Deb Roy: The birth of a word

415,815 views ใƒป 2011-03-14

TED


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืžืชืจื’ื: Yubal Masalker ืžื‘ืงืจ: Ido Dekkers
00:15
Imagine if you could record your life --
0
15260
4000
ื“ืžื™ื™ื ื• ืฉื”ื™ื™ืชื ื™ื›ื•ืœื™ื ืœื”ืงืœื™ื˜ ืืช ื—ื™ื™ื›ื --
00:19
everything you said, everything you did,
1
19260
3000
ื›ืœ ืžื” ืฉืืžืจืชื, ื›ืœ ืžื” ืฉืขืฉื™ืชื,
00:22
available in a perfect memory store at your fingertips,
2
22260
3000
ื–ืžื™ืŸ ื‘ืžื—ืกืŸ ื–ื™ื›ืจื•ืŸ ื”ื ื’ื™ืฉ ืœืงืฆื•ื•ืช ืืฆื‘ืขื•ืชื™ื›ื,
00:25
so you could go back
3
25260
2000
ื›ืš ืฉื”ื™ื™ืชื ื™ื›ื•ืœื™ื ืœื—ื–ื•ืจ ืื—ื•ืจื”
00:27
and find memorable moments and relive them,
4
27260
3000
ื•ืœืืชืจ ืจื’ืขื™ื ื‘ืœืชื™ ื ืฉื›ื—ื™ื ื•ืœื—ื™ื•ืช ืื•ืชื ืžื—ื“ืฉ,
00:30
or sift through traces of time
5
30260
3000
ืื• ืœื‘ืจื•ืจ ืžืชื•ืš ืขืงื‘ื•ืช ื‘ื–ืžืŸ
00:33
and discover patterns in your own life
6
33260
2000
ื•ืœื’ืœื•ืช ืชื‘ื ื™ื•ืช ื‘ื—ื™ื™ื›ื
00:35
that previously had gone undiscovered.
7
35260
3000
ืฉื”ื™ื• ื ืกืชืจื•ืช ืœืคื ื™-ื›ืŸ.
00:38
Well that's exactly the journey
8
38260
2000
ื–ื” ื‘ื“ื™ื•ืง ื”ืžืกืข
00:40
that my family began
9
40260
2000
ืฉืžืฉืคื—ืชื™ ื”ื—ืœื” ื‘ื•
00:42
five and a half years ago.
10
42260
2000
ืœืคื ื™ ื—ืžืฉ ื•ื—ืฆื™ ืฉื ื™ื.
00:44
This is my wife and collaborator, Rupal.
11
44260
3000
ื–ื•ื”ื™ ืื™ืฉืชื™ ื•ืฉื•ืชืคืชื™, ืจื•ืคืืœ.
00:47
And on this day, at this moment,
12
47260
2000
ื•ื‘ื™ื•ื ื–ื”, ื‘ืจื’ืข ื–ื”,
00:49
we walked into the house with our first child,
13
49260
2000
ืฆืขื“ื ื• ืืœ ืชื•ืš ื”ื‘ื™ืช ื‘ื™ื—ื“ ืขื ื‘ื ื™ื ื• ื”ืจืืฉื•ืŸ,
00:51
our beautiful baby boy.
14
51260
2000
ื‘ื ื™ื ื• ื”ืชื™ื ื•ืง ื”ื™ืคื”ืคื”.
00:53
And we walked into a house
15
53260
3000
ื•ืฆืขื“ื ื• ืืœ ืชื•ืš ื”ื‘ื™ืช
00:56
with a very special home video recording system.
16
56260
4000
ืขื ืžืขืจื›ืช ื‘ื™ืชื™ืช ืžืื•ื“ ืžื™ื•ื—ื“ืช ืœื”ืงืœื˜ืช-ื•ื™ื“ืื•.
01:07
(Video) Man: Okay.
17
67260
2000
(ื•ื™ื“ืื•) ื’ื‘ืจ: ื‘ืกื“ืจ.
01:10
Deb Roy: This moment
18
70260
1000
ื“ื‘ ืจื•ื™: ืจื’ืข ื–ื”
01:11
and thousands of other moments special for us
19
71260
3000
ื•ืืœืคื™ ืจื’ืขื™ื ืื—ืจื™ื ืฉื”ื ืžื™ื•ื—ื“ื™ื ืขื‘ื•ืจื ื•,
01:14
were captured in our home
20
74260
2000
ื”ื•ืงืœื˜ื• ื‘ื‘ื™ืชื™ื ื•
01:16
because in every room in the house,
21
76260
2000
ืžืคื ื™ ืฉื‘ื›ืœ ื—ื“ืจ ื‘ื‘ื™ืช,
01:18
if you looked up, you'd see a camera and a microphone,
22
78260
3000
ืื ื”ืกืชื›ืœืชื ืœืžืขืœื”, ื”ื™ื™ืชื ืจื•ืื™ื ืžืฆืœืžื” ื•ืžื™ืงืจื•ืคื•ืŸ,
01:21
and if you looked down,
23
81260
2000
ื•ืื ื”ืกืชื›ืœืชื ืœืžื˜ื”, ื”ื™ื™ืชื ืจื•ืื™ื
01:23
you'd get this bird's-eye view of the room.
24
83260
2000
ืืช ื”ื—ื“ืจ ื‘ืžื‘ื˜ ืฉืœ ืžืขื•ืฃ-ืฆื™ืคื•ืจ.
01:25
Here's our living room,
25
85260
3000
ื”ื ื” ื”ืกืœื•ืŸ ืฉืœื ื•,
01:28
the baby bedroom,
26
88260
3000
ื—ื“ืจ-ื”ืฉื™ื ื” ืœืชื™ื ื•ืง,
01:31
kitchen, dining room
27
91260
2000
ืžื˜ื‘ื—, ื—ื“ืจ-ืื•ื›ืœ
01:33
and the rest of the house.
28
93260
2000
ื•ืฉืืจ ื”ื‘ื™ืช.
01:35
And all of these fed into a disc array
29
95260
3000
ื•ื›ืœ ื–ื” ืžื•ื–ืŸ ืืœ ืชื•ืš ืžืขืจืš ืฉืœ ื“ื™ืกืงื™ื
01:38
that was designed for a continuous capture.
30
98260
3000
ืฉืชื•ื›ื ืŸ ืœื”ืงืœื˜ื” ืžืชืžืฉื›ืช ื•ืจืฆื•ืคื”.
01:41
So here we are flying through a day in our home
31
101260
3000
ื›ืืŸ ืื ื• ืขื•ื‘ืจื™ื ืขืœ ื™ื•ื ืฉืœื ื‘ื‘ื™ืชื ื•
01:44
as we move from sunlit morning
32
104260
3000
ื›ืืฉืจ ืื ื• ื ืขื™ื ืžื‘ื•ืงืจ ืฉื˜ื•ืฃ-ืฉืžืฉ
01:47
through incandescent evening
33
107260
2000
ื“ืจืš ืขืจื‘ ื–ื•ื”ืจ
01:49
and, finally, lights out for the day.
34
109260
3000
ื•ืœื‘ืกื•ืฃ, ื›ื™ื‘ื•ื™ ืื•ืจื•ืช ืœืื•ืชื• ื™ื•ื.
01:53
Over the course of three years,
35
113260
3000
ื‘ืžื”ืœืš ืฉืœื•ืฉ ืฉื ื™ื,
01:56
we recorded eight to 10 hours a day,
36
116260
2000
ื”ืงืœื˜ื ื• 8-10 ืฉืขื•ืช ื‘ื™ื•ื,
01:58
amassing roughly a quarter-million hours
37
118260
3000
ื“ื‘ืจ ืฉื”ืกืชื›ื ื‘ืขืจืš ื‘-250 ืืœืฃ ืฉืขื•ืช
02:01
of multi-track audio and video.
38
121260
3000
ืฉืœ ืื•ื“ื™ื• ื•ื•ื™ื“ืื• ืจื‘ื™-ืขืจื•ืฆื™ื.
02:04
So you're looking at a piece of what is by far
39
124260
2000
ืืชื ืžืกืชื›ืœื™ื ืขืœ ืžืฉื”ื• ืฉื ื›ื•ืŸ ืœืขื›ืฉื™ื•
02:06
the largest home video collection ever made.
40
126260
2000
ื”ื•ื ืื•ืกืฃ ื”ื•ื™ื“ืื• ื”ื‘ื™ืชื™ ื”ื’ื“ื•ืœ ื‘ื™ื•ืชืจ ืฉื ื•ืฆืจ ืื™-ืคืขื.
02:08
(Laughter)
41
128260
3000
(ืฆื—ื•ืง)
02:11
And what this data represents
42
131260
2000
ื•ืžื” ืฉื”ืžื™ื“ืข ื”ื–ื” ืžื™ื™ืฆื’
02:13
for our family at a personal level,
43
133260
4000
ืขื‘ื•ืจ ืžืฉืคื—ืชื ื• ื‘ืจืžื” ื”ืื™ืฉื™ืช,
02:17
the impact has already been immense,
44
137260
2000
ื”ืฉืคืขืชื• ื›ื‘ืจ ืขืฆื•ืžื”,
02:19
and we're still learning its value.
45
139260
3000
ื•ืื ื• ืขื“ื™ื™ืŸ ืœื•ืžื“ื™ื ืืช ืžืฉืžืขื•ืชื•.
02:22
Countless moments
46
142260
2000
ืื™ืŸ-ืกืคื•ืจ ืจื’ืขื™ื
02:24
of unsolicited natural moments, not posed moments,
47
144260
3000
ื˜ื‘ืขื™ื™ื, ืœื ืžืชื•ื›ื ื ื™ื, ืœืœื ื”ืฆื’ื•ืช,
02:27
are captured there,
48
147260
2000
ืžื•ืงืœื˜ื™ื ืฉื,
02:29
and we're starting to learn how to discover them and find them.
49
149260
3000
ื•ืื ื• ืžืชื—ื™ืœื™ื ืœืœืžื•ื“ ื›ื™ืฆื“ ืœืืชืจื ื•ืœื’ืœื•ืชื.
02:32
But there's also a scientific reason that drove this project,
50
152260
3000
ืื‘ืœ ื™ืฉื ื” ื’ื ืกื™ื‘ื” ืžื“ืขื™ืช ืฉื”ื ื™ืขื” ืžื™ื–ื ื–ื”,
02:35
which was to use this natural longitudinal data
51
155260
4000
ื•ื”ื™ื, ืœื”ืฉืชืžืฉ ื‘ืžื™ื“ืข ื”ื˜ื‘ืขื™ ื•ื”ืžืงื™ืฃ ื”ื–ื”
02:39
to understand the process
52
159260
2000
ื›ื“ื™ ืœื”ื‘ื™ืŸ ืืช ื”ืชื”ืœื™ืš
02:41
of how a child learns language --
53
161260
2000
ื‘ื• ื™ืœื“ ืœื•ืžื“ ืฉืคื” --
02:43
that child being my son.
54
163260
2000
ื‘ืžืงืจื” ื–ื” ื”ื™ืœื“ ื”ื•ื ื‘ื ื™.
02:45
And so with many privacy provisions put in place
55
165260
4000
ืชื•ืš ื ืงื™ื˜ืช ืืžืฆืขื™ ื–ื”ื™ืจื•ืช ืจื‘ื™ื ืœืฉืžื™ืจืช ื”ืคืจื˜ื™ื•ืช
02:49
to protect everyone who was recorded in the data,
56
169260
3000
ื›ื“ื™ ืœื”ื’ืŸ ืขืœ ื›ืœ ืื—ื“ ืฉื”ื•ืงืœื˜,
02:52
we made elements of the data available
57
172260
3000
ื—ืฉืคื ื• ื ืชื•ื ื™ื ืฉื•ื ื™ื
02:55
to my trusted research team at MIT
58
175260
3000
ื‘ืคื ื™ ืงื‘ื•ืฆืช ื”ืžื—ืงืจ ื”ืžื•ืกืžื›ืช ืฉืœื™ ื‘-MIT
02:58
so we could start teasing apart patterns
59
178260
3000
ื›ื“ื™ ืฉื ื•ื›ืœ ืœื”ืชื—ื™ืœ ืœืกืจื•ืง ืœืžืฆื™ืืช ืชื‘ื ื™ื•ืช
03:01
in this massive data set,
60
181260
3000
ื‘ืžืื’ืจ ื”ื ืชื•ื ื™ื ื”ืื“ื™ืจ ื”ื–ื”,
03:04
trying to understand the influence of social environments
61
184260
3000
ื‘ื ืกื•ืชื ื• ืœื”ื‘ื™ืŸ ืืช ื”ื”ืฉืคืขื” ืฉืœ ืกื‘ื™ื‘ื•ืช ื—ื‘ืจืชื™ื•ืช
03:07
on language acquisition.
62
187260
2000
ืขืœ ืจื›ื™ืฉืช ืฉืคื”.
03:09
So we're looking here
63
189260
2000
ืื ื• ืžืกืชื›ืœื™ื ื›ืืŸ
03:11
at one of the first things we started to do.
64
191260
2000
ืขืœ ืื—ื“ ื”ื“ื‘ืจื™ื ื”ืจืืฉื•ื ื™ื ืฉื”ืชื—ืœื ื• ืœืขืฉื•ืช.
03:13
This is my wife and I cooking breakfast in the kitchen,
65
193260
4000
ื›ืืŸ ืื™ืฉืชื™ ื•ืื ื™ ืžื›ื™ื ื™ื ืืจื•ื—ืช-ื‘ื•ืงืจ ื‘ืžื˜ื‘ื—.
03:17
and as we move through space and through time,
66
197260
3000
ื‘ืขื•ื“ื ื• ื ืขื™ื ื‘ืžืจื—ื‘ ื•ื‘ื–ืžืŸ,
03:20
a very everyday pattern of life in the kitchen.
67
200260
3000
ื“ืคื•ืก ื™ื•ืžื™ ื‘ื”ื—ืœื˜ ืฉืœ ื—ื™ื™ื ื‘ืžื˜ื‘ื—,
03:23
In order to convert
68
203260
2000
ื›ื“ื™ ืœื”ืคื•ืš
03:25
this opaque, 90,000 hours of video
69
205260
3000
ืืช 90,000 ืฉืขื•ืช ื”ื•ื™ื“ืื• ื”ืกืชื•ืžื•ืช ืœืžืฉื”ื• ืฉื ื•ื›ืœ
03:28
into something that we could start to see,
70
208260
2000
ืœื”ืชื—ื™ืœ ืœืื‘ื—ืŸ ื‘ื”ืŸ ืžืฉื”ื•,
03:30
we use motion analysis to pull out,
71
210260
2000
ืื ื• ืžืฉืชืžืฉื™ื ื‘ืื ืœื™ื–ืช ืชื ื•ืขื”,
03:32
as we move through space and through time,
72
212260
2000
ื›ืืฉืจ ืื ื• ื ืขื™ื ื“ืจืš ืžืจื—ื‘ ื•ื“ืจืš ื–ืžืŸ,
03:34
what we call space-time worms.
73
214260
3000
ืžื” ืฉื ืงืจื ืชื•ืœืขื•ืช ืžืจื—ื‘-ื–ืžืŸ.
03:37
And this has become part of our toolkit
74
217260
3000
ื•ื–ื” ื”ืคืš ืœื”ื™ื•ืช ื—ืœืง ืžืืจื’ื– ื”ื›ืœื™ื ืฉืœื ื•
03:40
for being able to look and see
75
220260
3000
ื›ื“ื™ ืฉื ื•ื›ืœ ืœื”ื‘ื™ื˜ ื•ืœืจืื•ืช
03:43
where the activities are in the data,
76
223260
2000
ื”ื™ื›ืŸ ื‘ื ืชื•ื ื™ื ื ืžืฆืืช ื”ืคืขื™ืœื•ืช,
03:45
and with it, trace the pattern of, in particular,
77
225260
3000
ื•ืื™ืชื”, ืœืืชืจ ืืช ื”ืชื‘ื ื™ืช, ื•ื‘ืžื™ื•ื—ื“,
03:48
where my son moved throughout the home,
78
228260
2000
ื‘ืื•ืชื ื”ืžืงื•ืžื•ืช ืฉื‘ื ื™ ื ืข ื‘ืจื—ื‘ื™ ื”ื‘ื™ืช,
03:50
so that we could focus our transcription efforts,
79
230260
3000
ื›ืš ืฉื ื•ื›ืœ ืœืžืงื“ ืืช ืžืืžืฆื™ ื”ื‘ืœืฉื ื•ืช ืฉืœื ื•,
03:53
all of the speech environment around my son --
80
233260
3000
ืืช ื›ืœืœ ืกื‘ื™ื‘ืช ื”ื“ื™ื‘ื•ืจ, ืกื‘ื™ื‘ ื‘ื ื™ --
03:56
all of the words that he heard from myself, my wife, our nanny,
81
236260
3000
ื›ืœ ื”ืžื™ืœื™ื ืฉื”ื•ื ืฉืžืข ืžืžื ื™, ืžืื™ืฉืชื™, ื”ืžื˜ืคืœืช,
03:59
and over time, the words he began to produce.
82
239260
3000
ื•ืขื ื”ื–ืžืŸ, ื”ืžื™ืœื™ื ืฉื”ื•ื ื”ื—ืœ ืœื™ืฆื•ืจ.
04:02
So with that technology and that data
83
242260
3000
ื›ืš ืฉืขื ื”ื˜ื›ื ื•ืœื•ื’ื™ื” ื”ื–ื• ื•ื”ื ืชื•ื ื™ื ื”ืœืœื•
04:05
and the ability to, with machine assistance,
84
245260
2000
ื•ื”ื™ื›ื•ืœืช, ื‘ืกื™ื•ืข ืžื›ื•ื ื”,
04:07
transcribe speech,
85
247260
2000
ืœืฉืขืชืง ื“ื™ื‘ื•ืจ,
04:09
we've now transcribed
86
249260
2000
ื”ืฆืœื—ื ื• ืœืฉืขืชืง
04:11
well over seven million words of our home transcripts.
87
251260
3000
ื™ื•ืชืจ ืžืฉื‘ืข ืžื™ืœื™ื•ืŸ ืžื™ืœื™ื ืžื”ื”ืงืœื˜ื•ืช.
04:14
And with that, let me take you now
88
254260
2000
ื•ื›ืขืช ืขื ื›ืœ ื–ื”, ืืงื— ืืชื›ื
04:16
for a first tour into the data.
89
256260
3000
ืœืกื™ื•ืจ ื”ืจืืฉื•ืŸ ื‘ื ืชื•ื ื™ื.
04:19
So you've all, I'm sure,
90
259260
2000
ืื ื™ ื‘ื˜ื•ื— ืฉื›ื•ืœื›ื
04:21
seen time-lapse videos
91
261260
2000
ืจืื™ืชื ืคืขื ืกืจื˜ื•ืŸ ืฆื™ืœื•ื-ืื™ื˜ื™
04:23
where a flower will blossom as you accelerate time.
92
263260
3000
ื‘ื• ืจื•ืื™ื ืคืจื— ืคื•ืจื— ื‘ื–ืžืŸ ืžื•ืืฅ.
04:26
I'd like you to now experience
93
266260
2000
ื”ื™ื™ืชื™ ืจื•ืฆื” ืฉืชืชื ืกื• ื›ืขืช
04:28
the blossoming of a speech form.
94
268260
2000
ื‘ืคืจื™ื—ื” ืฉืœ ืฆื•ืจืช ื“ื™ื‘ื•ืจ.
04:30
My son, soon after his first birthday,
95
270260
2000
ื‘ื ื™, ืžื™ื™ื“ ืœืื—ืจ ื™ื•ื-ื”ื•ืœื“ืชื• ื”ืจืืฉื•ืŸ,
04:32
would say "gaga" to mean water.
96
272260
3000
ื”ื™ื” ืื•ืžืจ "ื’ืื’ื”" ื‘ื”ืชื›ื•ื•ื ื• ืœืžื™ื.
04:35
And over the course of the next half-year,
97
275260
3000
ื•ื‘ืžื”ืœืš ื—ืฆื™ ืฉื ื” ื”ื‘ืื”,
04:38
he slowly learned to approximate
98
278260
2000
ื”ื•ื ืœืžื“ ื‘ื”ื“ืจื’ื” ืœื”ืชืงืจื‘
04:40
the proper adult form, "water."
99
280260
3000
ืืœ ื”ืฆื•ืจื” ื”ื”ื•ืœืžืช ืฉืœ ื‘ื•ื’ืจื™ื ,"ืžื™ื (ื•ื•ืื˜ืจ)".
04:43
So we're going to cruise through half a year
100
283260
2000
ืื– ืื ื• ืขื•ืžื“ื™ื ืœืฉื™ื™ื˜ ื“ืจืš ื—ืฆื™ ืฉื ื”
04:45
in about 40 seconds.
101
285260
2000
ื‘-40 ืฉื ื™ื•ืช.
04:47
No video here,
102
287260
2000
ืื™ืŸ ื›ืืŸ ื•ื™ื“ืื•,
04:49
so you can focus on the sound, the acoustics,
103
289260
3000
ืœื›ืŸ ืชื•ื›ืœื• ืœื”ืชืจื›ื– ื‘ืงื•ืœ, ื‘ืืงื•ืกื˜ื™ืงื”,
04:52
of a new kind of trajectory:
104
292260
2000
ืฉืœ ื ืชื™ื‘ ืžืกื•ื’ ื—ื“ืฉ:
04:54
gaga to water.
105
294260
2000
ืžื’ืื’ื” ืืœ ืžื™ื (ื•ื•ืื˜ืจ).
04:56
(Audio) Baby: Gagagagagaga
106
296260
12000
(ืงื•ืœ) ืชื™ื ื•ืง: ื’ื”ื’ื”ื’ื”ื’ื”ื’ื”
05:08
Gaga gaga gaga
107
308260
4000
ื’ื’ื” ื’ื’ื” ื’ื’ื”
05:12
guga guga guga
108
312260
5000
ื’ื•ื’ื” ื’ื•ื’ื” ื’ื•ื’ื”
05:17
wada gaga gaga guga gaga
109
317260
5000
ื•ืื“ื” ื’ื’ื” ื’ื’ื” ื’ื•ื’ื” ื’ื’ื”
05:22
wader guga guga
110
322260
4000
ื•ื•ืื“ืจ ื’ื•ื’ื” ื’ื•ื’ื”
05:26
water water water
111
326260
3000
ื•ื•ืื˜ืจ ื•ื•ืื˜ืจ ื•ื•ืื˜ืจ
05:29
water water water
112
329260
6000
ื•ื•ืื˜ืจ ื•ื•ืื˜ืจ ื•ื•ืื˜ืจ
05:35
water water
113
335260
4000
ื•ื•ืื˜ืจ ื•ื•ืื˜ืจ
05:39
water.
114
339260
2000
ื•ื•ืื˜ืจ.
05:41
DR: He sure nailed it, didn't he.
115
341260
2000
ื”ื•ื ืขืฉื” ืืช ื–ื”, ื ื›ื•ืŸ?
05:43
(Applause)
116
343260
7000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
05:50
So he didn't just learn water.
117
350260
2000
ื”ื•ื ืœื ืจืง ืœืžื“ ืžื™ื.
05:52
Over the course of the 24 months,
118
352260
2000
ื‘ืžื”ืœืš ืชืงื•ืคื” ืฉืœ 24 ื—ื•ื“ืฉื™ื,
05:54
the first two years that we really focused on,
119
354260
3000
ื‘ืฉื ืชื™ื™ื ื”ืจืืฉื•ื ื•ืช, ืฉื‘ื”ืŸ ื”ืชืจื›ื–ื ื•,
05:57
this is a map of every word he learned in chronological order.
120
357260
4000
ื–ื•ื”ื™ ืžืคื” ืฉืœ ื›ืœ ื”ืžื™ืœื™ื ืฉื”ื•ื ืœืžื“ ื‘ืกื“ืจ ื›ืจื•ื ื•ืœื•ื’ื™.
06:01
And because we have full transcripts,
121
361260
3000
ื•ืžืื—ืจ ื•ื™ืฉ ืœื ื• ืืช ื›ืœ ื”ืชืขืชื™ืงื™ื,
06:04
we've identified each of the 503 words
122
364260
2000
ื–ื™ื”ื™ื ื• ื›ืœ ืื—ืช ืžื”-503 ืžื™ืœื™ื
06:06
that he learned to produce by his second birthday.
123
366260
2000
ืฉื”ื•ื ืœืžื“ ืœื”ื’ื•ืช ืขื“ ื™ื•ื ื”ื•ืœื“ืชื• ื”ืฉื ื™.
06:08
He was an early talker.
124
368260
2000
ื”ื•ื ื”ืชื—ื™ืœ ืœื“ื‘ืจ ืžื•ืงื“ื.
06:10
And so we started to analyze why.
125
370260
3000
ื•ืœื›ืŸ ื”ืชื—ืœื ื• ืœื‘ื“ื•ืง ืœืžื”.
06:13
Why were certain words born before others?
126
373260
3000
ืžื“ื•ืข ืžื™ืœื™ื ืžืกื•ื™ื™ืžื•ืช ื ื•ืœื“ื• ืœืคื ื™ ืื—ืจื•ืช?
06:16
This is one of the first results
127
376260
2000
ื–ื•ื”ื™ ืื—ืช ื”ืชื•ืฆืื•ืช ื”ืจืืฉื•ื ื•ืช
06:18
that came out of our study a little over a year ago
128
378260
2000
ืฉื”ืชืงื‘ืœื” ืžืžื—ืงืจื ื• ืœืคื ื™ ืงืฆืช ื™ื•ืชืจ ืžืฉื ื”
06:20
that really surprised us.
129
380260
2000
ื•ืืฉืจ ืžืžืฉ ื”ืคืชื™ืขื” ืื•ืชื ื•.
06:22
The way to interpret this apparently simple graph
130
382260
3000
ื”ื“ืจืš ืœืคืจืฉ ื’ืจืฃ ืคืฉื•ื˜ ื–ื” ืœื›ืื•ืจื”
06:25
is, on the vertical is an indication
131
385260
2000
ื”ื™ื ืฉื”ืฆื™ืจ ื”ืื ื›ื™ ืžืชืืจ
06:27
of how complex caregiver utterances are
132
387260
3000
ืขื“ ื›ืžื” ืžื•ืจื›ื‘ื™ื ื”ื‘ื™ื˜ื•ื™ื™ื ืฉืœ ื”ืžื˜ืคืœืช
06:30
based on the length of utterances.
133
390260
2000
ื‘ื”ืชื‘ืกืก ืขืœ ืื•ืจืš ื”ื‘ื™ื˜ื•ื™ื™ื.
06:32
And the [horizontal] axis is time.
134
392260
3000
ื•ื”ืฆื™ืจ ื”ืื•ืคืงื™ ื–ื” ื”ื–ืžืŸ.
06:35
And all of the data,
135
395260
2000
ื•ืืช ื›ืœ ื”ื ืชื•ื ื™ื,
06:37
we aligned based on the following idea:
136
397260
3000
ืกื™ื“ืจื ื• ืœืคื™ ื”ืจืขื™ื•ืŸ ื”ื‘ื:
06:40
Every time my son would learn a word,
137
400260
3000
ื‘ื›ืœ ืคืขื ืฉื‘ื ื™ ืœืžื“ ืžื™ืœื”,
06:43
we would trace back and look at all of the language he heard
138
403260
3000
ื”ื™ื™ื ื• ืžืฉื—ื–ืจื™ื ืื—ื•ืจื” ื•ืžืกืชื›ืœื™ื ืขืœ ื›ืœ ื”ืฉืคื” ืฉื”ื•ื ืฉืžืข
06:46
that contained that word.
139
406260
2000
ื•ืืฉืจ ื›ืœืœื” ืืช ืื•ืชื” ืžื™ืœื”.
06:48
And we would plot the relative length of the utterances.
140
408260
4000
ื•ื”ื™ื™ื ื• ืžืฉืจื˜ื˜ื™ื ืืช ื”ืื•ืจืš ื”ื™ื—ืกื™ ืฉืœ ื”ื‘ื™ื˜ื•ื™ื™ื.
06:52
And what we found was this curious phenomena,
141
412260
3000
ื•ืžื” ืฉืžืฆืื ื• ื”ื™ืชื” ื”ืชื•ืคืขื” ื”ืžืกืงืจื ืช
06:55
that caregiver speech would systematically dip to a minimum,
142
415260
3000
ืฉื”ื“ื™ื‘ื•ืจ ืฉืœ ื”ืžื˜ืคืœืช ื”ื™ื” ืฉื•ืงืข ื‘ื”ืชืžื“ื” ืœืžื™ื ื™ืžื•ื,
06:58
making language as simple as possible,
143
418260
3000
ื•ื”ืฉืคื” ื”ื™ืชื” ื”ื•ืคื›ืช ืœืคืฉื•ื˜ื” ื›ื›ืœ ื”ืืคืฉืจ,
07:01
and then slowly ascend back up in complexity.
144
421260
3000
ื•ืื– ืขื•ืœื” ื‘ื”ื“ืจื’ื” ื‘ืจืžืช ื”ืžื•ืจื›ื‘ื•ืช.
07:04
And the amazing thing was
145
424260
2000
ื•ื”ื“ื‘ืจ ื”ืžื“ื”ื™ื ื”ื™ื”
07:06
that bounce, that dip,
146
426260
2000
ืฉืื•ืชื” ืงืคื™ืฆื”, ืื•ืชื” ืฉืงื™ืขื”,
07:08
lined up almost precisely
147
428260
2000
ื”ืชืื™ืžื• ื›ืžืขื˜ ื‘ืžื“ื•ื™ื™ืง
07:10
with when each word was born --
148
430260
2000
ืœืžื•ืขื“ ื”ื•ืœื“ืชื” ืฉืœ ืžื™ืœื” ื›ืœืฉื”ื™ --
07:12
word after word, systematically.
149
432260
2000
ืžื™ืœื” ืื—ืจ ืžื™ืœื”, ื‘ืื•ืคืŸ ืฉื™ื˜ืชื™.
07:14
So it appears that all three primary caregivers --
150
434260
2000
ื›ืš ืฉื ืจืื” ืฉื›ืœ ืฉืœื•ืฉืช ื”ืžื˜ืคืœื™ื --
07:16
myself, my wife and our nanny --
151
436260
3000
ืื ื™, ืื™ืฉืชื™ ื•ื”ืžื˜ืคืœืช --
07:19
were systematically and, I would think, subconsciously
152
439260
3000
ื”ื™ื• ื‘ืฉื™ื˜ืชื™ื•ืช, ื•ืื ื™ ืกื‘ื•ืจ, ื‘ืื•ืคืŸ ืชืช-ื”ื›ืจืชื™
07:22
restructuring our language
153
442260
2000
ื‘ื•ื ื™ื ืžื—ื“ืฉ ืืช ืฉืคืชื™ื ื•
07:24
to meet him at the birth of a word
154
444260
3000
ื›ื“ื™ ืœืชืžื•ืš ื‘ื• ื‘ื”ื•ืœื“ืชื” ืฉืœ ืžื™ืœื”
07:27
and bring him gently into more complex language.
155
447260
4000
ื•ืœื”ื‘ื™ืื• ื‘ืขื“ื™ื ื•ืช ืืœ ืชื•ืš ืฉืคื” ื™ื•ืชืจ ืžื•ืจื›ื‘ืช.
07:31
And the implications of this -- there are many,
156
451260
2000
ื•ื”ื”ืฉืœื›ื•ืช ืฉืœ ื–ื” -- ื”ืŸ ืจื‘ื•ืช,
07:33
but one I just want to point out,
157
453260
2000
ืื‘ืœ ื‘ืจืฆื•ื ื™ ืœื”ืฆื‘ื™ืข ืขืœ ืื—ืช,
07:35
is that there must be amazing feedback loops.
158
455260
3000
ื•ื”ื™ื ืฉื—ื™ื™ื‘ื•ืช ืœื”ืชืงื™ื™ื ืœื•ืœืื•ืช ืžืฉื•ื‘ ืžื•ืคืœืื•ืช.
07:38
Of course, my son is learning
159
458260
2000
ื›ืžื•ื‘ืŸ, ื‘ื ื™ ืœื•ืžื“
07:40
from his linguistic environment,
160
460260
2000
ืžื”ืกื‘ื™ื‘ื” ื”ืœืฉื•ื ื™ืช ืฉืœื•,
07:42
but the environment is learning from him.
161
462260
3000
ืื‘ืœ ื”ืกื‘ื™ื‘ื” ื’ื ืœื•ืžื“ืช ืžืžื ื•.
07:45
That environment, people, are in these tight feedback loops
162
465260
3000
ื‘ืื•ืชื” ืกื‘ื™ื‘ื”, ืื ืฉื™ื ื ืžืฆืื™ื ื‘ืชื•ืš ืœื•ืœืื•ืช ืžืฉื•ื‘ ื”ื“ื•ืงื•ืช ืืœื•
07:48
and creating a kind of scaffolding
163
468260
2000
ื•ื™ื•ืฆืจื™ื ืžื™ืŸ ืžืขืจื›ืช ืคื™ื’ื•ืžื™ื
07:50
that has not been noticed until now.
164
470260
3000
ืฉืขื“ ื”ื™ื•ื ืœื ืฉืžื• ืืœื™ื” ืœื‘.
07:54
But that's looking at the speech context.
165
474260
2000
ื–ื” ื›ืืฉืจ ืžื“ื•ื‘ืจ ื‘ื”ื™ื‘ื˜ ื”ื“ื™ื‘ื•ืจื™.
07:56
What about the visual context?
166
476260
2000
ืžื” ื‘ื ื•ื’ืข ืœื”ื™ื‘ื˜ ื”ื—ื–ื•ืชื™?
07:58
We're not looking at --
167
478260
2000
ืื™ื ื ื• ืžืกืชื›ืœื™ื ืขืœ --
08:00
think of this as a dollhouse cutaway of our house.
168
480260
2000
ืชื—ืฉื‘ื• ืฉื–ื” ื—ืชืš ื‘ื™ืช-ื‘ื•ื‘ื•ืช ืฉืœ ื‘ื™ืชื™ื ื•.
08:02
We've taken those circular fish-eye lens cameras,
169
482260
3000
ื ื˜ืœื ื• ืืช ืžืฆืœืžื•ืช ืขื“ืฉื•ืช ืขื™ืŸ-ื“ื’ ื”ืขื’ื•ืœื•ืช,
08:05
and we've done some optical correction,
170
485260
2000
ื•ืขืฉื™ื ื• ืงืฆืช ืชื™ืงื•ืŸ ืื•ืคื˜ื™,
08:07
and then we can bring it into three-dimensional life.
171
487260
4000
ื•ื›ืš ืื ื• ื™ื›ื•ืœื™ื ืœื”ื‘ื™ื ื–ืืช ืœืžืฆื‘ ืชืœืช-ืžื™ืžื“ื™ ื—ื™.
08:11
So welcome to my home.
172
491260
2000
ืื– ื‘ืจื•ื›ื™ื ื”ื‘ืื™ื ืœื‘ื™ืชื™.
08:13
This is a moment,
173
493260
2000
ื–ื”ื• ืจื’ืข,
08:15
one moment captured across multiple cameras.
174
495260
3000
ืจื’ืข ืื—ื“ ืฉื ืœื›ื“ ื‘ืืžืฆืขื•ืช ืžืกืคืจ ืžืฆืœืžื•ืช.
08:18
The reason we did this is to create the ultimate memory machine,
175
498260
3000
ื”ืกื™ื‘ื” ืฉืขืฉื™ื ื• ื–ืืช ื”ื™ืชื” ื›ื“ื™ ืœื™ืฆื•ืจ ืžื›ื•ื ื” ื‘ืขืœืช ื–ื™ื›ืจื•ืŸ ืื•ืœื˜ื™ืžื˜ื™ื‘ื™,
08:21
where you can go back and interactively fly around
176
501260
3000
ื‘ื” ื ื™ืชืŸ ืœื˜ื•ืก ื›ื” ื•ื›ื” ื‘ืื•ืคืŸ ืื™ื ื˜ืจืืงื˜ื™ื‘ื™
08:24
and then breathe video-life into this system.
177
504260
3000
ื•ืื– ืœื”ืคื™ื— ื—ื™ื™-ื•ื™ื“ืื• ืืœ ืชื•ืš ืžืขืจื›ืช ื–ื•.
08:27
What I'm going to do
178
507260
2000
ืžื” ืฉืื ื™ ื”ื•ืœืš ืœืขืฉื•ืช
08:29
is give you an accelerated view of 30 minutes,
179
509260
3000
ื”ื•ื ืœืชืช ืœื›ื ืžื‘ื˜ ืžื•ืืฅ ืฉืœ 30 ื“ืงื•ืช,
08:32
again, of just life in the living room.
180
512260
2000
ืฉื•ื‘, ืจืง ืฉืœ ื—ื™ื™ื ื‘ืกืœื•ืŸ.
08:34
That's me and my son on the floor.
181
514260
3000
ื–ื”ื• ื‘ื ื™ ืขืœ ื”ืจื™ืฆืคื”.
08:37
And there's video analytics
182
517260
2000
ื•ืฉื ื–ื” ืžื ืชื— ื•ื™ื“ืื•
08:39
that are tracking our movements.
183
519260
2000
ืืฉืจ ืขื•ืงื‘ ืื—ืจ ืชื ื•ืขื•ืชื™ื ื•.
08:41
My son is leaving red ink. I am leaving green ink.
184
521260
3000
ื‘ื ื™ ืžืฉืื™ืจ ืกื™ืžืŸ ืื“ื•ื ื•ืื ื™ ื™ืจื•ืง.
08:44
We're now on the couch,
185
524260
2000
ืขื›ืฉื™ื• ืื ื• ืขืœ ื”ืกืคื”,
08:46
looking out through the window at cars passing by.
186
526260
3000
ืžืชื‘ื•ื ื ื™ื ื“ืจืš ื”ื—ืœื•ืŸ ื‘ืžื›ื•ื ื™ื•ืช ื—ื•ืœืคื•ืช.
08:49
And finally, my son playing in a walking toy by himself.
187
529260
3000
ื•ืœื‘ืกื•ืฃ, ื‘ื ื™ ืžืฉื—ืง ื‘ืฆืขืฆื•ืข.
08:52
Now we freeze the action, 30 minutes,
188
532260
3000
ื›ืขืช ืื ื• ืžืงืคื™ืื™ื ืืช ื”ืชื ื•ืขื”, 30 ื“ืงื•ืช,
08:55
we turn time into the vertical axis,
189
535260
2000
ืื ื• ืžืกื•ื‘ื‘ื™ื ืืช ื”ื–ืžืŸ ืœืฆื™ืจ ืื ื›ื™,
08:57
and we open up for a view
190
537260
2000
ื•ืื ื• ืคื•ืชื—ื™ื ืœืฆืคื™ื™ื”
08:59
of these interaction traces we've just left behind.
191
539260
3000
ืืช ืขืงื‘ื•ืช ื”ืื™ื ื˜ืจืืงืฆื™ื” ืฉื”ืฉืืจื ื• ื–ื” ืขืชื” ืžืื—ื•ืจ.
09:02
And we see these amazing structures --
192
542260
3000
ื•ืื ื• ืจื•ืื™ื ืžื‘ื ื™ื ืžื“ื”ื™ืžื™ื --
09:05
these little knots of two colors of thread
193
545260
3000
ืœืฉื ื™ ื”ืงืฉืจื™ื ื”ืงื˜ื ื™ื ื”ืœืœื• ืฉืœ ืฉื ื™ ืงื•ื™ ืฆื‘ืขื™ื
09:08
we call "social hot spots."
194
548260
2000
ืื ื• ืงื•ืจืื™ื ื ืงื•ื“ื•ืช ืžืคื’ืฉ ื—ืžื•ืช.
09:10
The spiral thread
195
550260
2000
ืœืงื• ื”ื—ืœื–ื•ื ื™
09:12
we call a "solo hot spot."
196
552260
2000
ืื ื• ืงื•ืจืื™ื ื ืงื•ื“ื” ื—ืžื” ืกื•ืœื•.
09:14
And we think that these affect the way language is learned.
197
554260
3000
ื•ืื ื• ืกื‘ื•ืจื™ื ืฉื–ื” ืžืฉืคื™ืข ืขืœ ื”ืื•ืคืŸ ื‘ื• ืฉืคื” ื ืœืžื“ืช.
09:17
What we'd like to do
198
557260
2000
ืžื” ืฉื”ื™ื™ื ื• ืจื•ืฆื™ื ืœืขืฉื•ืช
09:19
is start understanding
199
559260
2000
ื–ื” ืœื”ืชื—ื™ืœ ืœื”ื‘ื™ืŸ
09:21
the interaction between these patterns
200
561260
2000
ืืช ื”ืื™ื ื˜ืจืืงืฆื™ื” ื‘ื™ืŸ ืฉืชื™ ืชื‘ื ื™ื•ืช ื”ืœืœื•
09:23
and the language that my son is exposed to
201
563260
2000
ื•ืืช ื”ืฉืคื” ืฉื‘ื ื™ ื—ืฉื•ืฃ ืืœื™ื”
09:25
to see if we can predict
202
565260
2000
ื›ื“ื™ ืœืจืื•ืช ืื ื ื™ืชืŸ ืœื—ื–ื•ืช
09:27
how the structure of when words are heard
203
567260
2000
ื›ื™ืฆื“ ื”ืžื‘ื ื” ืฉืœ ืžื•ืขื“ ืฉืžื™ืขืช ื”ืžื™ืœื™ื
09:29
affects when they're learned --
204
569260
2000
ืžืฉืคื™ืข ื›ืืฉืจ ื”ืŸ ื ืœืžื“ื•ืช --
09:31
so in other words, the relationship
205
571260
2000
ื‘ืžื™ืœื™ื ืื—ืจื•ืช,
09:33
between words and what they're about in the world.
206
573260
4000
ื”ืงืฉืจ ื‘ื™ืŸ ืžื™ืœื™ื ืœื‘ื™ืŸ ืžืฉืžืขื•ืชืŸ ื‘ืขื•ืœื ื”ื–ื”.
09:37
So here's how we're approaching this.
207
577260
2000
ื•ื›ืš ื ื™ื’ืฉื ื• ืœืคื™ืชืจื•ืŸ.
09:39
In this video,
208
579260
2000
ื‘ื•ื™ื“ืื• ื–ื”,
09:41
again, my son is being traced out.
209
581260
2000
ืฉื•ื‘, ื™ืฉ ืžืขืงื‘ ืื—ืจ ื‘ื ื™.
09:43
He's leaving red ink behind.
210
583260
2000
ื”ื•ื ืžืฉืื™ืจ ืกื™ืžืŸ ืื“ื•ื.
09:45
And there's our nanny by the door.
211
585260
2000
ื•ื”ื ื” ื”ืžื˜ืคืœืช ืœื™ื“ ื”ื“ืœืช.
09:47
(Video) Nanny: You want water? (Baby: Aaaa.)
212
587260
3000
(ื•ื™ื“ืื•) ืžื˜ืคืœืช: ืืชื” ืจื•ืฆื” ืžื™ื? (ืชื™ื ื•ืง: ืื”ื”ื”.)
09:50
Nanny: All right. (Baby: Aaaa.)
213
590260
3000
ืžื˜ืคืœืช: ื‘ืกื“ืจ. (ืชื™ื ื•ืง: ืื”ื”ื”.)
09:53
DR: She offers water,
214
593260
2000
ื“.ืจ.: ื”ื™ื ืžืฆื™ืขื” ืžื™ื,
09:55
and off go the two worms
215
595260
2000
ื•ื™ื•ืฆืื•ืช ืฉืชื™ ื”ืชื•ืœืขื•ืช
09:57
over to the kitchen to get water.
216
597260
2000
ืืœ ืขื‘ืจ ื”ืžื˜ื‘ื— ืœืงื—ืช ืžื™ื.
09:59
And what we've done is use the word "water"
217
599260
2000
ื•ืžื” ืฉืขืฉื™ื ื• ื–ื” ื”ืฉืชืžืฉื ื• ื‘ืžื™ืœื” "ืžื™ื"
10:01
to tag that moment, that bit of activity.
218
601260
2000
ื›ื“ื™ ืœืกืžืŸ ืืช ื”ืจื’ืข ื”ื–ื”, ืืช ืื•ืชื• ืฉื‘ื™ื‘ ืคืขื•ืœื”.
10:03
And now we take the power of data
219
603260
2000
ื•ื›ืขืช ืื ื• ืžื ืฆืœื™ื ืืช ื”ืขื•ืฆืžื” ืฉื‘ื ืชื•ื ื™ื
10:05
and take every time my son
220
605260
3000
ื•ื ื•ื˜ืœื™ื ืืช ื›ืœ ื”ืคืขืžื™ื ืฉื‘ื ื™
10:08
ever heard the word water
221
608260
2000
ืฉืžืข ืื™-ืคืขื ืืช ื”ืžื™ืœื” ืžื™ื
10:10
and the context he saw it in,
222
610260
2000
ื•ืืช ื”ื”ืงืฉืจ ื‘ื• ื”ื•ื ืจืื” ืื•ืชื”,
10:12
and we use it to penetrate through the video
223
612260
3000
ื•ืžืฉืชืžืฉื™ื ื‘ื–ื” ื›ื“ื™ ืœื—ื“ื•ืจ ื“ืจืš ื”ื•ื™ื“ืื•
10:15
and find every activity trace
224
615260
3000
ื•ืœืžืฆื•ื ื›ืœ ืขื™ืงื‘ื” ืฉืœ ืคืขื•ืœื”
10:18
that co-occurred with an instance of water.
225
618260
3000
ืืฉืจ ื”ืชืจื—ืฉื” ื‘ื•-ื–ืžื ื™ืช ืขื ืื™ืจื•ืข ื”ืžื™ื.
10:21
And what this data leaves in its wake
226
621260
2000
ื•ืžื” ืฉื”ื ืชื•ื ื™ื ืžืฉืื™ืจื™ื ื‘ืฉื•ื‘ืœ ืฉืœื”ื
10:23
is a landscape.
227
623260
2000
ื–ื” ืชื•ื•ืื™-ืฉื˜ื—.
10:25
We call these wordscapes.
228
625260
2000
ืื ื• ืงื•ืจืื™ื ืœื–ื” ืชื•ื•ืื™-ืžื™ืœื™ื.
10:27
This is the wordscape for the word water,
229
627260
2000
ื–ื”ื• ืชื•ื•ืื™-ื”ืžื™ืœื™ื ืขื‘ื•ืจ ื”ืžื™ืœื” ืžื™ื,
10:29
and you can see most of the action is in the kitchen.
230
629260
2000
ื ื™ืชืŸ ืœืจืื•ืช ืฉืจื•ื‘ ื”ืคืขื™ืœื•ืช ื”ื™ื ื‘ืžื˜ื‘ื—.
10:31
That's where those big peaks are over to the left.
231
631260
3000
ืฉื ื™ืฉื ื ืฉื ื™ ืฉื™ืื™ื ื’ื“ื•ืœื™ื ื‘ืฆื“ ืฉืžืืœ.
10:34
And just for contrast, we can do this with any word.
232
634260
3000
ืœืฉื ื”ืžื—ืฉื”, ื ื™ืชืŸ ืœืขืฉื•ืช ื–ืืช ืขื ื›ืœ ืžื™ืœื”.
10:37
We can take the word "bye"
233
637260
2000
ื ื•ื›ืœ ืœืงื—ืช ืืช ื”ืžื™ืœื” "ืฉืœื•ื"
10:39
as in "good bye."
234
639260
2000
ื›ืžื• ื‘"ื”ื™ื™ื” ืฉืœื•ื".
10:41
And we're now zoomed in over the entrance to the house.
235
641260
2000
ื•ืžืชื›ื ืกื™ื ืืœ ืขื‘ืจ ื”ื›ื ื™ืกื” ืฉืœ ื”ื‘ื™ืช.
10:43
And we look, and we find, as you would expect,
236
643260
3000
ืžืกืชื›ืœื™ื, ื•ืžื•ืฆืื™ื ื›ืžืฆื•ืคื”,
10:46
a contrast in the landscape
237
646260
2000
ืฉื™ื ื•ื™ ื‘ืชื•ื•ืื™-ืฉื˜ื—
10:48
where the word "bye" occurs much more in a structured way.
238
648260
3000
ื‘ื• ื”ืžื™ืœื” "ืฉืœื•ื" ืžืชืจื—ืฉืช ื”ืจื‘ื” ื™ื•ืชืจ ื‘ืื•ืคืŸ ืžื•ื‘ื ื”.
10:51
So we're using these structures
239
651260
2000
ื›ืš ืฉืื ื• ืžืฉืชืžืฉื™ื ื‘ืžื‘ื ื™ื ื”ืœืœื•
10:53
to start predicting
240
653260
2000
ื›ื“ื™ ืœื—ื–ื•ืช
10:55
the order of language acquisition,
241
655260
3000
ืืช ื”ืกื“ืจ ื‘ื• ื”ืฉืคื” ื ืจื›ืฉืช,
10:58
and that's ongoing work now.
242
658260
2000
ื•ื–ื• ื”ืขื‘ื•ื“ื” ืฉื ืขืฉื™ืช ื ื›ื•ืŸ ืœื”ื™ื•ื.
11:00
In my lab, which we're peering into now, at MIT --
243
660260
3000
ื‘ืžืขื‘ื“ื” ืฉืœื™, ืฉืื ื• ืžืฆื™ืฆื™ื ืืœื™ื” ืขื›ืฉื™ื•, ื‘-MIT --
11:03
this is at the media lab.
244
663260
2000
ื–ื” ื‘ืžืขื‘ื“ืช ืชืงืฉื•ืจืช.
11:05
This has become my favorite way
245
665260
2000
ื–ื• ื”ื“ืจืš ื”ืื”ื•ื‘ื” ืขืœื™ื™
11:07
of videographing just about any space.
246
667260
2000
ืœื”ืงืœื™ื˜ ืขืœ ื•ื™ื“ืื• ื—ืœืœ ื›ืœืฉื”ื•.
11:09
Three of the key people in this project,
247
669260
2000
ืฉืœื•ืฉืช ืื ืฉื™ ื”ืžืคืชื— ื‘ืžื™ื–ื ื–ื”,
11:11
Philip DeCamp, Rony Kubat and Brandon Roy are pictured here.
248
671260
3000
ืžืฆื•ืœืžื™ื ื›ืืŸ ืคื™ืœื™ืค ื“ื”-ืงืืžืค, ืจื•ื ื™ ืงื•ื‘ืื˜ ื•ื‘ืจื ื“ื•ืŸ ืจื•ื™.
11:14
Philip has been a close collaborator
249
674260
2000
ืคื™ืœื™ืค ื”ื•ื ืฉื•ืชืฃ ืงืจื•ื‘ ื‘ื™ืฆื™ืจืช
11:16
on all the visualizations you're seeing.
250
676260
2000
ื›ืœ ื”ื”ืžื—ืฉื•ืช ื”ื—ื–ื•ืชื™ื•ืช ืฉืืชื ืจื•ืื™ื.
11:18
And Michael Fleischman
251
678260
3000
ื•ืžื™ื™ืงืœ ืคืœื™ื™ืฉืžืŸ
11:21
was another Ph.D. student in my lab
252
681260
2000
ื”ื™ื” ื“ื•ืงื˜ื•ืจื ื˜ ื ื•ืกืฃ ื‘ืžืขื‘ื“ื”
11:23
who worked with me on this home video analysis,
253
683260
3000
ืืฉืจ ืขื‘ื“ ืื™ืชื™ ืขืœ ืื ืœื™ื–ืช ื•ื™ื“ืื• ื‘ื™ืชื™ ื–ื”,
11:26
and he made the following observation:
254
686260
3000
ื•ื”ื•ื ื”ื‘ื—ื™ืŸ ื‘ื“ื‘ืจ ื”ื‘ื:
11:29
that "just the way that we're analyzing
255
689260
2000
ืฉ"ื‘ื“ื™ื•ืง ื‘ืื•ืชื• ืื•ืคืŸ ื‘ื• ืื ื• ืžื ืชื—ื™ื
11:31
how language connects to events
256
691260
3000
ื›ื™ืฆื“ ืฉืคื” ืžืชืงืฉืจืช ืœืื™ืจื•ืขื™ื
11:34
which provide common ground for language,
257
694260
2000
ื”ืžืขื ื™ืงื™ื ื‘ืกื™ืก ืžืฉื•ืชืฃ ืœืฉืคื”,
11:36
that same idea we can take out of your home, Deb,
258
696260
4000
ื ื™ืชืŸ ืœืงื—ืช ืืช ืื•ืชื• ืจืขื™ื•ืŸ ืืœ ืžื—ื•ืฅ ืœื‘ื™ืชืš, ื“ื‘,
11:40
and we can apply it to the world of public media."
259
700260
3000
ื•ื ื™ืชืŸ ืœื™ื™ืฉืžื• ืœืขื•ืœื ืฉืœ ืชืงืฉื•ืจืช ื”ื”ืžื•ื ื™ื."
11:43
And so our effort took an unexpected turn.
260
703260
3000
ื•ื›ืš ืžืืžืฆื™ื ื• ืงื™ื‘ืœื• ืชืคื ื™ืช ืœื ืฆืคื•ื™ื”.
11:46
Think of mass media
261
706260
2000
ืชื—ืฉื‘ื• ืขืœ ืชืงืฉื•ืจืช ื”ืžื•ื ื™ื
11:48
as providing common ground
262
708260
2000
ื‘ืชื•ืจ ื›ื–ื• ื”ืžืขื ื™ืงื” ื‘ืกื™ืก ืžืฉื•ืชืฃ
11:50
and you have the recipe
263
710260
2000
ื•ื™ืฉ ืœื›ื ื”ืžืชื›ื•ืŸ
11:52
for taking this idea to a whole new place.
264
712260
3000
ืœืœืงื™ื—ืช ืจืขื™ื•ืŸ ื–ื” ืืœ ืžืงื•ื ืื—ืจ ืœื’ืžืจื™.
11:55
We've started analyzing television content
265
715260
3000
ื”ืชื—ืœื ื• ืœื ืชื— ืชื•ื›ืŸ ื˜ืœื•ื•ื™ื–ื™ื•ื ื™
11:58
using the same principles --
266
718260
2000
ืชื•ืš ืฉื™ืžื•ืฉ ื‘ืื•ืชื ื”ืขืงืจื•ื ื•ืช --
12:00
analyzing event structure of a TV signal --
267
720260
3000
ื ื™ืชื•ื— ืžื‘ื ื” ืื™ืจื•ืข ืฉืœ ืื•ืช ื˜ืœื•ื•ื™ื–ื™ื” --
12:03
episodes of shows,
268
723260
2000
ืคืจืงื™ ืกื“ืจื•ืช,
12:05
commercials,
269
725260
2000
ืคืจืกื•ืžื•ืช,
12:07
all of the components that make up the event structure.
270
727260
3000
ื›ืœ ื”ืžืจื›ื™ื‘ื™ื ืฉื™ื•ืฆืจื™ื ืžื‘ื ื” ืื™ืจื•ืข.
12:10
And we're now, with satellite dishes, pulling and analyzing
271
730260
3000
ื•ื›ื™ื•ื ืื ื• ืžื•ืฉื›ื™ื ื•ืžื ืชื—ื™ื ื‘ืขื–ืจืช ืฆืœื—ื•ืช ืœื•ื•ื™ื™ื ื™ื
12:13
a good part of all the TV being watched in the United States.
272
733260
3000
ื—ืœืง ื’ื“ื•ืœ ืžื›ืœ ืชื•ื›ื ื™ื•ืช ื”ื˜ืœื•ื•ื™ื–ื™ื” ื”ื ืฆืคื•ืช ื‘ืืจื”"ื‘.
12:16
And you don't have to now go and instrument living rooms with microphones
273
736260
3000
ื•ืื™ืŸ ืฆื•ืจืš ืœืœื›ืช ื•ืœืจืฉืช ืขื ืžื™ืงืจื•ืคื•ื ื™ื ืกืœื•ื ื™ื ื‘ื‘ืชื™ื
12:19
to get people's conversations,
274
739260
2000
ื›ื“ื™ ืœืงืœื•ื˜ ืฉื™ื—ื•ืช ื‘ื™ืŸ ืื ืฉื™ื,
12:21
you just tune into publicly available social media feeds.
275
741260
3000
ืืœื ืจืง ืœื”ืชื—ื‘ืจ ืœืžืฉื•ื‘ื™ื ื—ื‘ืจืชื™ื™ื ื”ื–ืžื™ื ื™ื ืœื›ืœ ื”ืฆื™ื‘ื•ืจ.
12:24
So we're pulling in
276
744260
2000
ื›ืš ืฉืื ื• ืžื•ืฉื›ื™ื
12:26
about three billion comments a month,
277
746260
2000
ื›-3 ืžื™ืœื™ืืจื“ ืชื’ื•ื‘ื•ืช ื‘ื—ื•ื“ืฉ.
12:28
and then the magic happens.
278
748260
2000
ื•ืื– ืžืชืจื—ืฉ ื”ืงืกื.
12:30
You have the event structure,
279
750260
2000
ื™ืฉ ืืช ืžื‘ื ื” ื”ืื™ืจื•ืข,
12:32
the common ground that the words are about,
280
752260
2000
ื”ื‘ืกื™ืก ื”ืžืฉื•ืชืฃ ืฉื”ืžื™ืœื™ื ื ืกื‘ื•ืช ืกื‘ื™ื‘ื•,
12:34
coming out of the television feeds;
281
754260
3000
ื”ื™ื•ืฆื ืžืชื•ืš ื”ื˜ืœื•ื•ื™ื–ื™ื”;
12:37
you've got the conversations
282
757260
2000
ื™ืฉ ืืช ื”ืฉื™ื—ื•ืช
12:39
that are about those topics;
283
759260
2000
ื”ืงืฉื•ืจื•ืช ื‘ืื•ืชื ื”ื ื•ืฉืื™ื;
12:41
and through semantic analysis --
284
761260
3000
ื•ื‘ืืžืฆืขื•ืช ื ื™ืชื•ื— ืกืžื ื˜ื™ --
12:44
and this is actually real data you're looking at
285
764260
2000
ื•ื‘ืขืฆื ืืœื” ื”ื ื”ื ืชื•ื ื™ื ื”ืืžื™ืชื™ื™ื ืฉืืชื ืจื•ืื™ื ืขื›ืฉื™ื•
12:46
from our data processing --
286
766260
2000
ืžืชื•ืš ืชื”ืœื™ืš ืขื™ื‘ื•ื“ ื”ื ืชื•ื ื™ื ืฉืœื ื• --
12:48
each yellow line is showing a link being made
287
768260
3000
ื›ืœ ืงื• ืฆื”ื•ื‘ ืžืกืžืœ ืงื™ืฉื•ืจ ืฉื ืขืฉื”
12:51
between a comment in the wild
288
771260
3000
ื‘ื™ืŸ ืชื’ื•ื‘ื” ื‘ืชื•ืš ื”ืกื‘ืš
12:54
and a piece of event structure coming out of the television signal.
289
774260
3000
ื•ืฉื‘ื™ื‘ ืฉืœ ืžื‘ื ื” ืื™ืจื•ืข ืืฉืจ ื™ื•ืฆื ืžืชื•ืš ืื•ืช ื”ื˜ืœื•ื•ื™ื–ื™ื”.
12:57
And the same idea now
290
777260
2000
ื•ื›ืขืช ื ื™ืชืŸ ืœื‘ื ื•ืช
12:59
can be built up.
291
779260
2000
ืืช ืื•ืชื• ืจืขื™ื•ืŸ
13:01
And we get this wordscape,
292
781260
2000
ื•ืื ื• ืžืงื‘ืœื™ื ืืช ืชื•ื•ืื™-ื”ืžื™ืœื™ื ื”ื–ื”,
13:03
except now words are not assembled in my living room.
293
783260
3000
ืืœื ืฉื›ืขืช ื”ืžื™ืœื™ื ืื™ื ืŸ ืžื•ืจื›ื‘ื•ืช ื‘ืกืœื•ืŸ ืฉืœื™.
13:06
Instead, the context, the common ground activities,
294
786260
4000
ื‘ืžืงื•ื ื–ื”, ื”ื”ืงืฉืจ, ืคืขื•ืœื•ืช ื”ื ื•ื’ืขื•ืช ืœื‘ืกื™ืก ื”ืžืฉื•ืชืฃ,
13:10
are the content on television that's driving the conversations.
295
790260
3000
ื”ื•ื ืชื•ื›ืŸ ื˜ืœื•ื•ื™ื–ื™ื•ื ื™ ืืฉืจ ืžื“ืจื‘ืŸ ืฉื™ื—ื•ืช.
13:13
And what we're seeing here, these skyscrapers now,
296
793260
3000
ื•ืžื” ืฉืื ื• ืจื•ืื™ื ื›ืืŸ, ื’ื•ืจื“ื™-ื”ืฉื—ืงื™ื ื”ืœืœื•,
13:16
are commentary
297
796260
2000
ื”ื ืคืจืฉื ื•ืช ื”ืงืฉื•ืจื”
13:18
that are linked to content on television.
298
798260
2000
ืœืชื•ื›ืŸ ื˜ืœื•ื•ื™ื–ื™ื•ื ื™.
13:20
Same concept,
299
800260
2000
ืื•ืชื• ืจืขื™ื•ืŸ,
13:22
but looking at communication dynamics
300
802260
2000
ืื‘ืœ ื‘ื”ืชื‘ื•ื ื ื•ืช ืขืœ ื“ื™ื ืžื™ืงื” ืชืงืฉื•ืจืชื™ืช
13:24
in a very different sphere.
301
804260
2000
ื‘ืชื—ื•ื ืฉื•ื ื” ืœื’ืžืจื™.
13:26
And so fundamentally, rather than, for example,
302
806260
2000
ื•ืœื›ืŸ ื‘ืื•ืคืŸ ื‘ืกื™ืกื™, ื‘ืžืงื•ื, ืœื“ื•ื’ืžื,
13:28
measuring content based on how many people are watching,
303
808260
3000
ืžื“ื™ื“ืช ืชื•ื›ืŸ ืœืคื™ ื›ืžื” ืื ืฉื™ื ืฆื•ืคื™ื ื‘ื•,
13:31
this gives us the basic data
304
811260
2000
ื”ื“ื‘ืจ ื”ื–ื” ื ื•ืชืŸ ืœื ื• ื ืชื•ื ื™ื ื‘ืกื™ืกื™ื™ื
13:33
for looking at engagement properties of content.
305
813260
3000
ื”ืžืืคืฉืจื™ื ืœื”ืกืชื›ืœ ืขืœ ืžื™ื“ืช ื”ื”ืชืขืกืงื•ืช ื‘ืชื•ื›ืŸ.
13:36
And just like we can look at feedback cycles
306
816260
3000
ื•ื‘ื“ื™ื•ืง ื›ืžื• ืฉืื ื• ื™ื›ื•ืœื™ื ืœื”ืกืชื›ืœ ืขืœ ืžืขื’ืœื™ื ืฉืœ ืžืฉื•ื‘ื™ื
13:39
and dynamics in a family,
307
819260
3000
ื•ื“ื™ื ืžื™ืงื” ื‘ืชื•ืš ืžืฉืคื—ื”,
13:42
we can now open up the same concepts
308
822260
3000
ื›ืขืช ืื ื• ื™ื›ื•ืœื™ื ืœืคืชื•ื— ืืช ืื•ืชื ื”ืจืขื™ื•ื ื•ืช
13:45
and look at much larger groups of people.
309
825260
3000
ื•ืœื”ืชื‘ื•ื ืŸ ืขืœ ืงื‘ื•ืฆื•ืช ื”ืจื‘ื” ื™ื•ืชืจ ื’ื“ื•ืœื•ืช ืฉืœ ืื ืฉื™ื.
13:48
This is a subset of data from our database --
310
828260
3000
ื–ื•ื”ื™ ืชืช-ืžืขืจื›ืช ืฉืœ ื ืชื•ื ื™ื ืžื‘ืกื™ืก ื”ื ืชื•ื ื™ื ืฉืœื ื• --
13:51
just 50,000 out of several million --
311
831260
3000
ืจืง 50,000 ืžืชื•ืš ื›ืžื” ืžื™ืœื™ื•ื ื™ื --
13:54
and the social graph that connects them
312
834260
2000
ื•ื”ื’ืจืฃ ื”ื—ื‘ืจืชื™ ื”ืžืงืฉืจ ืื•ืชื
13:56
through publicly available sources.
313
836260
3000
ื‘ืืžืฆืขื•ืช ืžืงื•ืจื•ืช ื”ื–ืžื™ื ื™ื ืœื›ืœ.
13:59
And if you put them on one plain,
314
839260
2000
ื•ืื ืชื ื™ื—ื• ืื•ืชื ื‘ืžื™ืฉื•ืจ ืื—ื“,
14:01
a second plain is where the content lives.
315
841260
3000
ื”ืžื™ืฉื•ืจ ื”ืฉื ื™ ื”ื•ื ื”ื™ื›ืŸ ืฉืžื•ื ื— ื”ืชื•ื›ืŸ.
14:04
So we have the programs
316
844260
3000
ื›ืš ืฉื™ืฉ ืœื ื• ืืช ื”ืชื•ื›ื ื™ื•ืช
14:07
and the sporting events
317
847260
2000
ื•ืื™ืจื•ืขื™ ื”ืกืคื•ืจื˜
14:09
and the commercials,
318
849260
2000
ื•ื”ืคืจืกื•ืžื•ืช,
14:11
and all of the link structures that tie them together
319
851260
2000
ื•ืืช ื›ืœ ืžื‘ื ื™ ื”ืงื™ืฉื•ืจ ืืฉืจ ืงื•ืฉืจื™ื ืื•ืชื ื™ื—ื“
14:13
make a content graph.
320
853260
2000
ื•ื”ื™ื•ืฆืจื™ื ืชืจืฉื™ื ืชื•ื›ืŸ.
14:15
And then the important third dimension.
321
855260
4000
ื•ืื– ืžื’ื™ืข ื”ืžื™ืžื“ ื”ืฉืœื™ืฉื™ ื”ื—ืฉื•ื‘.
14:19
Each of the links that you're seeing rendered here
322
859260
2000
ื›ืœ ืื—ื“ ืžื”ืงื™ืฉื•ืจื™ื ืฉืืชื ืจื•ืื™ื ื›ืืŸ
14:21
is an actual connection made
323
861260
2000
ื”ื•ื ืงืฉืจ ืืžื™ืชื™ ืฉื ื•ืฆืจ
14:23
between something someone said
324
863260
3000
ื‘ื™ืŸ ืžืฉื”ื• ืฉืžื™ืฉื”ื• ืืžืจ
14:26
and a piece of content.
325
866260
2000
ืœื‘ื™ืŸ ืงื˜ืข ืฉืœ ืชื•ื›ืŸ.
14:28
And there are, again, now tens of millions of these links
326
868260
3000
ื•ื”ื ื” ืฉื•ื‘, ืขื›ืฉื™ื• ืขืฉืจื•ืช ืžื™ืœื™ื•ื ื™ ืงื™ืฉื•ืจื™ื ื›ืืœื”
14:31
that give us the connective tissue of social graphs
327
871260
3000
ืฉื ื•ืชื ื™ื ืœื ื• ืืช ื”ืจื™ืงืžื” ื”ื—ื™ื‘ื•ืจื™ืช ืฉืœ ื’ืจืคื™ื ื—ื‘ืจืชื™ื™ื
14:34
and how they relate to content.
328
874260
3000
ื•ื”ืื•ืคืŸ ื‘ื• ื”ื ืงืฉื•ืจื™ื ืœืชื•ื›ืŸ.
14:37
And we can now start to probe the structure
329
877260
2000
ื•ื›ืขืช ืื ื• ื™ื›ื•ืœื™ื ืœื”ืชื—ื™ืœ ืœื—ืงื•ืจ ืืช ื”ืžื‘ื ื”
14:39
in interesting ways.
330
879260
2000
ื‘ื“ืจื›ื™ื ืžืขื ื™ื™ื ื•ืช.
14:41
So if we, for example, trace the path
331
881260
3000
ื›ืš ืฉืื ืื ื• ืœื“ื•ื’ืžื ืขื•ืงื‘ื™ื ืื—ืจ ื ืชื™ื‘
14:44
of one piece of content
332
884260
2000
ืฉืœ ืงื˜ืข ืชื•ื›ืŸ ืื—ื“
14:46
that drives someone to comment on it,
333
886260
2000
ืืฉืจ ื’ื•ืจื ืœืžื™ืฉื”ื• ืœื”ืขื™ืจ ืขืœื™ื•,
14:48
and then we follow where that comment goes,
334
888260
3000
ื•ืื– ืื ื• ืขื•ืงื‘ื™ื ืœื”ื™ื›ืŸ ืžื’ื™ืขื” ืื•ืชื” ื”ืขืจื”,
14:51
and then look at the entire social graph that becomes activated
335
891260
3000
ื•ืื—ืจ-ื›ืš ืžืกืชื›ืœื™ื ืขืœ ื›ืœ ื”ืชืจืฉื™ื ื”ื—ื‘ืจืชื™ ืฉืžื•ืคืขืœ
14:54
and then trace back to see the relationship
336
894260
3000
ื•ืœืื—ืจ-ืžื›ืŸ ืขื•ืงื‘ื™ื ื›ื“ื™ ืœืจืื•ืช ืืช ื”ืงืฉืจ
14:57
between that social graph and content,
337
897260
2000
ื‘ื™ืŸ ืื•ืชื• ืชืจืฉื™ื ื—ื‘ืจืชื™ ืœื‘ื™ืŸ ืชื•ื›ืŸ,
14:59
a very interesting structure becomes visible.
338
899260
2000
ืžื•ืคื™ืข ืžื•ืœ ืขื™ื ื™ื ื• ืžื‘ื ื” ืžืื•ื“ ืžืขื ื™ื™ืŸ.
15:01
We call this a co-viewing clique,
339
901260
2000
ืื ื• ืžื›ื ื™ื ืื•ืชื• ืžื•ืขื“ื•ืŸ ืฆืคื™ื™ื”-ืžืฉื•ืชืคืช,
15:03
a virtual living room if you will.
340
903260
3000
ืกืœื•ืŸ ื•ื™ืจื˜ื•ืืœื™ ืื ืชืจืฆื•.
15:06
And there are fascinating dynamics at play.
341
906260
2000
ื“ื™ื ืžื™ืงื” ืžืจืชืงืช ืžืžืœืืช ื›ืืŸ ืชืคืงื™ื“.
15:08
It's not one way.
342
908260
2000
ื–ื• ืœื ื“ืจืš ื—ื“-ืกื™ื˜ืจื™ืช.
15:10
A piece of content, an event, causes someone to talk.
343
910260
3000
ืงื˜ืข ืฉืœ ืชื•ื›ืŸ, ืื™ืจื•ืข, ื’ื•ืจื ืœืžื™ืฉื”ื• ืœื“ื‘ืจ.
15:13
They talk to other people.
344
913260
2000
ื”ื ืžื“ื‘ืจื™ื ืœืื ืฉื™ื ืื—ืจื™ื.
15:15
That drives tune-in behavior back into mass media,
345
915260
3000
ื•ื–ื” ื’ื•ืจื ืœื”ืชื ื”ื’ื•ืช ืžื›ื•ื•ื ื ืช ื”ื—ื•ื–ืจืช ืœืชืงืฉื•ืจืช ื”ืžื•ื ื™ืช,
15:18
and you have these cycles
346
918260
2000
ื•ืื– ืžืงื‘ืœื™ื ืืช ื”ืžืขื’ืœื™ื ื”ืœืœื•
15:20
that drive the overall behavior.
347
920260
2000
ืฉืžื ื™ืขื™ื ืืช ื”ื”ืชื ื”ื’ื•ืช ื”ื›ื•ืœืœืช.
15:22
Another example -- very different --
348
922260
2000
ื“ื•ื’ืžื ืื—ืจืช -- ืฉื•ื ื” ืžืื•ื“ --
15:24
another actual person in our database --
349
924260
3000
ืื“ื ืื—ืจ ื‘ื‘ืกื™ืก ื”ื ืชื•ื ื™ื ืฉืœื ื• --
15:27
and we're finding at least hundreds, if not thousands, of these.
350
927260
3000
ื•ืื ื• ืžื•ืฆืื™ื ืœืคื—ื•ืช ืžืื•ืช, ืื ืœื ืืœืคื™ื ื›ืืœื”.
15:30
We've given this person a name.
351
930260
2000
ื•ื ืชื ื• ืฉื ืœืื“ื ื–ื”.
15:32
This is a pro-amateur, or pro-am media critic
352
932260
3000
ื–ื”ื• ืžืงืฆื•ืขืŸ-ื—ื•ื‘ื‘, ืžื‘ืงืจ ืชืงืฉื•ืจืช
15:35
who has this high fan-out rate.
353
935260
3000
ืฉืœื“ืขื•ืชื™ื• ื™ืฉ ืคืจื™ืกื” ืจื—ื‘ื”.
15:38
So a lot of people are following this person -- very influential --
354
938260
3000
ื›ืš ืฉื”ืจื‘ื” ืื ืฉื™ื ืขื•ืงื‘ื™ื ืื—ืจ ืื“ื ื–ื” -- ืžืฉืคื™ืข ืžืื•ื“ --
15:41
and they have a propensity to talk about what's on TV.
355
941260
2000
ื•ื™ืฉ ืœื”ื ื ื˜ื™ื™ื” ืœื“ื‘ืจ ืขืœ ืžื” ืžืฉื•ื“ืจ ื‘ื˜ืœื•ื•ื™ื–ื™ื”.
15:43
So this person is a key link
356
943260
3000
ื›ืš ืฉืื“ื ื–ื” ื”ื•ื ื—ื•ืœื™ื™ืช ืžืคืชื—
15:46
in connecting mass media and social media together.
357
946260
3000
ื”ืžืงืฉืจืช ืชืงืฉื•ืจืช ื”ืžื•ื ื™ืช ื•ืชืงืฉื•ืจืช ื—ื‘ืจืชื™ืช ื‘ื™ื—ื“.
15:49
One last example from this data:
358
949260
3000
ื“ื•ื’ืžื ืื—ืช ืื—ืจื•ื ื” ืœื ืชื•ื ื™ื ืืœื”:
15:52
Sometimes it's actually a piece of content that is special.
359
952260
3000
ืœืคืขืžื™ื ื–ื” ื‘ืขืฆื ืงื˜ืข ืชื•ื›ืŸ ืฉื”ื™ื ื• ืžื™ื•ื—ื“.
15:55
So if we go and look at this piece of content,
360
955260
4000
ื›ืš ืฉืื ืžืกืชื›ืœื™ื ืขืœ ืงื˜ืข ืชื•ื›ืŸ ื–ื”,
15:59
President Obama's State of the Union address
361
959260
3000
ื ืื•ืžื• ืฉืœ ื”ื ืฉื™ื ืื•ื‘ืืžื” ืขืœ ืžืฆื‘ ื”ืื•ืžื”
16:02
from just a few weeks ago,
362
962260
2000
ืœืคื ื™ ืžืกืคืจ ืฉื‘ื•ืขื•ืช,
16:04
and look at what we find in this same data set,
363
964260
3000
ื•ืžืกืชื›ืœื™ื ืขืœ ืžื” ืฉืžื•ืฆืื™ื ื‘ืื•ืชื” ืžืขืจื›ืช ื ืชื•ื ื™ื,
16:07
at the same scale,
364
967260
3000
ื‘ืื•ืชื• ืงื ื”-ืžื™ื“ื”,
16:10
the engagement properties of this piece of content
365
970260
2000
ืžื™ื“ืช ื”ื”ืชืขืกืงื•ืช ื‘ืชื•ื›ืŸ ื”ื–ื”
16:12
are truly remarkable.
366
972260
2000
ื”ื™ื ื‘ืืžืช ืžืจืฉื™ืžื”.
16:14
A nation exploding in conversation
367
974260
2000
ืื•ืžื” ื”ืžืชืคื•ืฆืฆืช ื‘ืฉื™ื—ื•ืช
16:16
in real time
368
976260
2000
ื‘ื–ืžืŸ ืืžืช
16:18
in response to what's on the broadcast.
369
978260
3000
ื‘ืชื’ื•ื‘ื” ืœืžื” ืฉืžืฉื•ื“ืจ ื‘ืื•ืชื• ืจื’ืข.
16:21
And of course, through all of these lines
370
981260
2000
ื•ื›ืžื•ื‘ืŸ, ืขืœ-ื’ื‘ื™ ื›ืœ ื”ืฉื•ืจื•ืช ื”ืœืœื•
16:23
are flowing unstructured language.
371
983260
2000
ื–ื•ืจืžืช ืฉืคื” ืœื-ืžื•ื‘ื ื™ืช.
16:25
We can X-ray
372
985260
2000
ืื ื• ื™ื›ื•ืœื™ื ืœืฆืœื ืชืžื•ื ืช ืขื•ืžืง
16:27
and get a real-time pulse of a nation,
373
987260
2000
ื•ืœืงื‘ืœ ืืช ื“ื•ืคืง ื”ืื•ืžื” ื‘ื–ืžืŸ ืืžื™ืชื™,
16:29
real-time sense
374
989260
2000
ืชื—ื•ืฉื” ื‘ื–ืžืŸ ืืžื™ืชื™
16:31
of the social reactions in the different circuits in the social graph
375
991260
3000
ืฉืœ ื”ืชื’ื•ื‘ื•ืช ื”ืฆื™ื‘ื•ืจื™ื•ืช ื‘ืžืขื’ืœื™ื ื”ืฉื•ื ื™ื ื‘ืชืจืฉื™ื ื”ื—ื‘ืจืชื™
16:34
being activated by content.
376
994260
3000
ืืฉืจ ืžืชืขื•ืจืจื•ืช ื‘ื’ืœืœ ื”ืชื•ื›ืŸ.
16:37
So, to summarize, the idea is this:
377
997260
3000
ืœืกื™ื›ื•ื, ื”ืจืขื™ื•ืŸ ื”ื•ื ื›ื–ื”:
16:40
As our world becomes increasingly instrumented
378
1000260
3000
ื›ื›ืœ ืฉื”ืขื•ืœื ื”ื•ืคืš ืœืžืžื•ื›ืฉืจ ื™ื•ืชืจ
16:43
and we have the capabilities
379
1003260
2000
ื•ื™ืฉ ืœื ื• ื”ื™ื›ื•ืœื•ืช
16:45
to collect and connect the dots
380
1005260
2000
ืœืืกื•ืฃ ื•ืœื—ื‘ืจ ื‘ื™ืŸ
16:47
between what people are saying
381
1007260
2000
ืžื” ืฉืื ืฉื™ื ืื•ืžืจื™ื
16:49
and the context they're saying it in,
382
1009260
2000
ืœื‘ื™ืŸ ื”ื”ืงืฉืจ ืœื’ื‘ื™ื• ื”ื ืื•ืžืจื™ื ื–ืืช,
16:51
what's emerging is an ability
383
1011260
2000
ืžื” ืฉืขื•ืœื” ื•ืฆื•ืžื— ื”ื™ื ื”ื™ื›ื•ืœืช
16:53
to see new social structures and dynamics
384
1013260
3000
ืœื”ื‘ื—ื™ืŸ ื‘ืžื‘ื ื™ื ื—ื‘ืจืชื™ื™ื ื•ื“ื™ื ืžื™ืงื•ืช ื—ื“ืฉื™ื
16:56
that have previously not been seen.
385
1016260
2000
ืฉืœื ื ืจืื• ืงื•ื“ื ืœื›ืŸ.
16:58
It's like building a microscope or telescope
386
1018260
2000
ื–ื” ื›ืžื• ืœื‘ื ื•ืช ืžื™ืงืจื•ืกืงื•ืค ืื• ื˜ืœืกืงื•ืค
17:00
and revealing new structures
387
1020260
2000
ื•ืœื’ืœื•ืช ืžื‘ื ื™ื ื—ื“ืฉื™ื
17:02
about our own behavior around communication.
388
1022260
3000
ืฉืœ ื”ืชื ื”ื’ื•ืชื ื• ื”ื ื•ืฆืจื™ื ืกื‘ื™ื‘ ืชืงืฉื•ืจืช.
17:05
And I think the implications here are profound,
389
1025260
3000
ืื ื™ ืกื‘ื•ืจ ืฉื”ื”ืฉืœื›ื•ืช ื”ืŸ ืขืžื•ืงื•ืช,
17:08
whether it's for science,
390
1028260
2000
ื‘ื™ืŸ ืื ื–ื” ืœื’ื‘ื™ ืžื“ืข,
17:10
for commerce, for government,
391
1030260
2000
ืื• ืœื’ื‘ื™ ืžืกื—ืจ, ืื• ืœื’ื‘ื™ ืžืžืฉืœื”,
17:12
or perhaps most of all,
392
1032260
2000
ืื• ืื•ืœื™ ื™ื•ืชืจ ืžื”ื›ืœ,
17:14
for us as individuals.
393
1034260
3000
ืœื’ื‘ื™ื ื• ื›ื™ื—ื™ื“ื™ื.
17:17
And so just to return to my son,
394
1037260
3000
ืื– ืื ืœื—ื–ื•ืจ ืœื‘ื ื™,
17:20
when I was preparing this talk, he was looking over my shoulder,
395
1040260
3000
ื›ืืฉืจ ื”ื›ื ืชื™ ื”ืจืฆืื” ื–ื•, ื”ื•ื ื”ืกืชื›ืœ ืžืขื‘ืจ ืœื›ืชืคื™,
17:23
and I showed him the clips I was going to show to you today,
396
1043260
2000
ื•ื”ืจืืชื™ ืœื• ืืช ื”ืกืจื˜ื•ื ื™ื ืฉืื ื™ ื”ืจืืชื™ ืœื›ื ื›ืืŸ,
17:25
and I asked him for permission -- granted.
397
1045260
3000
ื•ื‘ื™ืงืฉืชื™ ืืช ืจืฉื•ืชื• -- ืฉื ื™ืชื ื”.
17:28
And then I went on to reflect,
398
1048260
2000
ื•ืื– ื”ืžืฉื›ืชื™ ืœื”ืจื”ืจ ื‘ืงื•ืœ ืจื,
17:30
"Isn't it amazing,
399
1050260
3000
"ื”ืื™ืŸ ื–ื” ืžื“ื”ื™ื,
17:33
this entire database, all these recordings,
400
1053260
3000
ืฉื›ืœ ื‘ืกื™ืก ื”ื ืชื•ื ื™ื ื”ื–ื”, ื›ืœ ื”ื”ืงืœื˜ื•ืช ื”ืืœื•,
17:36
I'm going to hand off to you and to your sister" --
401
1056260
2000
ืื ื™ ื”ื•ืœืš ืœืžืกื•ืจ ืœืš ื•ืœืื—ื•ืชืš,"
17:38
who arrived two years later --
402
1058260
3000
ืฉื”ื’ื™ืขื” ืœืขื•ืœื ืฉื ืชื™ื™ื ื™ื•ืชืจ ืžืื•ื—ืจ.
17:41
"and you guys are going to be able to go back and re-experience moments
403
1061260
3000
"ื•ืืชื ืชื”ื™ื• ืžืกื•ื’ืœื™ื ืœืฉื•ื‘ ืื—ื•ืจื” ื•ืœื—ื•ื•ืช ืžื—ื“ืฉ ืจื’ืขื™ื
17:44
that you could never, with your biological memory,
404
1064260
3000
ืฉืœื ื”ื™ื™ืชื ื™ื›ื•ืœื™ื ืืฃ ืคืขื, ืขื ื”ื–ื™ื›ืจื•ืŸ ื”ื‘ื™ื•ืœื•ื’ื™ ืฉืœื›ื,
17:47
possibly remember the way you can now?"
405
1067260
2000
ืœื–ื›ื•ืจ ื‘ืื•ืคืŸ ื‘ื• ืืชื ื™ื›ื•ืœื™ื ื›ืขืช."
17:49
And he was quiet for a moment.
406
1069260
2000
ื•ื”ื•ื ื”ืฉืชืชืง ืœืจื’ืข.
17:51
And I thought, "What am I thinking?
407
1071260
2000
ื•ืื ื™ ื—ืฉื‘ืชื™, "ืžื” ืื ื™ ื—ื•ืฉื‘ ืœืขืฆืžื™?
17:53
He's five years old. He's not going to understand this."
408
1073260
2000
ื”ื•ื ืจืง ื‘ืŸ 5. ื”ื•ื ื‘ื˜ื— ืœื ื™ื‘ื™ืŸ ืืช ื–ื”."
17:55
And just as I was having that thought, he looked up at me and said,
409
1075260
3000
ื•ื‘ื“ื™ื•ืง ื›ืืฉืจ ื”ืžื—ืฉื‘ื” ื”ื–ื• ืขื‘ืจื” ื‘ืžื•ื—ื™, ื”ื•ื ื”ืจื™ื ืืช ืžื‘ื˜ื• ื•ืืžืจ,
17:58
"So that when I grow up,
410
1078260
2000
"ืื– ื›ืืฉืจ ืื ื™ ืื’ื“ืœ,
18:00
I can show this to my kids?"
411
1080260
2000
ืื•ื›ืœ ืœื”ืจืื•ืช ืืช ื–ื” ืœื™ืœื“ื™ื™?"
18:02
And I thought, "Wow, this is powerful stuff."
412
1082260
3000
ื•ืื ื™ ื—ืฉื‘ืชื™ ืœืขืฆืžื™, "ื•ื•ืื•, ื–ื” ืžืฉื”ื• ืขื•ืฆืžืชื™."
18:05
So I want to leave you
413
1085260
2000
ืื ื›ืš, ื‘ืจืฆื•ื ื™ ืœื”ืฉืื™ืจ ืืชื›ื
18:07
with one last memorable moment
414
1087260
2000
ืขื ืจื’ืข ืื—ื“ ืื—ืจื•ืŸ ื‘ืœืชื™-ื ืฉื›ื—
18:09
from our family.
415
1089260
3000
ื”ืงืฉื•ืจ ื‘ืžืฉืคื—ืชื™.
18:12
This is the first time our son
416
1092260
2000
ื–ื•ื”ื™ ื”ืคืขื ื”ืจืืฉื•ื ื” ืฉื‘ื ื™
18:14
took more than two steps at once --
417
1094260
2000
ืขืฉื” ื™ื•ืชืจ ืžืฉื ื™ ืฆืขื“ื™ื ืจืฆื•ืฃ --
18:16
captured on film.
418
1096260
2000
ื•ืฆื•ืœื ื‘ืžืฆืœืžื”.
18:18
And I really want you to focus on something
419
1098260
3000
ื•ืื ื™ ืจื•ืฆื” ืฉืชืชืจื›ื–ื• ื‘ืžืฉื”ื•
18:21
as I take you through.
420
1101260
2000
ื‘ืขื•ื“ ืื ื™ ืžื•ื‘ื™ืœ ืืชื›ื.
18:23
It's a cluttered environment; it's natural life.
421
1103260
2000
ื–ื• ืกื‘ื™ื‘ื” ืžื‘ื•ืœื’ื ืช; ืืœื” ื—ื™ื™ื ืจื’ื™ืœื™ื.
18:25
My mother's in the kitchen, cooking,
422
1105260
2000
ืื™ืžื™ ืžื‘ืฉืœืช ื‘ืžื˜ื‘ื—,
18:27
and, of all places, in the hallway,
423
1107260
2000
ื•ืžื›ืœ ื”ืžืงื•ืžื•ืช, ื‘ืžืกื“ืจื•ืŸ,
18:29
I realize he's about to do it, about to take more than two steps.
424
1109260
3000
ืื ื™ ืžื’ืœื” ืฉื”ื•ื ืขื•ืžื“ ืœืขืฉื•ืช ื–ืืช, ืœื‘ืฆืข ื™ื•ืชืจ ืžืฉื ื™ ืฆืขื“ื™ื.
18:32
And so you hear me encouraging him,
425
1112260
2000
ื•ืœื›ืŸ ืฉื•ืžืขื™ื ืื•ืชื™ ืžืขื•ื“ื“ ืื•ืชื•,
18:34
realizing what's happening,
426
1114260
2000
ืชื•ืคืก ืžื” ืงื•ืจื” ื›ืืŸ,
18:36
and then the magic happens.
427
1116260
2000
ื•ืื– ื”ืงืกื ืžืชืจื—ืฉ.
18:38
Listen very carefully.
428
1118260
2000
ื”ืงืฉื™ื‘ื• ื”ื™ื˜ื‘.
18:40
About three steps in,
429
1120260
2000
ืœืื—ืจ ื›ืฉืœื•ืฉื” ืฆืขื“ื™ื,
18:42
he realizes something magic is happening,
430
1122260
2000
ื”ื•ื ืžื‘ื™ืŸ ืฉื™ืฉ ื›ืืŸ ืื™ื–ื” ืงืกื ืฉืงื•ืจื”.
18:44
and the most amazing feedback loop of all kicks in,
431
1124260
3000
ื•ืื– ืžืชืคืจืฅ ืคื ื™ืžื” ื”ืžืฉื•ื‘ ื”ื›ื™ ืžื“ื”ื™ื ืžื›ื•ืœื,
18:47
and he takes a breath in,
432
1127260
2000
ื”ื•ื ืฉื•ืืฃ ืคื ื™ืžื”,
18:49
and he whispers "wow"
433
1129260
2000
ื•ืื– ืœื•ื—ืฉ "ื•ื•ืื•"
18:51
and instinctively I echo back the same.
434
1131260
4000
ื•ื‘ืื•ืคืŸ ืื™ื ืกื˜ื™ื ืงื˜ื™ื‘ื™ ื—ื•ื–ืจ ืžืžื ื™ ืžื™ืŸ ื”ื“ ื–ื”ื”.
18:56
And so let's fly back in time
435
1136260
3000
ืื– ื”ื‘ื” ื ืคืœื™ื’ ืื—ื•ืจื” ื‘ื–ืžืŸ
18:59
to that memorable moment.
436
1139260
2000
ืืœ ืื•ืชื• ืจื’ืข ื‘ืœืชื™ ื ืฉื›ื—.
19:05
(Video) DR: Hey.
437
1145260
2000
(ืกืจื˜ื•ืŸ) ื“.ืจ.: ื”ื™ื™.
19:07
Come here.
438
1147260
2000
ื‘ื•ื ืœื›ืืŸ.
19:09
Can you do it?
439
1149260
3000
ื”ืื ืชื•ื›ืœ ืœืขืฉื•ืช ื–ืืช?
19:13
Oh, boy.
440
1153260
2000
ืื•ื™.
19:15
Can you do it?
441
1155260
3000
ื”ืื ืชื•ื›ืœ ืœืขืฉื•ืช ื–ืืช?
19:18
Baby: Yeah.
442
1158260
2000
ืชื™ื ื•ืง: ื•ื•ืื•.
19:20
DR: Ma, he's walking.
443
1160260
3000
ื“.ืจ.: (ื•ื•ืื•). ืื™ืžื, ื”ื•ื ื”ื•ืœืš.
19:24
(Laughter)
444
1164260
2000
(ืฆื—ื•ืง)
19:26
(Applause)
445
1166260
2000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
19:28
DR: Thank you.
446
1168260
2000
ื“.ืจ.: ืชื•ื“ื” ืœื›ื.
19:30
(Applause)
447
1170260
15000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)

Original video on YouTube.com
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7