1
00:00:00,539 --> 00:00:03,358
I know this is going to sound strange,

2
00:00:03,382 --> 00:00:08,365
but I think robots can inspire us
to be better humans.

3
00:00:09,245 --> 00:00:12,339
See, I grew up in Bethlehem, Pennsylvania,

4
00:00:12,363 --> 00:00:14,301
the home of Bethlehem Steel.

5
00:00:14,991 --> 00:00:17,103
My father was an engineer,

6
00:00:17,127 --> 00:00:21,446
and when I was growing up,
he would teach me how things worked.

7
00:00:21,470 --> 00:00:23,564
We would build projects together,

8
00:00:23,588 --> 00:00:26,356
like model rockets and slot cars.

9
00:00:26,732 --> 00:00:29,640
Here's the go-kart that we built together.

10
00:00:30,208 --> 00:00:31,952
That's me behind the wheel,

11
00:00:31,976 --> 00:00:34,200
with my sister and my best
friend at the time.

12
00:00:35,875 --> 00:00:37,801
And one day,

13
00:00:37,825 --> 00:00:40,908
he came home, when I was
about 10 years old,

14
00:00:40,932 --> 00:00:43,558
and at the dinner table, he announced

15
00:00:43,582 --> 00:00:47,294
that for our next project, 
we were going to build ...

16
00:00:47,318 --> 00:00:48,468
a robot.

17
00:00:49,604 --> 00:00:50,759
A robot.

18
00:00:51,120 --> 00:00:53,183
Now, I was thrilled about this,

19
00:00:53,207 --> 00:00:57,009
because at school,
there was a bully named Kevin,

20
00:00:57,033 --> 00:00:58,922
and he was picking on me,

21
00:00:58,946 --> 00:01:01,336
because I was the only
Jewish kid in class.

22
00:01:01,910 --> 00:01:04,283
So I couldn't wait to get
started to work on this,

23
00:01:04,307 --> 00:01:06,917
so I could introduce Kevin to my robot.

24
00:01:06,941 --> 00:01:07,971
(Laughter)

25
00:01:07,995 --> 00:01:12,238
(Robot noises)

26
00:01:17,552 --> 00:01:18,944
(Laughter)

27
00:01:18,968 --> 00:01:22,817
But that wasn't the kind of robot
my dad had in mind.

28
00:01:22,841 --> 00:01:23,909
(Laughter)

29
00:01:23,933 --> 00:01:27,565
See, he owned a chromium-plating company,

30
00:01:27,589 --> 00:01:33,462
and they had to move heavy steel parts
between tanks of chemicals.

31
00:01:33,486 --> 00:01:36,859
And so he needed
an industrial robot like this,

32
00:01:36,883 --> 00:01:38,986
that could basically do the heavy lifting.

33
00:01:39,684 --> 00:01:43,275
But my dad didn't get
the kind of robot he wanted, either.

34
00:01:43,870 --> 00:01:46,300
He and I worked on it for several years,

35
00:01:46,324 --> 00:01:51,069
but it was the 1970s, and the technology
that was available to amateurs

36
00:01:51,093 --> 00:01:52,419
just wasn't there yet.

37
00:01:53,992 --> 00:01:56,777
So Dad continued to do
this kind of work by hand.

38
00:01:57,595 --> 00:01:59,023
And a few years later,

39
00:01:59,730 --> 00:02:01,306
he was diagnosed with cancer.

40
00:02:03,822 --> 00:02:05,392
You see,

41
00:02:05,416 --> 00:02:08,012
what the robot we were trying
to build was telling him

42
00:02:08,036 --> 00:02:10,192
was not about doing the heavy lifting.

43
00:02:10,216 --> 00:02:11,367
It was a warning

44
00:02:11,391 --> 00:02:13,895
about his exposure to the toxic chemicals.

45
00:02:15,184 --> 00:02:17,477
He didn't recognize that at the time,

46
00:02:17,501 --> 00:02:19,187
and he contracted leukemia.

47
00:02:19,827 --> 00:02:21,719
And he died at the age of 45.

48
00:02:23,354 --> 00:02:25,130
I was devastated by this.

49
00:02:25,703 --> 00:02:28,465
And I never forgot the robot
that he and I tried to build.

50
00:02:30,195 --> 00:02:33,748
When I was at college, I decided
to study engineering, like him.

51
00:02:35,081 --> 00:02:39,629
And I went to Carnegie Mellon,
and I earned my PhD in robotics.

52
00:02:39,653 --> 00:02:41,562
I've been studying robots ever since.

53
00:02:42,721 --> 00:02:47,316
So what I'd like to tell you about
are four robot projects,

54
00:02:47,340 --> 00:02:50,389
and how they've inspired me
to be a better human.

55
00:02:54,206 --> 00:02:59,820
By 1993, I was a young professor at USC,

56
00:02:59,844 --> 00:03:02,662
and I was just building up
my own robotics lab,

57
00:03:02,686 --> 00:03:05,907
and this was the year
the World Wide Web came out.

58
00:03:06,487 --> 00:03:09,472
And I remember my students
were the ones who told me about it,

59
00:03:09,496 --> 00:03:11,813
and we would -- we were just amazed.

60
00:03:11,837 --> 00:03:15,220
We started playing with this,
and that afternoon,

61
00:03:15,244 --> 00:03:19,522
we realized that we could use
this new, universal interface

62
00:03:19,546 --> 00:03:23,974
to allow anyone in the world
to operate the robot in our lab.

63
00:03:25,339 --> 00:03:29,381
So, rather than have it fight
or do industrial work,

64
00:03:30,413 --> 00:03:32,943
we decided to build a planter,

65
00:03:32,967 --> 00:03:34,918
put the robot into the center of it,

66
00:03:34,942 --> 00:03:36,620
and we called it the Telegarden.

67
00:03:37,736 --> 00:03:42,253
And we had put a camera
in the gripper of the hand of the robot,

68
00:03:42,277 --> 00:03:45,208
and we wrote some
special scripts and software,

69
00:03:45,232 --> 00:03:47,266
so that anyone in the world could come in,

70
00:03:47,290 --> 00:03:49,074
and by clicking on the screen,

71
00:03:49,098 --> 00:03:52,719
they could move the robot around
and visit the garden.

72
00:03:53,416 --> 00:03:57,267
But we also set up some other software

73
00:03:57,291 --> 00:04:01,207
that lets you participate
and help us water the garden, remotely.

74
00:04:01,643 --> 00:04:03,932
And if you watered it a few times,

75
00:04:03,956 --> 00:04:06,231
we'd give you your own seed to plant.

76
00:04:07,167 --> 00:04:10,643
Now, this was an engineering project,

77
00:04:10,667 --> 00:04:14,779
and we published some papers
on the system design of it,

78
00:04:14,803 --> 00:04:17,448
but we also thought of it
as an art installation.

79
00:04:18,638 --> 00:04:20,787
It was invited, after the first year,

80
00:04:20,811 --> 00:04:23,831
by the Ars Electronica Museum in Austria,

81
00:04:23,855 --> 00:04:26,677
to have it installed in their lobby.

82
00:04:27,558 --> 00:04:31,545
And I'm happy to say, it remained
online there, 24 hours a day,

83
00:04:31,569 --> 00:04:33,400
for almost nine years.

84
00:04:34,369 --> 00:04:38,144
That robot was operated by more people

85
00:04:38,168 --> 00:04:40,065
than any other robot in history.

86
00:04:41,303 --> 00:04:42,930
Now, one day,

87
00:04:42,954 --> 00:04:46,302
I got a call out of the blue
from a student,

88
00:04:47,211 --> 00:04:51,369
who asked a very simple
but profound question.

89
00:04:52,361 --> 00:04:55,187
He said, "Is the robot real?"

90
00:04:56,599 --> 00:04:58,885
Now, everyone else had assumed it was,

91
00:04:58,909 --> 00:05:01,425
and we knew it was,
because we were working with it.

92
00:05:01,449 --> 00:05:02,847
But I knew what he meant,

93
00:05:02,871 --> 00:05:04,313
because it would be possible

94
00:05:04,337 --> 00:05:07,122
to take a bunch of pictures
of flowers in a garden

95
00:05:07,146 --> 00:05:10,785
and then, basically, index them
in a computer system,

96
00:05:10,809 --> 00:05:13,436
such that it would appear
that there was a real robot,

97
00:05:13,460 --> 00:05:14,610
when there wasn't.

98
00:05:15,088 --> 00:05:16,612
And the more I thought about it,

99
00:05:16,636 --> 00:05:20,215
I couldn't think of a good answer
for how he could tell the difference.

100
00:05:20,628 --> 00:05:23,425
This was right about the time
that I was offered a position

101
00:05:23,449 --> 00:05:24,750
here at Berkeley.

102
00:05:24,774 --> 00:05:26,419
And when I got here,

103
00:05:26,443 --> 00:05:28,805
I looked up Hubert Dreyfus,

104
00:05:28,829 --> 00:05:31,756
who's a world-renowned
professor of philosophy,

105
00:05:32,582 --> 00:05:35,312
And I talked with him
about this and he said,

106
00:05:35,336 --> 00:05:39,438
"This is one of the oldest
and most central problems in philosophy.

107
00:05:39,462 --> 00:05:43,479
It goes back to the Skeptics
and up through Descartes.

108
00:05:43,503 --> 00:05:46,685
It's the issue of epistemology,

109
00:05:46,709 --> 00:05:49,577
the study of how do we know
that something is true."

110
00:05:50,625 --> 00:05:52,625
So he and I started working together,

111
00:05:52,649 --> 00:05:55,520
and we coined a new term:
"telepistemology,"

112
00:05:56,757 --> 00:05:58,893
the study of knowledge at a distance.

113
00:05:59,303 --> 00:06:03,794
We invited leading artists,
engineers and philosophers

114
00:06:03,818 --> 00:06:05,137
to write essays about this,

115
00:06:05,161 --> 00:06:08,865
and the results are collected
in this book from MIT Press.

116
00:06:09,959 --> 00:06:12,055
So thanks to this student,

117
00:06:12,079 --> 00:06:15,398
who questioned what everyone else
had assumed to be true,

118
00:06:15,422 --> 00:06:19,399
this project taught me
an important lesson about life,

119
00:06:19,423 --> 00:06:22,410
which is to always question assumptions.

120
00:06:23,807 --> 00:06:25,983
Now, the second project
I'll tell you about

121
00:06:26,007 --> 00:06:27,999
grew out of the Telegarden.

122
00:06:28,023 --> 00:06:30,844
As it was operating, my students
and I were very interested

123
00:06:30,868 --> 00:06:33,209
in how people were interacting
with each other,

124
00:06:33,233 --> 00:06:35,290
and what they were doing with the garden.

125
00:06:35,314 --> 00:06:36,465
So we started thinking:

126
00:06:36,489 --> 00:06:38,457
what if the robot could leave the garden

127
00:06:38,481 --> 00:06:41,295
and go out into some other
interesting environment?

128
00:06:41,319 --> 00:06:44,060
Like, for example,
what if it could go to a dinner party

129
00:06:44,084 --> 00:06:45,545
at the White House?

130
00:06:45,569 --> 00:06:47,068
(Laughter)

131
00:06:48,401 --> 00:06:51,970
So, because we were interested
more in the system design

132
00:06:51,994 --> 00:06:54,937
and the user interface
than in the hardware,

133
00:06:54,961 --> 00:06:56,112
we decided that,

134
00:06:56,136 --> 00:07:00,623
rather than have a robot replace
the human to go to the party,

135
00:07:00,647 --> 00:07:02,877
we'd have a human replace the robot.

136
00:07:03,679 --> 00:07:05,139
We called it the Tele-Actor.

137
00:07:05,978 --> 00:07:07,986
We got a human,

138
00:07:08,010 --> 00:07:11,118
someone who's very
outgoing and gregarious,

139
00:07:11,142 --> 00:07:15,397
and she was outfitted with a helmet
with various equipment,

140
00:07:15,421 --> 00:07:16,705
cameras and microphones,

141
00:07:16,729 --> 00:07:19,794
and then a backpack with wireless
Internet connection.

142
00:07:20,882 --> 00:07:23,026
And the idea was that she could go

143
00:07:23,050 --> 00:07:25,500
into a remote and interesting environment,

144
00:07:25,524 --> 00:07:27,541
and then over the Internet,

145
00:07:27,565 --> 00:07:30,157
people could experience
what she was experiencing.

146
00:07:30,736 --> 00:07:33,707
So they could see what she was seeing,

147
00:07:33,731 --> 00:07:37,014
but then, more importantly,
they could participate,

148
00:07:37,038 --> 00:07:41,788
by interacting with each other
and coming up with ideas

149
00:07:41,812 --> 00:07:45,781
about what she should do next
and where she should go,

150
00:07:45,805 --> 00:07:48,011
and then conveying those
to the Tele-Actor.

151
00:07:49,069 --> 00:07:51,490
So we got a chance to take the Tele-Actor

152
00:07:51,514 --> 00:07:54,322
to the Webby Awards in San Francisco.

153
00:07:55,129 --> 00:07:58,369
And that year, Sam Donaldson was the host.

154
00:08:00,290 --> 00:08:02,980
Just before the curtain went
up, I had about 30 seconds

155
00:08:03,004 --> 00:08:06,746
to explain to Mr. Donaldson
what we were going to do.

156
00:08:08,119 --> 00:08:11,636
And I said, "The Tele-Actor
is going to be joining you onstage.

157
00:08:11,660 --> 00:08:14,014
This is a new experimental project,

158
00:08:14,038 --> 00:08:16,633
and people are watching her
on their screens,

159
00:08:16,657 --> 00:08:19,948
there's cameras involved
and there's microphones

160
00:08:19,972 --> 00:08:21,758
and she's got an earbud in her ear,

161
00:08:21,782 --> 00:08:24,099
and people over the network
are giving her advice

162
00:08:24,123 --> 00:08:25,281
about what to do next."

163
00:08:25,305 --> 00:08:26,700
And he said, "Wait a second.

164
00:08:27,806 --> 00:08:28,979
That's what I do."

165
00:08:29,003 --> 00:08:33,980
(Laughter)

166
00:08:34,004 --> 00:08:35,876
So he loved the concept,

167
00:08:35,900 --> 00:08:40,153
and when the Tele-Actor walked onstage,
she walked right up to him,

168
00:08:40,177 --> 00:08:42,636
and she gave him a big kiss
right on the lips.

169
00:08:42,660 --> 00:08:44,612
(Laughter)

170
00:08:44,636 --> 00:08:47,558
We were totally surprised --
we had no idea that would happen.

171
00:08:47,582 --> 00:08:50,200
And he was great, he just gave her
a big hug in return,

172
00:08:50,224 --> 00:08:51,881
and it worked out great.

173
00:08:51,905 --> 00:08:53,945
But that night, as we were packing up,

174
00:08:53,969 --> 00:08:58,524
I asked the Tele-Actor,
how did the Tele-Directors decide

175
00:08:58,548 --> 00:09:01,548
that they would give
a kiss to Sam Donaldson?

176
00:09:03,135 --> 00:09:04,592
And she said they hadn't.

177
00:09:05,274 --> 00:09:07,892
She said, when she was
just about to walk onstage,

178
00:09:07,916 --> 00:09:10,890
the Tele-Directors still were trying
to agree on what to do,

179
00:09:10,914 --> 00:09:14,014
and so she just walked onstage
and did what felt most natural.

180
00:09:14,038 --> 00:09:18,552
(Laughter)

181
00:09:18,576 --> 00:09:21,668
So, the success
of the Tele-Actor that night

182
00:09:21,692 --> 00:09:25,945
was due to the fact
that she was a wonderful actor.

183
00:09:25,969 --> 00:09:28,398
She knew when to trust her instincts.

184
00:09:28,422 --> 00:09:32,351
And so that project taught me
another lesson about life,

185
00:09:32,375 --> 00:09:36,022
which is that, when in doubt, improvise.

186
00:09:36,046 --> 00:09:37,712
(Laughter)

187
00:09:38,664 --> 00:09:43,583
Now, the third project
grew out of my experience

188
00:09:43,607 --> 00:09:45,456
when my father was in the hospital.

189
00:09:47,284 --> 00:09:50,637
He was undergoing a treatment --
chemotherapy treatments --

190
00:09:50,661 --> 00:09:54,931
and there's a related treatment
called brachytherapy,

191
00:09:54,955 --> 00:09:58,743
where tiny, radioactive seeds
are placed into the body

192
00:09:58,767 --> 00:10:00,501
to treat cancerous tumors.

193
00:10:01,572 --> 00:10:03,843
And the way it's done,
as you can see here,

194
00:10:03,867 --> 00:10:08,234
is that surgeons
insert needles into the body

195
00:10:08,258 --> 00:10:09,573
to deliver the seeds.

196
00:10:09,974 --> 00:10:13,672
And all these needles
are inserted in parallel.

197
00:10:14,445 --> 00:10:19,838
So it's very common that some
of the needles penetrate sensitive organs.

198
00:10:20,635 --> 00:10:27,608
And as a result, the needles damage
these organs, cause damage,

199
00:10:27,632 --> 00:10:30,041
which leads to trauma and side effects.

200
00:10:30,581 --> 00:10:32,225
So my students and I wondered:

201
00:10:32,249 --> 00:10:36,771
what if we could modify the system,

202
00:10:36,795 --> 00:10:39,420
so that the needles
could come in at different angles?

203
00:10:40,395 --> 00:10:42,238
So we simulated this;

204
00:10:42,262 --> 00:10:45,707
we developed some optimization
algorithms and we simulated this.

205
00:10:45,731 --> 00:10:46,882
And we were able to show

206
00:10:46,906 --> 00:10:49,468
that we are able to avoid
the delicate organs,

207
00:10:49,492 --> 00:10:54,545
and yet still achieve the coverage
of the tumors with the radiation.

208
00:10:55,313 --> 00:10:58,787
So now, we're working with doctors at UCSF

209
00:10:58,811 --> 00:11:01,078
and engineers at Johns Hopkins,

210
00:11:01,102 --> 00:11:05,064
and we're building a robot
that has a number of --

211
00:11:05,088 --> 00:11:07,687
it's a specialized design
with different joints

212
00:11:07,711 --> 00:11:12,034
that can allow the needles to come in
at an infinite variety of angles.

213
00:11:12,483 --> 00:11:16,205
And as you can see here,
they can avoid delicate organs

214
00:11:16,229 --> 00:11:18,893
and still reach the targets
they're aiming for.

215
00:11:20,019 --> 00:11:25,004
So, by questioning this assumption
that all the needles have to be parallel,

216
00:11:25,028 --> 00:11:27,654
this project also taught me
an important lesson:

217
00:11:28,114 --> 00:11:32,869
When in doubt, when your path
is blocked, pivot.

218
00:11:33,797 --> 00:11:37,637
And the last project
also has to do with medical robotics.

219
00:11:38,187 --> 00:11:41,757
And this is something
that's grown out of a system

220
00:11:41,781 --> 00:11:45,129
called the da Vinci surgical robot.

221
00:11:45,866 --> 00:11:48,310
And this is a commercially
available device.

222
00:11:48,334 --> 00:11:51,307
It's being used in over 2,000
hospitals around the world.

223
00:11:52,013 --> 00:11:56,386
The idea is it allows the surgeon
to operate comfortably

224
00:11:56,410 --> 00:11:58,187
in his own coordinate frame.

225
00:12:00,762 --> 00:12:06,531
Many of the subtasks in surgery are very
routine and tedious, like suturing,

226
00:12:06,555 --> 00:12:08,822
and currently, all of these are performed

227
00:12:08,846 --> 00:12:12,713
under the specific and immediate
control of the surgeon.

228
00:12:13,374 --> 00:12:15,642
So the surgeon becomes fatigued over time.

229
00:12:16,086 --> 00:12:17,522
And we've been wondering,

230
00:12:17,546 --> 00:12:21,917
what if we could program the robot
to perform some of these subtasks,

231
00:12:21,941 --> 00:12:23,322
and thereby free the surgeon

232
00:12:23,346 --> 00:12:26,581
to focus on the more complicated
parts of the surgery,

233
00:12:26,605 --> 00:12:29,405
and also cut down on the time
that the surgery would take

234
00:12:29,429 --> 00:12:32,268
if we could get the robot
to do them a little bit faster?

235
00:12:32,958 --> 00:12:36,380
Now, it's hard to program a robot
to do delicate things like this.

236
00:12:36,943 --> 00:12:41,149
But it turns out my colleague
Pieter Abbeel, who's here at Berkeley,

237
00:12:41,173 --> 00:12:46,653
has developed a new set of techniques
for teaching robots from example.

238
00:12:47,170 --> 00:12:49,913
So he's gotten robots to fly helicopters,

239
00:12:49,937 --> 00:12:53,191
do incredibly interesting,
beautiful acrobatics,

240
00:12:53,215 --> 00:12:55,182
by watching human experts fly them.

241
00:12:56,152 --> 00:12:58,058
So we got one of these robots.

242
00:12:58,082 --> 00:13:00,622
We started working with Pieter
and his students.

243
00:13:00,646 --> 00:13:04,826
And we asked a surgeon
to perform a task --

244
00:13:06,761 --> 00:13:07,912
with the robot.

245
00:13:07,936 --> 00:13:10,961
So what we're doing is asking
the surgeon to perform the task,

246
00:13:10,985 --> 00:13:13,021
and we record the motions of the robot.

247
00:13:13,045 --> 00:13:14,471
So here's an example.

248
00:13:14,495 --> 00:13:17,616
I'll use tracing out
a figure eight as an example.

249
00:13:18,170 --> 00:13:21,804
So here's what it looks like
when the robot --

250
00:13:21,828 --> 00:13:24,896
this is what the robot's path
looks like, those three examples.

251
00:13:24,920 --> 00:13:29,056
Now, those are much better
than what a novice like me could do,

252
00:13:29,080 --> 00:13:31,774
but they're still jerky and imprecise.

253
00:13:31,798 --> 00:13:34,739
So we record all these examples, the data,

254
00:13:34,763 --> 00:13:37,624
and then go through a sequence of steps.

255
00:13:38,278 --> 00:13:41,404
First, we use a technique
called dynamic time warping

256
00:13:41,428 --> 00:13:42,918
from speech recognition.

257
00:13:42,942 --> 00:13:46,279
And this allows us to temporally
align all of the examples.

258
00:13:46,918 --> 00:13:52,120
And then we apply Kalman filtering,
a technique from control theory,

259
00:13:52,144 --> 00:13:55,113
that allows us to statistically
analyze all the noise

260
00:13:55,137 --> 00:13:59,259
and extract the desired
trajectory that underlies them.

261
00:14:01,283 --> 00:14:03,833
Now we take those human demonstrations --

262
00:14:03,857 --> 00:14:05,532
they're all noisy and imperfect --

263
00:14:05,556 --> 00:14:08,154
and we extract from them
an inferred task trajectory

264
00:14:08,178 --> 00:14:10,704
and control sequence for the robot.

265
00:14:11,181 --> 00:14:13,497
We then execute that on the robot,

266
00:14:13,521 --> 00:14:15,546
we observe what happens,

267
00:14:15,570 --> 00:14:16,927
then we adjust the controls,

268
00:14:16,951 --> 00:14:19,747
using a sequence of techniques
called iterative learning.

269
00:14:21,129 --> 00:14:24,538
Then what we do is we increase
the velocity a little bit.

270
00:14:25,244 --> 00:14:28,379
We observe the results,
adjust the controls again,

271
00:14:29,340 --> 00:14:31,112
and observe what happens.

272
00:14:31,136 --> 00:14:33,303
And we go through this several rounds.

273
00:14:33,327 --> 00:14:34,508
And here's the result.

274
00:14:34,968 --> 00:14:36,706
That's the inferred task trajectory,

275
00:14:36,730 --> 00:14:39,951
and here's the robot
moving at the speed of the human.

276
00:14:39,975 --> 00:14:42,080
Here's four times the speed of the human.

277
00:14:42,477 --> 00:14:43,961
Here's seven times.

278
00:14:45,004 --> 00:14:49,829
And here's the robot operating
at 10 times the speed of the human.

279
00:14:50,762 --> 00:14:53,663
So we're able to get a robot
to perform a delicate task

280
00:14:53,687 --> 00:14:56,573
like a surgical subtask,

281
00:14:57,081 --> 00:14:58,962
at 10 times the speed of a human.

282
00:15:00,185 --> 00:15:02,445
So this project also,

283
00:15:02,469 --> 00:15:04,994
because of its involved
practicing and learning,

284
00:15:05,018 --> 00:15:06,780
doing something over and over again,

285
00:15:06,804 --> 00:15:09,381
this project also has a lesson, which is:

286
00:15:09,405 --> 00:15:11,809
if you want to do something well,

287
00:15:13,084 --> 00:15:17,524
there's no substitute
for practice, practice, practice.

288
00:15:21,186 --> 00:15:24,673
So these are four of the lessons
that I've learned from robots

289
00:15:24,697 --> 00:15:25,995
over the years.

290
00:15:27,312 --> 00:15:32,819
And the field of robotics
has gotten much better over time.

291
00:15:34,319 --> 00:15:36,595
Nowadays, high school students
can build robots,

292
00:15:36,619 --> 00:15:39,503
like the industrial robot
my dad and I tried to build.

293
00:15:40,675 --> 00:15:43,445
But, it's very -- now ...

294
00:15:43,954 --> 00:15:47,255
And now, I have a daughter,

295
00:15:47,857 --> 00:15:49,014
named Odessa.

296
00:15:49,673 --> 00:15:50,926
She's eight years old.

297
00:15:51,682 --> 00:15:53,331
And she likes robots, too.

298
00:15:53,970 --> 00:15:55,366
Maybe it runs in the family.

299
00:15:55,390 --> 00:15:56,630
(Laughter)

300
00:15:56,654 --> 00:15:58,749
I wish she could meet my dad.

301
00:16:00,303 --> 00:16:03,043
And now I get to teach her
how things work,

302
00:16:03,067 --> 00:16:05,297
and we get to build projects together.

303
00:16:05,321 --> 00:16:08,527
And I wonder what kind of lessons
she'll learn from them.

304
00:16:10,147 --> 00:16:14,154
Robots are the most human of our machines.

305
00:16:14,983 --> 00:16:17,937
They can't solve all
of the world's problems,

306
00:16:17,961 --> 00:16:21,484
but I think they have something
important to teach us.

307
00:16:22,276 --> 00:16:24,277
I invite all of you

308
00:16:24,301 --> 00:16:27,879
to think about the innovations
that you're interested in,

309
00:16:28,712 --> 00:16:30,955
the machines that you wish for.

310
00:16:31,660 --> 00:16:34,267
And think about
what they might be telling you.

311
00:16:35,211 --> 00:16:38,956
Because I have a hunch that many
of our technological innovations,

312
00:16:38,980 --> 00:16:40,651
the devices we dream about,

313
00:16:42,166 --> 00:16:45,015
can inspire us to be better humans.

314
00:16:45,984 --> 00:16:47,135
Thank you.

315
00:16:47,159 --> 00:16:49,071
(Applause)