Technology that knows what you're feeling | Poppy Crum

137,981 views ・ 2018-07-10

TED


Please double-click on the English subtitles below to play the video.

00:12
What happens when technology knows more about us than we do?
0
12881
4456
00:17
A computer now can detect our slightest facial microexpressions
1
17992
3672
00:21
and be able to tell the difference between a real smile and a fake one.
2
21688
3611
00:25
That's only the beginning.
3
25323
1734
00:27
Technology has become incredibly intelligent
4
27466
2865
00:30
and already knows a lot about our internal states.
5
30355
3400
00:34
And whether we like it or not,
6
34085
2286
00:36
we already are sharing parts of our inner lives
7
36395
3499
00:39
that's out of our control.
8
39918
1733
00:43
That seems like a problem,
9
43413
1421
00:44
because a lot of us like to keep what's going on inside
10
44858
3246
00:48
from what people actually see.
11
48128
1647
00:50
We want to have agency over what we share and what we don't.
12
50323
4420
00:55
We all like to have a poker face.
13
55473
2321
00:59
But I'm here to tell you that I think that's a thing of the past.
14
59584
3346
01:03
And while that might sound scary, it's not necessarily a bad thing.
15
63347
4770
01:09
I've spent a lot of time studying the circuits in the brain
16
69030
2770
01:11
that create the unique perceptual realities that we each have.
17
71824
3693
01:16
And now I bring that together
18
76110
1405
01:17
with the capabilities of current technology
19
77539
2062
01:19
to create new technology that does make us better,
20
79625
2537
01:22
feel more, connect more.
21
82186
1600
01:24
And I believe to do that,
22
84482
1786
01:26
we have to be OK losing some of our agency.
23
86292
2749
01:30
With some animals, it's really amazing,
24
90149
2523
01:32
and we get to see into their internal experiences.
25
92696
3474
01:36
We get this upfront look at the mechanistic interaction
26
96649
3722
01:40
between how they respond to the world around them
27
100395
2817
01:43
and the state of their biological systems.
28
103236
2008
01:45
This is where evolutionary pressures like eating, mating
29
105268
3809
01:49
and making sure we don't get eaten
30
109101
1762
01:50
drive deterministic behavioral responses to information in the world.
31
110887
4157
01:55
And we get to see into this window,
32
115806
2794
01:58
into their internal states and their biological experiences.
33
118624
3636
02:02
It's really pretty cool.
34
122284
1642
02:03
Now, stay with me for a moment -- I'm a violinist, not a singer.
35
123950
4103
02:08
But the spider's already given me a critical review.
36
128077
3590
02:16
(Video) (Singing in a low pitch)
37
136907
2060
02:19
(Singing in a middle pitch)
38
139868
2888
02:23
(Singing in a high pitch)
39
143800
2505
02:27
(Singing in a low pitch)
40
147069
1421
02:29
(Singing in a middle pitch)
41
149236
1600
02:31
(Singing in a high pitch)
42
151403
1777
02:33
(Laughter)
43
153204
1150
02:36
Poppy Crum: It turns out, some spiders tune their webs like violins
44
156387
3198
02:39
to resonate with certain sounds.
45
159609
2158
02:41
And likely, the harmonics of my voice as it went higher
46
161791
2771
02:44
coupled with how loud I was singing
47
164586
1730
02:46
recreated either the predatory call of an echolocating bat or a bird,
48
166340
4467
02:50
and the spider did what it should.
49
170831
1881
02:53
It predictively told me to bug off.
50
173300
2817
02:56
I love this.
51
176824
1150
02:58
The spider's responding to its external world
52
178546
3309
03:01
in a way that we get to see and know what's happening to its internal world.
53
181879
4350
03:07
Biology is controlling the spider's response;
54
187069
2206
03:09
it's wearing its internal state on its sleeve.
55
189299
2776
03:13
But us, humans --
56
193768
1655
03:16
we're different.
57
196184
1150
03:17
We like to think we have cognitive control over what people see, know and understand
58
197899
5735
03:23
about our internal states --
59
203658
1409
03:25
our emotions, our insecurities, our bluffs, our trials and tribulations --
60
205091
4303
03:29
and how we respond.
61
209418
1267
03:31
We get to have our poker face.
62
211927
2282
03:35
Or maybe we don't.
63
215799
1200
03:37
Try this with me.
64
217728
1182
03:38
Your eye responds to how hard your brain is working.
65
218934
2690
03:42
The response you're about to see is driven entirely by mental effort
66
222363
3230
03:45
and has nothing to do with changes in lighting.
67
225617
2635
03:48
We know this from neuroscience.
68
228276
1650
03:49
I promise, your eyes are doing the same thing as the subject in our lab,
69
229950
4560
03:54
whether you want them to or not.
70
234534
1734
03:56
At first, you'll hear some voices.
71
236292
2173
03:58
Try and understand them and keep watching the eye in front of you.
72
238489
3278
04:01
It's going to be hard at first,
73
241791
1498
04:03
one should drop out, and it should get really easy.
74
243313
2391
04:05
You're going to see the change in effort in the diameter of the pupil.
75
245728
3325
04:10
(Video) (Two overlapping voices talking)
76
250140
2567
04:12
(Single voice) Intelligent technology depends on personal data.
77
252731
2963
04:15
(Two overlapping voices talking)
78
255718
2446
04:18
(Single voice) Intelligent technology depends on personal data.
79
258188
2976
04:21
PC: Your pupil doesn't lie.
80
261680
1326
04:23
Your eye gives away your poker face.
81
263030
2400
04:25
When your brain's having to work harder,
82
265990
1913
04:27
your autonomic nervous system drives your pupil to dilate.
83
267927
2785
04:30
When it's not, it contracts.
84
270736
1555
04:32
When I take away one of the voices,
85
272680
1691
04:34
the cognitive effort to understand the talkers
86
274395
2262
04:36
gets a lot easier.
87
276681
1158
04:37
I could have put the two voices in different spatial locations,
88
277863
3000
04:40
I could have made one louder.
89
280887
1666
04:42
You would have seen the same thing.
90
282577
1738
04:45
We might think we have more agency over the reveal of our internal state
91
285006
4786
04:49
than that spider,
92
289816
1579
04:51
but maybe we don't.
93
291419
1266
04:53
Today's technology is starting to make it really easy
94
293021
2969
04:56
to see the signals and tells that give us away.
95
296014
2690
04:59
The amalgamation of sensors paired with machine learning
96
299109
3294
05:02
on us, around us and in our environments,
97
302427
2413
05:04
is a lot more than cameras and microphones tracking our external actions.
98
304864
4653
05:12
Our bodies radiate our stories
99
312529
2818
05:15
from changes in the temperature of our physiology.
100
315371
2666
05:18
We can look at these as infrared thermal images
101
318546
2261
05:20
showing up behind me,
102
320831
1160
05:22
where reds are hotter and blues are cooler.
103
322015
2070
05:24
The dynamic signature of our thermal response
104
324458
3183
05:27
gives away our changes in stress,
105
327665
3031
05:30
how hard our brain is working,
106
330720
2008
05:32
whether we're paying attention
107
332752
1936
05:34
and engaged in the conversation we might be having
108
334712
2627
05:37
and even whether we're experiencing a picture of fire as if it were real.
109
337363
4095
05:41
We can actually see people give off heat on their cheeks
110
341482
2643
05:44
in response to an image of flame.
111
344149
2200
05:48
But aside from giving away our poker bluffs,
112
348013
2929
05:50
what if dimensions of data from someone's thermal response
113
350966
4746
05:55
gave away a glow of interpersonal interest?
114
355736
2659
05:58
Tracking the honesty of feelings in someone's thermal image
115
358966
3532
06:02
might be a new part of how we fall in love and see attraction.
116
362522
3626
06:06
Our technology can listen, develop insights and make predictions
117
366172
3693
06:09
about our mental and physical health
118
369889
2095
06:12
just by analyzing the timing dynamics of our speech and language
119
372008
4000
06:16
picked up by microphones.
120
376032
1443
06:18
Groups have shown that changes in the statistics of our language
121
378038
3880
06:21
paired with machine learning
122
381942
1420
06:23
can predict the likelihood someone will develop psychosis.
123
383386
3161
06:27
I'm going to take it a step further
124
387442
1751
06:29
and look at linguistic changes and changes in our voice
125
389217
2587
06:31
that show up with a lot of different conditions.
126
391828
2239
06:34
Dementia, diabetes can alter the spectral coloration of our voice.
127
394091
4367
06:39
Changes in our language associated with Alzheimer's
128
399205
3119
06:42
can sometimes show up more than 10 years before clinical diagnosis.
129
402348
4365
06:47
What we say and how we say it tells a much richer story
130
407236
3960
06:51
than we used to think.
131
411220
1254
06:53
And devices we already have in our homes could, if we let them,
132
413022
4047
06:57
give us invaluable insight back.
133
417093
2134
06:59
The chemical composition of our breath
134
419998
2978
07:03
gives away our feelings.
135
423959
1354
07:06
There's a dynamic mixture of acetone, isoprene and carbon dioxide
136
426363
4478
07:10
that changes when our heart speeds up, when our muscles tense,
137
430865
3384
07:14
and all without any obvious change in our behaviors.
138
434809
2897
07:18
Alright, I want you to watch this clip with me.
139
438268
2738
07:21
Some things might be going on on the side screens,
140
441030
3119
07:24
but try and focus on the image in the front
141
444173
3777
07:27
and the man at the window.
142
447974
1463
07:31
(Eerie music)
143
451633
2658
07:39
(Woman screams)
144
459767
1437
07:50
PC: Sorry about that. I needed to get a reaction.
145
470692
2395
07:53
(Laughter)
146
473111
1785
07:55
I'm actually tracking the carbon dioxide you exhale in the room right now.
147
475412
4972
08:01
We've installed tubes throughout the theater,
148
481903
3293
08:05
lower to the ground, because CO2 is heavier than air.
149
485220
2595
08:07
But they're connected to a device in the back
150
487839
2667
08:10
that lets us measure, in real time, with high precision,
151
490530
3287
08:13
the continuous differential concentration of CO2.
152
493841
2922
08:17
The clouds on the sides are actually the real-time data visualization
153
497246
5508
08:22
of the density of our CO2.
154
502778
1998
08:25
You might still see a patch of red on the screen,
155
505374
3699
08:29
because we're showing increases with larger colored clouds,
156
509097
3705
08:32
larger colored areas of red.
157
512826
2196
08:35
And that's the point where a lot of us jumped.
158
515046
2559
08:38
It's our collective suspense driving a change in carbon dioxide.
159
518173
4915
08:43
Alright, now, watch this with me one more time.
160
523649
2722
08:46
(Cheerful music)
161
526395
2238
08:54
(Woman laughs)
162
534553
2137
09:05
PC: You knew it was coming.
163
545344
1349
09:06
But it's a lot different when we changed the creator's intent.
164
546717
3363
09:10
Changing the music and the sound effects
165
550776
2769
09:13
completely alter the emotional impact of that scene.
166
553569
3603
09:17
And we can see it in our breath.
167
557196
2134
09:20
Suspense, fear, joy
168
560196
2262
09:22
all show up as reproducible, visually identifiable moments.
169
562482
4507
09:27
We broadcast a chemical signature of our emotions.
170
567473
4151
09:35
It is the end of the poker face.
171
575249
2133
09:38
Our spaces, our technology will know what we're feeling.
172
578582
3566
09:42
We will know more about each other than we ever have.
173
582736
2785
09:45
We get a chance to reach in and connect to the experience and sentiments
174
585911
4307
09:50
that are fundamental to us as humans
175
590242
1742
09:52
in our senses, emotionally and socially.
176
592008
2410
09:55
I believe it is the era of the empath.
177
595482
2540
09:58
And we are enabling the capabilities that true technological partners can bring
178
598046
5222
10:03
to how we connect with each other and with our technology.
179
603292
3047
10:06
If we recognize the power of becoming technological empaths,
180
606363
3389
10:09
we get this opportunity
181
609776
1936
10:11
where technology can help us bridge the emotional and cognitive divide.
182
611736
4424
10:16
And in that way, we get to change how we tell our stories.
183
616680
2723
10:19
We can enable a better future for technologies like augmented reality
184
619427
3580
10:23
to extend our own agency and connect us at a much deeper level.
185
623031
4193
10:27
Imagine a high school counselor being able to realize
186
627625
2547
10:30
that an outwardly cheery student really was having a deeply hard time,
187
630196
3826
10:34
where reaching out can make a crucial, positive difference.
188
634046
3180
10:37
Or authorities, being able to know the difference
189
637766
3230
10:41
between someone having a mental health crisis
190
641020
2325
10:43
and a different type of aggression,
191
643369
1826
10:45
and responding accordingly.
192
645219
1800
10:47
Or an artist, knowing the direct impact of their work.
193
647609
3273
10:52
Leo Tolstoy defined his perspective of art
194
652173
2643
10:54
by whether what the creator intended
195
654840
1785
10:56
was experienced by the person on the other end.
196
656649
2586
10:59
Today's artists can know what we're feeling.
197
659259
2566
11:02
But regardless of whether it's art or human connection,
198
662204
3005
11:06
today's technologies will know and can know
199
666608
2802
11:09
what we're experiencing on the other side,
200
669434
2048
11:11
and this means we can be closer and more authentic.
201
671506
2618
11:14
But I realize a lot of us have a really hard time
202
674498
4293
11:18
with the idea of sharing our data,
203
678815
2267
11:21
and especially the idea that people know things about us
204
681673
3111
11:24
that we didn't actively choose to share.
205
684808
2321
11:28
Anytime we talk to someone,
206
688728
2216
11:31
look at someone
207
691946
1555
11:33
or choose not to look,
208
693525
1468
11:35
data is exchanged, given away,
209
695017
2647
11:38
that people use to learn,
210
698533
2205
11:40
make decisions about their lives and about ours.
211
700762
3267
11:45
I'm not looking to create a world where our inner lives are ripped open
212
705469
3968
11:49
and our personal data and our privacy given away
213
709461
2348
11:51
to people and entities where we don't want to see it go.
214
711833
2713
11:55
But I am looking to create a world
215
715117
2762
11:57
where we can care about each other more effectively,
216
717903
3408
12:01
we can know more about when someone is feeling something
217
721335
3060
12:04
that we ought to pay attention to.
218
724419
1872
12:06
And we can have richer experiences from our technology.
219
726800
3335
12:10
Any technology can be used for good or bad.
220
730887
2357
12:13
Transparency to engagement and effective regulation
221
733268
2412
12:15
are absolutely critical to building the trust for any of this.
222
735704
3120
12:20
But the benefits that "empathetic technology" can bring to our lives
223
740106
4834
12:24
are worth solving the problems that make us uncomfortable.
224
744964
3891
12:29
And if we don't, there are too many opportunities and feelings
225
749315
4025
12:33
we're going to be missing out on.
226
753364
1695
12:35
Thank you.
227
755083
1175
12:36
(Applause)
228
756282
2479
About this website

This site will introduce you to YouTube videos that are useful for learning English. You will see English lessons taught by top-notch teachers from around the world. Double-click on the English subtitles displayed on each video page to play the video from there. The subtitles scroll in sync with the video playback. If you have any comments or requests, please contact us using this contact form.

https://forms.gle/WvT1wiN1qDtmnspy7