TED2019
Janelle Shane: The danger of AI is weirder than you think
珍妮尔 · 尚恩: 人工智能的危险比你想象中更奇怪
Filmed:
Readability: 4.2
376,501 views
人工智能研究者珍妮尔 · 尚恩（Janelle Shane）认为，人工智能的危险不在于它会反抗我们，而是会严格按照我们的要求去做。在尝试解决人类提出的问题，如创造新的冰淇淋或者识别路上的车辆的时候，人工智能所做出的行为时而滑稽可笑，时而令人恐慌。通过这些分享，尚恩说明了为什么人工智能远未能媲美真正的大脑。
Janelle Shane - AI researcher
While moonlighting as a research scientist, Janelle Shane found fame documenting the often hilarious antics of AI algorithms. Full bio
While moonlighting as a research scientist, Janelle Shane found fame documenting the often hilarious antics of AI algorithms. Full bio
Double-click the English transcript below to play the video.
00:13
So, artificial intelligence
0
1765
3000
人工智能，
00:16
is known for disrupting
all kinds of industries.
all kinds of industries.
1
4789
3529
以能颠覆所有行业广为人知。
00:20
What about ice cream?
2
8961
2043
那冰淇淋呢？
00:23
What kind of mind-blowing
new flavors could we generate
new flavors could we generate
3
11879
3639
我们是否能利用先进的人工智能
00:27
with the power of an advanced
artificial intelligence?
artificial intelligence?
4
15542
2976
生成令人震惊的新口味呢？
00:31
So I teamed up with a group of coders
from Kealing Middle School
from Kealing Middle School
5
19011
4161
我和 Kealing 中学的程序员组了个队
00:35
to find out the answer to this question.
6
23196
2241
想要找到答案。
00:37
They collected over 1,600
existing ice cream flavors,
existing ice cream flavors,
7
25461
5081
他们收集了超过 1600 种
现有的冰淇淋口味，
现有的冰淇淋口味，
00:42
and together, we fed them to an algorithm
to see what it would generate.
to see what it would generate.
8
30566
5522
接着我们一起把这些口味输入
到算法中看看会有什么结果。
到算法中看看会有什么结果。
00:48
And here are some of the flavors
that the AI came up with.
that the AI came up with.
9
36112
3753
接下来给大家展示一些
人工智能所想到的口味。
人工智能所想到的口味。
00:52
[Pumpkin Trash Break]
10
40444
1471
【南瓜垃圾破裂】
00:53
(Laughter)
11
41939
1402
(笑声)
00:55
[Peanut Butter Slime]
12
43365
2469
【花生酱稀泥】
00:58
[Strawberry Cream Disease]
13
46822
1343
【草莓奶油病】
01:00
(Laughter)
14
48189
2126
(笑声)
01:02
These flavors are not delicious,
as we might have hoped they would be.
as we might have hoped they would be.
15
50339
4597
这些口味听起来并没有
我们想象中美味。
我们想象中美味。
01:06
So the question is: What happened?
16
54960
1864
所以问题来了：怎么回事？
01:08
What went wrong?
17
56848
1394
到底哪里出问题了？
01:10
Is the AI trying to kill us?
18
58266
1959
人工智能是想要干掉我们？
01:13
Or is it trying to do what we asked,
and there was a problem?
and there was a problem?
19
61027
4310
还是说它努力想要回应
我们的要求，但是却出问题了？
我们的要求，但是却出问题了？
01:18
In movies, when something
goes wrong with AI,
goes wrong with AI,
20
66567
2464
在电影中，当人工智能出了错，
01:21
it's usually because the AI has decided
21
69055
2712
通常是因为它们决定
01:23
that it doesn't want to obey
the humans anymore,
the humans anymore,
22
71791
2272
再也不要听从人类的指令，
01:26
and it's got its own goals,
thank you very much.
thank you very much.
23
74087
2623
它开始有了自己的目标，
不劳驾人类了。
不劳驾人类了。
01:29
In real life, though,
the AI that we actually have
the AI that we actually have
24
77266
3216
然而现实生活中，
我们现有的人工智能
我们现有的人工智能
01:32
is not nearly smart enough for that.
25
80506
1863
还没达到那样的水平。
01:34
It has the approximate computing power
26
82781
2982
它的计算能力大概跟
01:37
of an earthworm,
27
85787
1276
一条小虫子差不多，
01:39
or maybe at most a single honeybee,
28
87087
3403
又或者顶多只是一只小蜜蜂，
01:42
and actually, probably maybe less.
29
90514
2215
实际上可能更弱。
01:44
Like, we're constantly learning
new things about brains
new things about brains
30
92753
2594
我们持续从大脑学习到新事物，
01:47
that make it clear how much our AIs
don't measure up to real brains.
don't measure up to real brains.
31
95371
4360
使我们越来越清楚人工智能
与真正的大脑之间的距离。
与真正的大脑之间的距离。
01:51
So today's AI can do a task
like identify a pedestrian in a picture,
like identify a pedestrian in a picture,
32
99755
5663
现在人工智能所达到的大体就是
在图片中识别出行人的程度，
在图片中识别出行人的程度，
01:57
but it doesn't have a concept
of what the pedestrian is
of what the pedestrian is
33
105442
2983
但是它并没有
对于行人的概念，
对于行人的概念，
02:00
beyond that it's a collection
of lines and textures and things.
of lines and textures and things.
34
108449
4824
除此之外它所做的只是
收集线条，质地之类的信息。
收集线条，质地之类的信息。
02:05
It doesn't know what a human actually is.
35
113792
2521
但是它并不知道人类到底是什么。
02:08
So will today's AI
do what we ask it to do?
do what we ask it to do?
36
116822
3282
那么现在的人工智能
能否达到我们的要求？
能否达到我们的要求？
02:12
It will if it can,
37
120128
1594
能力允许的情况下它会，
02:13
but it might not do what we actually want.
38
121746
2726
但是它所做的可能
并不是我们真正想要的。
并不是我们真正想要的。
02:16
So let's say that you
were trying to get an AI
were trying to get an AI
39
124496
2415
假设你想要用人工智能
02:18
to take this collection of robot parts
40
126935
2619
利用一堆机器人的零件
02:21
and assemble them into some kind of robot
to get from Point A to Point B.
to get from Point A to Point B.
41
129578
4197
组装成一个机器人
从 A 点移动到 B 点。
从 A 点移动到 B 点。
02:25
Now, if you were going to try
and solve this problem
and solve this problem
42
133799
2481
如果你想要通过编写
一个传统的计算机程序
一个传统的计算机程序
02:28
by writing a traditional-style
computer program,
computer program,
43
136304
2351
来解决这个问题，
02:30
you would give the program
step-by-step instructions
step-by-step instructions
44
138679
3417
你需要输入一步步的指令，
02:34
on how to take these parts,
45
142120
1329
指示它怎样拿起零件，
02:35
how to assemble them
into a robot with legs
into a robot with legs
46
143473
2407
怎样把这些零件安装成
一个带脚的机器人，
一个带脚的机器人，
02:37
and then how to use those legs
to walk to Point B.
to walk to Point B.
47
145904
2942
以及如何用脚走到 B 点。
02:41
But when you're using AI
to solve the problem,
to solve the problem,
48
149441
2340
但是当你利用人工智能
来解决这个问题的时候，
来解决这个问题的时候，
02:43
it goes differently.
49
151805
1174
情况不太一样。
02:45
You don't tell it
how to solve the problem,
how to solve the problem,
50
153003
2382
你不用告诉它
要怎样解决问题，
要怎样解决问题，
02:47
you just give it the goal,
51
155409
1479
你只需要给它一个目标，
02:48
and it has to figure out for itself
via trial and error
via trial and error
52
156912
3262
它会通过试错
来解决这个问题，
来解决这个问题，
02:52
how to reach that goal.
53
160198
1484
来实现目标。
02:54
And it turns out that the way AI tends
to solve this particular problem
to solve this particular problem
54
162254
4102
结果是，貌似人工智能在
解决这一类问题的时候
解决这一类问题的时候
02:58
is by doing this:
55
166380
1484
会这么做：
02:59
it assembles itself into a tower
and then falls over
and then falls over
56
167888
3367
它把自己搭建成
一座塔然后倾倒，
一座塔然后倾倒，
03:03
and lands at Point B.
57
171279
1827
最后在 B 点落下。
03:05
And technically, this solves the problem.
58
173130
2829
从技术的层面上看，的确解决了问题。
03:07
Technically, it got to Point B.
59
175983
1639
从技术上来说的确到达了 B 点。
03:09
The danger of AI is not that
it's going to rebel against us,
it's going to rebel against us,
60
177646
4265
人工智能的危险
不在于它会反抗我们，
不在于它会反抗我们，
03:13
it's that it's going to do
exactly what we ask it to do.
exactly what we ask it to do.
61
181935
4274
而是它们会严格按照
我们的要求去做。
我们的要求去做。
03:18
So then the trick
of working with AI becomes:
of working with AI becomes:
62
186876
2498
所以和人工智能共事的技巧变成了：
03:21
How do we set up the problem
so that it actually does what we want?
so that it actually does what we want?
63
189398
3828
我们该如何设置问题才能让它
做我们真正想做的事？
做我们真正想做的事？
03:26
So this little robot here
is being controlled by an AI.
is being controlled by an AI.
64
194726
3306
这一台小机器人
由人工智能操控。
由人工智能操控。
03:30
The AI came up with a design
for the robot legs
for the robot legs
65
198056
2814
人工智能想到了一个
机器人脚部的设计，
机器人脚部的设计，
03:32
and then figured out how to use them
to get past all these obstacles.
to get past all these obstacles.
66
200894
4078
然后想到了如何
利用它们绕过障碍。
利用它们绕过障碍。
03:36
But when David Ha set up this experiment,
67
204996
2741
但是当大卫·哈
在做这个实验的时候，
在做这个实验的时候，
03:39
he had to set it up
with very, very strict limits
with very, very strict limits
68
207761
2856
他不得不对人工智能
容许搭建起来的脚
容许搭建起来的脚
03:42
on how big the AI
was allowed to make the legs,
was allowed to make the legs,
69
210641
3292
设立非常、非常严格的限制，
03:45
because otherwise ...
70
213957
1550
不然的话...
03:55
(Laughter)
71
223058
3931
（笑声）
04:00
And technically, it got
to the end of that obstacle course.
to the end of that obstacle course.
72
228563
3745
从技术上说，他的确
到达了障碍路线的终点。
到达了障碍路线的终点。
04:04
So you see how hard it is to get AI
to do something as simple as just walk.
to do something as simple as just walk.
73
232332
4942
现在我们知道了，仅仅是让人工智能
实现简单的行走就有多困难。
实现简单的行走就有多困难。
04:09
So seeing the AI do this,
you may say, OK, no fair,
you may say, OK, no fair,
74
237298
3820
当看到人工智能这么做的时候，
你可能会说，这不公平。
你可能会说，这不公平。
04:13
you can't just be
a tall tower and fall over,
a tall tower and fall over,
75
241142
2580
你不能只是变成
一座塔然后直接倒下，
一座塔然后直接倒下，
04:15
you have to actually, like,
use legs to walk.
use legs to walk.
76
243746
3435
你必须得用脚去走路，
04:19
And it turns out,
that doesn't always work, either.
that doesn't always work, either.
77
247205
2759
结果是，
那往往也不行。
那往往也不行。
04:21
This AI's job was to move fast.
78
249988
2759
这个人工智能的任务是快速移动。
04:25
They didn't tell it that it had
to run facing forward
to run facing forward
79
253115
3593
他们没有说它应该面向前方奔跑，
04:28
or that it couldn't use its arms.
80
256732
2258
也没有说不能使用它的手臂。
04:31
So this is what you get
when you train AI to move fast,
when you train AI to move fast,
81
259487
4618
这就是当你训练人工智能
快速移动时所能得到的结果，
快速移动时所能得到的结果，
04:36
you get things like somersaulting
and silly walks.
and silly walks.
82
264129
3534
你能得到的就是像这样的
空翻或者滑稽漫步。
空翻或者滑稽漫步。
04:39
It's really common.
83
267687
1400
太常见了。
04:41
So is twitching along the floor in a heap.
84
269667
3179
在地板上扭动前进
也是一样的结果。
也是一样的结果。
04:44
(Laughter)
85
272870
1150
（笑声）
04:47
So in my opinion, you know what
should have been a whole lot weirder
should have been a whole lot weirder
86
275241
3254
在我看来，更奇怪的
04:50
is the "Terminator" robots.
87
278519
1396
就是“终结者”机器人。
04:52
Hacking "The Matrix" is another thing
that AI will do if you give it a chance.
that AI will do if you give it a chance.
88
280256
3755
要是有可能的话，人工智能
还真会入侵“黑客帝国"。
还真会入侵“黑客帝国"。
04:56
So if you train an AI in a simulation,
89
284035
2517
如果你用仿真环境
训练一个人工智能的话，
训练一个人工智能的话，
04:58
it will learn how to do things like
hack into the simulation's math errors
hack into the simulation's math errors
90
286576
4113
它会学习如何入侵到
一个仿真环境中的数学错误里，
一个仿真环境中的数学错误里，
05:02
and harvest them for energy.
91
290713
2207
并从中获得能量。
05:04
Or it will figure out how to move faster
by glitching repeatedly into the floor.
by glitching repeatedly into the floor.
92
292944
5475
或者会计算出如何通过
不断地在地板上打滑来加快速度。
不断地在地板上打滑来加快速度。
05:10
When you're working with AI,
93
298443
1585
当你和人工智能一起工作的时候，
05:12
it's less like working with another human
94
300052
2389
不太像是在跟另一个人一起工作，
05:14
and a lot more like working
with some kind of weird force of nature.
with some kind of weird force of nature.
95
302465
3629
而更像是在和某种
奇怪的自然力量工作。
奇怪的自然力量工作。
05:18
And it's really easy to accidentally
give AI the wrong problem to solve,
give AI the wrong problem to solve,
96
306562
4623
一不小心就很容易让人工
智能去破解错误的问题，
智能去破解错误的问题，
05:23
and often we don't realize that
until something has actually gone wrong.
until something has actually gone wrong.
97
311209
4538
往往直到出现问题
我们才察觉到不妥。
我们才察觉到不妥。
05:28
So here's an experiment I did,
98
316242
2080
所以我做了这样的一个实验，
05:30
where I wanted the AI
to copy paint colors,
to copy paint colors,
99
318346
3182
我想要让人工智能
利用左边的颜色列表
利用左边的颜色列表
05:33
to invent new paint colors,
100
321552
1746
复制颜料颜色，
05:35
given the list like the ones
here on the left.
here on the left.
101
323322
2987
去创造新的颜色。
05:38
And here's what the AI
actually came up with.
actually came up with.
102
326798
3004
这就是人工智能想到的结果。
05:41
[Sindis Poop, Turdly, Suffer, Gray Pubic]
103
329826
3143
【辛迪斯粪便，如粪球般，
受难，灰色公众】
受难，灰色公众】
05:44
(Laughter)
104
332993
4230
（笑声）
05:51
So technically,
105
339177
1886
基本上，
05:53
it did what I asked it to.
106
341087
1864
它达到了我的要求。
05:54
I thought I was asking it for,
like, nice paint color names,
like, nice paint color names,
107
342975
3308
我以为我给出的要求是，
让它想出美好的颜色名，
让它想出美好的颜色名，
05:58
but what I was actually asking it to do
108
346307
2307
但是实际上我让它做的
06:00
was just imitate the kinds
of letter combinations
of letter combinations
109
348638
3086
只是单纯地模仿
字母的组合，
字母的组合，
06:03
that it had seen in the original.
110
351748
1905
那些它在输入中见到的字母组合。
06:05
And I didn't tell it anything
about what words mean,
about what words mean,
111
353677
3098
而且我并没有告诉它
这些单词的意思是什么，
这些单词的意思是什么，
06:08
or that there are maybe some words
112
356799
2560
或者告诉它也许有些单词
06:11
that it should avoid using
in these paint colors.
in these paint colors.
113
359383
2889
不能用来给颜色命名。
06:15
So its entire world
is the data that I gave it.
is the data that I gave it.
114
363141
3494
也就是说它的整个世界里
只有我给出的数据。
只有我给出的数据。
06:18
Like with the ice cream flavors,
it doesn't know about anything else.
it doesn't know about anything else.
115
366659
4028
正如让它发明冰淇淋的口味那样，
它除此之外一无所知。
它除此之外一无所知。
06:24
So it is through the data
116
372491
1638
也就是通过数据，
06:26
that we often accidentally tell AI
to do the wrong thing.
to do the wrong thing.
117
374153
4044
我们常常不小心
让人工智能做错事。
让人工智能做错事。
06:30
This is a fish called a tench.
118
378694
3032
有一种叫丁鲷的鱼，
06:33
And there was a group of researchers
119
381750
1815
一群研究者尝试过
06:35
who trained an AI to identify
this tench in pictures.
this tench in pictures.
120
383589
3874
训练人工智能去
识别图片里的丁鲷。
识别图片里的丁鲷。
06:39
But then when they asked it
121
387487
1296
但是当他们试图搞清
06:40
what part of the picture it was actually
using to identify the fish,
using to identify the fish,
122
388807
3426
它到底用了图片的
哪个部分去识别这种鱼，
哪个部分去识别这种鱼，
06:44
here's what it highlighted.
123
392257
1358
这是它所显示的部分。
06:47
Yes, those are human fingers.
124
395203
2189
没错，那些是人类的手指。
06:49
Why would it be looking for human fingers
125
397416
2059
为什么它会去识别人类的手指，
06:51
if it's trying to identify a fish?
126
399499
1921
而不是鱼呢？
06:54
Well, it turns out that the tench
is a trophy fish,
is a trophy fish,
127
402126
3164
因为丁鲷实际上是一种战利品鱼，
06:57
and so in a lot of pictures
that the AI had seen of this fish
that the AI had seen of this fish
128
405314
3811
所以人工智能在被训练时，
07:01
during training,
129
409149
1151
看过的大多数照片中
07:02
the fish looked like this.
130
410324
1490
鱼都长这样。
07:03
(Laughter)
131
411838
1635
（笑声）
07:05
And it didn't know that the fingers
aren't part of the fish.
aren't part of the fish.
132
413497
3330
而人工智能并不知道原来
手指并不是鱼的一部分。
手指并不是鱼的一部分。
07:10
So you see why it is so hard
to design an AI
to design an AI
133
418808
4120
现在你们应该能想象，
设计一个能真正懂得
设计一个能真正懂得
07:14
that actually can understand
what it's looking at.
what it's looking at.
134
422952
3319
自己在做什么的人工
智能是多么困难。
智能是多么困难。
07:18
And this is why designing
the image recognition
the image recognition
135
426295
2862
这也就是为什么
给无人驾驶汽车
给无人驾驶汽车
07:21
in self-driving cars is so hard,
136
429181
2067
设计图像识别技术那么困难，
07:23
and why so many self-driving car failures
137
431272
2205
导致无人驾驶失败的原因
07:25
are because the AI got confused.
138
433501
2885
就是，人工智能迷糊了。
07:28
I want to talk about an example from 2016.
139
436410
4008
接下来我想分享一个
发生在 2016 年的故事。
发生在 2016 年的故事。
07:32
There was a fatal accident when somebody
was using Tesla's autopilot AI,
was using Tesla's autopilot AI,
140
440442
4455
有人在使用特斯拉的
自动驾驶功能时发生了特大事故，
自动驾驶功能时发生了特大事故，
07:36
but instead of using it on the highway
like it was designed for,
like it was designed for,
141
444921
3414
因为这个人工智能是
为上高速路而设计的，
为上高速路而设计的，
07:40
they used it on city streets.
142
448359
2205
结果车主居然开到市内街道上。
07:43
And what happened was,
143
451239
1175
结果是，
一辆卡车突然出现在轿车前面，
而轿车没有刹车。
而轿车没有刹车。
07:44
a truck drove out in front of the car
and the car failed to brake.
and the car failed to brake.
144
452438
3396
07:48
Now, the AI definitely was trained
to recognize trucks in pictures.
to recognize trucks in pictures.
145
456507
4762
当然这个人工智能受过训练，
能识别图片中的卡车。
能识别图片中的卡车。
07:53
But what it looks like happened is
146
461293
2145
但是当时的情况看起来，
07:55
the AI was trained to recognize
trucks on highway driving,
trucks on highway driving,
147
463462
2931
人工智能接受的训练是
识别行驶在高速路上的卡车，
识别行驶在高速路上的卡车，
07:58
where you would expect
to see trucks from behind.
to see trucks from behind.
148
466417
2899
理论上你看到的应该是卡车的尾部，
08:01
Trucks on the side is not supposed
to happen on a highway,
to happen on a highway,
149
469340
3420
而侧面对着你的卡车
是不会出现在高速路上的，
是不会出现在高速路上的，
08:04
and so when the AI saw this truck,
150
472784
3455
所以当人工智能看到这辆卡车的时候，
08:08
it looks like the AI recognized it
as most likely to be a road sign
as most likely to be a road sign
151
476263
4827
可能把卡车认作一个路标，
08:13
and therefore, safe to drive underneath.
152
481114
2273
因此，它判断
从下面开过去是安全的。
从下面开过去是安全的。
08:16
Here's an AI misstep
from a different field.
from a different field.
153
484114
2580
接下来是人工智能在
另一个领域的错误示例。
另一个领域的错误示例。
08:18
Amazon recently had to give up
on a résumé-sorting algorithm
on a résumé-sorting algorithm
154
486718
3460
亚马逊最近不得不放弃
一个他们已经开发了一段时间
一个他们已经开发了一段时间
08:22
that they were working on
155
490202
1220
的简历分类的算法，
因为他们发现这个算法
竟然学会了歧视女性。
竟然学会了歧视女性。
08:23
when they discovered that the algorithm
had learned to discriminate against women.
had learned to discriminate against women.
156
491446
3908
原因是当他们把过去招聘人员的简历
08:27
What happened is they had trained it
on example résumés
on example résumés
157
495378
2716
08:30
of people who they had hired in the past.
158
498118
2242
用作人工智能的训练材料。
08:32
And from these examples, the AI learned
to avoid the résumés of people
to avoid the résumés of people
159
500384
4023
从这些素材中，人工智能学会了
怎样过滤一些应聘者的简历，
怎样过滤一些应聘者的简历，
那些上过女子大学的
08:36
who had gone to women's colleges
160
504431
2026
08:38
or who had the word "women"
somewhere in their resume,
somewhere in their resume,
161
506481
2806
或者是那些含有
“女性”字眼的简历，
“女性”字眼的简历，
08:41
as in, "women's soccer team"
or "Society of Women Engineers."
or "Society of Women Engineers."
162
509311
4576
比如说“女子足球队”
或者“女性工程师学会”。
或者“女性工程师学会”。
08:45
The AI didn't know that it wasn't supposed
to copy this particular thing
to copy this particular thing
163
513911
3974
人工智能并不知道自己
不应该复制他所见过的
不应该复制他所见过的
08:49
that it had seen the humans do.
164
517909
1978
人类这种特定的行为。
08:51
And technically, it did
what they asked it to do.
what they asked it to do.
165
519911
3177
从技术层面上说，
它的确按要求做到了。
它的确按要求做到了。
08:55
They just accidentally asked it
to do the wrong thing.
to do the wrong thing.
166
523112
2797
只是开发者不小心
下错了指令。
下错了指令。
08:58
And this happens all the time with AI.
167
526653
2895
这样的情况在人工智能领域屡见不鲜。
09:02
AI can be really destructive
and not know it.
and not know it.
168
530120
3591
人工智能破坏力惊人且不自知。
09:05
So the AIs that recommend
new content in Facebook, in YouTube,
new content in Facebook, in YouTube,
169
533735
5078
就如用于脸书和油管上
内容推荐的人工智能，
内容推荐的人工智能，
09:10
they're optimized to increase
the number of clicks and views.
the number of clicks and views.
170
538837
3539
它们被优化以增加
点击量和阅览量。
点击量和阅览量。
09:14
And unfortunately, one way
that they have found of doing this
that they have found of doing this
171
542400
3436
但是不幸的是，它们实现
目标的其中一个手段，
目标的其中一个手段，
09:17
is to recommend the content
of conspiracy theories or bigotry.
of conspiracy theories or bigotry.
172
545860
4503
就是推荐阴谋论或者偏执内容。
09:22
The AIs themselves don't have any concept
of what this content actually is,
of what this content actually is,
173
550902
5302
人工智能本身对这些内容没有概念，
09:28
and they don't have any concept
of what the consequences might be
of what the consequences might be
174
556228
3395
也根本不知道推荐这样的内容
09:31
of recommending this content.
175
559647
2109
会产生怎样的后果。
09:34
So, when we're working with AI,
176
562296
2011
所以当我们与人工智能
一起工作的时候，
一起工作的时候，
09:36
it's up to us to avoid problems.
177
564331
4182
我们有责任去规避问题。
09:40
And avoiding things going wrong,
178
568537
2323
规避可能出错的因素，
09:42
that may come down to
the age-old problem of communication,
the age-old problem of communication,
179
570884
4526
这也就带出一个
老生常谈的沟通问题，
老生常谈的沟通问题，
09:47
where we as humans have to learn
how to communicate with AI.
how to communicate with AI.
180
575434
3745
作为人类，我们要学习
怎样和人工智能沟通。
怎样和人工智能沟通。
09:51
We have to learn what AI
is capable of doing and what it's not,
is capable of doing and what it's not,
181
579203
4039
我们必须明白人工智能
能做什么，不能做什么，
能做什么，不能做什么，
09:55
and to understand that,
with its tiny little worm brain,
with its tiny little worm brain,
182
583266
3086
要明白，凭它们的那点小脑袋，
09:58
AI doesn't really understand
what we're trying to ask it to do.
what we're trying to ask it to do.
183
586376
4013
人工智能并不能完全明白
我们想让它们做什么。
我们想让它们做什么。
10:03
So in other words, we have
to be prepared to work with AI
to be prepared to work with AI
184
591148
3321
换言之，我们必须对与
人工智能共事做好准备，
人工智能共事做好准备，
10:06
that's not the super-competent,
all-knowing AI of science fiction.
all-knowing AI of science fiction.
185
594493
5258
这可不是科幻片里那些
全能全知的人工智能。
全能全知的人工智能。
10:11
We have to prepared to work with an AI
186
599775
2862
我们必须准备好跟
10:14
that's the one that we actually have
in the present day.
in the present day.
187
602661
2938
眼下存在的人工智能共事。
10:17
And present-day AI is plenty weird enough.
188
605623
4205
现在的人工智能还真的挺奇怪的。
10:21
Thank you.
189
609852
1190
谢谢。
10:23
(Applause)
190
611066
5225
（掌声）
ABOUT THE SPEAKER
Janelle Shane - AI researcherWhile moonlighting as a research scientist, Janelle Shane found fame documenting the often hilarious antics of AI algorithms.
Why you should listen
Janelle Shane's humor blog, AIweirdness.com, looks at, as she tells it, "the strange side of artificial intelligence." Her upcoming book, You Look Like a Thing and I Love You: How AI Works and Why It's Making the World a Weirder Place, uses cartoons and humorous pop-culture experiments to look inside the minds of the algorithms that run our world, making artificial intelligence and machine learning both accessible and entertaining.
According to Shane, she has only made a neural network-written recipe once -- and discovered that horseradish brownies are about as terrible as you might imagine.
Janelle Shane | Speaker | TED.com