Ken Goldberg: 4 lessons from robots about being human

14,457 views ใƒป 2015-07-15

TED


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

00:00
Translator: Morton Bast Reviewer: Thu-Huong Ha
0
0
7000
ืžืชืจื’ื: Shlomo Adam ืžื‘ืงืจ: Ido Dekkers
00:12
I know this is going to sound strange,
1
12539
2819
ืื ื™ ื™ื•ื“ืข ืฉื–ื” ื™ืฉืžืข ืœื›ื ืžื•ื–ืจ,
00:15
but I think robots can inspire us to be better humans.
2
15382
4983
ืื‘ืœ ืœื“ืขืชื™ ืจื•ื‘ื•ื˜ื™ื ื™ื›ื•ืœื™ื ืœื”ืฉืคื™ืข ืขืœื™ื ื•
ืœื”ื™ื•ืช ื‘ื ื™-ืื“ื ื˜ื•ื‘ื™ื ื™ื•ืชืจ.
ื’ื“ืœืชื™ ื‘ื‘ื™ืช-ืœื—ื ืฉื‘ืคื ืกื™ืœื‘ื ื™ื”,
00:21
See, I grew up in Bethlehem, Pennsylvania,
3
21245
3094
00:24
the home of Bethlehem Steel.
4
24363
1938
ื‘ื™ืชื” ืฉืœ "ืคืœื“ืช ื‘ื™ืช-ืœื—ื".
00:26
My father was an engineer,
5
26991
2112
ืื‘ื™ ื”ื™ื” ืžื”ื ื“ืก,
ื•ื‘ื™ืœื“ื•ืชื™ ื”ื•ื ืœื™ืžื“ ืื•ืชื™
00:29
and when I was growing up, he would teach me how things worked.
6
29127
4319
ืื™ืš ื”ื“ื‘ืจื™ื ืขื•ื‘ื“ื™ื.
ื”ื™ื™ื ื• ื‘ื•ื ื™ื ื‘ื™ื—ื“ ืžื™ื–ืžื™ื,
00:33
We would build projects together,
7
33470
2094
00:35
like model rockets and slot cars.
8
35588
2768
ื›ืžื• ื“ื’ืžื™ื ืฉืœ ื˜ื™ืœื™ื ื•ืžื›ื•ื ื™ื•ืช ืžืกืœื•ืœ.
00:38
Here's the go-kart that we built together.
9
38732
2908
ื”ื ื” ื”ืขื’ืœื” ืฉื‘ื ื™ื ื• ื‘ื™ื—ื“.
ื–ื” ืื ื™, ืžืื—ื•ืจื™ ื”ื”ื’ื”,
00:42
That's me behind the wheel,
10
42208
1744
00:43
with my sister and my best friend at the time.
11
43976
2224
ืขื ืื—ื•ืชื™ ื•ืขื ื”ื—ื‘ืจ ื”ื›ื™ ื˜ื•ื‘ ืฉืœื™ ืื–,
00:47
And one day,
12
47875
1926
ื•ื™ื•ื ืื—ื“,
00:49
he came home, when I was about 10 years old,
13
49825
3083
ื”ื•ื ื‘ื ื”ื‘ื™ืชื”, ื›ืฉื”ื™ื™ืชื™ ื‘ืŸ 10,
00:52
and at the dinner table, he announced
14
52932
2626
ื•ื‘ื–ืžืŸ ืืจื•ื—ืช ื”ืขืจื‘ ื”ื•ื ื”ื›ืจื™ื–
00:55
that for our next project, we were going to build ...
15
55582
3712
ืฉืขื‘ื•ืจ ื”ืžื™ื–ื ื”ื—ื“ืฉ ืฉืœื ื• ืื ื• ื ื‘ื ื” ืจื•ื‘ื•ื˜.
00:59
a robot.
16
59318
1150
01:01
A robot.
17
61604
1155
ืจื•ื‘ื•ื˜.
ืื ื™ ื ื•ืจื ื”ืชืจื’ืฉืชื™,
01:03
Now, I was thrilled about this,
18
63120
2063
01:05
because at school, there was a bully named Kevin,
19
65207
3802
ื›ื™ ื‘ื‘ื™ืช ื”ืกืคืจ
ื”ื™ื” ื‘ืจื™ื•ืŸ ื‘ืฉื ืงื•ื•ื™ืŸ,
ืฉื”ื™ื” ื ื˜ืคืœ ืืœื™
01:09
and he was picking on me,
20
69033
1889
01:10
because I was the only Jewish kid in class.
21
70946
2390
ื›ื™ ื”ื™ื™ืชื™ ื”ื™ืœื“ ื”ื™ื”ื•ื“ื™ ื”ื™ื—ื™ื“ ื‘ื›ื™ืชื”.
01:13
So I couldn't wait to get started to work on this,
22
73910
2373
ืื– ื—ื™ื›ื™ืชื™ ื‘ืงื•ืฆืจ-ืจื•ื— ืœื”ืชื—ื™ืœ ืœืขื‘ื•ื“ ืขืœ ื–ื”
01:16
so I could introduce Kevin to my robot.
23
76307
2610
ื›ื“ื™ ืฉืื•ื›ืœ ืœื”ืจืื•ืช ืœืงื•ื•ื™ืŸ ืืช ื”ืจื•ื‘ื•ื˜ ืฉืœื™. [ืฆื—ื•ืง]
01:18
(Laughter)
24
78941
1030
01:19
(Robot noises)
25
79995
4243
[ืจืขืฉื™ ืจื•ื‘ื•ื˜]
01:29
(Laughter)
26
89552
1392
01:30
But that wasn't the kind of robot my dad had in mind.
27
90968
3849
ืื‘ืœ ืœื ืขืœ ืจื•ื‘ื•ื˜ ื›ื–ื” ืื‘ื™ ื—ืฉื‘.
01:34
(Laughter)
28
94841
1068
01:35
See, he owned a chromium-plating company,
29
95933
3632
ืชื‘ื™ื ื•, ื”ื™ืชื” ืœื• ื—ื‘ืจื” ืœืฆื™ืคื•ื™ื™ ื›ืจื•ื,
01:39
and they had to move heavy steel parts between tanks of chemicals.
30
99589
5873
ื•ื”ื™ื” ืขืœื™ื”ื ืœื”ืขื‘ื™ืจ
ื—ืœืงื™ ืคืœื“ื” ื›ื‘ื“ื™ื ื‘ื™ืŸ ืžื™ื›ืœื™ื ืฉืœ ื›ื™ืžื™ืงืœื™ื,
ื›ืš ืฉื”ื•ื ื ื–ืงืง ืœืจื•ื‘ื•ื˜ ืชืขืฉื™ื™ืชื™ ื›ืžื• ื–ื”
01:45
And so he needed an industrial robot like this,
31
105486
3373
01:48
that could basically do the heavy lifting.
32
108883
2103
ืฉื™ื›ื•ืœ ืœื‘ืฆืข ืืช ื”ื”ืจืžื” ื”ืงืฉื”.
01:51
But my dad didn't get the kind of robot he wanted, either.
33
111684
3591
ืื‘ืœ ื’ื ืื‘ื™ ืœื ืงื™ื‘ืœ ืืช ื”ืจื•ื‘ื•ื˜ ื‘ื• ืจืฆื”.
01:55
He and I worked on it for several years,
34
115870
2430
ื”ื•ื ื•ืื ื™ ืขื‘ื“ื ื• ืขืœ ื–ื” ื‘ืžืฉืš ื›ืžื” ืฉื ื™ื,
01:58
but it was the 1970s, and the technology that was available to amateurs
35
118324
4745
ืื‘ืœ ื–ื” ื”ื™ื” ื‘ืฉื ื•ืช ื”-70,
ื•ื”ื˜ื›ื ื•ืœื•ื’ื™ื” ืฉื—ื•ื‘ื‘ื™ื ื™ื›ื•ืœื™ื ืœื”ืฉื™ื’
ืขื•ื“ ืœื ื”ื™ืชื” ืงื™ื™ืžืช.
02:03
just wasn't there yet.
36
123093
1326
02:05
So Dad continued to do this kind of work by hand.
37
125992
2785
ืื– ืื‘ื ื”ืžืฉื™ืš ืœืขืฉื•ืช ื™ื“ื ื™ืช ืืช ื”ืขื‘ื•ื“ื” ื”ื”ื™ื,
02:09
And a few years later,
38
129595
1428
ื•ืื—ืจื™ ื›ืžื” ืฉื ื™ื,
02:11
he was diagnosed with cancer.
39
131730
1576
ื”ื•ื ืื•ื‘ื—ืŸ ื›ื—ื•ืœื” ืกืจื˜ืŸ.
02:15
You see,
40
135822
1570
ืืชื ืžื‘ื™ื ื™ื, ื”ืœืงื— ืฉืœื• ืžืŸ ื”ืจื•ื‘ื•ื˜ ืฉื ื™ืกื™ื ื• ืœื‘ื ื•ืช
02:17
what the robot we were trying to build was telling him
41
137416
2596
ืœื ื”ื™ื” ืงืฉื•ืจ ื‘ื”ืจืžืช ื“ื‘ืจื™ื ื›ื‘ื“ื™ื.
02:20
was not about doing the heavy lifting.
42
140036
2156
02:22
It was a warning
43
142216
1151
ืืœื ืื–ื”ืจื” ืžืคื ื™ ื”ื—ืฉื™ืคื” ืœื›ื™ืžื™ืงืœื™ื ื”ืจืขื™ืœื™ื.
02:23
about his exposure to the toxic chemicals.
44
143391
2504
ื”ื•ื ืœื ื—ืฉื‘ ืขืœ ื›ืš ื‘ื–ืžื ื•,
02:27
He didn't recognize that at the time,
45
147184
2293
02:29
and he contracted leukemia.
46
149501
1686
ื•ื”ื•ื ื—ืœื” ื‘ืกืจื˜ืŸ-ื“ื,
02:31
And he died at the age of 45.
47
151827
1892
ื•ื ืคื˜ืจ ื‘ื’ื™ืœ 45.
ื”ื™ื™ืชื™ ืžื•ื›ื”-ืฆืขืจ,
02:35
I was devastated by this.
48
155354
1776
02:37
And I never forgot the robot that he and I tried to build.
49
157703
2762
ื•ืžืขื•ืœื ืœื ืฉื›ื—ืชื™ ืืช ื”ืจื•ื‘ื•ื˜ ืฉื”ื•ื ื•ืื ื™ ื ื™ืกื™ื ื• ืœื‘ื ื•ืช.
02:42
When I was at college, I decided to study engineering, like him.
50
162195
3553
ื›ืฉื”ื™ื™ืชื™ ื‘ืงื•ืœื’', ื”ื—ืœื˜ืชื™ ืœืœืžื•ื“ ื”ื ื“ืกื”, ื›ืžื•ื”ื•.
ื”ืœื›ืชื™ ืœืื•ื ' "ืงืจื’ื ื™ ืžืœื•ืŸ" ื•ืขืฉื™ืชื™ ื“ื•ืงื˜ื•ืจื˜ ื‘ืจื•ื‘ื•ื˜ื™ืงื”.
02:47
And I went to Carnegie Mellon, and I earned my PhD in robotics.
51
167081
4548
02:51
I've been studying robots ever since.
52
171653
1909
ื•ืžืื– ื•ืขื“ ื”ื™ื•ื ืื ื™ ื—ื•ืงืจ ืืช ื ื•ืฉื ื”ืจื•ื‘ื•ื˜ื™ื.
02:54
So what I'd like to tell you about are four robot projects,
53
174721
4595
ืื– ืื ื™ ืจื•ืฆื” ืœืกืคืจ ืœื›ื
ืขืœ ืืจื‘ืขื” ืžื™ื–ืžื™ ืจื•ื‘ื•ื˜ื™ืงื”
ื•ืื™ืš ื”ื ื ืชื ื• ืœื™ ื”ืฉืจืื” ืœื”ื™ื•ืช ืื“ื ื˜ื•ื‘ ื™ื•ืชืจ.
02:59
and how they've inspired me to be a better human.
54
179340
3049
03:06
By 1993, I was a young professor at USC,
55
186206
5614
ื‘-1993 ื›ื‘ืจ ื”ื™ื™ืชื™ ืžืจืฆื” ืฆืขื™ืจ ื‘ืื•ื ื™ื‘ืจืกื™ื˜ืช ื“ืจื•ื-ืงืœื™ืคื•ืจื ื™ื”,
03:11
and I was just building up my own robotics lab,
56
191844
2818
ื•ื‘ื“ื™ื•ืง ืื– ื”ืงืžืชื™ ืžืขื‘ื“ืช ืจื•ื‘ื•ื˜ื™ืงื” ืžืฉืœื™,
03:14
and this was the year the World Wide Web came out.
57
194686
3221
ื•ื‘ืื•ืชื” ื”ืฉื ื” ื”ื•ืคื™ืขื” ืจืฉืช ื”ืื™ื ื˜ืจื ื˜.
03:18
And I remember my students were the ones who told me about it,
58
198487
2985
ื•ืื ื™ ื–ื•ื›ืจ ืฉื”ืกื˜ื•ื“ื ื˜ื™ื ืฉืœื™ ื”ื™ื• ืืœื”
ืฉืกื™ืคืจื• ืœื™ ืขืœ ื›ืš.
03:21
and we would -- we were just amazed.
59
201496
2317
ื•ื”ื™ื™ื ื• -- ื”ื™ื™ื ื• ืžืœืื™ ืคืœื™ืื”.
03:23
We started playing with this, and that afternoon,
60
203837
3383
ื”ืชื—ืœื ื• ืœื”ืฉืชืขืฉืข ืขื ื–ื”, ื•ื‘ืื•ืชื• ืขืจื‘,
03:27
we realized that we could use this new, universal interface
61
207244
4278
ื”ื‘ื ื• ืฉืื ื• ื™ื›ื•ืœื™ื ืœื”ืฉืชืžืฉ ื‘ืžืžืฉืง ื”ื—ื“ืฉ ื•ื”ืื•ื ื™ื‘ืจืกืœื™ ื”ื–ื”
03:31
to allow anyone in the world to operate the robot in our lab.
62
211546
4428
ืœืืคืฉืจ ืœื›ืœ ืื“ื ื‘ืขื•ืœื
ืœื”ืคืขื™ืœ ืืช ื”ืจื•ื‘ื•ื˜ ืฉื‘ืžืขื‘ื“ื” ืฉืœื ื•.
ืื– ื‘ืžืงื•ื ืจื•ื‘ื•ื˜ ืœื•ื—ื ืื• ืจื•ื‘ื•ื˜ ืชืขืฉื™ื™ืชื™,
03:37
So, rather than have it fight or do industrial work,
63
217339
4042
03:42
we decided to build a planter,
64
222413
2530
ื”ื—ืœื˜ื ื• ืœื‘ื ื•ืช ื’ื ืŸ,
03:44
put the robot into the center of it,
65
224967
1951
ื”ืฆื‘ื ื• ืืช ื”ืจื•ื‘ื•ื˜ ื‘ืืžืฆืข,
03:46
and we called it the Telegarden.
66
226942
1678
ืงืจืื ื• ืœื–ื” "ื˜ืœ-ื’ืŸ".
03:49
And we had put a camera in the gripper of the hand of the robot,
67
229736
4517
ื”ื ื—ื ื• ืžืฆืœืžื” ื‘ืžืœืคืช ืฉืœ ื™ื“ื•
ืฉืœ ื”ืจื•ื‘ื•ื˜, ื•ื›ืชื‘ื ื• ื›ืžื” ืงื‘ืฆื™-ืคืงื•ื“ื•ืช ืžื™ื•ื—ื“ื™ื
03:54
and we wrote some special scripts and software,
68
234277
2931
ื•ืงืฆืช ืชื•ื›ื ื”, ื›ืš ืฉื›ืœ ืื—ื“ ื‘ืขื•ืœื ื™ื›ื•ืœ ืœื’ืฉืช
03:57
so that anyone in the world could come in,
69
237232
2034
ื•ืข"ื™ ืงืœื™ืง ืขืœ ื”ืžืกืš
03:59
and by clicking on the screen,
70
239290
1784
04:01
they could move the robot around and visit the garden.
71
241098
3621
ืœื”ื ื™ืข ืืช ื”ืจื•ื‘ื•ื˜ ื‘ืžืจื—ื‘
ื•ืœื‘ืงืจ ื‘ื’ื™ื ื”.
04:05
But we also set up some other software
72
245416
3851
ืื‘ืœ ืคืจื˜ ืœื›ืš, ื‘ืขื–ืจืช ืชื•ื›ื ื” ื ื•ืกืคืช,
04:09
that lets you participate and help us water the garden, remotely.
73
249291
3916
ืื™ืคืฉืจื ื• ืœื›ื ืœื”ืฉืชืชืฃ ื•ืœืขื–ื•ืจ ืœื ื• ืœื”ืฉืงื•ืช ืืช ื”ื’ืŸ
ืžืจื—ื•ืง, ื•ืžื™ ืฉื”ืฉืงื” ื›ืžื” ืคืขืžื™ื,
04:13
And if you watered it a few times,
74
253643
2289
04:15
we'd give you your own seed to plant.
75
255956
2275
ื–ื›ื” ืœืฉืชื•ืœ ื–ืจืข ืžืฉืœื•.
04:19
Now, this was an engineering project,
76
259167
3476
ื–ื” ื”ื™ื” ืžื™ื–ื, ืžื™ื–ื ื”ื ื“ืกื™,
04:22
and we published some papers on the system design of it,
77
262667
4112
ื•ืคื™ืจืกืžื ื• ื›ืžื” ืžืืžืจื™ื ืœื’ื‘ื™ ื”ืชื›ื ื•ืŸ,
ืชื›ื ื•ืŸ ื”ืžืขืจื›ื•ืช ืฉืœื•, ืื‘ืœ ืจืื™ื ื• ื‘ื• ื’ื
04:26
but we also thought of it as an art installation.
78
266803
2645
ืžื™ืฆื‘ ืืžื ื•ืชื™.
04:30
It was invited, after the first year,
79
270638
2149
ื”ื•ื ื”ื•ื–ืžืŸ, ืื—ืจื™ ื”ืฉื ื” ื”ืจืืฉื•ื ื”,
04:32
by the Ars Electronica Museum in Austria,
80
272811
3020
ืข"ื™ ื”"ืืจืก ืืœืงื˜ืจื•ื ื™ืงื” ืžื•ื–ื™ืื•ื" ื”ืื•ืกื˜ืจื™
04:35
to have it installed in their lobby.
81
275855
2822
ืœื”ื™ื•ืช ืžื•ืชืงืŸ ืืฆืœื ื‘ื›ื ื™ืกื”,
ื•ืื ื™ ืฉืžื— ืœื•ืžืจ ืฉื”ื•ื ื ื•ืชืจ ืฉื, ืžืงื•ื•ืŸ,
04:39
And I'm happy to say, it remained online there, 24 hours a day,
82
279558
3987
24 ืฉืขื•ืช ื‘ื™ืžืžื”, ื‘ืžืฉืš ื›ืžืขื˜ ืชืฉืข ืฉื ื™ื.
04:43
for almost nine years.
83
283569
1831
04:46
That robot was operated by more people
84
286369
3775
ื”ืจื•ื‘ื•ื˜ ื”ื–ื” ื”ื•ืคืขืœ ืข"ื™ ื™ื•ืชืจ ืื ืฉื™ื
04:50
than any other robot in history.
85
290168
1897
ืžื›ืœ ืจื•ื‘ื•ื˜ ืื—ืจ ื‘ื”ื™ืกื˜ื•ืจื™ื”.
04:53
Now, one day,
86
293303
1627
ื•ื™ื•ื ืื—ื“
04:54
I got a call out of the blue from a student,
87
294954
3348
ืงื™ื‘ืœืชื™ ืฉื™ื—ืช ื˜ืœืคื•ืŸ ื‘ืœืชื™-ืฆืคื•ื™ื”
ืžืกื˜ื•ื“ื ื˜ ืื—ื“,
04:59
who asked a very simple but profound question.
88
299211
4158
ืฉื”ืฆื™ื’ ืœื™ ืฉืืœื” ืคืฉื•ื˜ื” ืžืื“ ืืš ืžืขืžื™ืงื”.
ื”ื•ื ืฉืืœ, "ื”ืื ื”ืจื•ื‘ื•ื˜ ืืžื™ืชื™?"
05:04
He said, "Is the robot real?"
89
304361
2826
05:08
Now, everyone else had assumed it was,
90
308599
2286
ื›ืœ ื”ื™ืชืจ ื”ื ื™ื—ื• ืฉื›ืŸ.
05:10
and we knew it was, because we were working with it.
91
310909
2516
ื•ืื ื• ื™ื“ืขื ื• ื–ืืช ื›ื™ ืขื‘ื“ื ื• ืื™ืชื•.
05:13
But I knew what he meant,
92
313449
1398
ืื‘ืœ ืื ื™ ื”ื‘ื ืชื™ ืืช ื›ื•ื•ื ืชื•,
05:14
because it would be possible
93
314871
1442
ื›ื™ ืืคืฉืจ ื‘ื”ื—ืœื˜ ืœืฆืœื ื›ืžื” ืชืžื•ื ื•ืช
05:16
to take a bunch of pictures of flowers in a garden
94
316337
2785
ืฉืœ ืคืจื—ื™ื ื‘ื’ื™ื ื” ื•ืœืืจื’ืŸ ืื•ืชื
05:19
and then, basically, index them in a computer system,
95
319146
3639
ื‘ืžืขืจื›ืช ืžื™ื—ืฉื•ื‘ ื‘ืื•ืคืŸ ืฉื™ื™ืจืื”
05:22
such that it would appear that there was a real robot,
96
322809
2627
ื›ืื™ืœื• ืฉืขื•ืฉื” ื–ืืช ืจื•ื‘ื•ื˜ ืืžื™ืชื™, ื‘ืฉืขื” ืฉื‘ืขืฆื ืื™ืŸ ื›ื–ื”.
05:25
when there wasn't.
97
325460
1150
05:27
And the more I thought about it,
98
327088
1524
ื•ื›ื›ืœ ืฉื—ืฉื‘ืชื™ ืขืœ ื›ืš, ืœื ืขืœืชื” ื‘ื“ืขืชื™
05:28
I couldn't think of a good answer for how he could tell the difference.
99
328636
3579
ืชืฉื•ื‘ื” ื˜ื•ื‘ื”, ืื™ืš ื”ื•ื ื™ื›ื•ืœ ืœื–ื”ื•ืช ืืช ื”ื”ื‘ื“ืœ.
ื‘ืขืจืš ืื– ื”ื•ืฆืขื” ืœื™ ืžืฉืจื”
05:32
This was right about the time that I was offered a position
100
332628
2797
ื›ืืŸ ื‘ื‘ืจืงืœื™.
05:35
here at Berkeley.
101
335449
1301
05:36
And when I got here,
102
336774
1645
ื›ืฉื”ื’ืขืชื™ ืœื›ืืŸ, ื—ื™ืคืฉืชื™ ืืช ื™ื•ื‘ืจื˜ ื“ืจื™ื™ืคื•ืก,
05:38
I looked up Hubert Dreyfus,
103
338443
2362
05:40
who's a world-renowned professor of philosophy,
104
340829
2927
ืคืจื•ืคืกื•ืจ ืœืคื™ืœื•ืกื•ืคื™ื” ื‘ืขืœ ืคืจืกื•ื ืขื•ืœืžื™,
05:44
And I talked with him about this and he said,
105
344582
2730
ื•ืฉื—ื—ืชื™ ืื™ืชื• ืขืœ ื›ืš, ื•ื”ื•ื ืืžืจ,
"ื–ื• ืื—ืช ื”ื‘ืขื™ื•ืช ื”ื›ื™ ื•ืชื™ืงื•ืช ื•ืžืจื›ื–ื™ื•ืช ื‘ืคื™ืœื•ืกื•ืคื™ื”.
05:47
"This is one of the oldest and most central problems in philosophy.
106
347336
4102
ื”ื™ื ื”ื•ืฆื’ื” ื›ื‘ืจ ืข"ื™ "ืื’ื•ื“ืช ื”ืกืคืงื ื™ื"
05:51
It goes back to the Skeptics and up through Descartes.
107
351462
4017
ื•ืขื“ ืœื“ืงืืจื˜.
05:55
It's the issue of epistemology,
108
355503
3182
ื–ื• ืกื•ื’ื™ื” ืžืชื—ื•ื ื”ืืคื™ืกื˜ืžื•ืœื•ื’ื™ื”,
05:58
the study of how do we know that something is true."
109
358709
2868
ืฉื—ื•ืงืจืช ืื™ืš ืื ื• ื™ื•ื“ืขื™ื ืฉืžืฉื”ื• ื”ื™ื ื• ืืžื™ืชื™."
06:02
So he and I started working together,
110
362625
2000
ืื– ื”ื•ื ื•ืื ื™ ื”ืชื—ืœื ื• ืœืขื‘ื•ื“ ื‘ื™ื—ื“,
06:04
and we coined a new term: "telepistemology,"
111
364649
2871
ื•ื˜ื‘ืขื ื• ืžื•ืฉื’ ื—ื“ืฉ: "ื˜ืœ-ืคื™ืกื˜ืžื•ืœื•ื’ื™ื”",
ื—ืงืจ ื”ื™ื“ื™ืขื” ืžืจื—ื•ืง.
06:08
the study of knowledge at a distance.
112
368757
2136
06:11
We invited leading artists, engineers and philosophers
113
371303
4491
ื”ื–ืžื ื• ืกื˜ื•ื“ื ื˜ื™ื, ืžื”ื ื“ืกื™ื ื•ืคื™ืœื•ืกื•ืคื™ื ื‘ื•ืœื˜ื™ื
ืœื›ืชื•ื‘ ืขืœ ื›ืš ืžืืžืจื™ื,
06:15
to write essays about this,
114
375818
1319
06:17
and the results are collected in this book from MIT Press.
115
377161
3704
ื•ื”ืชื•ืฆืื•ืช ื ืืกืคื• ื‘ืกืคืจ
ืฉื”ื•ืฆื ืœืื•ืจ ืข"ื™ "ืื-ืื™ื™-ื˜ื™ ืคืจืก".
06:21
So thanks to this student,
116
381959
2096
ืื– ื”ื•ื“ื•ืช ืœืื•ืชื• ืกื˜ื•ื“ื ื˜ ืฉื”ื˜ื™ืœ ืกืคืง
06:24
who questioned what everyone else had assumed to be true,
117
384079
3319
ื‘ืžื” ืฉื›ื•ืœื ืงื™ื‘ืœื• ื›ืžื•ื‘ืŸ ืžืืœื™ื•,
06:27
this project taught me an important lesson about life,
118
387422
3977
ืžื™ื–ื ื–ื” ืœื™ืžื“ ืื•ืชื™ ืœืงื— ื—ืฉื•ื‘ ื‘ื ื•ื’ืข ืœื—ื™ื™ื,
06:31
which is to always question assumptions.
119
391423
2987
ืฉื”ื•ื ืชืžื™ื“ ืœืคืงืคืง ื‘ื”ื ื—ื•ืช-ื™ืกื•ื“.
06:35
Now, the second project I'll tell you about
120
395807
2176
ื”ืžื™ื–ื ื”ืฉื ื™ ืขืœื™ื• ืืกืคืจ ืœื›ื
06:38
grew out of the Telegarden.
121
398007
1992
ืฆืžื— ืžื”"ื˜ืœ-ื’ืŸ"
06:40
As it was operating, my students and I were very interested
122
400023
2821
ื‘ืžื”ืœืš ื”ืคืขื™ืœื•ืช ืฉืœื•, ื”ืกื˜ื•ื“ื ื˜ื™ื ื•ืื ื™ ื”ืชืขื ื™ื™ื ื• ืžืื“
06:42
in how people were interacting with each other,
123
402868
2341
ืœืจืื•ืช ืื™ืš ืื ืฉื™ื ื™ืฆืจื• ืื™ื ื˜ืจืืงืฆื™ื•ืช ื‘ื™ื ื™ื”ื
06:45
and what they were doing with the garden.
124
405233
2057
ื•ืžื” ื”ื ืขืฉื• ื‘ื’ืŸ.
ืื– ื”ืชื—ืœื ื• ืœื—ืฉื•ื‘, ืžื” ืื ื”ืจื•ื‘ื•ื˜
06:47
So we started thinking:
125
407314
1151
06:48
what if the robot could leave the garden
126
408489
1968
ื™ื™ืฆื ืžื”ื’ืŸ ื•ื™ื™ืœืš
06:50
and go out into some other interesting environment?
127
410481
2814
ืœืื™ื–ื• ืกื‘ื™ื‘ื” ืžืขื ื™ื™ื ืช ืื—ืจืช?
06:53
Like, for example, what if it could go to a dinner party
128
413319
2741
ืœืžืฉืœ, ื™ื‘ื•ื ืœืกืขื•ื“ื” ื—ื’ื™ื’ื™ืช
ื‘ื‘ื™ืช ื”ืœื‘ืŸ? [ืฆื—ื•ืง]
06:56
at the White House?
129
416084
1461
06:57
(Laughter)
130
417569
1499
07:00
So, because we were interested more in the system design
131
420401
3569
ืื– ื‘ื’ืœืœ ืฉื”ืชืขื ื™ื™ื ื• ื™ื•ืชืจ ื‘ืชื›ื ื•ืŸ ื”ืžืขืจื›ืช
07:03
and the user interface than in the hardware,
132
423994
2943
ื•ื‘ืžืžืฉืง ื”ืžืฉืชืžืฉ ืžืืฉืจ ื‘ื—ื•ืžืจื”,
07:06
we decided that,
133
426961
1151
ื”ื—ืœื˜ื ื• ืฉื‘ืžืงื•ื
07:08
rather than have a robot replace the human to go to the party,
134
428136
4487
ืฉืจื•ื‘ื•ื˜ ื™ื™ืœืš ืœืื™ืจื•ืข ื‘ืžืงื•ื ื‘ืŸ-ืื“ื,
07:12
we'd have a human replace the robot.
135
432647
2230
ื™ื”ื™ื” ืื“ื ืฉื™ื—ืœื™ืฃ ืืช ื”ืจื•ื‘ื•ื˜.
07:15
We called it the Tele-Actor.
136
435679
1460
ืงืจืื ื• ืœื• "ื˜ืœ-ืฉื—ืงืŸ".
07:17
We got a human,
137
437978
2008
ื”ืฉื’ื ื• ื‘ืŸ-ืื“ื,
07:20
someone who's very outgoing and gregarious,
138
440010
3108
ืžื™ืฉื”ื™ ืžื•ื—ืฆื ืช ื•ื—ื‘ืจื•ืชื™ืช ืžืื“,
07:23
and she was outfitted with a helmet with various equipment,
139
443142
4255
ื”ืชืื™ืžื• ืœื” ืงืกื“ื”
ืขื ื›ืœ ืžื™ื ื™ ืคืจื™ื˜ื™ ืฆื™ื•ื“, ืžืฆืœืžื•ืช ื•ืžื™ืงืจื•ืคื•ื ื™ื,
07:27
cameras and microphones,
140
447421
1284
07:28
and then a backpack with wireless Internet connection.
141
448729
3065
ื•ืชืจืžื™ืœ-ื’ื‘ ืขื ื—ื™ื‘ื•ืจ ืื™ื ื˜ืจื ื˜ ืืœื—ื•ื˜ื™,
07:32
And the idea was that she could go
142
452882
2144
ื•ื”ืจืขื™ื•ืŸ ื”ื™ื” ืฉื”ื™ื ืชื•ื›ืœ ืœืœื›ืช ืœืื™ื–ื• ืกื‘ื™ื‘ื” ืžืจื•ื—ืงืช
07:35
into a remote and interesting environment,
143
455050
2450
ื•ืžืขื ื™ื™ื ืช, ื•ื“ืจืš ื”ืื™ื ื˜ืจื ื˜,
07:37
and then over the Internet,
144
457524
2017
07:39
people could experience what she was experiencing.
145
459565
2592
ืื ืฉื™ื ื™ื•ื›ืœื• ืœื—ื•ื•ืช ืืช ื”ื—ื•ื•ื™ื•ืช ืฉืœื”,
07:42
So they could see what she was seeing,
146
462736
2971
ืœืจืื•ืช ืืช ืžื” ืฉื”ื™ื ืจื•ืื”,
07:45
but then, more importantly, they could participate,
147
465731
3283
ื•ื™ื•ืชืจ ื—ืฉื•ื‘, ื™ื•ื›ืœื• ืœื”ืฉืชืชืฃ,
07:49
by interacting with each other and coming up with ideas
148
469038
4750
ืข"ื™ ืื™ื ื˜ืจืืงืฆื™ื•ืช ื‘ื™ื ื™ื”ื,
ื•ืœื”ืขืœื•ืช ืจืขื™ื•ื ื•ืช ืœื’ื‘ื™ ืžื” ืฉื”ื™ื ืชืขืฉื” ื‘ื”ืžืฉืš
07:53
about what she should do next and where she should go,
149
473812
3969
ื•ืœืืŸ ื›ื“ืื™ ืฉืชืœืš,
07:57
and then conveying those to the Tele-Actor.
150
477805
2206
ื•ื”ื ื™ืขื‘ื™ืจื• ืื•ืชื ืœื˜ืœ-ืฉื—ืงื ื™ืช.
08:01
So we got a chance to take the Tele-Actor
151
481069
2421
ื ืคืœื” ืœื™ื“ื™ื ื• ื”ื–ื“ืžื ื•ืช ืœืงื—ืช ืืช ื”ื˜ืœ-ืฉื—ืงื ื™ืช
08:03
to the Webby Awards in San Francisco.
152
483514
2808
ืœื˜ืงืก ืคืจืกื™ "ื•ื•ื‘ื™" ื‘ืกืŸ-ืคืจื ืฆื™ืกืงื•,
08:07
And that year, Sam Donaldson was the host.
153
487129
3240
ื•ื‘ืื•ืชื” ืฉื ื”, ืกื ื“ื•ื ืœื“ืกื•ืŸ ื”ื™ื” ื”ืžื ื—ื”.
ืžืžืฉ ืœืคื ื™ ืฉื”ืžืกืš ืขืœื”, ื”ื™ื• ืœื™ ื›-30 ืฉื ื™ื•ืช
08:12
Just before the curtain went up, I had about 30 seconds
154
492290
2690
ืœื”ืกื‘ื™ืจ ืœืžืจ ื“ื•ื ืœื“ืกื•ืŸ ืžื” ืื ื• ืขื•ืžื“ื™ื ืœืขืฉื•ืช,
08:15
to explain to Mr. Donaldson what we were going to do.
155
495004
3742
ื•ืืžืจืชื™, "ื”ื˜ืœ-ืฉื—ืงื ื™ืช
08:20
And I said, "The Tele-Actor is going to be joining you onstage.
156
500119
3517
ืชืฆื˜ืจืฃ ืืœื™ืš ืขืœ ื”ื‘ืžื”,
08:23
This is a new experimental project,
157
503660
2354
ื–ื”ื• ืžื™ื–ื ื ืกื™ื•ื ื™ ื—ื“ืฉ,
ืื ืฉื™ื ื™ืฆืคื• ื‘ื” ืขืœ ื”ืžืกื›ื™ื ืฉืœื”ื,
08:26
and people are watching her on their screens,
158
506038
2595
08:28
there's cameras involved and there's microphones
159
508657
3291
ื•ื™ืฉ ืœื”-- ื™ืฉ ื‘ืขืกืง ื›ืœ ืžื™ื ื™ ืžืฆืœืžื•ืช ื•ืžื™ืงืจื•ืคื•ื ื™ื
ื•ื™ืฉ ืœื” ืื•ื–ื ื™ื”,
08:31
and she's got an earbud in her ear,
160
511972
1786
08:33
and people over the network are giving her advice
161
513782
2317
ื•ืื ืฉื™ื ืžื”ืื™ื ื˜ืจื ื˜ ื™ืชื ื• ืœื” ืขืฆื•ืช
ืžื” ืœืขืฉื•ืช ื”ืœืื”."
08:36
about what to do next."
162
516123
1158
ื•ื”ื•ื ืืžืจ, "ืจืง ืจื’ืข,
08:37
And he said, "Wait a second.
163
517305
1395
08:39
That's what I do."
164
519806
1173
ื–ื• ื”ืขื‘ื•ื“ื” ืฉืœื™." [ืฆื—ื•ืง]
08:41
(Laughter)
165
521003
4977
08:46
So he loved the concept,
166
526004
1872
ื”ืจืขื™ื•ืŸ ืžืฆื ื—ืŸ ื‘ืขื™ื ื™ื•,
08:47
and when the Tele-Actor walked onstage, she walked right up to him,
167
527900
4253
ื•ื›ืฉื”ื˜ืœ-ืฉื—ืงื ื™ืช ืขืœืชื” ืœื‘ืžื”,
ื”ื™ื ื ื™ื’ืฉื” ื™ืฉืจ ืืœื™ื•, ื•ื”ืขื ื™ืงื” ืœื• ื ืฉื™ืงื” ื’ื“ื•ืœื”
08:52
and she gave him a big kiss right on the lips.
168
532177
2459
ืขืœ ื”ืฉืคืชื™ื™ื. [ืฆื—ื•ืง]
08:54
(Laughter)
169
534660
1952
08:56
We were totally surprised -- we had no idea that would happen.
170
536636
2922
ื”ื™ื™ื ื• ืžื•ืคืชืขื™ื ืœื’ืžืจื™.
ืœื ื”ื™ื” ืœื ื• ืžื•ืฉื’ ืฉื–ื” ื”ื•ืœืš ืœืงืจื•ืช.
08:59
And he was great, he just gave her a big hug in return,
171
539582
2618
ื•ื”ื•ื ื”ืชื ื”ื’ ืœืžื•ืคืช. ื”ื•ื ืคืฉื•ื˜ ื”ื—ื–ื™ืจ ืœื” ื‘ื—ื™ื‘ื•ืง ื’ื“ื•ืœ,
09:02
and it worked out great.
172
542224
1657
ื•ื”ื›ืœ ืขื‘ื“ ืžืขื•ืœื”.
09:03
But that night, as we were packing up,
173
543905
2040
ืื‘ืœ ื‘ืื•ืชื• ืœื™ืœื”, ื›ืฉืืจื–ื ื•,
09:05
I asked the Tele-Actor, how did the Tele-Directors decide
174
545969
4555
ืฉืืœืชื™ ืืช ื”ื˜ืœ-ืฉื—ืงื ื™ืช ืื™ืš ื”ื˜ืœ-ื‘ืžืื™ื
ื”ื—ืœื™ื˜ื• ืขืœ ื ืฉื™ืงื” ืœืกื ื“ื•ื ืœื“ืกื•ืŸ?
09:10
that they would give a kiss to Sam Donaldson?
175
550548
3000
09:15
And she said they hadn't.
176
555135
1457
ื”ื™ื ืขื ืชื” ืฉื”ื ืœื,
09:17
She said, when she was just about to walk onstage,
177
557274
2618
ืฉื›ืืฉืจ ื”ื™ื ืขืžื“ื” ืœืฆืืช ืœื‘ืžื”,
09:19
the Tele-Directors still were trying to agree on what to do,
178
559916
2974
ื”ื˜ืœ-ื‘ืžืื™ื ืขื“ื™ื™ืŸ ื ื™ืกื• ืœื”ื’ื™ืข ืœื”ืกื›ืžื” ืžื” ืขื•ืฉื™ื,
09:22
and so she just walked onstage and did what felt most natural.
179
562914
3100
ืื– ื”ื™ื ืคืฉื•ื˜ ื™ืฆืื” ืœื‘ืžื”
ื•ืขืฉืชื” ืžื” ืฉื ืจืื” ืœื” ื”ื›ื™ ื˜ื‘ืขื™. [ืฆื—ื•ืง]
09:26
(Laughter)
180
566038
4514
09:30
So, the success of the Tele-Actor that night
181
570576
3092
ื›ืš ืฉื”ื”ืฆืœื—ื” ืฉืœ ื”ื˜ืœ-ืฉื—ืงื ื™ืช ื‘ืื•ืชื• ืขืจื‘,
09:33
was due to the fact that she was a wonderful actor.
182
573692
4253
ื ื‘ืขื” ืžื›ืš ืฉื”ื™ื ื”ื™ืชื” ืฉื—ืงื ื™ืช ืžืขื•ืœื”
09:37
She knew when to trust her instincts.
183
577969
2429
ื”ื™ื ื™ื“ืขื” ืžืชื™ ืœื‘ื˜ื•ื— ื‘ื—ื•ืฉื™ื ืฉืœื”,
09:40
And so that project taught me another lesson about life,
184
580422
3929
ื•ื›ืš ื”ืžื™ื–ื ื”ื–ื” ืœื™ืžื“ ืื•ืชื™ ืœืงื— ื ื•ืกืฃ ืขืœ ื”ื—ื™ื™ื,
09:44
which is that, when in doubt, improvise.
185
584375
3647
ื•ื”ื•ื: ื›ืฉืืชื” ื‘ืกืคืง, ืชืืœืชืจ. [ืฆื—ื•ืง]
09:48
(Laughter)
186
588046
1666
09:50
Now, the third project grew out of my experience
187
590664
4919
ื”ืžื™ื–ื ื”ืฉืœื™ืฉื™ ื”ืชืคืชื—
ืžื—ื•ื•ื™ื•ืชื™ ื›ืฉืื‘ื™ ื”ื™ื” ื‘ื‘ื™ืช-ื—ื•ืœื™ื.
09:55
when my father was in the hospital.
188
595607
1849
ื”ื•ื ืขื‘ืจ ื˜ื™ืคื•ืœื™ื,
09:59
He was undergoing a treatment -- chemotherapy treatments --
189
599284
3353
ื˜ื™ืคื•ืœื™ ื›ื™ืžื•ืชืจืคื™ื”, ื•ืื—ื“ ื”ื˜ื™ืคื•ืœื™ื
10:02
and there's a related treatment called brachytherapy,
190
602661
4270
ืงืจื•ื™ "ื‘ืจื›ื™ืชืจืคื™ื”", ื•ื‘ื• ื–ืจืขื™ื ื–ืขื™ืจื™ื ื•ืจื“ื™ื•ืืงื˜ื™ื‘ื™ื™ื
10:06
where tiny, radioactive seeds are placed into the body
191
606955
3788
ืžื•ื›ื ืกื™ื ืœื’ื•ืฃ ื›ื“ื™ ืœื˜ืคืœ ื‘ื’ื™ื“ื•ืœื™ื ืกืจื˜ื ื™ื™ื.
10:10
to treat cancerous tumors.
192
610767
1734
10:13
And the way it's done, as you can see here,
193
613572
2271
ื•ื–ื” ื ืขืฉื” ื›ืš, ื›ืคื™ ืฉืืชื ืจื•ืื™ื ื›ืืŸ,
10:15
is that surgeons insert needles into the body
194
615867
4367
ืฉื”ืžื ืชื—ื™ื ืžื—ื“ื™ืจื™ื ืœื’ื•ืฃ ืžื—ื˜ื™ื
10:20
to deliver the seeds.
195
620258
1315
ื›ื“ื™ ืœื”ื›ื ื™ืก ืืช ื”ื–ืจืขื™ื, ื•ื›ืœ ื–ื”,
10:21
And all these needles are inserted in parallel.
196
621974
3698
ื›ืœ ื”ืžื—ื˜ื™ื ืžื•ื—ื“ืจื•ืช ื‘ืžืงื‘ื™ืœ,
10:26
So it's very common that some of the needles penetrate sensitive organs.
197
626445
5393
ื›ืš ืฉืงื•ืจื” ื”ืจื‘ื” ืฉื›ืžื” ืžื”ืžื—ื˜ื™ื
ืคื•ื’ืขื•ืช ื‘ืื‘ืจื™ื ืจื’ื™ืฉื™ื, ื•ื›ืชื•ืฆืื” ืžื›ืš
10:32
And as a result, the needles damage these organs, cause damage,
198
632635
6973
ื”ืžื—ื˜ื™ื ื’ื•ืจืžื•ืช ื ื–ืง ืœืื‘ืจื™ื ืืœื”,
10:39
which leads to trauma and side effects.
199
639632
2409
ืฉืžื‘ื™ื ืœืคืฆื™ืขื” ื•ืœืชื•ืคืขื•ืช ืœื•ื•ืื™.
10:42
So my students and I wondered:
200
642581
1644
ืื– ื”ืกื˜ื•ื“ื ื˜ื™ื ืฉืœื™ ื•ืื ื™ ืชื”ื™ื ื• ื”ืื ืื ื• ื™ื›ื•ืœื™ื
10:44
what if we could modify the system,
201
644249
4522
ืœืฉื ื•ืช ืืช ื”ืžืขืจื›ืช
10:48
so that the needles could come in at different angles?
202
648795
2625
ื›ืš ืฉื”ืžื—ื˜ื™ื ื™ื•ื—ื“ืจื• ื‘ื–ื•ื•ื™ื•ืช ืฉื•ื ื•ืช?
10:52
So we simulated this;
203
652395
1843
ื‘ื™ืฆืขื ื• ื”ื“ืžื™ื”, ื•ืคื™ืชื—ื ื•
10:54
we developed some optimization algorithms and we simulated this.
204
654262
3445
ื›ืžื” ืืœื’ื•ืจื™ืชืžื™ื ืฉืœ ืžื™ื˜ื•ื‘ ื•ื”ืจืฆื ื• ื”ื“ืžื™ื” ืฉืœ ื–ื”,
10:57
And we were able to show
205
657731
1151
ื•ื”ืฆืœื—ื ื• ืœื”ื•ื›ื™ื— ืฉืื ื• ื™ื›ื•ืœื™ื ืœื”ื™ืžื ืข
10:58
that we are able to avoid the delicate organs,
206
658906
2562
ืžืœืคื’ื•ืข ื‘ืื‘ืจื™ื ืขื“ื™ื ื™ื ื•ืขื“ื™ื™ืŸ ืœื”ืฉื™ื’ ืืช ืžืœื•ื ื”ื›ื™ืกื•ื™
11:01
and yet still achieve the coverage of the tumors with the radiation.
207
661492
5053
ืฉืœ ื”ื’ื™ื“ื•ืœื™ื ื‘ืขื–ืจืช ื”ื”ืงืจื ื”.
11:07
So now, we're working with doctors at UCSF
208
667313
3474
ืื– ื›ืขืช ืื ื• ืขื•ื‘ื“ื™ื ืขื ืจื•ืคืื™ื ื‘ืื•ื ื™ื‘ืจืกื™ื˜ืช ืงืœื™ืคื•ืจื ื™ื”
11:10
and engineers at Johns Hopkins,
209
670811
2267
ื•ืžื”ื ื“ืกื™ื ื‘"ื’'ื•ืŸ ื”ื•ืคืงื™ื ืก"
11:13
and we're building a robot that has a number of --
210
673102
3962
ืขืœ ื‘ื ื™ื” ืฉืœ ืจื•ื‘ื•ื˜ ืฉื™ืฉ ืœื• ื›ืžื”--
11:17
it's a specialized design with different joints
211
677088
2599
ื–ื”ื• ืขื™ืฆื•ื‘ ื™ื™ืขื•ื“ื™ ื‘ืขืœ ืžื™ืคืจืงื™ื ืฉื•ื ื™ื ืฉื™ื›ื•ืœื™ื ืœืืคืฉืจ
11:19
that can allow the needles to come in at an infinite variety of angles.
212
679711
4323
ืœืžื—ื˜ื™ื ืœื—ื“ื•ืจ ื‘ืžื’ื•ื•ืŸ ืื™ื ืกื•ืคื™ ืฉืœ ื–ื•ื•ื™ื•ืช,
11:24
And as you can see here, they can avoid delicate organs
213
684483
3722
ื•ื›ืคื™ ืฉืืชื ืจื•ืื™ื ื›ืืŸ, ื”ืŸ ื™ื›ื•ืœื•ืช ืœื”ื™ืžื ืข ืžืคื’ื™ืขื” ื‘ืื‘ืจื™ื ืขื“ื™ื ื™ื
ื•ืขื“ื™ื™ืŸ ืœื”ื’ื™ืข ืœืžื˜ืจื•ืชื™ื”ืŸ.
11:28
and still reach the targets they're aiming for.
214
688229
2664
ืื– ื‘ื›ืš ืฉื”ื˜ืœื ื• ืกืคืง ื‘ื”ื ื—ืช ื”ื™ืกื•ื“ ืฉื›ืœ ื”ืžื—ื˜ื™ื
11:32
So, by questioning this assumption that all the needles have to be parallel,
215
692019
4985
ืฆืจื™ื›ื•ืช ืœื”ื™ื•ืช ืžืงื‘ื™ืœื•ืช, ื’ื ื”ืžื™ื–ื ื”ื–ื” ืœื™ืžื“ ืื•ืชื™
11:37
this project also taught me an important lesson:
216
697028
2626
ืœืงื— ื—ืฉื•ื‘: ื›ืฉืืชื” ื‘ืกืคืง --
11:40
When in doubt, when your path is blocked, pivot.
217
700114
4755
ื›ืฉื”ื“ืจืš ืฉืœืš ื—ืกื•ืžื”, ืขืฉื” ืชืคื ื™ืช.
11:45
And the last project also has to do with medical robotics.
218
705797
3840
ื•ื”ืžื™ื–ื ื”ืื—ืจื•ืŸ ืงืฉื•ืจ ื’ื ื”ื•ื ืœืจื•ื‘ื•ื˜ื™ืงื” ืจืคื•ืื™ืช.
11:50
And this is something that's grown out of a system
219
710187
3570
ื•ื”ืคืขื ืžื“ื•ื‘ืจ ื‘ืžืฉื”ื• ืฉืฆืžื— ืžืชื•ืš ืžืขืจื›ืช ื‘ืฉื
11:53
called the da Vinci surgical robot.
220
713781
3348
ื”ืจื•ื‘ื•ื˜ ื”ื›ื™ืจื•ืจื’ื™ "ื“ื”-ื•ื™ื ืฆ'ื™",
11:57
And this is a commercially available device.
221
717866
2444
ืฉื”ื•ื ืžืชืงืŸ ื–ืžื™ืŸ ืžืกื—ืจื™ืช,
12:00
It's being used in over 2,000 hospitals around the world.
222
720334
2973
ืžืฉืชืžืฉื™ื ื‘ื• ื‘ื™ื•ืชืจ ืž-2,000 ื‘ืชื™-ื—ื•ืœื™ื ื‘ืขื•ืœื,
ื•ื”ืจืขื™ื•ืŸ ื”ื•ื ืœืืคืฉืจ ืœืžื ืชื—
12:04
The idea is it allows the surgeon to operate comfortably
223
724013
4373
ืœืคืขื•ืœ ื‘ื ื•ื—ื•ืช ื‘ื’ื‘ื•ืœื•ืช ื”ืงื•ืื•ืจื“ื™ื ืฆื™ื” ืฉืœื•,
12:08
in his own coordinate frame.
224
728410
1777
ืืš ืจื‘ื•ืช ืžืžื˜ืœื•ืช ื”ืžืฉื ื” ื‘ื›ื™ืจื•ืจื’ื™ื”
12:12
Many of the subtasks in surgery are very routine and tedious, like suturing,
225
732762
5769
ื”ืŸ ืžืื“ ืฉื’ืจืชื™ื•ืช ื•ืžืฉืขืžืžื•ืช, ื›ืžื• ื”ืชืคื™ืจื”,
12:18
and currently, all of these are performed
226
738555
2267
ื•ืœืคื™ ืฉืขื”, ื›ื•ืœืŸ ืžื‘ื•ืฆืขื•ืช
12:20
under the specific and immediate control of the surgeon.
227
740846
3867
ืชื—ืช ืคื™ืงื•ื—ื• ื”ืกืคืฆื™ืคื™ ื•ื”ืžื™ื™ื“ื™ ืฉืœ ื”ืžื ืชื—,
12:25
So the surgeon becomes fatigued over time.
228
745374
2268
ื›ืš ืฉืขื ื”ื–ืžืŸ, ื”ืžื ืชื— ืžืชืขื™ื™ืฃ.
ื•ืื ื—ื ื• ืชื”ื™ื ื•
12:28
And we've been wondering,
229
748086
1436
12:29
what if we could program the robot to perform some of these subtasks,
230
749546
4371
ืื ื ื•ื›ืœ ืœืชื›ื ืช ืืช ื”ืจื•ื‘ื•ื˜
ืœื‘ืฆืข ื›ืžื” ืžืžื˜ืœื•ืช-ืžืฉื ื” ืืœื”,
12:33
and thereby free the surgeon
231
753941
1381
ื•ื‘ื›ืš ืœืฉื—ืจืจ ืืช ื”ืžื ืชื—, ื•ื”ื•ื ื™ื•ื›ืœ ืœื”ืชืžืงื“
12:35
to focus on the more complicated parts of the surgery,
232
755346
3235
ื‘ื—ืœืงื™ื ื”ืžืกื•ื‘ื›ื™ื ื™ื•ืชืจ ืฉืœ ื”ื ื™ืชื•ื—,
12:38
and also cut down on the time that the surgery would take
233
758605
2800
ื•ื’ื ืœื”ืคื—ื™ืช ืืช ืžืฉืš ื”ื ื™ืชื•ื—
12:41
if we could get the robot to do them a little bit faster?
234
761429
2839
ืื ื ื•ื›ืœ ืœื’ืจื•ื ืœืจื•ื‘ื•ื˜ ืœื‘ืฆืข ืื•ืชื ืžืขื˜ ื™ื•ืชืจ ืžื”ืจ.
12:44
Now, it's hard to program a robot to do delicate things like this.
235
764958
3422
ืงืฉื” ืœืชื›ื ืช ืจื•ื‘ื•ื˜ ืœื‘ืฆืข ืคืขื•ืœื•ืช ืขื“ื™ื ื•ืช ื›ืืœื”,
ืื‘ืœ ื”ืชื‘ืจืจ ืฉืขืžื™ืชื™, ืคื™ื˜ืจ ืื‘ื™ืœ,
12:48
But it turns out my colleague Pieter Abbeel, who's here at Berkeley,
236
768943
4206
ืฉืขื•ื‘ื“ ื›ืืŸ ื‘ื‘ืจืงืœื™, ืคื™ืชื—
12:53
has developed a new set of techniques for teaching robots from example.
237
773173
5480
ืžืขืจืš ื—ื“ืฉ ืฉืœ ื˜ื›ื ื™ืงื•ืช ืœื”ื“ืจื›ืช ืจื•ื‘ื•ื˜ื™ื ืœืคื™ ื“ื•ื’ืžื”.
12:59
So he's gotten robots to fly helicopters,
238
779170
2743
ื•ื›ืš ื”ื•ื ื’ืจื ืœืจื•ื‘ื•ื˜ื™ื ืœื”ื˜ื™ืก ืžืกื•ืงื™ื,
13:01
do incredibly interesting, beautiful acrobatics,
239
781937
3254
ืœื‘ืฆืข ืืงืจื•ื‘ื˜ื™ืงื” ืžืขื ื™ื™ื ืช ื•ื™ืคื” ืœื”ืคืœื™ื,
13:05
by watching human experts fly them.
240
785215
1967
ืžืชื•ืš ืฆืคื™ื” ื‘ืžื•ืžื—ื™ื ืื ื•ืฉื™ื™ื ืฉืžื˜ื™ืกื™ื ืื•ืชื.
ื”ืฉื’ื ื• ืืช ืื—ื“ ื”ืจื•ื‘ื•ื˜ื™ื ื”ืืœื”.
13:08
So we got one of these robots.
241
788152
1906
13:10
We started working with Pieter and his students.
242
790082
2540
ื”ืชื—ืœื ื• ืœืขื‘ื•ื“ ืขื ืคื™ื˜ืจ ื•ื”ืกื˜ื•ื“ื ื˜ื™ื ืฉืœื•
13:12
And we asked a surgeon to perform a task --
243
792646
4180
ื•ื‘ื™ืงืฉื ื• ืžืžื ืชื— ืื—ื“ ืœื‘ืฆืข
ืžื˜ืœื” ืžืกื•ื™ืžืช, ื•ืžื” ืฉืขืฉื™ื ื•, ืขื ื”ืจื•ื‘ื•ื˜,
13:18
with the robot.
244
798761
1151
13:19
So what we're doing is asking the surgeon to perform the task,
245
799936
3025
ืื ื• ืžื‘ืงืฉื™ื ืžื”ืจื•ื‘ื•ื˜--
ื”ืžื ืชื— ืžื‘ืฆืข ืืช ื”ืžื˜ืœื”,
13:22
and we record the motions of the robot.
246
802985
2036
ื•ืื ื• ืžืงืœื™ื˜ื™ื ืืช ืชื ื•ืขื•ืช ื”ืจื•ื‘ื•ื˜.
13:25
So here's an example.
247
805045
1426
ื”ื ื” ื“ื•ื’ืžื”. ืืฉืชืžืฉ ื‘ืฆื•ืจืช ื”ืกืคืจื” 8,
13:26
I'll use tracing out a figure eight as an example.
248
806495
3121
ื•ืืขืงื•ื‘ ืื—ืจื™ ืฆื•ืจืช ื”-8 ื›ื“ื•ื’ืžื”.
ื›ืš ื–ื” ื ืจืื” ื›ืฉื”ืจื•ื‘ื•ื˜--
13:30
So here's what it looks like when the robot --
249
810170
3634
13:33
this is what the robot's path looks like, those three examples.
250
813828
3068
ื›ืš ื ืจืื” ื”ืชื•ื•ืื™ ืฉืœ ื”ืจื•ื‘ื•ื˜,
ืืœื• ื”ืŸ 3 ื“ื•ื’ืžืื•ืช.
13:36
Now, those are much better than what a novice like me could do,
251
816920
4136
ื”ืŸ ื˜ื•ื‘ื•ืช ื‘ื”ืจื‘ื” ืžืžื” ืฉื—ื•ื‘ื‘ืŸ ื›ืžื•ื ื™
ื”ื™ื” ื™ื›ื•ืœ ืœืขืฉื•ืช, ืืš ื”ืŸ ืขื“ื™ื™ืŸ ืงื•ืคืฆื ื™ื•ืช ื•ืœื ืžื“ื•ื™ืงื•ืช.
13:41
but they're still jerky and imprecise.
252
821080
2694
13:43
So we record all these examples, the data,
253
823798
2941
ืื– ืื ื• ืžืชืขื“ื™ื ืืช ื›ืœ ื”ื“ื•ื’ืžืื•ืช ื”ืืœื”, ืืช ื›ืœ ื”ื ืชื•ื ื™ื,
ื•ืื– ืžื‘ืฆืขื™ื ืกื“ืจื” ืฉืœ ืฉืœื‘ื™ื.
13:46
and then go through a sequence of steps.
254
826763
2861
ืจืืฉื™ืช ืื ื• ืžืฉืชืžืฉื™ื ื‘ื˜ื›ื ื™ืงื” ืฉืงืจื•ื™ื” "ืคื™ืชื•ืœ ื–ืžืŸ ื“ื™ื ืžื™",
13:50
First, we use a technique called dynamic time warping
255
830278
3126
13:53
from speech recognition.
256
833428
1490
ืฉืงืฉื•ืจื” ื‘ื–ื™ื”ื•ื™ ื“ื™ื‘ื•ืจ, ื•ื–ื” ืžืืคืฉืจ ืœื ื•
13:54
And this allows us to temporally align all of the examples.
257
834942
3337
ืœืืจื’ืŸ ืืช ื›ืœ ื”ื“ื•ื’ืžืื•ืช ืžื‘ื—ื™ื ืช ื”ื–ืžื ื™ื,
13:58
And then we apply Kalman filtering, a technique from control theory,
258
838918
5202
ื•ืื– ืื ื• ืžื™ื™ืฉืžื™ื ืืช "ืกื™ื ื•ืŸ ืงืœืžืŸ",
ื˜ื›ื ื™ืงื” ืžืชื™ืื•ืจื™ื™ืช ื”ืฉืœื™ื˜ื”, ืฉืžืืคืฉืจืช ืœื ื•
14:04
that allows us to statistically analyze all the noise
259
844144
2969
ืœื ืชื— ื‘ืื•ืคืŸ ืกื˜ื˜ื™ืกื˜ื™ ืืช ื›ืœ ื”ืจืขืฉ
ื•ืœื—ืœืฅ ืžืชื•ื›ื• ืืช ื”ืžืกืœื•ืœ ื”ืจืฆื•ื™.
14:07
and extract the desired trajectory that underlies them.
260
847137
4122
14:13
Now we take those human demonstrations --
261
853283
2550
ืื– ื›ืขืช ืื ื• ืœื•ืงื—ื™ื
14:15
they're all noisy and imperfect --
262
855857
1675
ืืช ื”ื”ื“ื’ืžื•ืช ื”ืื ื•ืฉื™ื•ืช ื”ืืœื”, ืฉื›ื•ืœืŸ ืจื•ืขืฉื•ืช ื•ื‘ืœืชื™-ืžื•ืฉืœืžื•ืช,
14:17
and we extract from them an inferred task trajectory
263
857556
2598
ื•ืžื—ืœืฆื™ื ืžื”ืŸ ืžืกืœื•ืœ ืžื˜ืœื” ืžืฉื•ืขืจ
14:20
and control sequence for the robot.
264
860178
2526
ื•ืจืฆืฃ ื‘ืงืจื” ืขื‘ื•ืจ ื”ืจื•ื‘ื•ื˜.
14:23
We then execute that on the robot,
265
863181
2316
ืœืื—ืจ ืžื›ืŸ ืื ื• ืžืจื™ืฆื™ื ื–ืืช ืขืœ ื”ืจื•ื‘ื•ื˜,
14:25
we observe what happens,
266
865521
2025
ืฆื•ืคื™ื ื‘ืžื” ืฉืงื•ืจื”,
14:27
then we adjust the controls,
267
867570
1357
ื•ืื—ืจ ืžื›ื•ื•ื ื ื™ื ืืช ื”ื‘ืงืจื•ืช ื‘ืขื–ืจืช ืจืฆืฃ ื˜ื›ื ื™ืงื•ืช
14:28
using a sequence of techniques called iterative learning.
268
868951
2796
ื”ืงืจื•ื™ "ืœื™ืžื•ื“ ื“ืจืš ื”ื™ืฉื ื•ืช".
14:33
Then what we do is we increase the velocity a little bit.
269
873129
3409
ื‘ืฉืœื‘ ื”ื‘ื ืื ื• ืžืขืœื™ื ืžืขื˜ ืืช ื”ืžื”ื™ืจื•ืช.
14:37
We observe the results, adjust the controls again,
270
877244
3135
ืฆื•ืคื™ื ื‘ืชื•ืฆืื•ืช, ืžื›ื•ื•ื ื ื™ื ืฉื•ื‘ ืืช ื”ื‘ืงืจื•ืช,
ื•ืฆื•ืคื™ื ื‘ืžื” ืฉืงื•ืจื”.
14:41
and observe what happens.
271
881340
1772
14:43
And we go through this several rounds.
272
883136
2167
ืื ื• ืžื‘ืฆืขื™ื ื–ืืช ื‘ืžืฉืš ืžืกืคืจ ืžื—ื–ื•ืจื™ื.
14:45
And here's the result.
273
885327
1181
ื•ื”ื ื” ื”ืชื•ืฆืื”.
14:46
That's the inferred task trajectory,
274
886968
1738
ื–ื”ื• ืžืกืœื•ืœ ื”ืžื˜ืœื” ื”ืžืฉื•ืขืจ,
14:48
and here's the robot moving at the speed of the human.
275
888730
3221
ื•ื”ื ื” ื”ืจื•ื‘ื•ื˜ ื ืข ื‘ืžื”ื™ืจื•ืช ืื“ื.
14:51
Here's four times the speed of the human.
276
891975
2105
ื›ืืŸ ื–ื” ืคื™ 4 ืžื”ื™ืจื•ืช ืื“ื.
14:54
Here's seven times.
277
894477
1484
ื›ืืŸ - ืคื™ 7.
14:57
And here's the robot operating at 10 times the speed of the human.
278
897004
4825
ื•ื›ืืŸ ื”ืจื•ื‘ื•ื˜ ืคื•ืขืœ ื‘ืžื”ื™ืจื•ืช ื’ื“ื•ืœื” ืคื™ 10
ืžืžื”ื™ืจื•ืช ื”ืื“ื.
15:02
So we're able to get a robot to perform a delicate task
279
902762
2901
ื›ืš ืฉืื ื• ื™ื›ื•ืœื™ื ืœื’ืจื•ื ืœืจื•ื‘ื•ื˜ ืœื‘ืฆืข ืžื˜ืœื” ืขื“ื™ื ื”
15:05
like a surgical subtask,
280
905687
2886
ื›ืžื• ืžื˜ืœื” ื›ื™ืจื•ืจื’ื™ืช ืžืฉื ื™ืช,
15:09
at 10 times the speed of a human.
281
909081
1881
ื‘ืžื”ื™ืจื•ืช ื’ื‘ื•ื”ื” ืคื™ 10 ืžืžื”ื™ืจื•ืช ืื ื•ืฉื™ืช.
15:12
So this project also,
282
912185
2260
ืื– ื’ื ื”ืžื™ื–ื ื”ื–ื”, ื‘ื’ืœืœ ื”ืชืจื’ื•ืœ ื•ื”ืœืžื™ื“ื”
15:14
because of its involved practicing and learning,
283
914469
2525
ื”ืžืขื•ืจื‘ื™ื ื‘ื•, ื‘ื™ืฆื•ืข ืฉืœ ืžืฉื”ื• ืฉื•ื‘ ื•ืฉื•ื‘,
15:17
doing something over and over again,
284
917018
1762
15:18
this project also has a lesson, which is:
285
918804
2577
ื’ื ื‘ืžื™ื–ื ื”ื–ื” ื˜ืžื•ืŸ ืœืงื—, ื•ื”ื•ื,
15:21
if you want to do something well,
286
921405
2404
ืื ืจื•ืฆื™ื ืœืขืฉื•ืช ืžืฉื”ื• ื”ื™ื˜ื‘,
ืื™ืŸ ืชื—ืœื™ืฃ ืœืชืจื’ื•ืœ, ืชืจื’ื•ืœ, ืชืจื’ื•ืœ.
15:25
there's no substitute for practice, practice, practice.
287
925084
4440
ื•ืืœื” ืืจื‘ืขื” ืžื”ืœืงื—ื™ื ืฉืœืžื“ืชื™
15:33
So these are four of the lessons that I've learned from robots
288
933186
3487
ืžื”ืจื•ื‘ื•ื˜ื™ื ื‘ืžืจื•ืฆืช ื”ืฉื ื™ื,
15:36
over the years.
289
936697
1298
ื•ื”ืจื•ื‘ื•ื˜ื™ืงื”, ืชื—ื•ื ื”ืจื•ื‘ื•ื˜ื™ืงื”, ื”ืฉืชืคืจ ื‘ื”ืจื‘ื”
15:39
And the field of robotics has gotten much better over time.
290
939312
5507
ืขื ื”ื–ืžืŸ.
15:46
Nowadays, high school students can build robots,
291
946319
2276
ื›ื™ื•ื ืชืœืžื™ื“ื™ ืชื™ื›ื•ืŸ ื™ื›ื•ืœื™ื ืœื‘ื ื•ืช ืจื•ื‘ื•ื˜ื™ื
15:48
like the industrial robot my dad and I tried to build.
292
948619
2884
ื›ืžื• ื”ืจื•ื‘ื•ื˜ื™ื ื”ืชืขืฉื™ื™ืชื™ื™ื ืฉืื‘ื™ ื•ืื ื™ ื ื™ืกื™ื ื• ืœื‘ื ื•ืช.
15:52
But, it's very -- now ...
293
952675
2770
ื•ื”ื™ื•ื ื™ืฉ ืœื™ ื‘ืช,
15:55
And now, I have a daughter,
294
955954
3301
15:59
named Odessa.
295
959857
1157
ื‘ืฉื ืื•ื“ืกื”.
16:01
She's eight years old.
296
961673
1253
ื”ื™ื ื‘ืช ืฉืžื•ื ื”.
16:03
And she likes robots, too.
297
963682
1649
ื•ื’ื ื”ื™ื ืื•ื”ื‘ืช ืจื•ื‘ื•ื˜ื™ื.
16:05
Maybe it runs in the family.
298
965970
1396
ืื•ืœื™ ื–ื” ื’ื ื˜ื™. [ืฆื—ื•ืง]
16:07
(Laughter)
299
967390
1240
16:08
I wish she could meet my dad.
300
968654
2095
ื”ืœื•ื•ืื™ ื•ื™ื›ืœื” ืœืคื’ื•ืฉ ืืช ืื‘ื™.
16:12
And now I get to teach her how things work,
301
972303
2740
ื•ื›ืขืช ืชื•ืจื™ ืœืœืžื“ ืื•ืชื” ืื™ืš ื”ื“ื‘ืจื™ื ืคื•ืขืœื™ื,
16:15
and we get to build projects together.
302
975067
2230
ื•ืื ื• ื–ื•ื›ื™ื ืœื‘ื ื•ืช ืžื™ื–ืžื™ื ื‘ื™ื—ื“, ื•ืื ื™ ืชื•ื”ื”
16:17
And I wonder what kind of lessons she'll learn from them.
303
977321
3206
ืžื”ื ื”ืœืงื—ื™ื ืฉื”ื™ื ืขืชื™ื“ื” ืœืœืžื•ื“ ืžื”ื.
16:22
Robots are the most human of our machines.
304
982147
4007
ื”ืจื•ื‘ื•ื˜ื™ื ื”ื ื”ื›ื™ ืื ื•ืฉื™ื™ื
ืžื›ืœ ื”ืžื›ื•ื ื•ืช ืฉืœื ื•.
16:26
They can't solve all of the world's problems,
305
986983
2954
ื”ื ืœื ื™ื›ื•ืœื™ื ืœืคืชื•ืจ ืืช ื›ืœ ื‘ืขื™ื•ืช ื”ืขื•ืœื,
16:29
but I think they have something important to teach us.
306
989961
3523
ืืš ืœื“ืขืชื™ ื™ืฉ ืœื”ื ืžืฉื”ื• ื—ืฉื•ื‘ ืœืœืžื“ ืื•ืชื ื•.
16:34
I invite all of you
307
994276
2001
ืื ื™ ืžื–ืžื™ืŸ ืืช ื›ื•ืœื›ื ืœื—ืฉื•ื‘ ืขืœ ื”ื—ื™ื“ื•ืฉื™ื
16:36
to think about the innovations that you're interested in,
308
996301
3578
ืฉืžืขื ื™ื™ื ื™ื ืืชื›ื,
16:40
the machines that you wish for.
309
1000712
2243
ืขืœ ื”ืžื›ื•ื ื•ืช ืฉื”ื™ื™ืชื ืจื•ืฆื™ื ืœืจืื•ืช,
16:43
And think about what they might be telling you.
310
1003660
2607
ื•ืœื—ืฉื•ื‘ ืขืœ ืžื” ืขืฉื•ื™ ืœื”ื™ื•ืช ื”ืžืกืจ ืฉืœื”ืŸ ืขื‘ื•ืจื›ื,
ื›ื™ ื™ืฉ ืœื™ ื”ืจื’ืฉื”
16:47
Because I have a hunch that many of our technological innovations,
311
1007211
3745
ืฉืจื‘ื™ื ืžื”ื—ื™ื“ื•ืฉื™ื ื”ื˜ื›ื ื•ืœื•ื’ื™ื™ื ืฉืœื ื•,
16:50
the devices we dream about,
312
1010980
1671
ืžื”ืžื›ืฉื™ืจื™ื ืฉืื ื• ื—ื•ืœืžื™ื ืขืœื™ื”ื,
ื™ื›ื•ืœื™ื ืœืชืช ืœื ื• ื”ืฉืจืื” ืœื”ื™ื•ืช ื™ืฆื•ืจื™-ืื ื•ืฉ ื˜ื•ื‘ื™ื ื™ื•ืชืจ.
16:54
can inspire us to be better humans.
313
1014166
2849
16:57
Thank you.
314
1017984
1151
ืชื•ื“ื” ืœื›ื. [ืžื—ื™ืื•ืช ื›ืคื™ื™ื]
16:59
(Applause)
315
1019159
1912
ื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