ํ•ด์„œ
08:28
so it's kind of invisible and worthless
231
508321
1855
ํŽ˜์ด์Šค๋ถ ๊ฐ™์€ ์‚ฌ์ดํŠธ๋‚˜
08:30
to sites like Facebook
232
510176
1431
์ œ3์ž๊ฐ€ ์•„์˜ˆ ์ถ”์ ๋„ ๋ชปํ•˜๊ณ 
08:31
or third party services that access it,
233
511607
2629
์“ธ ์ˆ˜๋„ ์—†๊ฒŒ ๋ง์ž…๋‹ˆ๋‹ค.
08:34
but that select users who the person who posted it
234
514236
3247
ํ•˜์ง€๋งŒ ์—…๋กœ๋“œํ•œ ์‚ฌ์šฉ์ž๊ฐ€ ์ง€์ •ํ•œ ์‚ฌ๋žŒ๋งŒ
08:37
want to see it have access to see it.
235
517483
2670
๊ทธ๊ฒƒ์„ ๋ณด๊ฒŒ ํ•ด ์ฃผ๋Š” ๊ฒƒ์ด์ฃ .
08:40
This is all super exciting research
236
520153
2166
์ง€์ ์ธ ๊ด€์ ์—์„œ ๋ณด๋ฉด
์ด๊ฒƒ์€ ๋Œ€๋‹จํžˆ ์„ค๋ ˆ๋Š” ์—ฐ๊ตฌ์ž…๋‹ˆ๋‹ค.
08:42
from an intellectual perspective,
237
522319
1620
08:43
and so scientists are going to be willing to do it.
238
523939
1859
๊ณผํ•™์ž๋“ค์€ ๊ธฐ๊บผ์ด ํ•˜๋ ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค.
08:45
So that gives us an advantage over the law side.
239
525798
3610
๊ทธ๋ž˜์„œ ๋ฒ•๋ฅ ์ ์ธ ์ธก๋ฉด์—์„œ ์šฐ๋ฆฌ์—๊ฒŒ ์œ ๋ฆฌํ•˜๋„๋ก ๋ง์ž…๋‹ˆ๋‹ค.
08:49
One of the problems that people bring up
240
529408
1725
์ œ๊ฐ€ ์ด๋Ÿฐ ๋ง์„ ํ• ๋•Œ
08:51
when I talk about this is, they say,
241
531133
1595
์‚ฌ๋žŒ๋“ค์ด ์ œ๊ธฐํ•˜๋Š” ๋ฌธ์ œ๊ฐ€ ์žˆ๋Š”๋ฐ์š”,
08:52
you know, if people start keeping all this data private,
242
532728
2646
"๋ชจ๋‘๊ฐ€ ๋ฐ์ดํƒ€๋ฅผ ๋น„๋ฐ€๋กœ ํ•œ๋‹ค๋ฉด
08:55
all those methods that you've been developing
243
535374
2113
๋‹น์‹ ์ด ์˜ˆ์ธกํ•˜๋ ค๊ณ  ๊ฐœ๋ฐœํ•˜๋Š” ๋ฐฉ๋ฒ•๋“ค์ด
08:57
to predict their traits are going to fail.
244
537487
2653
๋‹ค ์‹คํŒจํ•  ๊ฒ๋‹ˆ๋‹ค." ๋ผ๊ณ  ๋ง์ด์—์š”.
09:00
And I say, absolutely, and for me, that's success,
245
540140
3520
๊ทธ๋Ÿผ ์ €๋Š”, "๋งž์•„์š”, ๊ทธ๋Ÿผ ์„ฑ๊ณตํ•œ ๊ฑฐ์—์š”" ๋ผ๊ณ ์š”.
09:03
because as a scientist,
246
543660
1786
์ €๋Š” ๊ณผํ•™์ž๋กœ์„œ
09:05
my goal is not to infer information about users,
247
545446
3688
์‚ฌ์šฉ์ž๋“ค์— ๋Œ€ํ•œ ์ •๋ณด๋ฅผ ์ถ”์ธกํ•˜๋Š”๊ฒŒ ์•„๋‹ˆ๋ผ
09:09
it's to improve the way people interact online.
248
549134
2767
์˜จ๋ผ์ธ์—์„œ ๋” ์ข‹์€ ๋ฐฉ๋ฒ•์œผ๋กœ ๊ต๋ฅ˜ํ•˜๋„๋ก ํ•˜๋Š”๊ฒŒ ๋ชฉํ‘œ์ด๋‹ˆ๊นŒ์š”.
09:11
And sometimes that involves inferring things about them,
249
551901
3218
์–ด๋–ค ๋•Œ๋Š” ๊ทธ๋“ค์— ๋Œ€ํ•œ ์ •๋ณด๋ฅผ ์ถ”์ธกํ•˜๋Š” ์ผ์ด ์ƒ๊น๋‹ˆ๋‹ค.
09:15
but if users don't want me to use that data,
250
555119
3022
ํ•˜์ง€๋งŒ ์‚ฌ์šฉ์ž๊ฐ€ ์›์น˜ ์•Š์œผ๋ฉด,
09:18
I think they should have the right to do that.
251
558141
2038
์ œ ์ƒ๊ฐ์—” ๊ทธ๋“ค์—๊ฒŒ ๊ทธ๋Ÿด ๊ถŒ๋ฆฌ๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.
09:20
I want users to be informed and consenting
252
560179
2651
์šฐ๋ฆฌ๊ฐ€ ๊ฐœ๋ฐœํ•œ ๋ฐฉ๋ฒ•๋“ค์„
09:22
users of the tools that we develop.
253
562830
2112
์‚ฌ์šฉ์ž๋“ค์—๊ฒŒ ์•Œ๋ ค์ฃผ๊ณ  ๋™์˜๋ฅผ ๋ฐ›๊ณ  ์‹ถ์Šต๋‹ˆ๋‹ค.
09:24
And so I think encouraging this kind of science
254
564942
2952
์ด๋Ÿฌํ•œ ๊ณผํ•™์„ ์žฅ๋ คํ•˜๊ณ 
09:27
and supporting researchers
255
567894
1346
์‚ฌ์šฉ์ž๋“ค์—๊ฒŒ
09:29
who want to cede some of that control back to users
256
569240
3023
ํ†ต์ œ๊ถŒ์„ ๋Œ๋ ค์ฃผ๊ณ  ์†Œ์…œ๋ฏธ๋””์–ด ํšŒ์‚ฌ๋“ค๋กœ๋ถ€ํ„ฐ
09:32
and away from the social media companies
257
572263
2311
๋ฉ€์–ด์ง€๋„๋ก ํ•˜๋Š” ์—ฐ๊ตฌ์ž๋“ค์„ ์ง€์›ํ•˜๋Š” ๊ฒƒ์ด
09:34
means that going forward, as these tools evolve
258
574574
2671
์ด ๊ฐœ๋ฐœ๋„๊ตฌ๋“ค์ด ๋ณ€ํ™”ํ•˜๊ณ  ๋ฐœ์ „ํ•˜๋ฉด์„œ
09:37
and advance,
259
577245
1476
์•ž์œผ๋กœ ๋‚˜์•„๊ฐ€๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
09:38
means that we're going to have an educated
260
578721
1414
๋˜ํ•œ ์ง€์‹์ด ์žˆ๊ณ 
09:40
and empowered user base,
261
580135
1694
๊ถŒ๋ฆฌ๊ฐ€ ์žˆ๋Š” ์‚ฌ์šฉ์ž๋“ค์ด ์ƒ๊ธด๋‹ค๋Š” ๊ฒƒ์ด์ฃ .
09:41
and I think all of us can agree
262
581829
1100
์ €๋Š” ์•ž์œผ๋กœ ๋‚˜์•„๊ฐ€๊ธฐ ์œ„ํ•ด
09:42
that that's a pretty ideal way to go forward.
263
582929
2564
๋ชจ๋‘๊ฐ€ ๋™์˜ํ•  ๋งŒํ•œ ๊ฒƒ์ด๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
09:45
Thank you.
264
585493
2184
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.
09:47
(Applause)
265
587677
3080
(๋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