How we can teach computers to make sense of our emotions | Raphael Arar

64,666 views ・ 2018-04-24

TED


請雙擊下方英文字幕播放視頻。

00:00
Translator: Ivana Korom Reviewer: Joanna Pietrulewicz
0
0
7000
譯者: Lilian Chiu 審譯者: Yanyan Hong
00:13
I consider myself one part artist and one part designer.
1
13760
4680
我自認一部分是藝術家, 一部分是設計師。
00:18
And I work at an artificial intelligence research lab.
2
18480
3160
我在一間人工智能研究實驗室工作。
00:22
We're trying to create technology
3
22720
1696
我們在試圖創造的技術,
00:24
that you'll want to interact with in the far future.
4
24440
3296
是在遙遠的未來 你會想要和它互動的技術。
00:27
Not just six months from now, but try years and decades from now.
5
27760
4640
不只是現在算起六個月之後, 而是數年、數十年之後。
00:33
And we're taking a moonshot
6
33120
1616
我們做了個很大膽的猜測,
00:34
that we'll want to be interacting with computers
7
34760
2456
猜想我們將來會想要和電腦
00:37
in deeply emotional ways.
8
37240
2120
用帶有深度感情的方式來互動。
00:40
So in order to do that,
9
40280
1456
為了做到這一點,
00:41
the technology has to be just as much human as it is artificial.
10
41760
4480
這項技術不能只是很人工, 也要很人性。
00:46
It has to get you.
11
46920
2256
它必須要能懂你。
00:49
You know, like that inside joke that'll have you and your best friend
12
49200
3336
就像只有你和你最好的朋友之間 才懂的「圈內笑話」
00:52
on the floor, cracking up.
13
52560
1936
能讓你們笑到在地上打滾。
00:54
Or that look of disappointment that you can just smell from miles away.
14
54520
4560
或是你遠遠就能嗅到的 那種失望表情。
01:00
I view art as the gateway to help us bridge this gap between human and machine:
15
60560
6040
我把藝術視為是一種途徑, 能協助我們在人、機之間搭起橋樑:
01:07
to figure out what it means to get each other
16
67280
3136
找出「了解彼此」是什麼意思,
01:10
so that we can train AI to get us.
17
70440
2760
這樣才能訓練人工智能來了解我們。
01:13
See, to me, art is a way to put tangible experiences
18
73920
3816
對我而言,藝術是一種 可以把有形的經驗
01:17
to intangible ideas, feelings and emotions.
19
77760
3240
放到無形的想法、感受, 及情緒中的方式。
01:21
And I think it's one of the most human things about us.
20
81800
2600
我認為它是我們 最有人性的事物之一。
01:25
See, we're a complicated and complex bunch.
21
85480
2936
我們是很複雜又難懂的物種。
01:28
We have what feels like an infinite range of emotions,
22
88440
3136
我們的情緒範圍感覺 起來似乎是無限的,
01:31
and to top it off, we're all different.
23
91600
2496
更麻煩的是,我們每個人都不同。
01:34
We have different family backgrounds,
24
94120
2296
我們有不同的家庭背景、
01:36
different experiences and different psychologies.
25
96440
3080
不同的經驗,以及不同的心理。
01:40
And this is what makes life really interesting.
26
100240
2720
正是因為如此,人生才有意思。
01:43
But this is also what makes working on intelligent technology
27
103440
3496
但也是因為如此,
發展人工智能才極度困難。
01:46
extremely difficult.
28
106960
1600
01:49
And right now, AI research, well,
29
109640
3456
而現在,人工智能研究,嗯,
01:53
it's a bit lopsided on the tech side.
30
113120
2016
它有點不平衡,傾向技術面。
01:55
And that makes a lot of sense.
31
115160
2136
那是非常合理的。
01:57
See, for every qualitative thing about us --
32
117320
2456
對於人類的質性成分──
01:59
you know, those parts of us that are emotional, dynamic and subjective --
33
119800
4456
我們有很多情緒、 動態、主觀的部分──
02:04
we have to convert it to a quantitative metric:
34
124280
3136
我們得將之轉換為量性的度量:
02:07
something that can be represented with facts, figures and computer code.
35
127440
4360
可以用事實、數字, 和電腦程式來呈現的東西。
02:13
The issue is, there are many qualitative things
36
133000
3376
問題在於,有許多質性的東西
02:16
that we just can't put our finger on.
37
136400
1960
是我們實在無法認出來的。
02:20
So, think about hearing your favorite song for the first time.
38
140400
3200
試想看看,當你第一次 聽見你最喜歡的歌曲時,
02:25
What were you doing?
39
145200
1200
你在做什麼?
02:28
How did you feel?
40
148000
1200
你的感受是什麼?
02:30
Did you get goosebumps?
41
150720
1360
你有起雞皮疙瘩嗎?
02:33
Or did you get fired up?
42
153240
1640
你有充滿激情嗎?
02:36
Hard to describe, right?
43
156400
1200
很難形容,對吧?
02:38
See, parts of us feel so simple,
44
158800
2096
我們的一些部分感覺起來很簡單,
02:40
but under the surface, there's really a ton of complexity.
45
160920
3656
但在表面底下, 是相當可觀的複雜度。
02:44
And translating that complexity to machines
46
164600
2936
要把那種複雜翻譯給機器了解,
02:47
is what makes them modern-day moonshots.
47
167560
2856
讓這個目標變得非常高難度。
02:50
And I'm not convinced that we can answer these deeper questions
48
170440
4176
我不相信我們能夠只用零和一來回答
02:54
with just ones and zeros alone.
49
174640
1480
這些較深的問題,
02:57
So, in the lab, I've been creating art
50
177120
1936
所以,在實驗室中, 我一直在創造藝術,
02:59
as a way to help me design better experiences
51
179080
2456
藝術協助我為最先進的技術
03:01
for bleeding-edge technology.
52
181560
2096
設計出更佳的體驗。
03:03
And it's been serving as a catalyst
53
183680
1736
它一直是一種催化劑,
03:05
to beef up the more human ways that computers can relate to us.
54
185440
3840
用來讓電腦用更人性的 方式和我們相處。
03:10
Through art, we're tacking some of the hardest questions,
55
190000
2696
我們透過藝術來處理 一些最困難的問題,
03:12
like what does it really mean to feel?
56
192720
2360
比如「去感覺」究竟是什麼意思?
03:16
Or how do we engage and know how to be present with each other?
57
196120
4080
或者我們要如何與人互動, 如何在彼此相處時不是心不在焉?
03:20
And how does intuition affect the way that we interact?
58
200800
4000
直覺又如何影響我們互動的方式?
03:26
So, take for example human emotion.
59
206440
2056
以人類情緒為例。
03:28
Right now, computers can make sense of our most basic ones,
60
208520
3256
現在,電腦能夠理解 我們最基本的情緒,
03:31
like joy, sadness, anger, fear and disgust,
61
211800
3696
比如喜悅、悲傷、 憤怒、恐懼,及作噁,
03:35
by converting those characteristics to math.
62
215520
3000
做法是將那些特徵轉換為數學。
03:39
But what about the more complex emotions?
63
219400
2536
但遇到更複雜的情緒怎麼辦?
03:41
You know, those emotions
64
221960
1216
你們知道的,
03:43
that we have a hard time describing to each other?
65
223200
2376
有些情緒我們都很難對彼此形容?
03:45
Like nostalgia.
66
225600
1440
像是鄉愁。
03:47
So, to explore this, I created a piece of art, an experience,
67
227640
3936
為了探究這一點,我創作出了 一件藝術作品,一項體驗,
03:51
that asked people to share a memory,
68
231600
2096
我請大家分享一段記憶,
03:53
and I teamed up with some data scientists
69
233720
2136
我和一些資料科學家合作,
03:55
to figure out how to take an emotion that's so highly subjective
70
235880
3576
想辦法把一種高度主觀性的情緒
03:59
and convert it into something mathematically precise.
71
239480
3200
轉換成很精確的數學。
04:03
So, we created what we call a nostalgia score
72
243840
2136
於是我們創造出了鄉愁分數,
04:06
and it's the heart of this installation.
73
246000
2216
它是這項裝置的核心。
04:08
To do that, the installation asks you to share a story,
74
248240
3056
做法是這樣的:這項裝置 會先請你分享一個故事,
04:11
the computer then analyzes it for its simpler emotions,
75
251320
3256
接著,電腦會分析 這個故事中比較簡單的情緒,
04:14
it checks for your tendency to use past-tense wording
76
254600
2656
它會檢查你是否傾向 用比較多過去式修辭,
04:17
and also looks for words that we tend to associate with nostalgia,
77
257280
3336
也會尋找我們談鄉愁時 比較會用到的字眼,
04:20
like "home," "childhood" and "the past."
78
260640
3040
比如「家」、「童年」,和「過去」。
04:24
It then creates a nostalgia score
79
264760
2056
接著,它就會算出鄉愁分數,
04:26
to indicate how nostalgic your story is.
80
266840
2736
表示你的故事有多麼具有鄉愁。
04:29
And that score is the driving force behind these light-based sculptures
81
269600
4136
那個分數,就是這些以光線為 基礎之雕塑背後的驅動力,
04:33
that serve as physical embodiments of your contribution.
82
273760
3896
這些雕塑就是將你的貢獻 給具體化的結果。
04:37
And the higher the score, the rosier the hue.
83
277680
3216
分數越高,色調就會越偏玫瑰色。
04:40
You know, like looking at the world through rose-colored glasses.
84
280920
3936
就像是戴上玫瑰色的 眼鏡來看世界。
04:44
So, when you see your score
85
284880
2616
所以,當你看到你的分數
04:47
and the physical representation of it,
86
287520
2656
以及它的實體代表呈現之後,
04:50
sometimes you'd agree and sometimes you wouldn't.
87
290200
2936
有時你可能可以認同,有時不能。
04:53
It's as if it really understood how that experience made you feel.
88
293160
3480
它就像是真正去了解 那體驗帶給你什麼感覺。
04:57
But other times it gets tripped up
89
297400
2216
但其他時候,這裝置會犯錯,
04:59
and has you thinking it doesn't understand you at all.
90
299640
2560
讓你認為它完全不了解你。
05:02
But the piece really serves to show
91
302680
1896
但這件作品實際上呈現出
05:04
that if we have a hard time explaining the emotions that we have to each other,
92
304600
4056
如果我們都很難向彼此 解釋我們的某些情緒,
05:08
how can we teach a computer to make sense of them?
93
308680
2360
我們要如何教電腦理解那些情緒?
05:12
So, even the more objective parts about being human are hard to describe.
94
312360
3576
即使是人類比較客觀的 部分,也很難形容。
05:15
Like, conversation.
95
315960
1240
比如,對談。
05:17
Have you ever really tried to break down the steps?
96
317880
2736
你可曾真正試過把它拆解成步驟?
05:20
So think about sitting with your friend at a coffee shop
97
320640
2656
試想一下,你和朋友 坐在咖啡廳裡,
05:23
and just having small talk.
98
323320
1320
只是隨意閒聊。
05:25
How do you know when to take a turn?
99
325160
1720
怎麼知道何時要換人說話?
05:27
How do you know when to shift topics?
100
327440
1840
怎麼知道何時要換主題?
05:29
And how do you even know what topics to discuss?
101
329960
2720
甚至,怎麼知道要聊什麼主題?
05:33
See, most of us don't really think about it,
102
333560
2096
我們大部分人並不會去思考這些,
05:35
because it's almost second nature.
103
335680
1656
因為我們早就習慣成自然。
05:37
And when we get to know someone, we learn more about what makes them tick,
104
337360
3496
當我們漸漸了解一個人, 就會更清楚什麼會打動他,
05:40
and then we learn what topics we can discuss.
105
340880
2376
然後就會知道我們能討論什麼話題。
05:43
But when it comes to teaching AI systems how to interact with people,
106
343280
3656
但若要教導人工智能系統 怎麼和人互動,
05:46
we have to teach them step by step what to do.
107
346960
2856
我們得要教它們每一個步驟。
05:49
And right now, it feels clunky.
108
349840
3496
現在,感覺還很笨拙。
05:53
If you've ever tried to talk with Alexa, Siri or Google Assistant,
109
353360
4136
如果你曾經試過和 Alexa、 Siri,或 Google 助理說話,
05:57
you can tell that it or they can still sound cold.
110
357520
4200
你就知道它們聽起來還是很冰冷。
06:02
And have you ever gotten annoyed
111
362440
1656
你是否曾被它們惹惱,
06:04
when they didn't understand what you were saying
112
364120
2256
因為它們聽不懂你在說什麼,
06:06
and you had to rephrase what you wanted 20 times just to play a song?
113
366400
3840
你得要換二十種說法, 只為播放一首歌?
06:11
Alright, to the credit of the designers, realistic communication is really hard.
114
371440
4896
好吧,設計師是很辛苦的, 真實的溝通真的很難。
06:16
And there's a whole branch of sociology,
115
376360
2136
而社會學還有一整個分支,
06:18
called conversation analysis,
116
378520
1936
就叫做談話分析,
06:20
that tries to make blueprints for different types of conversation.
117
380480
3136
試圖為不同類型的對談繪製出藍圖。
06:23
Types like customer service or counseling, teaching and others.
118
383640
4080
像客服、諮詢、教學和其他類型。
06:28
I've been collaborating with a conversation analyst at the lab
119
388880
2936
我在實驗室和一位對談分析師合作,
06:31
to try to help our AI systems hold more human-sounding conversations.
120
391840
4696
試圖協助我們的人工智慧系統 在進行對談時聽起來更像人類。
06:36
This way, when you have an interaction with a chatbot on your phone
121
396560
3176
這麼一來,當你用手機 和聊天機器人互動,
06:39
or a voice-based system in the car,
122
399760
1856
或和車上的語音系統互動時,
06:41
it sounds a little more human and less cold and disjointed.
123
401640
3320
它聽起來會更像人一點, 不那麼冰冷,不那麼沒條理。
06:46
So I created a piece of art
124
406360
1336
我創作了一件藝術作品,
06:47
that tries to highlight the robotic, clunky interaction
125
407720
2816
試圖強調出機器式、笨拙的互動,
06:50
to help us understand, as designers,
126
410560
1976
來協助我們設計師了解
06:52
why it doesn't sound human yet and, well, what we can do about it.
127
412560
4576
為什麼它聽起來還不像人, 以及對此我們能怎麼辦。
06:57
The piece is called Bot to Bot
128
417160
1456
這件作品叫機器人對機器人,
06:58
and it puts one conversational system against another
129
418640
2936
它讓一個交談系統 和另一個交談系統互動,
07:01
and then exposes it to the general public.
130
421600
2536
接著讓系統接觸一般民眾。
07:04
And what ends up happening is that you get something
131
424160
2496
最後發生的狀況就是, 你得到一種產物,
07:06
that tries to mimic human conversation,
132
426680
1896
它會試圖模仿人類交談,
07:08
but falls short.
133
428600
1896
但達不到標準。
07:10
Sometimes it works and sometimes it gets into these, well,
134
430520
2736
有時它行得通,
有時它會陷入誤解的迴圈當中。
07:13
loops of misunderstanding.
135
433280
1536
07:14
So even though the machine-to-machine conversation can make sense,
136
434840
3096
雖然機器對機器的交談是有意義的,
07:17
grammatically and colloquially,
137
437960
2016
在文法上和口語上皆是如此,
07:20
it can still end up feeling cold and robotic.
138
440000
3216
但它最後的感覺 可能還是很冰冷、很機械化。
07:23
And despite checking all the boxes, the dialogue lacks soul
139
443240
4016
儘管所有的條件都做到了, 對話還是沒有靈魂,
07:27
and those one-off quirks that make each of us who we are.
140
447280
3136
缺乏讓我們每個人 獨一無二的個人特色。
07:30
So while it might be grammatically correct
141
450440
2056
所以,即使文法是正確的,
07:32
and uses all the right hashtags and emojis,
142
452520
2656
所有「#」和表情符號的 應用也都沒錯,
07:35
it can end up sounding mechanical and, well, a little creepy.
143
455200
4136
聽起來還是很機械, 且還有一點詭異。
07:39
And we call this the uncanny valley.
144
459360
2336
我們稱之為恐怖谷。
07:41
You know, that creepiness factor of tech
145
461720
1936
技術中的怪異因子,
07:43
where it's close to human but just slightly off.
146
463680
2856
很接近人類,但又還差那麼一點。
07:46
And the piece will start being
147
466560
1456
這件作品將開始成為一種方式,
07:48
one way that we test for the humanness of a conversation
148
468040
3216
讓我們能夠測試對話的人性
07:51
and the parts that get lost in translation.
149
471280
2160
和翻譯中迷失的部分。
07:54
So there are other things that get lost in translation, too,
150
474560
2856
在翻譯中還有其他的東西會遺失,
07:57
like human intuition.
151
477440
1616
比如人類直覺。
07:59
Right now, computers are gaining more autonomy.
152
479080
2776
目前,電腦越來越自動化。
08:01
They can take care of things for us,
153
481880
1736
它們能為我們處理事情,
08:03
like change the temperature of our houses based on our preferences
154
483640
3176
比如根據我們的偏好 幫我們改變房子裡的室溫,
08:06
and even help us drive on the freeway.
155
486840
2160
甚至協助我們在高速公路上開車。
08:09
But there are things that you and I do in person
156
489560
2496
但有些我們個人會做的事情,
08:12
that are really difficult to translate to AI.
157
492080
2760
很難翻譯讓人工智能理解。
08:15
So think about the last time that you saw an old classmate or coworker.
158
495440
4360
想想看你上回碰見一位 老同學或同事的情況。
08:21
Did you give them a hug or go in for a handshake?
159
501080
2480
你是給他一個擁抱,還是握個手?
08:24
You probably didn't think twice
160
504800
1496
你可能根本不用思考,
08:26
because you've had so many built up experiences
161
506320
2336
因為你有許多既有經驗,
08:28
that had you do one or the other.
162
508680
2000
會影響你決定做前者或後者。
08:31
And as an artist, I feel that access to one's intuition,
163
511440
3456
身為藝術家,我覺得 正是因為會使用直覺,
08:34
your unconscious knowing,
164
514920
1416
潛意識中所知道的事,
08:36
is what helps us create amazing things.
165
516360
3056
才讓我們能創造出了不起的事物。
08:39
Big ideas, from that abstract, nonlinear place in our consciousness
166
519440
4056
大點子是來自於我們意識中 那塊抽象、非線性的地方,
08:43
that is the culmination of all of our experiences.
167
523520
2960
我們所有經驗的最高點。
08:47
And if we want computers to relate to us and help amplify our creative abilities,
168
527840
4656
若我們想讓電腦能和我們相處, 並協助強化我們的創意能力,
08:52
I feel that we'll need to start thinking about how to make computers be intuitive.
169
532520
3896
我覺得我們得要開始思考 如何讓電腦能有直覺。
08:56
So I wanted to explore how something like human intuition
170
536440
3096
所以,我想要探究的是, 像人類直覺這類東西
08:59
could be directly translated to artificial intelligence.
171
539560
3456
如何能被直接翻譯成 人工智慧能懂的語言。
09:03
And I created a piece that explores computer-based intuition
172
543040
3216
我創作了一件作品, 來探究實體空間中
09:06
in a physical space.
173
546280
1320
以電腦為基礎的直覺。
09:08
The piece is called Wayfinding,
174
548480
1696
這件作品叫做「找路」,
09:10
and it's set up as a symbolic compass that has four kinetic sculptures.
175
550200
3936
它的設計是個象徵性的羅盤, 具有四個動態雕塑。
09:14
Each one represents a direction,
176
554160
2056
每一個都代表一個方向,
09:16
north, east, south and west.
177
556240
2120
北、東、南、西。
09:19
And there are sensors set up on the top of each sculpture
178
559080
2696
每個雕塑上都裝有感測器,
09:21
that capture how far away you are from them.
179
561800
2256
能知道你離它們有多遠。
09:24
And the data that gets collected
180
564080
1816
資料會被收集起來,
09:25
ends up changing the way that sculptures move
181
565920
2136
最後會改變雕塑移動的方式,
09:28
and the direction of the compass.
182
568080
2040
以及羅盤的方向。
09:31
The thing is, the piece doesn't work like the automatic door sensor
183
571360
3656
重點是,這件作品並不是像 自動門的感測器那樣運作,
09:35
that just opens when you walk in front of it.
184
575040
2656
當你走到自動門前時它就會打。
09:37
See, your contribution is only a part of its collection of lived experiences.
185
577720
5056
它在收集活體的經驗, 而你的貢獻只是其中一部分。
09:42
And all of those experiences affect the way that it moves.
186
582800
4056
所有經驗都會影響它移動的方式。
09:46
So when you walk in front of it,
187
586880
1736
所以,當你走到它前方時,
09:48
it starts to use all of the data
188
588640
1976
它會開始用所有資料,
09:50
that it's captured throughout its exhibition history --
189
590640
2616
所有它在展示期間 所捕捉到的資料──
09:53
or its intuition --
190
593280
1816
可以說是它的直覺──
09:55
to mechanically respond to you based on what it's learned from others.
191
595120
3560
根據它從其他人身上學到的, 來對你做出機械式的反應。
09:59
And what ends up happening is that as participants
192
599480
2536
最後的結果是,我們參與者開始
10:02
we start to learn the level of detail that we need
193
602040
2816
學到許多細節, 我們需要這些細節,
10:04
in order to manage expectations
194
604880
2016
才能管理來自人類
10:06
from both humans and machines.
195
606920
2776
以及機器的期望。
10:09
We can almost see our intuition being played out on the computer,
196
609720
3616
我們幾乎可以看見 我們的直覺在電腦上呈現出來,
10:13
picturing all of that data being processed in our mind's eye.
197
613360
3600
描繪出我們心靈之眼所看見的 所有被處理的資料。
10:17
My hope is that this type of art
198
617560
1656
我希望這類藝術
10:19
will help us think differently about intuition
199
619240
2416
能夠協助我們對直覺 有不同的想法,
10:21
and how to apply that to AI in the future.
200
621680
2280
及未來要如何 把它用在人工智能上。
10:24
So these are just a few examples of how I'm using art to feed into my work
201
624480
3936
這些只是幾個例子,
用來說明我這個 人工智能設計師兼研究者,
10:28
as a designer and researcher of artificial intelligence.
202
628440
3096
如何把藝術注入工作中。
10:31
And I see it as a crucial way to move innovation forward.
203
631560
3496
我認為,要讓創新再向前邁進, 這種方式十分關鍵。
10:35
Because right now, there are a lot of extremes when it comes to AI.
204
635080
4376
因為,目前談到人工智能時, 會有許多極端想法。
10:39
Popular movies show it as this destructive force
205
639480
2816
熱門電影把人工智能 呈現成毀滅性的力量,
10:42
while commercials are showing it as a savior
206
642320
3056
廣告又把它呈現成救星,
10:45
to solve some of the world's most complex problems.
207
645400
2936
用來解決世界上最困難的一些問題。
10:48
But regardless of where you stand,
208
648360
2536
但,不論你的立場為何,
10:50
it's hard to deny that we're living in a world
209
650920
2176
很難否認,我們所居住的世界
10:53
that's becoming more and more digital by the second.
210
653120
2576
每一秒鐘都在變得越來越數位。
10:55
Our lives revolve around our devices, smart appliances and more.
211
655720
4600
我們的生活圍繞著我們的裝置、 智慧設備等等在轉動。
11:01
And I don't think this will let up any time soon.
212
661400
2320
我不認為近期內這會減緩下來。
11:04
So, I'm trying to embed more humanness from the start.
213
664400
3736
所以,我試著打從一開始 就嵌入更多的人性。
11:08
And I have a hunch that bringing art into an AI research process
214
668160
5136
我有預感,將藝術帶入
人工智能的研究過程是一種方法。
11:13
is a way to do just that.
215
673320
1896
11:15
Thank you.
216
675240
1216
謝謝。
11:16
(Applause)
217
676480
2280
(掌聲)
關於本網站

本網站將向您介紹對學習英語有用的 YouTube 視頻。 您將看到來自世界各地的一流教師教授的英語課程。 雙擊每個視頻頁面上顯示的英文字幕,從那裡播放視頻。 字幕與視頻播放同步滾動。 如果您有任何意見或要求,請使用此聯繫表與我們聯繫。

https://forms.gle/WvT1wiN1qDtmnspy7