TED@BCG Milan
Margaret Mitchell: How we can build AI to help humans, not hurt us
玛格丽特‧米切尔: 如何构建人工智能来帮助，而非伤害我们
Filmed:
Readability: 4.6
1,154,210 views
作为谷歌的研究科学家，玛格丽特·米切尔帮助开发电脑，它们能够沟通所看到和理解的事情。她警示，如今我们潜意识地将差距，盲点和偏见编码到人工智能中——我们应该考虑今天创造的技术对未来意味着什么。米切尔说：“我们现在所看到的是人工智能进化过程中的一个快照。如果我们希望人工智能以一种帮助人类的方式发展，那么我们需要定义目标和策略，来开通这这条路径。”
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
我致力于协助电脑和
我们周围世界的沟通。
我们周围世界的沟通。
是有很多方法可以做到这一点，
00:17
There are a lot of ways to do this,
1
5754
1793
我喜欢专注于协助电脑
00:19
and I like to focus on helping computers
2
7571
2592
去谈论它们看到和理解的内容。
00:22
to talk about what they see
and understand.
and understand.
3
10187
2874
00:25
Given a scene like this,
4
13514
1571
鉴于这样的情景，
一个现代的计算机视觉演算法，
00:27
a modern computer-vision algorithm
5
15109
1905
可以告诉你，有一个女人，
还有一只狗。
还有一只狗。
00:29
can tell you that there's a woman
and there's a dog.
and there's a dog.
6
17038
3095
它可以告诉你，那个女人在微笑。
00:32
It can tell you that the woman is smiling.
7
20157
2706
它甚至可以告诉你，
这只狗非常可爱。
这只狗非常可爱。
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
我处理这个问题
00:38
I work on this problem
9
26784
1349
思考人类如何理解和与世界共处。
00:40
thinking about how humans
understand and process the world.
understand and process the world.
10
28157
4212
00:45
The thoughts, memories and stories
11
33577
2952
那些思想，记忆和故事
00:48
that a scene like this
might evoke for humans.
might evoke for humans.
12
36553
2818
在这样的场景中，
可能会唤起人类的注意。
可能会唤起人类的注意。
00:51
All the interconnections
of related situations.
of related situations.
13
39395
4285
所有关连情况的相互联系。
00:55
Maybe you've seen
a dog like this one before,
a dog like this one before,
14
43704
3126
也许你以前见过这样的狗，
00:58
or you've spent time
running on a beach like this one,
running on a beach like this one,
15
46854
2969
或者你曾经花时间，
在这样的沙滩上跑步，
在这样的沙滩上跑步，
01:01
and that further evokes thoughts
and memories of a past vacation,
and memories of a past vacation,
16
49847
4778
并进一步唤起
过去假期的记忆和想法，
过去假期的记忆和想法，
01:06
past times to the beach,
17
54649
1920
以前去海滩的时候，
01:08
times spent running around
with other dogs.
with other dogs.
18
56593
2603
花在与其他狗儿，
跑来跑去的时间。
跑来跑去的时间。
01:11
One of my guiding principles
is that by helping computers to understand
is that by helping computers to understand
19
59688
5207
我的指导原则之一，
是通过帮助电脑了解
是通过帮助电脑了解
01:16
what it's like to have these experiences,
20
64919
2896
这是什么样的经历，
01:19
to understand what we share
and believe and feel,
and believe and feel,
21
67839
5176
从而了解我们所相信的
和感受的共通点，
和感受的共通点，
01:26
then we're in a great position
to start evolving computer technology
to start evolving computer technology
22
74094
4310
那么我们就有能力开始
不断发展计算机技术，
不断发展计算机技术，
01:30
in a way that's complementary
with our own experiences.
with our own experiences.
23
78428
4587
以一种与我们经验互补的方式。
01:35
So, digging more deeply into this,
24
83539
3387
因此，深入挖掘这一点，
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
我几年前开始致力于帮助电脑
产生类似人类的故事，
产生类似人类的故事，
从图像序列。
01:44
from sequences of images.
26
92879
1666
01:47
So, one day,
27
95427
1904
所以，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
我正在用电脑工作时，询问它
对澳大利亚之行的看法。
对澳大利亚之行的看法。
01:54
It took a look at the pictures,
and it saw a koala.
and it saw a koala.
29
102768
2920
它看了看图片，
看到一只树袋熊。
看到一只树袋熊。
01:58
It didn't know what the koala was,
30
106236
1643
它不知道树袋熊是什么，
01:59
but it said it thought
it was an interesting-looking creature.
it was an interesting-looking creature.
31
107903
2999
但电脑表示它认为树袋熊
看起来是很有趣的生物。
看起来是很有趣的生物。
02:04
Then I shared with it a sequence of images
about a house burning down.
about a house burning down.
32
112053
4004
然后我与电脑分享一系列
关于房屋烧毁的图像。
关于房屋烧毁的图像。
02:09
It took a look at the images and it said,
33
117704
3285
电脑看了一下图片，它说：
02:13
"This is an amazing view!
This is spectacular!"
This is spectacular!"
34
121013
3500
「这是个惊人的景观！
这很壮观！」
这很壮观！」
02:17
It sent chills down my spine.
35
125450
2095
它使我的脊背发冷。
02:20
It saw a horrible, life-changing
and life-destroying event
and life-destroying event
36
128983
4572
电脑看到一个可怕的，
改变生活和毁灭生命的事件
改变生活和毁灭生命的事件
02:25
and thought it was something positive.
37
133579
2382
并认为这是积极的事情。
02:27
I realized that it recognized
the contrast,
the contrast,
38
135985
3441
我意识到电脑认识到
红色和黄色的对比，
02:31
the reds, the yellows,
39
139450
2699
02:34
and thought it was something
worth remarking on positively.
worth remarking on positively.
40
142173
3078
并认为这是值得积极评价的事情。
部分原因是因为
02:37
And part of why it was doing this
41
145928
1615
02:39
was because most
of the images I had given it
of the images I had given it
42
147577
2945
我输入电脑的大部分
02:42
were positive images.
43
150546
1840
是积极的图像。
02:44
That's because people
tend to share positive images
tend to share positive images
44
152903
3658
那是因为人们谈论自己的经历时，
倾向于分享积极的图像。
02:48
when they talk about their experiences.
45
156585
2190
你上次在葬礼上看到自拍照
是什么时候？
是什么时候？
02:51
When was the last time
you saw a selfie at a funeral?
you saw a selfie at a funeral?
46
159267
2541
02:55
I realized that,
as I worked on improving AI
as I worked on improving AI
47
163434
3095
我意识到在改进人工智能的过程中，
02:58
task by task, dataset by dataset,
48
166553
3714
是任务归任务，数据集归数据集，
03:02
that I was creating massive gaps,
49
170291
2897
在电脑的理解上，
创造出巨大的差距，缺陷和盲点。
03:05
holes and blind spots
in what it could understand.
in what it could understand.
50
173212
3999
03:10
And while doing so,
51
178307
1334
在这样做的同时，
03:11
I was encoding all kinds of biases.
52
179665
2483
我正在为各种偏见编码。
03:15
Biases that reflect a limited viewpoint,
53
183029
3318
偏见反映有限观点，
03:18
limited to a single dataset --
54
186371
2261
源于一个数据集——
03:21
biases that can reflect
human biases found in the data,
human biases found in the data,
55
189283
3858
其中就反映着人类同样的，
03:25
such as prejudice and stereotyping.
56
193165
3104
比如成见和刻板印象。
03:29
I thought back to the evolution
of the technology
of the technology
57
197554
3057
我回想起那技术的发展，
03:32
that brought me to where I was that day --
58
200635
2502
让我达到那天我所处的境地——
03:35
how the first color images
59
203966
2233
第一张彩色的图像
03:38
were calibrated against
a white woman's skin,
a white woman's skin,
60
206223
3048
是针对一位白人女性的
皮肤颜色进行校准的，
皮肤颜色进行校准的，
03:41
meaning that color photography
was biased against black faces.
was biased against black faces.
61
209665
4145
这意味着彩色摄影对黑脸有偏差。
03:46
And that same bias, that same blind spot
62
214514
2925
而同样的偏差，那个盲点
03:49
continued well into the '90s.
63
217463
1867
继续带入了 90 年代。
03:51
And the same blind spot
continues even today
continues even today
64
219701
3154
同样的盲点即使在今天，
03:54
in how well we can recognize
different people's faces
different people's faces
65
222879
3698
仍然存在于在面部识别技术应用中，
怎样辨识不同人物的脸。
03:58
in facial recognition technology.
66
226601
2200
04:01
I though about the state of the art
in research today,
in research today,
67
229323
3143
在今天的研究中，
我想到了最先进的技术，
我想到了最先进的技术，
04:04
where we tend to limit our thinking
to one dataset and one problem.
to one dataset and one problem.
68
232490
4514
都倾向于将我们的想法，
限制在一个数据集和一个问题上。
限制在一个数据集和一个问题上。
04:09
And that in doing so, we were creating
more blind spots and biases
more blind spots and biases
69
237688
4881
而这样做，我们正在创造
更多的盲点和偏见，
更多的盲点和偏见，
04:14
that the AI could further amplify.
70
242593
2277
会在使用人工智能时
被进一步放大。
被进一步放大。
04:17
I realized then
that we had to think deeply
that we had to think deeply
71
245712
2079
那时我意识到我们必须深思，
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
我们今天发明创造技术，
在五年到十年之后会怎样被看待 。
在五年到十年之后会怎样被看待 。
04:25
Humans evolve slowly,
with time to correct for issues
with time to correct for issues
73
253990
3142
在人类与环境互动作用中，
人类用时间纠正问题，
所以进化缓慢。
所以进化缓慢。
04:29
in the interaction of humans
and their environment.
and their environment.
74
257156
3534
04:33
In contrast, artificial intelligence
is evolving at an incredibly fast rate.
is evolving at an incredibly fast rate.
75
261276
5429
人工智能相比之下，正在以
令人难以置信的速度发展。
令人难以置信的速度发展。
这意味着它确实很重要，
04:39
And that means that it really matters
76
267013
1773
04:40
that we think about this
carefully right now --
carefully right now --
77
268810
2317
我们现在要仔细考虑这一点 ——
04:44
that we reflect on our own blind spots,
78
272180
3008
反思自己的盲点，
04:47
our own biases,
79
275212
2317
及偏见，
04:49
and think about how that's informing
the technology we're creating
the technology we're creating
80
277553
3857
并考虑这些偏见是如何影响
我们现在创造的技术，
我们现在创造的技术，
04:53
and discuss what the technology of today
will mean for tomorrow.
will mean for tomorrow.
81
281434
3902
并讨论今天的技术，
对未来意味着什么。
对未来意味着什么。
04:58
CEOs and scientists have weighed in
on what they think
on what they think
82
286593
3191
CEO和科学家，
已经权衡了他们的想法，
已经权衡了他们的想法，
05:01
the artificial intelligence technology
of the future will be.
of the future will be.
83
289808
3325
关于未来的人工智能发展。
05:05
Stephen Hawking warns that
84
293157
1618
斯蒂芬·霍金警告说：
05:06
"Artificial intelligence
could end mankind."
could end mankind."
85
294799
3007
「人工智能会使人类灭亡。」
05:10
Elon Musk warns
that it's an existential risk
that it's an existential risk
86
298307
2683
伊隆‧马斯克警告
这是一种存在的风险，
这是一种存在的风险，
05:13
and one of the greatest risks
that we face as a civilization.
that we face as a civilization.
87
301014
3574
也是我们作为文明社会;
要面临的最大风险之一。
要面临的最大风险之一。
05:17
Bill Gates has made the point,
88
305665
1452
比尔‧盖茨指出：
05:19
"I don't understand
why people aren't more concerned."
why people aren't more concerned."
89
307141
3185
「我不明白为什么人们
对人工智能不更忧虑。」
对人工智能不更忧虑。」
05:23
But these views --
90
311412
1318
但是这些观点——
05:25
they're part of the story.
91
313618
1734
只是故事的一部分。
05:28
The math, the models,
92
316079
2420
那数学，模型，
05:30
the basic building blocks
of artificial intelligence
of artificial intelligence
93
318523
3070
这些人工智能的基本组成部分，
05:33
are something that we call access
and all work with.
and all work with.
94
321617
3135
是我们都可以取得和并使用的。
05:36
We have open-source tools
for machine learning and intelligence
for machine learning and intelligence
95
324776
3785
我们有向大众开放的源代码工具
来学习机器，
来学习机器，
05:40
that we can contribute to.
96
328585
1734
并同时作出自己的贡献。
05:42
And beyond that,
we can share our experience.
we can share our experience.
97
330919
3340
除此之外，我们也可以
分享我们的经验。
分享我们的经验。
我们可以分享在技术方面
及其与我们的关系，
及其与我们的关系，
05:46
We can share our experiences
with technology and how it concerns us
with technology and how it concerns us
98
334760
3468
和如何令我们雀跃的地方。
05:50
and how it excites us.
99
338252
1467
我们可以讨论我们所爱的东西。
05:52
We can discuss what we love.
100
340251
1867
我们可以与预见的将来进行沟通，
05:55
We can communicate with foresight
101
343244
2031
05:57
about the aspects of technology
that could be more beneficial
that could be more beneficial
102
345299
4857
关于技术方面这可能会更有益，
06:02
or could be more problematic over time.
103
350180
2600
或随着时间的推移，
可能会出现更多的问题。
可能会出现更多的问题。
06:05
If we all focus on opening up
the discussion on AI
the discussion on AI
104
353799
4143
如果我们都专注于
开放对于人工智能的讨论
开放对于人工智能的讨论
06:09
with foresight towards the future,
105
357966
1809
展望未来，
06:13
this will help create a general
conversation and awareness
conversation and awareness
106
361093
4270
这将有助于创造一个
常规的对话和意识，
常规的对话和意识，
06:17
about what AI is now,
107
365387
2513
关于人工智能是什么？
它能成为什么？
06:21
what it can become
108
369212
2001
以及我们需要做的所有事情，
06:23
and all the things that we need to do
109
371237
1785
06:25
in order to enable that outcome
that best suits us.
that best suits us.
110
373046
3753
以实现最适合我们的结果。
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
我们已经在今天使用的技术中
看到和了解这一点。
看到和了解这一点。
06:33
We use smart phones
and digital assistants and Roombas.
and digital assistants and Roombas.
112
381767
3880
我们使用智能手机，数码助理
和自动吸尘器。
和自动吸尘器。
06:38
Are they evil?
113
386457
1150
它们邪恶吗？
06:40
Maybe sometimes.
114
388268
1547
也许有时是。
06:42
Are they beneficial?
115
390664
1333
他们有益吗？
06:45
Yes, they're that, too.
116
393005
1533
是的，他们也是。
06:48
And they're not all the same.
117
396236
1761
它们并不完全相同。
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
在那里你已经看到了未来的光芒。
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
未来将继续从我们现在
建立和创造的东西开始。
建立和创造的东西开始。
06:59
We set into motion that domino effect
120
407165
2642
我们启动了多米诺骨牌效应，
07:01
that carves out AI's evolutionary path.
121
409831
2600
这就揭开了人工智能的进化通道
07:05
In our time right now,
we shape the AI of tomorrow.
we shape the AI of tomorrow.
122
413173
2871
在我们的时代，塑造了
明天的人工智能。
明天的人工智能。
07:08
Technology that immerses us
in augmented realities
in augmented realities
123
416566
3699
让我们能沉浸在增强现实的技术中，
07:12
bringing to life past worlds.
124
420289
2566
使过去的世界复活，
07:15
Technology that helps people
to share their experiences
to share their experiences
125
423844
4312
当人们沟通有困难时，
07:20
when they have difficulty communicating.
126
428180
2262
科技就帮助他们分享彼此的经验。
07:23
Technology built on understanding
the streaming visual worlds
the streaming visual worlds
127
431323
4532
建立于在线视觉媒体的科技，
07:27
used as technology for self-driving cars.
128
435879
3079
可被用在汽车自动驾驶上。
07:32
Technology built on understanding images
and generating language,
and generating language,
129
440490
3413
科技基于图像的理解而产生语言，
07:35
evolving into technology that helps people
who are visually impaired
who are visually impaired
130
443927
4063
能演变成协助视障人士的技术，
07:40
be better able to access the visual world.
131
448014
2800
帮助他们更好地拥有视觉世界。
07:42
And we also see how technology
can lead to problems.
can lead to problems.
132
450838
3261
我们也看到科技
如何导致一些问题。
如何导致一些问题。
07:46
We have technology today
133
454885
1428
我们今天有科技
07:48
that analyzes physical
characteristics we're born with --
characteristics we're born with --
134
456337
3835
分析我们出生的身体特征 ——
07:52
such as the color of our skin
or the look of our face --
or the look of our face --
135
460196
3272
比如我们皮肤的颜色
还是我们脸上的表情
还是我们脸上的表情
07:55
in order to determine whether or not
we might be criminals or terrorists.
we might be criminals or terrorists.
136
463492
3804
以确定我们是否罪犯或恐怖分子。
07:59
We have technology
that crunches through our data,
that crunches through our data,
137
467688
2905
我们拥有处理数据的技术，
08:02
even data relating
to our gender or our race,
to our gender or our race,
138
470617
2896
处理关于性别或种族的数据，
08:05
in order to determine whether or not
we might get a loan.
we might get a loan.
139
473537
2865
以确定我们是否可以获得贷款。
08:09
All that we see now
140
477494
1579
我们现在看到的所有东西，
08:11
is a snapshot in the evolution
of artificial intelligence.
of artificial intelligence.
141
479097
3617
只是人工智能演变过程中的
快照。
快照。
08:15
Because where we are right now,
142
483763
1778
因为我们现在所处的地方，
08:17
is within a moment of that evolution.
143
485565
2238
是演变中的一个时刻。
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
这意味着我们现在所做的，
将会影响事情的往后发展，
将会影响事情的往后发展，
并延至未来的世界。
08:24
and in the future.
145
492516
1200
08:26
If we want AI to evolve
in a way that helps humans,
in a way that helps humans,
146
494063
3951
如果我们希望人工智能
能协助人类的方式进化，
能协助人类的方式进化，
08:30
then we need to define
the goals and strategies
the goals and strategies
147
498038
2801
我们就需要确定策略和目标，
08:32
that enable that path now.
148
500863
1733
马上开通那条路径。
08:35
What I'd like to see is something
that fits well with humans,
that fits well with humans,
149
503680
3738
我想看到的是适合人类的
08:39
with our culture and with the environment.
150
507442
2800
文化和环境的发展方向。
08:43
Technology that aids and assists
those of us with neurological conditions
those of us with neurological conditions
151
511435
4484
科技能帮助我们治愈神经系统疾病
08:47
or other disabilities
152
515943
1721
或其它残疾的患者，
08:49
in order to make life
equally challenging for everyone.
equally challenging for everyone.
153
517688
3216
让他们与每个人一样，
让生活同样具有挑战性。
让生活同样具有挑战性。
08:54
Technology that works
154
522097
1421
科技的运作
不会考量你的特征
不会考量你的特征
08:55
regardless of your demographics
or the color of your skin.
or the color of your skin.
155
523542
3933
或皮肤颜色。
09:00
And so today, what I focus on
is the technology for tomorrow
is the technology for tomorrow
156
528383
4742
我今天关注的是
明日和十年后的科技，
09:05
and for 10 years from now.
157
533149
1733
09:08
AI can turn out in many different ways.
158
536530
2634
人工智能可以以许多不同的方式出现。
09:11
But in this case,
159
539688
1225
但在这种情况下，
09:12
it isn't a self-driving car
without any destination.
without any destination.
160
540937
3328
它并不是没有任何目的地的
无人驾驶车。
无人驾驶车。
09:16
This is the car that we are driving.
161
544884
2400
这是我们能驾驶同时控制的汽车。
09:19
We choose when to speed up
and when to slow down.
and when to slow down.
162
547953
3595
我们选择何时加速和何时减速。
09:23
We choose if we need to make a turn.
163
551572
2400
我们选择是否需要转弯。
09:26
We choose what the AI
of the future will be.
of the future will be.
164
554868
3000
我们选择未来的人工智能会是什么。
会有一个广阔的竞技场。
09:31
There's a vast playing field
165
559186
1337
09:32
of all the things that artificial
intelligence can become.
intelligence can become.
166
560547
2965
容许人工智能可以成为所有的东西。
它会变成很多不同的东西。
09:36
It will become many things.
167
564064
1800
09:39
And it's up to us now,
168
567694
1732
现在取决于
09:41
in order to figure out
what we need to put in place
what we need to put in place
169
569450
3061
我们要弄清楚所需要实施的
09:44
to make sure the outcomes
of artificial intelligence
of artificial intelligence
170
572535
3807
以确保人工智能的结果
09:48
are the ones that will be
better for all of us.
better for all of us.
171
576366
3066
是对所有人类都会更好。
09:51
Thank you.
172
579456
1150
谢谢。
09:52
(Applause)
173
580630
2187
（掌声）
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