TED@BCG Milan
Margaret Mitchell: How we can build AI to help humans, not hurt us
瑪格莉特．米契爾: 如何建立人工智慧來協助，而不是傷害人類
Filmed:
Readability: 4.6
1,154,210 views
身為 Google 的研究科學家，瑪格莉特．米契爾協助開發能夠溝通看見什麼、了解什麼的電腦。她說了一個警惕的故事，談到我們下意識編碼到人工智慧中的隔閡、盲點、偏見，並請大家思考我們現今所創造的科技在未來的意義是什麼。她說：「我們現在所看見的一切，都是人工智慧演進的約略寫照。如果我們想讓人工智慧以對人類有助益的方式演進，那麼現在就得要定義目標和策略，來讓那條路成為可能。」
Margaret Mitchell - AI research scientist
Margaret Mitchell is a senior research scientist in Google's Research & Machine Intelligence group, working on artificial intelligence. Full bio
Margaret Mitchell is a senior research scientist in Google's Research & Machine Intelligence group, working on artificial intelligence. Full bio
Double-click the English transcript below to play the video.
00:13
I work on helping computers
communicate about the world around us.
communicate about the world around us.
0
1381
4015
我的工作是在協助電腦
和我們周遭的世界交流。
和我們周遭的世界交流。
00:17
There are a lot of ways to do this,
1
5754
1793
要這麼做，有很多方式，
00:19
and I like to focus on helping computers
2
7571
2592
我喜歡聚焦在協助電腦
00:22
to talk about what they see
and understand.
and understand.
3
10187
2874
來談它們看見什麼、了解什麼。
00:25
Given a scene like this,
4
13514
1571
給予一個這樣的情景，
00:27
a modern computer-vision algorithm
5
15109
1905
一個現代電腦視覺演算法
00:29
can tell you that there's a woman
and there's a dog.
and there's a dog.
6
17038
3095
就能告訴你情景中
有一個女子和一隻狗。
有一個女子和一隻狗。
00:32
It can tell you that the woman is smiling.
7
20157
2706
它能告訴你，女子在微笑。
00:34
It might even be able to tell you
that the dog is incredibly cute.
that the dog is incredibly cute.
8
22887
3873
它甚至可能可以告訴你，
那隻狗相當可愛。
那隻狗相當可愛。
00:38
I work on this problem
9
26784
1349
我處理這個問題時
00:40
thinking about how humans
understand and process the world.
understand and process the world.
10
28157
4212
腦中想的是人類如何
了解和處理這個世界。
了解和處理這個世界。
00:45
The thoughts, memories and stories
11
33577
2952
對人類而言，這樣的情景有可能會
00:48
that a scene like this
might evoke for humans.
might evoke for humans.
12
36553
2818
喚起什麼樣的思想、記憶、故事。
00:51
All the interconnections
of related situations.
of related situations.
13
39395
4285
相關情況的所有相互連結。
00:55
Maybe you've seen
a dog like this one before,
a dog like this one before,
14
43704
3126
也許你以前看過像這樣的狗，
00:58
or you've spent time
running on a beach like this one,
running on a beach like this one,
15
46854
2969
或者你曾花時間在
像這樣的海灘上跑步，
像這樣的海灘上跑步，
01:01
and that further evokes thoughts
and memories of a past vacation,
and memories of a past vacation,
16
49847
4778
那就會進一步喚起一個
過去假期的思想和記憶，
過去假期的思想和記憶，
01:06
past times to the beach,
17
54649
1920
過去在海灘的時光，
01:08
times spent running around
with other dogs.
with other dogs.
18
56593
2603
花在和其他狗兒到處奔跑的時間。
01:11
One of my guiding principles
is that by helping computers to understand
is that by helping computers to understand
19
59688
5207
我的指導原則之一，
是要協助電腦了解
是要協助電腦了解
01:16
what it's like to have these experiences,
20
64919
2896
有這些經驗是什麼樣的感覺，
01:19
to understand what we share
and believe and feel,
and believe and feel,
21
67839
5176
去了解我們共有什麼、
相信什麼、感受到什麼，
相信什麼、感受到什麼，
01:26
then we're in a great position
to start evolving computer technology
to start evolving computer technology
22
74094
4310
那麼，我們就有很好的機會
可以開始讓電腦科技
可以開始讓電腦科技
01:30
in a way that's complementary
with our own experiences.
with our own experiences.
23
78428
4587
以一種和我們自身經驗
互補的方式來演化。
互補的方式來演化。
01:35
So, digging more deeply into this,
24
83539
3387
所以，我更深入研究這主題，
01:38
a few years ago I began working on helping
computers to generate human-like stories
computers to generate human-like stories
25
86950
5905
幾年前，我開始努力協助
電腦產生出像人類一樣的故事，
電腦產生出像人類一樣的故事，
01:44
from sequences of images.
26
92879
1666
從一連串的影像來產生。
01:47
So, one day,
27
95427
1904
有一天，
01:49
I was working with my computer to ask it
what it thought about a trip to Australia.
what it thought about a trip to Australia.
28
97355
4622
我在問我的電腦，它對於
去澳洲旅行有什麼想法。
去澳洲旅行有什麼想法。
01:54
It took a look at the pictures,
and it saw a koala.
and it saw a koala.
29
102768
2920
它看了照片，看見一隻無尾熊。
01:58
It didn't know what the koala was,
30
106236
1643
它不知道無尾熊是什麼，
01:59
but it said it thought
it was an interesting-looking creature.
it was an interesting-looking creature.
31
107903
2999
但它說它認為這是一隻
看起來很有趣的生物。
看起來很有趣的生物。
02:04
Then I shared with it a sequence of images
about a house burning down.
about a house burning down.
32
112053
4004
我和它分享了一系列的圖片，
都和房子被燒毀有關。
都和房子被燒毀有關。
02:09
It took a look at the images and it said,
33
117704
3285
它看了那些圖片，說：
02:13
"This is an amazing view!
This is spectacular!"
This is spectacular!"
34
121013
3500
「這是好棒的景色！這好壯觀！」
02:17
It sent chills down my spine.
35
125450
2095
它讓我背脊發涼。
02:20
It saw a horrible, life-changing
and life-destroying event
and life-destroying event
36
128983
4572
它看著一個會改變人生、
摧毀生命的可怕事件，
摧毀生命的可怕事件，
02:25
and thought it was something positive.
37
133579
2382
卻以為它是很正面的東西。
02:27
I realized that it recognized
the contrast,
the contrast,
38
135985
3441
我了解到，它能認出對比反差、
02:31
the reds, the yellows,
39
139450
2699
紅色、黃色，
02:34
and thought it was something
worth remarking on positively.
worth remarking on positively.
40
142173
3078
然後就認為它是
值得正面評論的東西。
值得正面評論的東西。
02:37
And part of why it was doing this
41
145928
1615
它這麼做的部分原因，
02:39
was because most
of the images I had given it
of the images I had given it
42
147577
2945
是因為我給它的大多數圖片
02:42
were positive images.
43
150546
1840
都是正面的圖片。
02:44
That's because people
tend to share positive images
tend to share positive images
44
152903
3658
那是因為人在談論他們的經驗時，
02:48
when they talk about their experiences.
45
156585
2190
本來就傾向會分享正面的圖片。
02:51
When was the last time
you saw a selfie at a funeral?
you saw a selfie at a funeral?
46
159267
2541
你上次看到在葬禮上的
自拍照是何時？
自拍照是何時？
02:55
I realized that,
as I worked on improving AI
as I worked on improving AI
47
163434
3095
我了解到，當我努力
在改善人工智慧，
在改善人工智慧，
02:58
task by task, dataset by dataset,
48
166553
3714
一個任務一個任務、一個資料集
一個資料集地改善，
一個資料集地改善，
03:02
that I was creating massive gaps,
49
170291
2897
結果我卻在「它能了解什麼」上
03:05
holes and blind spots
in what it could understand.
in what it could understand.
50
173212
3999
創造出了大量的隔閡、
漏洞，以及盲點。
漏洞，以及盲點。
03:10
And while doing so,
51
178307
1334
這麼做的時候，
03:11
I was encoding all kinds of biases.
52
179665
2483
我是在把各種偏見做編碼。
03:15
Biases that reflect a limited viewpoint,
53
183029
3318
這些偏見反映出受限的觀點，
03:18
limited to a single dataset --
54
186371
2261
受限於單一資料集──
03:21
biases that can reflect
human biases found in the data,
human biases found in the data,
55
189283
3858
這些偏見能反應出
在資料中的人類偏見，
在資料中的人類偏見，
03:25
such as prejudice and stereotyping.
56
193165
3104
比如偏袒以及刻板印象。
03:29
I thought back to the evolution
of the technology
of the technology
57
197554
3057
我回頭去想一路帶我走到
03:32
that brought me to where I was that day --
58
200635
2502
那個時點的科技演化──
03:35
how the first color images
59
203966
2233
第一批彩色影像如何
03:38
were calibrated against
a white woman's skin,
a white woman's skin,
60
206223
3048
根據一個白種女子的皮膚來做校準，
03:41
meaning that color photography
was biased against black faces.
was biased against black faces.
61
209665
4145
這表示，彩色照片對於
黑皮膚臉孔是有偏見的。
黑皮膚臉孔是有偏見的。
03:46
And that same bias, that same blind spot
62
214514
2925
同樣的偏見，同樣的盲點，
03:49
continued well into the '90s.
63
217463
1867
持續湧入了九○年代。
03:51
And the same blind spot
continues even today
continues even today
64
219701
3154
而同樣的盲點甚至持續到現今，
03:54
in how well we can recognize
different people's faces
different people's faces
65
222879
3698
出現在我們對於不同人的
臉部辨識能力中，
臉部辨識能力中，
03:58
in facial recognition technology.
66
226601
2200
在人臉辨識技術中。
04:01
I though about the state of the art
in research today,
in research today,
67
229323
3143
我思考了現今在研究上發展水平，
04:04
where we tend to limit our thinking
to one dataset and one problem.
to one dataset and one problem.
68
232490
4514
我們傾向會把我們的思路限制
在一個資料集或一個問題上。
在一個資料集或一個問題上。
04:09
And that in doing so, we were creating
more blind spots and biases
more blind spots and biases
69
237688
4881
這麼做時，我們就會
創造出更多盲點和偏見，
創造出更多盲點和偏見，
04:14
that the AI could further amplify.
70
242593
2277
它們可能會被人工智慧給放大。
04:17
I realized then
that we had to think deeply
that we had to think deeply
71
245712
2079
那時，我了解到，我們必須要
04:19
about how the technology we work on today
looks in five years, in 10 years.
looks in five years, in 10 years.
72
247815
5519
深入思考我們現今努力發展的科技，
在五年、十年後會是什麼樣子。
在五年、十年後會是什麼樣子。
04:25
Humans evolve slowly,
with time to correct for issues
with time to correct for issues
73
253990
3142
人類演進很慢，有時間可以去修正
04:29
in the interaction of humans
and their environment.
and their environment.
74
257156
3534
在人類互動以及其環境中的議題。
04:33
In contrast, artificial intelligence
is evolving at an incredibly fast rate.
is evolving at an incredibly fast rate.
75
261276
5429
相對的，人工智慧的
演進速度非常快。
演進速度非常快。
04:39
And that means that it really matters
76
267013
1773
那就意味著，很重要的是
04:40
that we think about this
carefully right now --
carefully right now --
77
268810
2317
我們現在要如何仔細思考這件事──
04:44
that we reflect on our own blind spots,
78
272180
3008
我們要反省我們自己的盲點，
04:47
our own biases,
79
275212
2317
我們自己的偏見，
04:49
and think about how that's informing
the technology we're creating
the technology we're creating
80
277553
3857
並想想它們帶給我們所
創造出的科技什麼樣的資訊，
創造出的科技什麼樣的資訊，
04:53
and discuss what the technology of today
will mean for tomorrow.
will mean for tomorrow.
81
281434
3902
並討論現今的科技在
將來代表的是什麼意涵。
將來代表的是什麼意涵。
04:58
CEOs and scientists have weighed in
on what they think
on what they think
82
286593
3191
對於未來的人工智慧
應該是什麼樣子，
應該是什麼樣子，
05:01
the artificial intelligence technology
of the future will be.
of the future will be.
83
289808
3325
執行長和科學家的意見
是很有份量的。
是很有份量的。
05:05
Stephen Hawking warns that
84
293157
1618
史蒂芬霍金警告過：
05:06
"Artificial intelligence
could end mankind."
could end mankind."
85
294799
3007
「人工智慧可能終結人類。」
05:10
Elon Musk warns
that it's an existential risk
that it's an existential risk
86
298307
2683
伊隆馬斯克警告過，
它是個生存風險，
它是個生存風險，
05:13
and one of the greatest risks
that we face as a civilization.
that we face as a civilization.
87
301014
3574
也是我們人類文明所
面臨最大的風險之一。
面臨最大的風險之一。
05:17
Bill Gates has made the point,
88
305665
1452
比爾蓋茲有個論點：
05:19
"I don't understand
why people aren't more concerned."
why people aren't more concerned."
89
307141
3185
「我不了解為什麼
人們不更關心一點。」
人們不更關心一點。」
05:23
But these views --
90
311412
1318
但，這些看法──
05:25
they're part of the story.
91
313618
1734
它們是故事的一部分。
05:28
The math, the models,
92
316079
2420
數學、模型，
05:30
the basic building blocks
of artificial intelligence
of artificial intelligence
93
318523
3070
人工智慧的基礎材料
05:33
are something that we call access
and all work with.
and all work with.
94
321617
3135
是我們所有人都能夠取得並使用的。
05:36
We have open-source tools
for machine learning and intelligence
for machine learning and intelligence
95
324776
3785
我們有機器學習和智慧用的
開放原始碼工具，
開放原始碼工具，
05:40
that we can contribute to.
96
328585
1734
我們都能對其做出貢獻。
05:42
And beyond that,
we can share our experience.
we can share our experience.
97
330919
3340
在那之外，我們可以
分享我們的經驗。
分享我們的經驗。
05:46
We can share our experiences
with technology and how it concerns us
with technology and how it concerns us
98
334760
3468
分享關於科技、它如何影響我們、
05:50
and how it excites us.
99
338252
1467
它如何讓我們興奮的經驗。
05:52
We can discuss what we love.
100
340251
1867
我們可以討論我們所愛的。
05:55
We can communicate with foresight
101
343244
2031
我們能帶著遠見來交流，
05:57
about the aspects of technology
that could be more beneficial
that could be more beneficial
102
345299
4857
談談關於科技有哪些面向，
隨著時間發展可能可以更有助益，
或可能產生問題。
或可能產生問題。
06:02
or could be more problematic over time.
103
350180
2600
06:05
If we all focus on opening up
the discussion on AI
the discussion on AI
104
353799
4143
若我們都能把焦點放在
開放地帶著對未來的遠見
來討論人工智慧，
來討論人工智慧，
06:09
with foresight towards the future,
105
357966
1809
06:13
this will help create a general
conversation and awareness
conversation and awareness
106
361093
4270
這就能創造出一般性的談話和意識，
06:17
about what AI is now,
107
365387
2513
關於人工智慧現在是什麼樣子、
06:21
what it can become
108
369212
2001
它未來可以變成什麼樣子，
06:23
and all the things that we need to do
109
371237
1785
以及所有我們需要做的事，
06:25
in order to enable that outcome
that best suits us.
that best suits us.
110
373046
3753
以產生出最適合我們的結果。
06:29
We already see and know this
in the technology that we use today.
in the technology that we use today.
111
377490
3674
我們已經在現今我們所使用的
科技中看見這一點了。
科技中看見這一點了。
06:33
We use smart phones
and digital assistants and Roombas.
and digital assistants and Roombas.
112
381767
3880
我們用智慧手機、數位助理，
以及掃地機器人。
以及掃地機器人。
06:38
Are they evil?
113
386457
1150
它們邪惡嗎？
06:40
Maybe sometimes.
114
388268
1547
也許有時候。
06:42
Are they beneficial?
115
390664
1333
它們有助益嗎？
06:45
Yes, they're that, too.
116
393005
1533
是的，這也是事實。
06:48
And they're not all the same.
117
396236
1761
且它們並非全都一樣的。
06:50
And there you already see
a light shining on what the future holds.
a light shining on what the future holds.
118
398489
3540
你們已經看到未來
可能性的一絲光芒。
可能性的一絲光芒。
06:54
The future continues on
from what we build and create right now.
from what we build and create right now.
119
402942
3619
未來延續的基礎，是我們
現在所建立和創造的。
現在所建立和創造的。
06:59
We set into motion that domino effect
120
407165
2642
我們開始了骨牌效應，
07:01
that carves out AI's evolutionary path.
121
409831
2600
刻劃出了人工智慧的演進路徑。
07:05
In our time right now,
we shape the AI of tomorrow.
we shape the AI of tomorrow.
122
413173
2871
我們在現在這時代
形塑未來的人工智慧。
形塑未來的人工智慧。
07:08
Technology that immerses us
in augmented realities
in augmented realities
123
416566
3699
讓我們能沉浸入擴增實境中的科技，
07:12
bringing to life past worlds.
124
420289
2566
讓過去的世界又活了過來。
07:15
Technology that helps people
to share their experiences
to share their experiences
125
423844
4312
協助人們在溝通困難時
07:20
when they have difficulty communicating.
126
428180
2262
還能分享經驗的科技。
07:23
Technology built on understanding
the streaming visual worlds
the streaming visual worlds
127
431323
4532
立基在了解串流
視覺世界之上的科技，
視覺世界之上的科技，
07:27
used as technology for self-driving cars.
128
435879
3079
被用來當作自動駕駛汽車的科技。
07:32
Technology built on understanding images
and generating language,
and generating language,
129
440490
3413
立基在了解圖像和產生語言的科技，
07:35
evolving into technology that helps people
who are visually impaired
who are visually impaired
130
443927
4063
演進成協助視覺損傷者的科技，
07:40
be better able to access the visual world.
131
448014
2800
讓他們更能進入視覺的世界。
07:42
And we also see how technology
can lead to problems.
can lead to problems.
132
450838
3261
我們也看到了科技如何導致問題。
07:46
We have technology today
133
454885
1428
現今，我們有科技
07:48
that analyzes physical
characteristics we're born with --
characteristics we're born with --
134
456337
3835
能夠分析我們天生的身體特徵──
07:52
such as the color of our skin
or the look of our face --
or the look of our face --
135
460196
3272
比如膚色或面部的外觀──
07:55
in order to determine whether or not
we might be criminals or terrorists.
we might be criminals or terrorists.
136
463492
3804
可以用來判斷我們是否
有可能是罪犯或恐怖份子。
有可能是罪犯或恐怖份子。
07:59
We have technology
that crunches through our data,
that crunches through our data,
137
467688
2905
我們有科技能夠分析我們的資料，
08:02
even data relating
to our gender or our race,
to our gender or our race,
138
470617
2896
甚至和我們的性別
或種族相關的資料，
或種族相關的資料，
08:05
in order to determine whether or not
we might get a loan.
we might get a loan.
139
473537
2865
來決定我們的貸款是否能被核准。
08:09
All that we see now
140
477494
1579
我們現在所看見的一切，
08:11
is a snapshot in the evolution
of artificial intelligence.
of artificial intelligence.
141
479097
3617
都是人工智慧演進的約略寫照。
08:15
Because where we are right now,
142
483763
1778
因為我們現在所處的位置，
08:17
is within a moment of that evolution.
143
485565
2238
是在那演進的一個時刻當中。
08:20
That means that what we do now
will affect what happens down the line
will affect what happens down the line
144
488690
3802
那表示，我們現在
所做的，會影響到後續
所做的，會影響到後續
08:24
and in the future.
145
492516
1200
未來發生的事。
08:26
If we want AI to evolve
in a way that helps humans,
in a way that helps humans,
146
494063
3951
如果我們想讓人工智慧的
演進方式是對人類有助益的，
演進方式是對人類有助益的，
08:30
then we need to define
the goals and strategies
the goals and strategies
147
498038
2801
那麼我們現在就得要
定義目標和策略，
定義目標和策略，
08:32
that enable that path now.
148
500863
1733
來讓那條路成為可能。
08:35
What I'd like to see is something
that fits well with humans,
that fits well with humans,
149
503680
3738
我想要看見的東西是要能夠
和人類、我們的文化，
和人類、我們的文化，
08:39
with our culture and with the environment.
150
507442
2800
及我們的環境能非常符合的東西。
08:43
Technology that aids and assists
those of us with neurological conditions
those of us with neurological conditions
151
511435
4484
這種科技要能夠幫助和協助
有神經系統疾病
有神經系統疾病
08:47
or other disabilities
152
515943
1721
或其他殘疾者的人，
08:49
in order to make life
equally challenging for everyone.
equally challenging for everyone.
153
517688
3216
讓人生對於每個人的
挑戰程度是平等的。
挑戰程度是平等的。
08:54
Technology that works
154
522097
1421
這種科技的運作
08:55
regardless of your demographics
or the color of your skin.
or the color of your skin.
155
523542
3933
不會考量你的
人口統計資料或你的膚色。
人口統計資料或你的膚色。
09:00
And so today, what I focus on
is the technology for tomorrow
is the technology for tomorrow
156
528383
4742
所以，現今我著重的是明日的科技
09:05
and for 10 years from now.
157
533149
1733
和十年後的科技。
09:08
AI can turn out in many different ways.
158
536530
2634
產生人工智慧的方式相當多。
09:11
But in this case,
159
539688
1225
但在這個情況中，
09:12
it isn't a self-driving car
without any destination.
without any destination.
160
540937
3328
它並不是沒有目的地的
自動駕駛汽車。
自動駕駛汽車。
09:16
This is the car that we are driving.
161
544884
2400
它是我們在開的汽車。
09:19
We choose when to speed up
and when to slow down.
and when to slow down.
162
547953
3595
我們選擇何時要加速何時要減速。
09:23
We choose if we need to make a turn.
163
551572
2400
我們選擇是否要轉彎。
09:26
We choose what the AI
of the future will be.
of the future will be.
164
554868
3000
我們選擇將來的
人工智慧會是哪一種。
人工智慧會是哪一種。
09:31
There's a vast playing field
165
559186
1337
人工智慧能夠
09:32
of all the things that artificial
intelligence can become.
intelligence can become.
166
560547
2965
變成各式各樣的東西。
09:36
It will become many things.
167
564064
1800
它會變成許多東西。
09:39
And it's up to us now,
168
567694
1732
現在，決定權在我們，
09:41
in order to figure out
what we need to put in place
what we need to put in place
169
569450
3061
我們要想清楚我們得要準備什麼，
09:44
to make sure the outcomes
of artificial intelligence
of artificial intelligence
170
572535
3807
來確保人工智慧的結果
09:48
are the ones that will be
better for all of us.
better for all of us.
171
576366
3066
會是對所有人都更好的結果。
09:51
Thank you.
172
579456
1150
謝謝。
09:52
(Applause)
173
580630
2187
（掌聲）
ABOUT THE SPEAKER
Margaret Mitchell - AI research scientistMargaret Mitchell is a senior research scientist in Google's Research & Machine Intelligence group, working on artificial intelligence.
Why you should listen
Margaret Mitchell's research involves vision-language and grounded language generation, focusing on how to evolve artificial intelligence towards positive goals. Her work combines computer vision, natural language processing, social media as well as many statistical methods and insights from cognitive science. Before Google, Mitchell was a founding member of Microsoft Research's "Cognition" group, focused on advancing artificial intelligence, and a researcher in Microsoft Research's Natural Language Processing group.
Margaret Mitchell | Speaker | TED.com