ABOUT THE SPEAKER
Jennifer Healey - Research scientist
A research scientist at Intel, Jennifer Healey develops the mobile internet devices of the future.

Why you should listen

Jennifer Healey imagines a future where computers and smartphones are capable of being sensitive to human emotions and where cars are able to talk to each other, and thus keep their drivers away from accidents. A scientist at Intel Corporation Research Labs, she researches devices and systems that would allow for these major innovations.

Healey holds PhD from MIT in electrical engineering and computer science. While there, she pioneered “Affective Computing” with Rosalind Picard and developed the first wearable computer with physiological sensors and a video camera that allows the wearer to track their daily activities and how they feel while doing them. From there, she moved to IBM where she worked on the next generation of multi-modal interactive smartphones and helped architect the "Interaction Mark-Up language" that allows users to switch from voice to speech input seamlessly.

Healey has also used her interest in embedded devices in the field of healthcare. While an instructor at Harvard Medical School and at Beth Israel Deaconess Medical Center, she worked on new ways to use heart rate to predict cardiac health. She then joined HP Research in Cambridge to further develop wearable sensors for health monitoring and continued this research when she joined Intel Digital Health.

More profile about the speaker
Jennifer Healey | Speaker | TED.com
TED@Intel

Jennifer Healey: If cars could talk, accidents might be avoidable

Jennifer Healey: 如果車子會說話, 就能避免事故發生

Filmed:
908,454 views

開車就像是坐在一顆玻璃泡泡內, 關上門, 踩油門, 只能依靠你的雙眼來引導方向 -- 即使自己只能看到前後幾部車輛。那麼如果車子可以彼此分享各自的位置和速度等資訊, 然後運用預估模型幫每一部車去計算安全路徑呢? Jennifer Healey 帶我們想像一個沒有事故的世界。 (攝於TED@Intel)
- Research scientist
A research scientist at Intel, Jennifer Healey develops the mobile internet devices of the future. Full bio

Double-click the English transcript below to play the video.

00:12
Let's face面對 it:
0
703
1914
面對現實吧
00:14
Driving駕駛 is dangerous危險.
1
2617
2445
開車是危險的
00:17
It's one of the things that we don't like to think about,
2
5062
3098
這是我們不願去想的事
00:20
but the fact事實 that religious宗教 icons圖標 and good luck運氣 charms魅力
3
8160
3652
但現實中我們在世界各地看到
00:23
show顯示 up on dashboards儀表板 around the world世界
4
11812
4790
汽車儀器板上的宗教畫像和平安符
00:28
betrays原形畢露 the fact事實 that we know this to be true真正.
5
16602
4137
卻揭露我們其實知道開車危險
00:32
Car汽車 accidents事故 are the leading領導 cause原因 of death死亡
6
20739
3594
車禍是美國16至19歲青年
00:36
in people ages年齡 16 to 19 in the United聯合的 States狀態 --
7
24333
4170
死亡的主因
00:40
leading領導 cause原因 of death死亡 --
8
28503
2843
車禍是導致死亡的主因
00:43
and 75 percent百分 of these accidents事故 have nothing to do
9
31346
3863
而當中百分之七十五的車禍
00:47
with drugs毒品 or alcohol.
10
35209
2285
和毒品或酒精無關
00:49
So what happens發生?
11
37494
2261
到底發生了什麼?
00:51
No one can say for sure, but I remember記得 my first accident事故.
12
39755
4219
沒人能確實說出
我記得我第一次車禍
00:55
I was a young年輕 driver司機 out on the highway高速公路,
13
43974
3803
當時我還年輕 在高速公路上
00:59
and the car汽車 in front面前 of me, I saw the brake制動 lights燈火 go on.
14
47777
2258
看到前面車輛的煞車燈亮了
01:02
I'm like, "Okay, all right, this guy is slowing減緩 down,
15
50035
1800
我心想﹕好吧 這傢伙慢了下來
01:03
I'll slow down too."
16
51835
1282
那我也跟著慢下來吧
01:05
I step on the brake制動.
17
53117
1926
我跟著踩煞車
01:07
But no, this guy isn't slowing減緩 down.
18
55043
2254
但我錯了 他不是減慢速度
01:09
This guy is stopping停止, dead stop, dead stop on the highway高速公路.
19
57297
3178
而是停車 竟然在高速公路上停車
01:12
It was just going 65 -- to zero?
20
60475
2540
從時速六十五英哩降到... 零?
01:15
I slammed抨擊 on the brakes剎車.
21
63015
1520
我急踩剎車
01:16
I felt the ABSABS kick in, and the car汽車 is still going,
22
64535
3059
ABS系統啟動了 但車仍向前駛
01:19
and it's not going to stop, and I know it's not going to stop,
23
67594
2696
車繼續向前 我知道車不會停了
01:22
and the air空氣 bag deploys展開時, the car汽車 is totaled總計,
24
70290
2939
氣囊彈出 車子毀了
01:25
and fortunately幸好, no one was hurt傷害.
25
73229
3557
幸好沒有傷亡
01:28
But I had no idea理念 that car汽車 was stopping停止,
26
76786
4211
我完全沒想到那輛車會停
01:32
and I think we can do a lot better than that.
27
80997
3645
我認為我們可以做得更好
01:36
I think we can transform轉變 the driving主動 experience經驗
28
84642
4145
我們可以透過讓車之間對話
01:40
by letting出租 our cars汽車 talk to each other.
29
88787
3879
改變駕駛的方式
01:44
I just want you to think a little bit
30
92666
1424
大家想一想
01:46
about what the experience經驗 of driving主動 is like now.
31
94090
2888
現在我們是怎樣開車的
01:48
Get into your car汽車. Close the door. You're in a glass玻璃 bubble泡沫.
32
96978
4028
上車 關上門 在一個玻璃安全室裡
01:53
You can't really directly sense the world世界 around you.
33
101006
2916
你無法直接感受周遭的世界
01:55
You're in this extended擴展 body身體.
34
103922
2181
在這個擴大了的軀殼裡
01:58
You're tasked任務 with navigating導航 it down
35
106103
2163
你的任務是
02:00
partially-seen部分見過 roadways道路,
36
108266
2056
在看不清路面的情況下
02:02
in and amongst其中包括 other metal金屬 giants豪門, at super-human超人類 speeds速度.
37
110322
4424
在其他高速行駛的車輛間行駛
02:06
Okay? And all you have to guide指南 you are your two eyes眼睛.
38
114746
4480
只有你的眼睛能幫助你
02:11
Okay, so that's all you have,
39
119226
1762
對 就只有你的眼睛
02:12
eyes眼睛 that weren't really designed設計 for this task任務,
40
120988
1735
我們的眼睛不是用來完成這任務的
02:14
but then people ask you to do things like,
41
122723
3751
可是人們總要求你做這樣的事
02:18
you want to make a lane車道 change更改,
42
126474
1549
例如你要換條車道
02:20
what's the first thing they ask you do?
43
128023
2321
第一件要做的事是什麼?
02:22
Take your eyes眼睛 off the road. That's right.
44
130344
3095
不看路面的情況 是的
02:25
Stop looking where you're going, turn,
45
133439
2096
不去看你前進的方向 然後轉彎
02:27
check your blind spot,
46
135535
2018
檢查是不是有盲點
02:29
and drive駕駛 down the road without looking where you're going.
47
137553
3471
不看路 一直向前開
02:33
You and everyone大家 else其他. This is the safe安全 way to drive駕駛.
48
141024
3135
在座各位和其他人
都會認為這是安全駕駛
02:36
Why do we do this? Because we have to,
49
144159
2241
為什麼要這樣做? 因為我們必須
02:38
we have to make a choice選擇, do I look here or do I look here?
50
146400
2579
必須選擇 是要看這邊 還是那邊
02:40
What's more important重要?
51
148979
1521
更重要的是甚麼?
02:42
And usually平時 we do a fantastic奇妙 job工作
52
150500
2711
通常我們都能在路上
02:45
picking選擇 and choosing選擇 what we attend出席 to on the road.
53
153211
3769
好好挑選我們要走的路
02:48
But occasionally偶爾 we miss小姐 something.
54
156980
3650
但偶爾我們也會出小差池
02:52
Occasionally偶爾 we sense something wrong錯誤 or too late晚了.
55
160630
4461
有時候我們意識到做錯決定 已經太遲
02:57
In countless無數 accidents事故, the driver司機 says,
56
165091
1988
無數的車禍 司機都說
02:59
"I didn't see it coming未來."
57
167079
2308
沒看到有車駛過來
03:01
And I believe that. I believe that.
58
169387
3281
我相信這種說法 非常相信
03:04
We can only watch so much.
59
172668
2925
我們能看到的有限
03:07
But the technology技術 exists存在 now that can help us improve提高 that.
60
175593
5144
不過現今科技可以幫助我們
03:12
In the future未來, with cars汽車 exchanging交換 data數據 with each other,
61
180737
4296
將來只要車輛間能互相交換資訊
03:17
we will be able能夠 to see not just three cars汽車 ahead
62
185033
3928
我們不僅可以看到前面三部輛車
03:20
and three cars汽車 behind背後, to the right and left,
63
188961
1594
還能看到後面三輛車 右邊的 左邊的
03:22
all at the same相同 time, bird's鳥類 eye view視圖,
64
190555
3166
全都看到 就像在空中俯瞰
03:25
we will actually其實 be able能夠 to see into those cars汽車.
65
193721
3128
我們能夠實際觀察其他車
03:28
We will be able能夠 to see the velocity速度 of the car汽車 in front面前 of us,
66
196849
2371
我們可以知道前面車輛的速度
03:31
to see how fast快速 that guy's傢伙 going or stopping停止.
67
199220
3240
其他司機開得多快或是否要停車
03:34
If that guy's傢伙 going down to zero, I'll know.
68
202460
4510
他要停車 我也會知道
03:38
And with computation計算 and algorithms算法 and predictive預測 models楷模,
69
206970
3859
透過電腦計算、規則系統和預知模擬
03:42
we will be able能夠 to see the future未來.
70
210829
3273
我們可以預知未來
03:46
You may可能 think that's impossible不可能.
71
214102
1556
或許大家會覺得不可能
03:47
How can you predict預測 the future未來? That's really hard.
72
215658
2731
我們怎能預知未來 這太難了
03:50
Actually其實, no. With cars汽車, it's not impossible不可能.
73
218389
3619
其實這不難 車輛可以做到
03:54
Cars汽車 are three-dimensional三維 objects對象
74
222008
2732
車是立體的
03:56
that have a fixed固定 position位置 and velocity速度.
75
224740
2332
有固定的位置和速度
03:59
They travel旅行 down roads道路.
76
227072
1631
在路上行駛
04:00
Often經常 they travel旅行 on pre-published預發布 routes路線.
77
228703
2412
通常在規劃好的路上行駛
04:03
It's really not that hard to make reasonable合理 predictions預測
78
231115
3938
要合理地預測 車子在短時間內
04:07
about where a car's汽車 going to be in the near future未來.
79
235053
2864
會怎樣 並不困難
04:09
Even if, when you're in your car汽車
80
237917
2002
即使你駕駛時
04:11
and some motorcyclist摩托車手 comes -- bshoombshoom! --
81
239919
1994
遇到幾台機車 咻 !
04:13
85 miles英里 an hour小時 down, lane-splitting車道分割 --
82
241913
2296
以時速85英哩呼嘯而過 車子紛紛讓道
04:16
I know you've had this experience經驗 --
83
244209
2547
相信大家都有這經驗吧
04:18
that guy didn't "just come out of nowhere無處."
84
246756
2603
那人可不是憑空出現
04:21
That guy's傢伙 been on the road probably大概 for the last half hour小時.
85
249359
3643
他說不定都已經在路上跑了半小時
04:25
(Laughter笑聲)
86
253002
1190
[笑聲]
04:26
Right? I mean, somebody's某人的 seen看到 him.
87
254192
3589
是吧?我是說 一定有人看到他
04:29
Ten, 20, 30 miles英里 back, someone's誰家 seen看到 that guy,
88
257781
2768
10哩前 20哩前 30哩前, 一定有人看到他
04:32
and as soon不久 as one car汽車 sees看到 that guy
89
260549
2384
而只要有一部車看到他
04:34
and puts看跌期權 him on the map地圖, he's on the map地圖 --
90
262933
2231
把他放到地圖上 大家都會知道
04:37
position位置, velocity速度,
91
265164
2176
他的位置和速度
04:39
good estimate估計 he'll地獄 continue繼續 going 85 miles英里 an hour小時.
92
267340
2321
估計他會以每小時85英哩速度持續前進
04:41
You'll你會 know, because your car汽車 will know, because
93
269661
2184
你的車子預報 你便會知道這訊息
04:43
that other car汽車 will have whispered低聲道 something in his ear,
94
271845
2275
其他車也會對司機碎碎念
04:46
like, "By the way, five minutes分鐘,
95
274120
1923
就像: 順便一提 再過5分鐘
04:48
motorcyclist摩托車手, watch out."
96
276043
2775
便會有機車 要小心
04:50
You can make reasonable合理 predictions預測 about how cars汽車 behave表現.
97
278818
2703
你可以合理地預測車在路的情況
04:53
I mean, they're Newtonian牛頓 objects對象.
98
281521
1365
我是說 車子也遵從牛頓定律
04:54
That's very nice不錯 about them.
99
282886
2909
還好是這樣
04:57
So how do we get there?
100
285795
3034
那 我們要如何走到那一步?
05:00
We can start開始 with something as simple簡單
101
288829
2266
我們可以開始分享我們車與車的位置
05:03
as sharing分享 our position位置 data數據 between之間 cars汽車,
102
291095
2870
從這件小事做起
05:05
just sharing分享 GPS全球定位系統.
103
293965
1892
就只是分享 GPS (全球定位系統)
05:07
If I have a GPS全球定位系統 and a camera相機 in my car汽車,
104
295857
2444
如果我的車子裝有 GPS 和鏡頭
05:10
I have a pretty漂亮 precise精確 idea理念 of where I am
105
298301
2231
我就能精確地知道我所在的位置
05:12
and how fast快速 I'm going.
106
300532
1732
以及我是以多快的速度前進
05:14
With computer電腦 vision視力, I can estimate估計 where
107
302264
1657
有電腦輔助
我還可以估算我旁邊哪裡有車
05:15
the cars汽車 around me are, sort分類 of, and where they're going.
108
303921
3537
以及他們要往哪裡去
05:19
And same相同 with the other cars汽車.
109
307458
970
其他車也可以知道
05:20
They can have a precise精確 idea理念 of where they are,
110
308428
1814
他們可以知道自己身在何方
05:22
and sort分類 of a vague模糊 idea理念 of where the other cars汽車 are.
111
310242
2146
也能大約了解其他車子的位置
05:24
What happens發生 if two cars汽車 share分享 that data數據,
112
312388
3231
那麼如果兩部車子分享資訊, 會如何呢?
05:27
if they talk to each other?
113
315619
1955
假設它們可以對話呢?
05:29
I can tell you exactly究竟 what happens發生.
114
317574
2778
我可以告訴你會發生什麼事
05:32
Both models楷模 improve提高.
115
320352
2339
兩部車都獲得行進間的改善
05:34
Everybody每個人 wins.
116
322691
2055
是雙贏
05:36
Professor教授 Bob短發 Wang and his team球隊
117
324746
2577
Bob Wang 教授和他的團隊
05:39
have doneDONE computer電腦 simulations模擬 of what happens發生
118
327323
2738
做了一個電腦模擬
看下列情況會產生什麼結果
05:42
when fuzzy模糊 estimates估計 combine結合, even in light traffic交通,
119
330061
3431
當集合一些粗估的資料
用在行車順暢的情況下,
05:45
when cars汽車 just share分享 GPS全球定位系統 data數據,
120
333492
2624
讓車與車只是交換 GPS 資訊
05:48
and we've我們已經 moved移動 this research研究 out of the computer電腦 simulation模擬
121
336116
2513
我們將這個研究帶離電腦模擬
05:50
and into robot機器人 test測試 beds that have the actual實際 sensors傳感器
122
338629
3027
帶到裝有感應器的機器實驗上
05:53
that are in cars汽車 now on these robots機器人:
123
341656
3133
把原本放車子的裝置, 用到機器身上
05:56
stereo立體聲 cameras相機, GPS全球定位系統,
124
344789
1838
音響, 鏡頭, GPS
05:58
and the two-dimensional二維 laser激光 range範圍 finders發現者
125
346627
1874
還有平面雷射測距儀
06:00
that are common共同 in backup備用 systems系統.
126
348501
2240
這些都是車子的標準配備
06:02
We also attach連接 a discrete離散的 short-range短距離 communication通訊 radio無線電,
127
350741
4484
我們還裝上一台短距的無線電機台
06:07
and the robots機器人 talk to each other.
128
355225
1909
讓機器可以相互溝通
06:09
When these robots機器人 come at each other,
129
357134
1539
當兩部機器朝彼此前進
06:10
they track跟踪 each other's其他 position位置 precisely恰恰,
130
358673
2971
它們可以追蹤到彼此的精確位置
06:13
and they can avoid避免 each other.
131
361644
2737
因此可避免互撞
06:16
We're now adding加入 more and more robots機器人 into the mix混合,
132
364381
3226
我們現在在這個實驗中加入越來越多的機器
06:19
and we encountered遇到 some problems問題.
133
367607
1471
然後我們遇到了一些狀況
06:21
One of the problems問題, when you get too much chatter喋喋不休,
134
369078
2359
其一便是, 太吵了
當接收到太多瑣碎的言語
06:23
it's hard to process處理 all the packets, so you have to prioritize優先,
135
371437
3728
就很難去處理所有的資訊封包
這時就得選擇優先順序
06:27
and that's where the predictive預測 model模型 helps幫助 you.
136
375165
2357
而預知模型系統
在這個時候就派上用場了
06:29
If your robot機器人 cars汽車 are all tracking追踪 the predicted預料到的 trajectories軌跡,
137
377522
4372
如果機器人車子
一直行駛在可預期的軌道上
06:33
you don't pay工資 as much attention注意 to those packets.
138
381894
1767
就不用費心去解讀那些資訊封包
06:35
You prioritize優先 the one guy
139
383661
1703
首先被挑出的這個人
06:37
who seems似乎 to be going a little off course課程.
140
385364
1333
往往就是有點偏移行進路線
06:38
That guy could be a problem問題.
141
386697
2526
那麼這個人就有可能是個麻煩
06:41
And you can predict預測 the new trajectory彈道.
142
389223
3002
然後你就可以再規劃一條新的軌道
06:44
So you don't only know that he's going off course課程, you know how.
143
392225
2763
所以你不只知道他要偏離航道了
還知道他要怎麼偏
06:46
And you know which哪一個 drivers司機 you need to alert警報 to get out of the way.
144
394988
3725
也能知道遇到那個駕駛要提高警覺, 離遠一點
06:50
And we wanted to do -- how can we best最好 alert警報 everyone大家?
145
398713
2633
我們希望 -- 我們要如何警告其他人?
06:53
How can these cars汽車 whisper耳語, "You need to get out of the way?"
146
401346
3183
這些車要怎麼小聲警告:
"你得趕緊離開這裡"
06:56
Well, it depends依靠 on two things:
147
404529
1517
嗯, 這得要兩個條件配合
06:58
one, the ability能力 of the car汽車,
148
406046
2169
一, 車子的性能
07:00
and second第二 the ability能力 of the driver司機.
149
408215
3217
二, 駕駛本身的技巧
07:03
If one guy has a really great car汽車,
150
411432
1505
就算這個人擁有一部很讚的車
07:04
but they're on their phone電話 or, you know, doing something,
151
412937
2925
但是他如果邊開邊做其他事
像講手機之類的
07:07
they're not probably大概 in the best最好 position位置
152
415862
1930
他們可能就無法在緊急狀態中
07:09
to react應對 in an emergency.
153
417792
2970
做出最佳的反應
07:12
So we started開始 a separate分離 line of research研究
154
420762
1665
因此我們又著手另一項實驗
07:14
doing driver司機 state modeling造型.
155
422427
2551
做有關駕駛者的狀態模型
07:16
And now, using運用 a series系列 of three cameras相機,
156
424978
2329
我們使用三個一組的攝像機
07:19
we can detect檢測 if a driver司機 is looking forward前鋒,
157
427307
2270
來偵測這個駕駛人是看前面
07:21
looking away, looking down, on the phone電話,
158
429577
2860
看別的地方, 看下面, 還是講手機
07:24
or having a cup杯子 of coffee咖啡.
159
432437
3061
或喝咖啡
07:27
We can predict預測 the accident事故
160
435498
2070
我們可以預知事故
07:29
and we can predict預測 who, which哪一個 cars汽車,
161
437568
3651
我們可以知道是誰以及哪一部車子
07:33
are in the best最好 position位置 to move移動 out of the way
162
441219
3486
最好離開目前的道路
07:36
to calculate計算 the safest最安全 route路線 for everyone大家.
163
444705
3009
並且幫每一個人規劃最安全的路徑
07:39
Fundamentally從根本上, these technologies技術 exist存在 today今天.
164
447714
4635
最重要的是
這些科技在今天都已經有了
07:44
I think the biggest最大 problem問題 that we face面對
165
452349
2824
我認為眼前最大的困難是
07:47
is our own擁有 willingness願意 to share分享 our data數據.
166
455173
3013
大家是否願意做資訊分享
07:50
I think it's a very disconcerting令人不安 notion概念,
167
458186
2631
我想這樣的概念或許讓人有點不安
07:52
this idea理念 that our cars汽車 will be watching觀看 us,
168
460817
2386
因為我們的車子會監視著我們
07:55
talking about us to other cars汽車,
169
463203
3371
還跟別的車子打小報告
07:58
that we'll be going down the road in a sea of gossip八卦.
170
466574
3427
簡直就像航行在一片流言蜚語的汪洋中
08:02
But I believe it can be doneDONE in a way that protects保護 our privacy隱私,
171
470001
3897
但是我相信, 我們還是可以
在保有隱私的情況下執行
08:05
just like right now, when I look at your car汽車 from the outside,
172
473898
3741
好比現在, 如果我從外面看你的車
08:09
I don't really know about you.
173
477639
2363
我不會知道你是何許人物
08:12
If I look at your license執照 plate盤子 number,
174
480002
1137
就算看著你的牌照號碼
08:13
I don't really know who you are.
175
481139
1886
我也不會知道你是誰
08:15
I believe our cars汽車 can talk about us behind背後 our backs.
176
483025
4249
我相信我們的車子會在我們背後說長道短
08:19
(Laughter笑聲)
177
487274
2975
[笑聲]
08:22
And I think it's going to be a great thing.
178
490249
3185
但我認為那將會是件好事
08:25
I want you to consider考慮 for a moment時刻
179
493434
1650
花幾分鐘想想
08:27
if you really don't want the distracted分心 teenager青少年 behind背後 you
180
495084
4118
你是不是真的不希望
那煩人的青少年在你背後
08:31
to know that you're braking制動,
181
499202
2120
知道你要煞車
08:33
that you're coming未來 to a dead stop.
182
501322
2924
知道你要停下來
08:36
By sharing分享 our data數據 willingly甘心,
183
504246
2741
倘若各位願意分享
08:38
we can do what's best最好 for everyone大家.
184
506987
2812
我們可以幫每個人做到最好
08:41
So let your car汽車 gossip八卦 about you.
185
509799
3076
所以 讓你的車去東家長西家短吧
08:44
It's going to make the roads道路 a lot safer更安全.
186
512875
3038
這能讓道路使用更為安全
08:47
Thank you.
187
515913
1791
謝謝各位
08:49
(Applause掌聲)
188
517704
4985
[掌聲]
Translated by KA WAI WONG
Reviewed by Emma Chiang

▲Back to top

ABOUT THE SPEAKER
Jennifer Healey - Research scientist
A research scientist at Intel, Jennifer Healey develops the mobile internet devices of the future.

Why you should listen

Jennifer Healey imagines a future where computers and smartphones are capable of being sensitive to human emotions and where cars are able to talk to each other, and thus keep their drivers away from accidents. A scientist at Intel Corporation Research Labs, she researches devices and systems that would allow for these major innovations.

Healey holds PhD from MIT in electrical engineering and computer science. While there, she pioneered “Affective Computing” with Rosalind Picard and developed the first wearable computer with physiological sensors and a video camera that allows the wearer to track their daily activities and how they feel while doing them. From there, she moved to IBM where she worked on the next generation of multi-modal interactive smartphones and helped architect the "Interaction Mark-Up language" that allows users to switch from voice to speech input seamlessly.

Healey has also used her interest in embedded devices in the field of healthcare. While an instructor at Harvard Medical School and at Beth Israel Deaconess Medical Center, she worked on new ways to use heart rate to predict cardiac health. She then joined HP Research in Cambridge to further develop wearable sensors for health monitoring and continued this research when she joined Intel Digital Health.

More profile about the speaker
Jennifer Healey | Speaker | TED.com