TEDWomen 2010
Heather Knight: Silicon-based comedy
Filmed:
Readability: 3.9
880,023 views
In this first-of-its-kind demo, Heather Knight introduces Data, a robotic stand-up comedian that does much more than rattle off one-liners -- it gathers audience feedback (using software co-developed with Scott Satkin and Varun Ramakrishna at CMU) and tunes its act as the crowd responds. Is this thing on?
Heather Knight - Roboticist
Heather Knight runs Marilyn Monrobot, which creates socially intelligent robot performances and sensor-based electronic art. (and she helped build the amazing Rube Goldberg machine for OK Go’s “This Too Shall Pass.") Full bio
Heather Knight runs Marilyn Monrobot, which creates socially intelligent robot performances and sensor-based electronic art. (and she helped build the amazing Rube Goldberg machine for OK Go’s “This Too Shall Pass.") Full bio
Double-click the English transcript below to play the video.
00:15
Some of the greatest innovations
0
0
2000
00:17
and developments in the world
1
2000
3000
00:20
often happen at the intersection of two fields.
2
5000
2000
00:22
So tonight I'd like to tell you
3
7000
2000
00:24
about the intersection that I'm most excited about at this very moment,
4
9000
3000
00:27
which is entertainment and robotics.
5
12000
2000
00:29
So if we're trying to make robots
6
14000
2000
00:31
that can be more expressive
7
16000
2000
00:33
and that can connect better with us in society,
8
18000
2000
00:35
maybe we should look to some of the human professionals
9
20000
3000
00:38
of artificial emotion and personality
10
23000
2000
00:40
that occur in the dramatic arts.
11
25000
3000
00:43
I'm also interested in creating new technologies for the arts
12
28000
3000
00:46
and to attract people
13
31000
2000
00:48
to science and technology.
14
33000
3000
00:51
Some people in the last decade or two
15
36000
3000
00:54
have started creating artwork with technology.
16
39000
3000
00:57
With my new venture, Marilyn Monrobot,
17
42000
2000
00:59
I would like to use art to create tech.
18
44000
3000
01:05
(Laughter)
19
50000
2000
01:07
So we're based in New York City.
20
52000
2000
01:09
And if you're a performer that wants to collaborate
21
54000
2000
01:11
with an adorable robot,
22
56000
2000
01:13
or if you have a robot that needs entertainment representation,
23
58000
3000
01:16
please contact me, the Bot-Agent.
24
61000
3000
01:19
The bot, our rising celebrity,
25
64000
2000
01:21
also has his own Twitter account: @robotinthewild.
26
66000
3000
01:24
I'd like to introduce you to one of our first robots, Data.
27
69000
3000
01:27
He's named after the Star Trek character.
28
72000
2000
01:29
I think he's going to be super popular.
29
74000
2000
01:31
We've got the robot --
30
76000
2000
01:33
in his head is a database of a lot of jokes.
31
78000
3000
01:36
Now each of these jokes is labeled with certain attributes.
32
81000
3000
01:39
So it knows something about the subject; it knows about the length.
33
84000
2000
01:41
It knows how much it's moving.
34
86000
2000
01:43
And so it's going to try to watch your response.
35
88000
2000
01:45
I actually have no idea what my robot is going to do today.
36
90000
3000
01:48
(Laughter)
37
93000
2000
01:50
It can also learn from you
38
95000
2000
01:52
about the quality of its jokes
39
97000
2000
01:54
and cater things, sort of like Netflix-style,
40
99000
2000
01:56
over longer-term
41
101000
2000
01:58
to different communities or audiences,
42
103000
2000
02:00
children versus adults, different cultures.
43
105000
3000
02:03
You can learn something from the robot
44
108000
2000
02:05
about the community that you're in.
45
110000
2000
02:07
And also I can use each one of you as the acting coach
46
112000
3000
02:10
to our future robot companions.
47
115000
3000
02:13
Some of you in this middle section --
48
118000
2000
02:15
you have red/green paddles.
49
120000
3000
02:18
If you like what's going on,
50
123000
3000
02:21
show the green.
51
126000
2000
02:23
If you don't like the subject or the performance,
52
128000
2000
02:25
you can hold the red.
53
130000
2000
02:27
Now don't be shy.
54
132000
2000
02:29
It's just a robot.
55
134000
2000
02:31
It doesn't have feelings ... yet.
56
136000
2000
02:33
(Laughter)
57
138000
2000
02:35
And the rest of you,
58
140000
2000
02:37
you still count, you still matter.
59
142000
2000
02:39
There's also a microphone
60
144000
2000
02:41
that's listening to the aggregate laughter and applause
61
146000
2000
02:43
and booing -- I hope not --
62
148000
3000
02:46
to help make some of its next decisions.
63
151000
3000
02:53
Right, so,
64
158000
2000
02:55
let the robot stand-up comedy begin.
65
160000
3000
03:05
Data: Hello TEDWomen.
66
170000
2000
03:07
It's an honor to be here.
67
172000
2000
03:09
(Laughter)
68
174000
2000
03:11
(Applause)
69
176000
6000
03:18
You guys are looking good out there.
70
183000
2000
03:20
Ready for some jokes?
71
185000
3000
03:23
Audience: Yeah.
72
188000
2000
03:25
Data: Here's the first one.
73
190000
2000
03:28
Right, so,
74
193000
2000
03:30
a doctor says to his patient,
75
195000
2000
03:32
"I have bad news and worse news.
76
197000
3000
03:35
The bad news is that you only have 24 hours to live."
77
200000
4000
03:39
"That's terrible," said the patient.
78
204000
3000
03:42
"How can the news possibly be worse?"
79
207000
3000
03:45
"I've been trying to contact you since yesterday."
80
210000
3000
03:48
(Laughter)
81
213000
2000
03:50
(Applause)
82
215000
2000
03:52
The Swiss have an interesting army. Five hundred years without a war.
83
217000
3000
03:55
Heather Knight: He's talking about the Swiss.
84
220000
2000
03:57
Data: Pretty lucky for them.
85
222000
2000
03:59
Ever see that little Swiss Army knife
86
224000
2000
04:01
they have to fight with?
87
226000
3000
04:04
"Come on buddy, I have the toe clippers right here.
88
229000
3000
04:07
You get past me,
89
232000
2000
04:09
the guy behind me has a spoon."
90
234000
2000
04:11
(Laughter)
91
236000
2000
04:13
HK: He's a French robot.
92
238000
2000
04:15
Data: A couple of New Jersey hunters are out in the woods.
93
240000
3000
04:18
One of them falls to the ground.
94
243000
3000
04:21
He does not seem to be breathing.
95
246000
2000
04:23
The other guy whips out his cell phone
96
248000
2000
04:25
and calls 911.
97
250000
3000
04:28
He gasps to the operator, "My friend is dead.
98
253000
3000
04:31
What can I do?"
99
256000
2000
04:33
The operator says, "Just take it easy.
100
258000
2000
04:35
I can help.
101
260000
2000
04:37
First, let's make sure he's dead."
102
262000
3000
04:40
There is a silence,
103
265000
2000
04:42
and the operator hears a shot.
104
267000
2000
04:44
The guy's voice comes back on the line,
105
269000
2000
04:46
"Okay, now what?"
106
271000
3000
04:49
(Laughter)
107
274000
2000
04:51
(Applause)
108
276000
2000
04:53
Question: Why is television called a medium?
109
278000
2000
04:55
Anyone?
110
280000
2000
04:57
Because it's neither rare nor well done.
111
282000
3000
05:00
But to be completely honest with you,
112
285000
3000
05:03
I kind of love television.
113
288000
2000
05:05
Any of you like television? Audience: Yes.
114
290000
2000
05:07
Data: I find it incredibly educational.
115
292000
3000
05:10
Actually, as soon as someone turns it on,
116
295000
3000
05:13
I go into the other room and read.
117
298000
2000
05:15
(Laughter)
118
300000
5000
05:20
That's all for now.
119
305000
2000
05:22
Was that okay for my first time?
120
307000
2000
05:24
(Applause)
121
309000
9000
05:33
You've been a great audience.
122
318000
2000
05:35
Thank you.
123
320000
2000
05:37
HK: Yay.
124
322000
2000
05:39
(Applause)
125
324000
4000
05:43
So this is actually the first time
126
328000
2000
05:45
we've ever done live audience feedback to a performance.
127
330000
3000
05:48
So thank you all for being a part of it.
128
333000
2000
05:50
There's a lot more to come.
129
335000
3000
05:53
And we hope to learn a lot about robot expression.
130
338000
2000
05:55
Thank you very much.
131
340000
2000
05:57
(Applause)
132
342000
2000
ABOUT THE SPEAKER
Heather Knight - RoboticistHeather Knight runs Marilyn Monrobot, which creates socially intelligent robot performances and sensor-based electronic art. (and she helped build the amazing Rube Goldberg machine for OK Go’s “This Too Shall Pass.")
Why you should listen
Heather Knight is conducting her doctoral research at the intersection of robotics and entertainment at Carnegie Mellon’s Robotics Institute. She runs Marilyn Monrobot in New York, where she and her cohort create “charismatic machine performances.” As she describes it: “In one example, robot and human actors will perform the same script and basic set of actions, but each time vary the emotional and social content of those interactions. This variation will ... allow the general public to craft and hone robot personalities.”
Her installations have been featured at the Smithsonian-Cooper Hewitt Design Museum, LACMA, SIGGRAPH, PopTech and the Fortezza da Basso in Florence, Italy. Her past work also includes robotics and instrumentation at NASA’s Jet Propulsion Laboratory, interactive installations with Syyn Labs (including the award- winning Rube Goldberg machine music video with OK Go), and sensor design at Aldebaran Robotics. She was recently named Assistant Director of Robotics at Humanity+.
Heather Knight | Speaker | TED.com