TEDxPortland
Maurice Conti: The incredible inventions of intuitive AI
繆萊斯．康提: 直覺型人工智慧的驚人發明
Filmed:
Readability: 4.6
6,173,221 views
如果你在一個設計工具上安裝了神經系統，那你會得到什麼東西呢？電腦增進了我們的思考與想像能力、機械人系統可以提出（建造出）前所未見、具設計感的全新橋樑、汽車、無人機......等等，這些都可經由它們自己自造完成。讓我們與未來學家繆萊斯．康提 (Maurice Conti) 一起進行這場擴增年代的冒險之旅，並一窺機械人與人類肩並肩工作，共同來成就一個單憑一己之力是無法完成的美麗新世界
Maurice Conti - Designer, futurist
Maurice Conti explores new partnerships between technology, nature and humanity. Full bio
Maurice Conti explores new partnerships between technology, nature and humanity. Full bio
Double-click the English transcript below to play the video.
00:12
How many of you are creatives,
0
735
2289
你們有多少人是創意人、
00:15
designers, engineers,
entrepreneurs, artists,
entrepreneurs, artists,
1
3048
3624
設計師、工程師、企業家、藝術家？
00:18
or maybe you just have
a really big imagination?
a really big imagination?
2
6696
2387
或者你只是有無遠弗屆的想像力？
00:21
Show of hands? (Cheers)
3
9107
1848
請舉一下手？（歡呼聲）
00:22
That's most of you.
4
10979
1181
現場大部分人都是。
00:25
I have some news for us creatives.
5
13334
2294
我有一些消息要給我們的創意人。
00:28
Over the course of the next 20 years,
6
16714
2573
接下來的 20 年，
00:33
more will change around
the way we do our work
the way we do our work
7
21471
2973
很多我們工作的方式，
00:37
than has happened in the last 2,000.
8
25382
2157
將會遠遠不同於過去的 2000 年。
00:40
In fact, I think we're at the dawn
of a new age in human history.
of a new age in human history.
9
28511
4628
實際上，我認為我們正處在
人類歷史新世代的黎明。
人類歷史新世代的黎明。
00:45
Now, there have been four major historical
eras defined by the way we work.
eras defined by the way we work.
10
33645
4761
人類工作的方式，
有四個主要的歷史階段。
有四個主要的歷史階段。
00:51
The Hunter-Gatherer Age
lasted several million years.
lasted several million years.
11
39404
3275
人類歷經了幾百萬年的
狩獵採集時代。
狩獵採集時代。
00:55
And then the Agricultural Age
lasted several thousand years.
lasted several thousand years.
12
43163
3576
然後經歷了幾千年的農業時代。
00:59
The Industrial Age lasted
a couple of centuries.
a couple of centuries.
13
47195
3490
工業時代則延續了幾世紀。
01:02
And now the Information Age
has lasted just a few decades.
has lasted just a few decades.
14
50709
4287
而目前的資訊時代才走了幾十年。
01:07
And now today, we're on the cusp
of our next great era as a species.
of our next great era as a species.
15
55020
5220
如今，身為人類的我們，
即將邁入下一個偉大的時代。
即將邁入下一個偉大的時代。
01:13
Welcome to the Augmented Age.
16
61296
2680
歡迎來到「擴增時代」。
01:16
In this new era, your natural human
capabilities are going to be augmented
capabilities are going to be augmented
17
64000
3693
在這個新時代，
人類天生的能力將會被強化擴增，
人類天生的能力將會被強化擴增，
01:19
by computational systems
that help you think,
that help you think,
18
67717
3068
電腦計算系統將幫助你思考、
01:22
robotic systems that help you make,
19
70809
2186
機械人系統協助你製造、
01:25
and a digital nervous system
20
73019
1648
遠超過你自然感官強度的
數位神經系統，
數位神經系統，
01:26
that connects you to the world
far beyond your natural senses.
far beyond your natural senses.
21
74691
3690
能夠讓你與全世界接軌。
01:31
Let's start with cognitive augmentation.
22
79437
1942
我們先從「認知擴增」談起。
01:33
How many of you are augmented cyborgs?
23
81403
2200
現場有多少人是「強化的半機械人」？
01:36
(Laughter)
24
84133
2650
（笑聲）
01:38
I would actually argue
that we're already augmented.
that we're already augmented.
25
86807
2821
其實我想說的是，
我們都已經被強化、擴增了。
我們都已經被強化、擴增了。
01:42
Imagine you're at a party,
26
90288
1504
想像你正在參加一場派對，
01:43
and somebody asks you a question
that you don't know the answer to.
that you don't know the answer to.
27
91816
3520
有人問了你一個
你不知道如何回答的問題。
你不知道如何回答的問題。
01:47
If you have one of these,
in a few seconds, you can know the answer.
in a few seconds, you can know the answer.
28
95360
3760
如果你有這個，只要幾秒鐘，
你就會得到答案。
你就會得到答案。
01:51
But this is just a primitive beginning.
29
99869
2299
但這也只是剛開始而已。
01:54
Even Siri is just a passive tool.
30
102863
3331
甚至 Siri 也只是個被動工具。
01:58
In fact, for the last
three-and-a-half million years,
three-and-a-half million years,
31
106660
3381
實際上，在過去的 350 萬年，
02:02
the tools that we've had
have been completely passive.
have been completely passive.
32
110065
3109
我們所有的工具都是被動的。
02:06
They do exactly what we tell them
and nothing more.
and nothing more.
33
114203
3655
它們只會照我們的指令去做，
僅此而已。
僅此而已。
02:09
Our very first tool only cut
where we struck it.
where we struck it.
34
117882
3101
我們最早使用的工具，
遵循一個口令一個動作的指示。
遵循一個口令一個動作的指示。
02:13
The chisel only carves
where the artist points it.
where the artist points it.
35
121822
3040
藝術家指哪裡，雕刻刀就雕刻哪裡。
02:17
And even our most advanced tools
do nothing without our explicit direction.
do nothing without our explicit direction.
36
125343
5641
即使最先進的工具，如果沒有
我們明確的指令也不會工作。
我們明確的指令也不會工作。
02:23
In fact, to date, and this
is something that frustrates me,
is something that frustrates me,
37
131008
3181
說真的，時到今日，
有件事仍讓我感覺很挫敗，
有件事仍讓我感覺很挫敗，
02:26
we've always been limited
38
134213
1448
我們一直以來都被限制在
02:27
by this need to manually
push our wills into our tools --
push our wills into our tools --
39
135685
3501
「需要動手將我們的意念
傳達給工具」的這種迷思框框中——
傳達給工具」的這種迷思框框中——
02:31
like, manual,
literally using our hands,
literally using our hands,
40
139210
2297
就是得動手去做，即使有了電腦
還是得靠雙手。
還是得靠雙手。
02:33
even with computers.
41
141531
1428
02:36
But I'm more like Scotty in "Star Trek."
42
144072
2463
但我還是比較喜歡當
<<星際迷航>>裡的史考迪。
<<星際迷航>>裡的史考迪。
02:38
(Laughter)
43
146559
1850
（笑聲）
02:40
I want to have a conversation
with a computer.
with a computer.
44
148433
2146
我也想跟電腦對話。
02:42
I want to say, "Computer,
let's design a car,"
let's design a car,"
45
150603
2970
當我說，「電腦，
我們來設計一輛車吧！」
我們來設計一輛車吧！」
02:45
and the computer shows me a car.
46
153597
1539
然後電腦就會顯示一輛車給我看。
02:47
And I say, "No, more fast-looking,
and less German,"
and less German,"
47
155160
2608
然後我說：「不，要拉風一點，
德國味兒少一點。」
德國味兒少一點。」
02:49
and bang, the computer shows me an option.
48
157792
2163
接著「蹦」，
電腦給了我一個新選擇。
電腦給了我一個新選擇。
02:51
(Laughter)
49
159979
1865
（笑聲）
02:54
That conversation might be
a little ways off,
a little ways off,
50
162208
2306
這樣的對話可能有點不切實際，
02:56
probably less than many of us think,
51
164538
2665
也許沒有我們認為的那麼不切實際，
02:59
but right now,
52
167227
1763
但現在，
03:01
we're working on it.
53
169014
1151
我們正在做這件事。
這些工具將帶領我們大躍進，
從被動轉為衍生。
從被動轉為衍生。
03:02
Tools are making this leap
from being passive to being generative.
from being passive to being generative.
54
170189
4033
03:06
Generative design tools
use a computer and algorithms
use a computer and algorithms
55
174831
3308
「衍生設計工具 」
是利用電腦及演算法，
是利用電腦及演算法，
03:10
to synthesize geometry
56
178163
2608
合成出幾何結構，
03:12
to come up with new designs
all by themselves.
all by themselves.
57
180795
2754
產製出新的設計圖，
全部都是它們自己構思出來的。
全部都是它們自己構思出來的。
03:15
All it needs are your goals
and your constraints.
and your constraints.
58
183996
2748
你只需要設定目標及限制條件。
03:18
I'll give you an example.
59
186768
1408
我給各位舉個例子。
03:20
In the case of this aerial drone chassis,
60
188200
2788
就拿這個無人機底盤為例，
03:23
all you would need to do
is tell it something like,
is tell it something like,
61
191012
2626
你唯一要做的，
就是告訴它你的需求，
就是告訴它你的需求，
03:25
it has four propellers,
62
193662
1273
像是，你要四個螺旋槳的，
03:26
you want it to be
as lightweight as possible,
as lightweight as possible,
63
194959
2131
它越輕越好，
03:29
and you need it to be
aerodynamically efficient.
aerodynamically efficient.
64
197114
2270
空氣動力學表現效率佳的。
03:31
Then what the computer does
is it explores the entire solution space:
is it explores the entire solution space:
65
199408
4914
電腦做的，就是探索
所有可能的解決方案：
所有可能的解決方案：
03:36
every single possibility that solves
and meets your criteria --
and meets your criteria --
66
204346
3927
每一個能解決且符合你標準的
可能方案——
可能方案——
03:40
millions of them.
67
208297
1442
有上百萬個。
03:41
It takes big computers to do this.
68
209763
1975
這需要大型電腦才能做到。
03:43
But it comes back to us with designs
69
211762
1955
但它回饋給我們的設計方案，
03:45
that we, by ourselves,
never could've imagined.
never could've imagined.
70
213741
3143
是我們單憑自己無法想像出來的
設計方案。
設計方案。
03:49
And the computer's coming up
with this stuff all by itself --
with this stuff all by itself --
71
217326
2912
電腦憑藉著自己的能力
做出這些東西——
做出這些東西——
03:52
no one ever drew anything,
72
220262
1678
我們人類沒有動筆畫任何東西，
03:53
and it started completely from scratch.
73
221964
2086
完全是它自己從頭、從零畫起的。
03:57
And by the way, it's no accident
74
225038
2387
順便一提，這可不是偶然......
無人機的機體長的像飛鼠的骨盆，
03:59
that the drone body looks just like
the pelvis of a flying squirrel.
the pelvis of a flying squirrel.
75
227449
3481
04:03
(Laughter)
76
231287
2007
（笑聲）
04:06
It's because the algorithms
are designed to work
are designed to work
77
234040
2302
那是因為演算法的計算模式，
04:08
the same way evolution does.
78
236366
1637
是遵循生物演化模式而設計的。
04:10
What's exciting is we're starting
to see this technology
to see this technology
79
238715
2660
令人興奮的是，
我們開始見證這樣的科技
我們開始見證這樣的科技
04:13
out in the real world.
80
241399
1159
在現實世界中實現。
04:14
We've been working with Airbus
for a couple of years
for a couple of years
81
242582
2452
我們與空中巴士
（歐洲最大飛機製造商）
（歐洲最大飛機製造商）
合作開發未來的概念機
已經好幾年了，
已經好幾年了，
04:17
on this concept plane for the future.
82
245058
1909
04:18
It's a ways out still.
83
246991
2070
這計畫目前還在進行。
04:21
But just recently we used
a generative-design AI
a generative-design AI
84
249085
3780
但最近，我們用了
衍生設計的人工智慧
衍生設計的人工智慧
04:24
to come up with this.
85
252889
1807
做出了這一個。
04:27
This is a 3D-printed cabin partition
that's been designed by a computer.
that's been designed by a computer.
86
255609
5153
這是一個 3D 列印的客艙隔間板，
由一台電腦所設計。
由一台電腦所設計。
04:32
It's stronger than the original
yet half the weight,
yet half the weight,
87
260786
2824
它比原款式還要堅固，
但重量只有原本的一半，
但重量只有原本的一半，
04:35
and it will be flying
in the Airbus A320 later this year.
in the Airbus A320 later this year.
88
263634
3146
今年稍晚，它將跟
A320 空中巴士一起飛上天。
A320 空中巴士一起飛上天。
所以現在電腦會主動生成、衍生了；
04:39
So computers can now generate;
89
267405
1559
04:40
they can come up with their own solutions
to our well-defined problems.
to our well-defined problems.
90
268988
4595
它們可以對界定明確的問題，
給出自己的答案。
給出自己的答案。
04:46
But they're not intuitive.
91
274677
1310
但它們並不是靠直覺做事。
04:48
They still have to start from scratch
every single time,
every single time,
92
276011
3086
它們還是每次都得從頭開始，
04:51
and that's because they never learn.
93
279121
2565
因為它們不會學習。
04:54
Unlike Maggie.
94
282368
1766
不像瑪姬。
04:56
(Laughter)
95
284158
1581
（笑聲）
瑪姬其實比我們最先進的
設計工具都還要聰明。
設計工具都還要聰明。
04:57
Maggie's actually smarter
than our most advanced design tools.
than our most advanced design tools.
96
285763
3297
05:01
What do I mean by that?
97
289467
1440
這是什麼意思？
05:02
If her owner picks up that leash,
98
290931
1590
如果狗主人拿起狗鍊，
05:04
Maggie knows with a fair
degree of certainty
degree of certainty
99
292545
2068
瑪姬就知道有相當的確定性，
05:06
it's time to go for a walk.
100
294637
1404
主人要帶她去散步了。
05:08
And how did she learn?
101
296065
1185
她是怎麼知道的？
05:09
Well, every time the owner picked up
the leash, they went for a walk.
the leash, they went for a walk.
102
297274
3324
因為，每當主人拿起狗鍊，
他們就會一起去散步。
他們就會一起去散步。
05:12
And Maggie did three things:
103
300622
1878
瑪姬會做三件事：
05:14
she had to pay attention,
104
302524
1869
她必須專注、
05:16
she had to remember what happened
105
304417
2082
必須記得發生過什麼事、
05:18
and she had to retain and create
a pattern in her mind.
a pattern in her mind.
106
306523
4017
必須在腦中記憶並產生一個模式。
05:23
Interestingly, that's exactly what
107
311429
2095
有趣的是，這正是電腦科學家
05:25
computer scientists
have been trying to get AIs to do
have been trying to get AIs to do
108
313548
2523
過去 60 年來，一直嘗試
要讓人工智慧做的事。
要讓人工智慧做的事。
05:28
for the last 60 or so years.
109
316095
1859
05:30
Back in 1952,
110
318683
1349
回想一下 1952 年，
05:32
they built this computer
that could play Tic-Tac-Toe.
that could play Tic-Tac-Toe.
111
320056
3801
科學家建立了這一台電腦，
它會玩井字遊戲。
它會玩井字遊戲。
05:37
Big deal.
112
325081
1160
真了不起。
05:39
Then 45 years later, in 1997,
113
327029
3000
45 年後，1997 年，
05:42
Deep Blue beats Kasparov at chess.
114
330053
2472
深藍擊敗了當時的西洋棋世界冠軍
卡司帕洛夫，
卡司帕洛夫，
05:46
2011, Watson beats these two
humans at Jeopardy,
humans at Jeopardy,
115
334046
4968
2011年，華生 (IBM電腦)
在<<危機邊緣>>擊敗這兩個人，
在<<危機邊緣>>擊敗這兩個人，
05:51
which is much harder for a computer
to play than chess is.
to play than chess is.
116
339038
2928
對電腦來說，這比下棋難多了。
05:53
In fact, rather than working
from predefined recipes,
from predefined recipes,
117
341990
3812
事實上，華生並不是從
預先定義的題庫中來找答案，
預先定義的題庫中來找答案，
05:57
Watson had to use reasoning
to overcome his human opponents.
to overcome his human opponents.
118
345826
3323
它必須使用推理
來擊敗它的人類對手。
來擊敗它的人類對手。
06:02
And then a couple of weeks ago,
119
350393
2439
就在幾個禮拜前，
06:04
DeepMind's AlphaGo beats
the world's best human at Go,
the world's best human at Go,
120
352856
4262
DeepMind 的阿爾法圍棋
擊敗了世界圍棋冠軍，
擊敗了世界圍棋冠軍，
06:09
which is the most difficult
game that we have.
game that we have.
121
357142
2212
而圍棋是我們人類最複雜的遊戲。
06:11
In fact, in Go, there are more
possible moves
possible moves
122
359378
2896
事實上，圍棋走法的可能性
06:14
than there are atoms in the universe.
123
362298
2024
超過全宇宙的原子數量。
06:18
So in order to win,
124
366210
1826
所以為了取得勝利，
06:20
what AlphaGo had to do
was develop intuition.
was develop intuition.
125
368060
2618
阿爾法圍棋必須學會使用直覺。
06:23
And in fact, at some points,
AlphaGo's programmers didn't understand
AlphaGo's programmers didn't understand
126
371098
4110
實際上，有些下法，
阿爾法圍棋的程式人員也不懂
阿爾法圍棋的程式人員也不懂
06:27
why AlphaGo was doing what it was doing.
127
375232
2286
為什麼阿爾法圍棋要那樣下。
06:31
And things are moving really fast.
128
379451
1660
世界變化真快。
06:33
I mean, consider --
in the space of a human lifetime,
in the space of a human lifetime,
129
381135
3227
我的意思是，想像一下——
在人類壽命這麼長的時間裡，
在人類壽命這麼長的時間裡，
06:36
computers have gone from a child's game
130
384386
2233
電腦已經從小孩子的遊戲
06:39
to what's recognized as the pinnacle
of strategic thought.
of strategic thought.
131
387920
3048
發展到策略思考的頂尖水平。
06:43
What's basically happening
132
391999
2417
電腦基本上的發展，
06:46
is computers are going
from being like Spock
from being like Spock
133
394440
3310
已經從史巴克大副進化到
06:49
to being a lot more like Kirk.
134
397774
1949
寇克艦長。
06:51
(Laughter)
135
399747
3618
（笑聲）
06:55
Right? From pure logic to intuition.
136
403389
3424
對吧？從純粹的邏輯運算
到直覺判斷。
到直覺判斷。
07:00
Would you cross this bridge?
137
408184
1743
你們會跨過這座橋嗎？
07:02
Most of you are saying, "Oh, hell no!"
138
410609
2323
大部分人應該都說，
「喔，打死我也不要！」
「喔，打死我也不要！」
07:04
(Laughter)
139
412956
1308
（笑聲）
07:06
And you arrived at that decision
in a split second.
in a split second.
140
414288
2657
你瞬間就可以做出這個決定。
07:08
You just sort of knew
that bridge was unsafe.
that bridge was unsafe.
141
416969
2428
你就是隱約知道那座橋並不安全。
07:11
And that's exactly the kind of intuition
142
419421
1989
這種直覺判斷，
07:13
that our deep-learning systems
are starting to develop right now.
are starting to develop right now.
143
421434
3568
就是目前我們深度學習系統
正在發展的能力。
正在發展的能力。
07:17
Very soon, you'll literally be able
144
425722
1707
很快的，各位就可以
07:19
to show something you've made,
you've designed,
you've designed,
145
427453
2206
把你製作、設計出來的東西
07:21
to a computer,
146
429683
1153
拿給電腦評判，
07:22
and it will look at it and say,
147
430860
1489
然後它看完後會說，
07:24
"Sorry, homie, that'll never work.
You have to try again."
You have to try again."
148
432373
2823
「抱歉，兄弟，這東西行不通，
你再試試別的吧！」
你再試試別的吧！」
07:27
Or you could ask it if people
are going to like your next song,
are going to like your next song,
149
435854
3070
或者你可以問它，
人們會不會喜歡你的新歌？
人們會不會喜歡你的新歌？
07:31
or your next flavor of ice cream.
150
439773
2063
或者你冰淇淋的新口味？
07:35
Or, much more importantly,
151
443549
2579
再或者，更重要的，
07:38
you could work with a computer
to solve a problem
to solve a problem
152
446152
2364
你可以跟電腦一起解決
我們從未面臨過的問題。
07:40
that we've never faced before.
153
448540
1637
07:42
For instance, climate change.
154
450201
1401
例如，氣候變遷問題。
07:43
We're not doing a very
good job on our own,
good job on our own,
155
451626
2020
我們自己沒有做得很好的事，
07:45
we could certainly use
all the help we can get.
all the help we can get.
156
453670
2245
我們當然可以利用身邊
各種資源來幫忙解決。
各種資源來幫忙解決。
07:47
That's what I'm talking about,
157
455939
1458
這就是我接下來要談的，
07:49
technology amplifying
our cognitive abilities
our cognitive abilities
158
457421
2555
科技強化了我們的認知能力，
07:52
so we can imagine and design things
that were simply out of our reach
that were simply out of our reach
159
460000
3552
所以我們可以想像並設計出，
當我們還未具有強化擴增能力時
當我們還未具有強化擴增能力時
07:55
as plain old un-augmented humans.
160
463576
2559
所未能創造出來的東西。
07:59
So what about making
all of this crazy new stuff
all of this crazy new stuff
161
467984
2941
那麼，製造這些我們即將發明設計的
08:02
that we're going to invent and design?
162
470949
2441
瘋狂新產品會如何呢？
08:05
I think the era of human augmentation
is as much about the physical world
is as much about the physical world
163
473952
4093
我認為在人類擴增的時代，
現實世界
現實世界
及虛擬智慧領域
與其皆有不分軒輊的重要相關性。
與其皆有不分軒輊的重要相關性。
08:10
as it is about the virtual,
intellectual realm.
intellectual realm.
164
478069
3065
08:13
How will technology augment us?
165
481833
1921
科技將會如何強化我們？
08:16
In the physical world, robotic systems.
166
484261
2473
在現實世界，就是機械人系統。
08:19
OK, there's certainly a fear
167
487620
1736
沒錯，很多人擔心，
08:21
that robots are going to take
jobs away from humans,
jobs away from humans,
168
489380
2488
機械人會搶走人類的工作，
08:23
and that is true in certain sectors.
169
491892
1830
在某些領域，確實是如此。
08:26
But I'm much more interested in this idea
170
494174
2878
但，我對以下的想法比較有興趣，
08:29
that humans and robots working together
are going to augment each other,
are going to augment each other,
171
497076
5010
就是，人類與機械人
將會一起工作並互相強化，
將會一起工作並互相強化，
08:34
and start to inhabit a new space.
172
502110
2058
並開創出一種新的共生空間。
08:36
This is our applied research lab
in San Francisco,
in San Francisco,
173
504192
2362
這是我們在舊金山的
應用研究實驗室，
應用研究實驗室，
08:38
where one of our areas of focus
is advanced robotics,
is advanced robotics,
174
506578
3142
我們專研的領域之一就是
高階機械人，
高階機械人，
08:41
specifically, human-robot collaboration.
175
509744
2511
特別是人機合作的領域。
08:45
And this is Bishop, one of our robots.
176
513034
2759
這是畢夏普，
我們其中的一個機器人。
我們其中的一個機器人。
08:47
As an experiment, we set it up
177
515817
1789
在實驗裡，我將它設定為
08:49
to help a person working in construction
doing repetitive tasks --
doing repetitive tasks --
178
517630
3460
在建築領域中，
幫助人類做重複性的工作——
幫助人類做重複性的工作——
08:53
tasks like cutting out holes for outlets
or light switches in drywall.
or light switches in drywall.
179
521984
4194
比如說，在石牆上打出一個
插座孔或電燈開關孔。
插座孔或電燈開關孔。
08:58
(Laughter)
180
526202
2466
（笑聲）
09:01
So, Bishop's human partner
can tell what to do in plain English
can tell what to do in plain English
181
529877
3111
所以，畢夏普的人類夥伴
就可以用簡單的英語和手勢
就可以用簡單的英語和手勢
09:05
and with simple gestures,
182
533012
1305
告訴它該做什麼，
09:06
kind of like talking to a dog,
183
534341
1447
有點像是在跟狗狗說話。
09:07
and then Bishop executes
on those instructions
on those instructions
184
535812
2143
然後畢夏普會以完美的準確度
09:09
with perfect precision.
185
537979
1892
執行人類所下達的指令。
09:11
We're using the human
for what the human is good at:
for what the human is good at:
186
539895
2989
我們讓人類做人類擅長的事，像是：
09:14
awareness, perception and decision making.
187
542908
2333
需要意識力、洞察力、做決策的工作。
09:17
And we're using the robot
for what it's good at:
for what it's good at:
188
545265
2240
我們讓機械人做機械人擅長的事，
像是：
像是：
09:19
precision and repetitiveness.
189
547529
1748
準確度及重複性的工作。
09:22
Here's another cool project
that Bishop worked on.
that Bishop worked on.
190
550252
2367
畢夏普還有另一個很酷的專案。
09:24
The goal of this project,
which we called the HIVE,
which we called the HIVE,
191
552643
3075
這個專案的目標，
我們稱它為<<蜂巢>>，
我們稱它為<<蜂巢>>，
09:27
was to prototype the experience
of humans, computers and robots
of humans, computers and robots
192
555742
3851
主要目標是把人類、電腦、
機械人的經驗結合起來，
機械人的經驗結合起來，
09:31
all working together to solve
a highly complex design problem.
a highly complex design problem.
193
559617
3220
一起工作解決極複雜的設計問題。
09:35
The humans acted as labor.
194
563793
1451
人類的工作是
09:37
They cruised around the construction site,
they manipulated the bamboo --
they manipulated the bamboo --
195
565268
3473
在建築基地巡邏監工
並熟練地操作竹子——
並熟練地操作竹子——
09:40
which, by the way,
because it's a non-isomorphic material,
because it's a non-isomorphic material,
196
568765
2756
順便一提，因為每一根竹子的
材料性質都不一樣，
材料性質都不一樣，
09:43
is super hard for robots to deal with.
197
571545
1874
所以機械人操作起來非常困難。
09:45
But then the robots
did this fiber winding,
did this fiber winding,
198
573443
2022
但機械人做的是彎曲竹子的纖維，
09:47
which was almost impossible
for a human to do.
for a human to do.
199
575489
2451
這種事人類幾乎做不來。
09:49
And then we had an AI
that was controlling everything.
that was controlling everything.
200
577964
3621
然後我們讓一台人工智慧
來控制所有的東西。
來控制所有的東西。
09:53
It was telling the humans what to do,
telling the robots what to do
telling the robots what to do
201
581609
3290
它會告訴人類要做什麼，
告訴機械人要做什麼，
告訴機械人要做什麼，
09:56
and keeping track of thousands
of individual components.
of individual components.
202
584923
2915
並且對成千上萬個部件
進行持續的追蹤。
進行持續的追蹤。
09:59
What's interesting is,
203
587862
1180
有趣的是，
10:01
building this pavilion
was simply not possible
was simply not possible
204
589066
3141
要建造出這樣的亭狀建築物，
10:04
without human, robot and AI
augmenting each other.
augmenting each other.
205
592231
4524
如果沒有人類、機械、人工智慧的
互補強化，根本不可能做得出來。
互補強化，根本不可能做得出來。
10:09
OK, I'll share one more project.
This one's a little bit crazy.
This one's a little bit crazy.
206
597890
3320
好，我再分享一個專案，
這個有點瘋狂。
這個有點瘋狂。
10:13
We're working with Amsterdam-based artist
Joris Laarman and his team at MX3D
Joris Laarman and his team at MX3D
207
601234
4468
我們與阿姆斯特丹的藝術家
尤爾斯‧拉曼和他的 MX3D 團隊，
尤爾斯‧拉曼和他的 MX3D 團隊，
10:17
to generatively design
and robotically print
and robotically print
208
605726
2878
正使用衍生性設計
與機械列印的方式，
與機械列印的方式，
10:20
the world's first autonomously
manufactured bridge.
manufactured bridge.
209
608628
2995
打造世界第一座機械人自造的橋梁。
10:24
So, Joris and an AI are designing
this thing right now, as we speak,
this thing right now, as we speak,
210
612315
3685
所以，就在我們談話的這一刻，
尤爾斯正和人工智慧一起
在阿姆斯特丹設計這座橋梁。
在阿姆斯特丹設計這座橋梁。
10:28
in Amsterdam.
211
616024
1172
10:29
And when they're done,
we're going to hit "Go,"
we're going to hit "Go,"
212
617220
2321
等他們設計完成後，
我們就會按下「啟動」開關，
我們就會按下「啟動」開關，
10:31
and robots will start 3D printing
in stainless steel,
in stainless steel,
213
619565
3311
讓機械人開始
用不鏽鋼 3D 列印出橋梁，
用不鏽鋼 3D 列印出橋梁，
10:34
and then they're going to keep printing,
without human intervention,
without human intervention,
214
622900
3283
在沒有人類的介入幫忙下，
它們會持續地列印
它們會持續地列印
10:38
until the bridge is finished.
215
626207
1558
直到橋樑完工為止。
10:41
So, as computers are going
to augment our ability
to augment our ability
216
629099
2928
所以，電腦將強化
我們的想像及設計新事物的能力，
10:44
to imagine and design new stuff,
217
632051
2150
10:46
robotic systems are going to help us
build and make things
build and make things
218
634225
2895
機械人系統將協助我們製造
10:49
that we've never been able to make before.
219
637144
2084
我們以前無法製造的東西。
10:52
But what about our ability
to sense and control these things?
to sense and control these things?
220
640347
4160
但是我們感知和控制
這些東西的能力呢？
這些東西的能力呢？
10:56
What about a nervous system
for the things that we make?
for the things that we make?
221
644531
4031
我們製成東西的神經系統
又如何呢？
又如何呢？
11:00
Our nervous system,
the human nervous system,
the human nervous system,
222
648586
2512
我們的神經系統，人類的神經系統，
11:03
tells us everything
that's going on around us.
that's going on around us.
223
651122
2311
可以告訴我們周遭發生的每一件事。
11:06
But the nervous system of the things
we make is rudimentary at best.
we make is rudimentary at best.
224
654186
3684
但這些東西的神經系統，
最多只能算「尚未成熟」。
最多只能算「尚未成熟」。
11:09
For instance, a car doesn't tell
the city's public works department
the city's public works department
225
657894
3563
比如說，車輛本身
不會主動通告市政府的工部門，
不會主動通告市政府的工部門，
11:13
that it just hit a pothole at the corner
of Broadway and Morrison.
of Broadway and Morrison.
226
661481
3130
說它在經過百老匯和
莫里森轉角口時撞到水坑。
莫里森轉角口時撞到水坑。
11:16
A building doesn't tell its designers
227
664635
2032
建築物本身不會告知它的設計師，
11:18
whether or not the people inside
like being there,
like being there,
228
666691
2684
裡面的居民是否喜歡住在那裏，
11:21
and the toy manufacturer doesn't know
229
669399
3010
玩具製造商也不知道
他們的玩具
現在是跟誰在玩、在哪玩、
現在是跟誰在玩、在哪玩、
11:24
if a toy is actually being played with --
230
672433
2007
11:26
how and where and whether
or not it's any fun.
or not it's any fun.
231
674464
2539
是不是玩的很開心。
11:29
Look, I'm sure that the designers
imagined this lifestyle for Barbie
imagined this lifestyle for Barbie
232
677620
3814
我確定設計師在設計芭比時，
一定想像過芭比的生活方式。
11:33
when they designed her.
233
681458
1224
11:34
(Laughter)
234
682706
1447
（笑聲）
11:36
But what if it turns out that Barbie's
actually really lonely?
actually really lonely?
235
684177
2906
但要是芭比變的很孤單怎麼辦？
11:39
(Laughter)
236
687107
3147
（笑聲）
11:43
If the designers had known
237
691266
1288
如果設計師知道
11:44
what was really happening
in the real world
in the real world
238
692578
2107
他們設計的東西，
在真實世界裡發生了什麼事，
在真實世界裡發生了什麼事，
11:46
with their designs -- the road,
the building, Barbie --
the building, Barbie --
239
694709
2583
像是道路、建築物、芭比——
11:49
they could've used that knowledge
to create an experience
to create an experience
240
697316
2694
那他們就可以運用所獲得的訊息，
為使用者創造出更好的使用體驗。
11:52
that was better for the user.
241
700034
1400
11:53
What's missing is a nervous system
242
701458
1791
我們欠缺的就是一個
11:55
connecting us to all of the things
that we design, make and use.
that we design, make and use.
243
703273
3709
可以連結所有我們設計、製造、
使用事物的神經系統。
使用事物的神經系統。
11:59
What if all of you had that kind
of information flowing to you
of information flowing to you
244
707915
3555
如果大家在真實世界，能收到
自己創造的東西所回饋的資訊，
自己創造的東西所回饋的資訊，
12:03
from the things you create
in the real world?
in the real world?
245
711494
2183
那會如何呢？
12:07
With all of the stuff we make,
246
715432
1451
所有我們製造的東西，
12:08
we spend a tremendous amount
of money and energy --
of money and energy --
247
716907
2435
我們花了很多錢跟精力 ──
實際上光是去年，
大約就有兩兆美金 ──
大約就有兩兆美金 ──
12:11
in fact, last year,
about two trillion dollars --
about two trillion dollars --
248
719366
2376
去說服人們購買我們製造的東西。
12:13
convincing people to buy
the things we've made.
the things we've made.
249
721766
2854
12:16
But if you had this connection
to the things that you design and create
to the things that you design and create
250
724644
3388
但如果你所設計製造出來的東西
能連結傳送給你回饋的訊息，
能連結傳送給你回饋的訊息，
12:20
after they're out in the real world,
251
728056
1727
不管是在它們上市以後，
12:21
after they've been sold
or launched or whatever,
or launched or whatever,
252
729807
3614
或是在賣出或發表以後，
12:25
we could actually change that,
253
733445
1620
我們就可以改變既有的銷售模式，
12:27
and go from making people want our stuff,
254
735089
3047
從說服人們來購買我們的產品，
12:30
to just making stuff that people
want in the first place.
want in the first place.
255
738160
3434
轉變成我們第一時間就做出
人們真正需要的東西。
人們真正需要的東西。
12:33
The good news is, we're working
on digital nervous systems
on digital nervous systems
256
741618
2787
好消息是，我們正在研發的
這套數位神經系統
這套數位神經系統
12:36
that connect us to the things we design.
257
744429
2801
能連結我們與我們所設計的產品。
12:40
We're working on one project
258
748365
1627
我們與在洛杉磯的
12:42
with a couple of guys down in Los Angeles
called the Bandito Brothers
called the Bandito Brothers
259
750016
3712
邦帝圖兄弟公司和他們的團隊
正合作進行一個專案。
12:45
and their team.
260
753752
1407
12:47
And one of the things these guys do
is build insane cars
is build insane cars
261
755183
3433
這幾個人做的其中一件事
就是製造「瘋狂賽車」，
就是製造「瘋狂賽車」，
12:50
that do absolutely insane things.
262
758640
2873
他們做的東西真的很瘋狂。
12:54
These guys are crazy --
263
762905
1450
這些人真的是瘋了──
12:56
(Laughter)
264
764379
1036
（笑聲）
12:57
in the best way.
265
765439
1403
不過是用最厲害的方式。
13:00
And what we're doing with them
266
768993
1763
我們跟他們一起合作的模式，
13:02
is taking a traditional race-car chassis
267
770780
2440
就是將傳統的賽車底盤
13:05
and giving it a nervous system.
268
773244
1585
安裝神經系統。
13:06
So we instrumented it
with dozens of sensors,
with dozens of sensors,
269
774853
3058
所以，我們在底盤
安裝了好幾組感應器，
安裝了好幾組感應器，
13:09
put a world-class driver behind the wheel,
270
777935
2635
然後請一位世界級車手來駕駛，
13:12
took it out to the desert
and drove the hell out of it for a week.
and drove the hell out of it for a week.
271
780594
3357
把車送到沙漠連續開它個一禮拜。
13:15
And the car's nervous system
captured everything
captured everything
272
783975
2491
之後車子的神經系統
就能紀錄到車子發生的所有反應。
13:18
that was happening to the car.
273
786490
1482
13:19
We captured four billion data points;
274
787996
2621
我們抓到了 40 億個資料點；
13:22
all of the forces
that it was subjected to.
that it was subjected to.
275
790641
2310
所有底盤所承受的壓力數據。
13:24
And then we did something crazy.
276
792975
1659
然後我們做了一些瘋狂的事。
13:27
We took all of that data,
277
795268
1500
我們把所有的資料
13:28
and plugged it into a generative-design AI
we call "Dreamcatcher."
we call "Dreamcatcher."
278
796792
3736
連接到一個叫做「捕夢者」的
衍生設計人工智慧上 。
衍生設計人工智慧上 。
13:33
So what do get when you give
a design tool a nervous system,
a design tool a nervous system,
279
801270
3964
所以當你把神經系統
安裝到設計工具上，
安裝到設計工具上，
13:37
and you ask it to build you
the ultimate car chassis?
the ultimate car chassis?
280
805258
2882
並請它幫你建造一個
終極汽車底盤時，你會得到什麼？
終極汽車底盤時，你會得到什麼？
13:40
You get this.
281
808723
1973
你會得到這個。
13:44
This is something that a human
could never have designed.
could never have designed.
282
812293
3713
這是人類永遠無法設計出的東西。
13:48
Except a human did design this,
283
816707
1888
如果真有人這樣設計過，
13:50
but it was a human that was augmented
by a generative-design AI,
by a generative-design AI,
284
818619
4309
那個人一定也是透過
衍生設計的人工智慧強化、
衍生設計的人工智慧強化、
13:54
a digital nervous system
285
822952
1231
數位神經系統的強化、
13:56
and robots that can actually
fabricate something like this.
fabricate something like this.
286
824207
3005
和機械人一起合作，
才做得出來的東西。
才做得出來的東西。
13:59
So if this is the future,
the Augmented Age,
the Augmented Age,
287
827680
3595
所以，如果擴增時代
就是我們的未來，
就是我們的未來，
14:03
and we're going to be augmented
cognitively, physically and perceptually,
cognitively, physically and perceptually,
288
831299
4261
而我們的認知、體格、知覺
都將被強化、擴增，
都將被強化、擴增，
14:07
what will that look like?
289
835584
1408
那會是怎樣的世界？
14:09
What is this wonderland going to be like?
290
837576
3321
那會是個什麼樣的美麗新世界？
14:12
I think we're going to see a world
291
840921
1709
我認為我們即將見證這麼一個世界，
14:14
where we're moving
from things that are fabricated
from things that are fabricated
292
842654
3068
一個東西從製造出來的變成
14:17
to things that are farmed.
293
845746
1445
「種 」出來的世界。
14:20
Where we're moving from things
that are constructed
that are constructed
294
848159
3453
一個東西從建造出來的
14:23
to that which is grown.
295
851636
1704
變成自己「長 」出來的世界。
14:26
We're going to move from being isolated
296
854134
2188
我們將從自我隔離
14:28
to being connected.
297
856346
1610
轉變成相互交流。
14:30
And we'll move away from extraction
298
858634
2411
我們也將從奪取者
14:33
to embrace aggregation.
299
861069
1873
變成相互擁抱的給予者。
14:35
I also think we'll shift
from craving obedience from our things
from craving obedience from our things
300
863967
3767
我也認為，我們將會從冀望產品
順從我們的指令，
順從我們的指令，
14:39
to valuing autonomy.
301
867758
1641
轉變成重視其自主性。
14:42
Thanks to our augmented capabilities,
302
870510
1905
由於我們的擴增強化能力，
14:44
our world is going to change dramatically.
303
872439
2377
我們的世界將會有劇烈的變化。
14:47
We're going to have a world
with more variety, more connectedness,
with more variety, more connectedness,
304
875576
3246
我們的世界會變得更多元、
更加連通、
更加連通、
14:50
more dynamism, more complexity,
305
878846
2287
更有活力、更多複雜的變化、
14:53
more adaptability and, of course,
306
881157
2318
更有適應力、當然
14:55
more beauty.
307
883499
1217
也會更美麗。
14:57
The shape of things to come
308
885231
1564
未來世界的雛型
14:58
will be unlike anything
we've ever seen before.
we've ever seen before.
309
886819
2290
是我們前所未見的。
15:01
Why?
310
889133
1159
為什麼？
15:02
Because what will be shaping those things
is this new partnership
is this new partnership
311
890316
3755
因為形塑這個世界的
將會是科技、自然與人類的
新結盟關係。
新結盟關係。
15:06
between technology, nature and humanity.
312
894095
3670
15:11
That, to me, is a future
well worth looking forward to.
well worth looking forward to.
313
899279
3804
對我而言，那樣的未來
是值得我們期待的。
是值得我們期待的。
15:15
Thank you all so much.
314
903107
1271
非常感謝各位。
15:16
(Applause)
315
904402
5669
（掌聲）
ABOUT THE SPEAKER
Maurice Conti - Designer, futuristMaurice Conti explores new partnerships between technology, nature and humanity.
Why you should listen
Maurice Conti is a designer, futurist and innovator. He's worked with startups, government agencies, artists and corporations to explore the things that will matter to us in the future, and to design solutions to get us there.
Conti is currently Chief Innovation Officer at Alpha -- Europe's first moonshot factory, powered by Telefónica. Conti and his team are responsible for coming up with the ideas, prototypes and proofs of concepts that will go on to become full-blown moonshots at Alpha: projects that will affect 100 million people or more, be a force for good on the planet and grow into billion-euro businesses.
Previously, Conti was Director of Applied Research & Innovation at Autodesk where built and led Autodesk's Applied Research Lab. Conti's work focuses on applied machine learning, advanced robotics, augmented and virtual realities, and the future of work, cities and mobility.
Conti is also an explorer of geographies and cultures. He has circumnavigated the globe once and been half-way around twice. In 2009 he was awarded the Medal for Exceptional Bravery at Sea by the United Nations, the New Zealand Bravery Medal and a US Coast Guard Citation for Bravery for risking his own life to save three shipwrecked sailors.
Conti lives in Barcelona, Spain, and travels around the world speaking to groups about innovation, technology trends, the future, and high adventure.
Maurice Conti | Speaker | TED.com