AI-Generated Creatures That Stretch the Boundaries of Imagination | Sofia Crespo | TED

44,342 views ・ 2022-11-30

TED


μ•„λž˜ μ˜λ¬Έμžλ§‰μ„ λ”λΈ”ν΄λ¦­ν•˜μ‹œλ©΄ μ˜μƒμ΄ μž¬μƒλ©λ‹ˆλ‹€.

λ²ˆμ—­: SA H κ²€ν† : Hyeryung Kim
00:04
I'd like to start by asking you to imagine a color
0
4543
4379
μš°λ¦¬κ°€ ν•œλ²ˆλ„ 보지 λͺ»ν–ˆλ˜ 색깔을 상상해 λ³΄λ©΄μ„œ
00:08
that you've never seen before.
1
8922
2544
이 강연을 μ‹œμž‘ν•˜κ² μŠ΅λ‹ˆλ‹€.
00:12
Just for a second give this a try.
2
12217
2252
잠깐 μ‹œλ„ν•΄λ³΄μ„Έμš”.
00:14
Can you actually visualize a color that you've never been able to perceive?
3
14928
4463
μ—¬λŸ¬λΆ„μ΄ μ ˆλŒ€ 인식할 수 μ—†μ—ˆλ˜ 색을 μ‹€μ œλ‘œ μ‹œκ°ν™”ν•  수 μžˆμœΌμ‹ κ°€μš”?
00:20
I never seem to get tired of trying this
4
20892
2878
쉽지 μ•Šμ€ λ„μ „μ΄λΌλŠ” 것을 잘 μ•Œμ§€λ§Œ
00:23
although I know it's not an easy challenge.
5
23770
2795
μ €λŠ” 이것을 μ–Έμ œ μ‹œλ„ν•΄λ„ μ§ˆλ¦¬μ§€ μ•ŠμŠ΅λ‹ˆλ‹€.
00:26
And the thing is,
6
26606
1210
그리고 μ€‘μš”ν•œ 건,
00:27
we can't imagine something without drawing upon our experiences.
7
27857
4880
μš°λ¦¬λŠ” 슀슀둜의 κ²½ν—˜μ„ ν™œμš©ν•΄μ•Όλ§Œ μ–΄λ–€ 것을 상상할 수 μžˆλ‹€λŠ” 것이죠.
00:33
A color we haven't yet seen
8
33655
2002
μš°λ¦¬κ°€ 인식 κ°€λŠ₯ν•œ μŠ€νŽ™νŠΈλŸΌ 밖에 μžˆλŠ” 색은
00:35
outside the spectrum we can perceive
9
35699
2836
00:38
is outside our ability to conjure up.
10
38535
2502
μš°λ¦¬κ°€ 생각해 λ‚Ό 수 μžˆλŠ” λŠ₯λ ₯ 밖에 μžˆμŠ΅λ‹ˆλ‹€.
00:42
It's almost like there's a boundary to our imagination
11
42372
3212
마치 우리의 상상λ ₯에 ν•œκ³„κ°€ μžˆλŠ” 것과 κ°™μ£ .
00:45
where all the colors we can imagine
12
45625
2002
μš°λ¦¬κ°€ 상상할 수 μžˆλŠ” λͺ¨λ“  색은
00:47
can only be various shades of other colors we have previously seen.
13
47627
4505
이전에 λ³΄μ•˜λ˜ λ‹€λ₯Έ μƒ‰μ˜ λ‹€μ–‘ν•œ 음영일 μˆ˜λ°–μ— μ—†μœΌλ‹ˆκΉŒμš”.
00:52
Yet we know for a fact
14
52841
2210
ν•˜μ§€λ§Œ μš°λ¦¬λŠ” μ•Œκ³  있죠.
00:55
that those color frequencies outside our visible spectrum are there.
15
55051
4463
μš°λ¦¬κ°€ λ³Ό 수 μžˆλŠ” μŠ€νŽ™νŠΈλŸΌ 밖에도 μƒ‰μ˜ μ£ΌνŒŒμˆ˜λŠ” μ‘΄μž¬ν•œλ‹€λŠ” μ‚¬μ‹€μ„μš”.
01:00
And scientists believe that there are species
16
60223
5339
그리고 κ³Όν•™μžλ“€μ€
인간이 κ°–κ³  μžˆλŠ” μ„Έ 가지 μƒ‰μ˜ κ΄‘μˆ˜μš©μ²΄λ³΄λ‹€
01:05
that have many more photo receptors
17
65562
3503
더 λ§Žμ€ κ΄‘μˆ˜μš©μ²΄λ₯Ό 가지고 μžˆλŠ” 쒅듀도 μ‘΄μž¬ν•œλ‹€κ³  λ―ΏμŠ΅λ‹ˆλ‹€.
01:09
than just the three color ones we humans have.
18
69107
4880
01:13
Which, by the way,
19
73987
1835
ν•œνŽΈ, λͺ¨λ“  인간이 세상을 같은 λ°©μ‹μœΌλ‘œ λ³΄λŠ” 것은 μ•„λ‹™λ‹ˆλ‹€.
01:15
not all humans see the world in the same way.
20
75864
3295
01:19
Some of us are colorblind to various degrees,
21
79200
4505
μš°λ¦¬λ“€ 쀑 μΌλΆ€λŠ” λ‹€μ–‘ν•œ μ •λ„λ‘œ μƒ‰λ§Ήμ΄λΌμ„œ
01:23
and very often we don't even agree on small things,
22
83705
4713
λ•Œλ•Œλ‘œ μ•„μ£Ό μ‚¬μ†Œν•œ 것듀에 λŒ€ν•΄μ„œλ„ 의견이 κ°ˆλ¦½λ‹ˆλ‹€.
01:28
like if a dress on the internet is blue and black or white and gold.
23
88460
5005
μΈν„°λ„·μƒμ˜ λ“œλ ˆμŠ€κ°€ νŒŒλž‘κ³Ό 검정인지 흰색과 κΈˆμƒ‰μΈμ§€ 같은 κ²ƒλ“€μ²˜λŸΌμš”.
01:34
But my favorite creature, one of my favorite creatures,
24
94215
4380
κ·ΈλŸ¬λ‚˜ μ œκ°€ κ°€μž₯ μ’‹μ•„ν•˜λŠ” 생물 쀑 ν•˜λ‚˜μΈ
01:38
is the peacock mantis shrimp,
25
98595
2335
κ΄‘λŒ€μ‚¬λ§ˆκ·€ μƒˆμš°λŠ”
01:40
which is estimated to have 12 to 16 photo receptors.
26
100972
5297
12κ°œμ—μ„œ 16개의 κ΄‘μˆ˜μš©μ²΄λ₯Ό 가지고 μžˆλ‹€κ³  μ•Œλ €μ Έ μžˆμŠ΅λ‹ˆλ‹€.
01:46
And that indicates the world to them might look so much more colorful.
27
106269
5798
κ·Έλ“€μ—κ²ŒλŠ” 이 세상이 훨씬 더 λ‹€μ±„λ‘­κ²Œ 보이겠죠.
01:54
So what about artificial intelligence?
28
114194
2836
κ·Έλ ‡λ‹€λ©΄ 인곡 지λŠ₯은 μ–΄λ–¨κΉŒμš”?
01:57
Can AI help us see beyond our human capabilities?
29
117947
3838
인간이 자기 λŠ₯λ ₯ μ΄μƒμœΌλ‘œ λ³Ό 수 있게 인곡 지λŠ₯이 λ„μšΈ 수 μžˆμ„κΉŒμš”?
02:03
Well, I've been working with AI for the past five years,
30
123203
3670
μ €λŠ” μ§€λ‚œ 5λ…„ λ™μ•ˆ 인곡 지λŠ₯κ³Ό κ΄€λ ¨λœ 일을 ν•΄μ™”κ³ 
02:06
and in my experience, it can see within the data it gets fed.
31
126873
4630
제 κ²½ν—˜μƒ 인곡 지λŠ₯은 제곡된 자료 λ‚΄μ—μ„œλ§Œ λ³Ό 수 μžˆμŠ΅λ‹ˆλ‹€.
02:12
But then you might be wondering, OK,
32
132420
2586
κ·Έλ ‡λ‹€λ©΄ μ—¬κΈ°μ„œ μ—¬λŸ¬λΆ„μ€ κΆκΈˆν•˜μ‹€ κ²λ‹ˆλ‹€.
02:15
if AI can't help imagine anything new,
33
135006
3337
인곡 지λŠ₯이 μƒˆλ‘œμš΄ 것을 μƒμƒν•˜λŠ”λ° 도움이 λ˜μ§€ μ•ŠλŠ”λ‹€λ©΄
02:18
why would an artist see any point in using it?
34
138385
2794
μ™œ μ˜ˆμˆ κ°€λ“€μ€ ꡳ이 인곡 지λŠ₯을 μ΄μš©ν•˜λ €κ³  ν• κΉŒμš”?
02:21
And my answer to that is because I think that it can help augment our creativity
35
141930
4254
μ €λŠ” 인곡 지λŠ₯이 우리의 창의λ ₯을 ν–₯μƒμ‹œν‚¨λ‹€κ³  μƒκ°ν•©λ‹ˆλ‹€.
02:26
as there's value in creating combinations of known elements to form new ones.
36
146226
6173
이미 μ•Œλ €μ§„ μš”μ†Œλ“€μ„ μ‘°ν•©ν•˜μ—¬ μƒˆλ‘œμš΄ 것을 μ°½μ‘°ν•  수 있기 λ•Œλ¬Έμ΄μ£ .
02:33
And this boundary of what we can imagine based on what we have experienced
37
153858
5422
그리고 우리의 κ²½ν—˜μ— κ·Όκ±°ν•΄ 상상할 수 μžˆλŠ” κ²ƒμ˜ 경계가
02:39
is the place that I have been exploring.
38
159322
2419
λ°”λ‘œ μ œκ°€ νƒκ΅¬ν•΄μ˜¨ μž₯μ†Œμž…λ‹ˆλ‹€.
02:42
For me, it started with jellyfish on a screen at an aquarium
39
162200
5005
μ €μ˜ κ²½μš°μ—λŠ”, μˆ˜μ‘±κ΄€μ˜ 슀크린 속 ν•΄νŒŒλ¦¬λ₯Ό
02:47
and wearing those old 3D glasses, which I hope you remember,
40
167247
3712
μ—¬λŸ¬λΆ„μ΄ κΈ°μ–΅ν–ˆμœΌλ©΄ ν•˜λŠ” 였래된 3D μ•ˆκ²½μ„ μ“°κ³  λ³΄λ©΄μ„œ μ‹œμž‘ν–ˆμŠ΅λ‹ˆλ‹€.
02:51
the ones with the blue and red lens.
41
171000
1919
νŒŒλž€μƒ‰κ³Ό 빨간색 λ Œμ¦ˆκ°€ μžˆλŠ” μ•ˆκ²½ λ§μž…λ‹ˆλ‹€.
02:53
And this experience made me want to recreate their textures.
42
173336
4213
이 κ²½ν—˜μ„ 톡해 μ €λŠ” κ·Έ ν•΄νŒŒλ¦¬λ“€μ˜ μ§ˆκ°μ„ μž¬ν˜„ν•˜κ³  μ‹Άμ–΄μ‘ŒμŠ΅λ‹ˆλ‹€.
02:57
But not just that,
43
177882
1168
ν•˜μ§€λ§Œ 그뿐 μ•„λ‹ˆλΌ,
02:59
I also wanted to create new jellyfish
44
179092
1960
μ €λŠ” 전에 보지 λͺ»ν–ˆλ˜ μƒˆλ‘œμš΄ ν•΄νŒŒλ¦¬λ“€μ„ λ§Œλ“€μ–΄λ‚΄κ³  μ‹Άμ—ˆμŠ΅λ‹ˆλ‹€.
03:01
that I hadn't seen before, like these.
45
181052
2878
μ΄κ²ƒμ²˜λŸΌμš”.
03:04
And what started with jellyfish,
46
184556
1543
그리고 ν•΄νŒŒλ¦¬λ‘œ μ‹œμž‘ν•œ 것이
03:06
very quickly escalated to other sea creatures
47
186099
3587
λ‹€λ₯Έ λ°”λ‹€ μƒλ¬Όλ“€λ‘œ 맀우 λΉ λ₯΄κ²Œ ν™•μž₯λ˜μ—ˆμ£ .
03:09
like sea anemone, coral and fish.
48
189686
4254
말미잘, μ‚°ν˜Έ, λ¬Όκ³ κΈ° 같은 κ²ƒλ“€λ‘œμš”.
03:14
And then from there came amphibians, birds and insects.
49
194524
5756
그런 λ‹€μŒ μ–‘μ„œλ₯˜, μ‘°λ₯˜, 그리고 κ³€μΆ©λ“€λ‘œ μ΄μ–΄μ‘ŒμŠ΅λ‹ˆλ‹€.
03:20
And this became a series called β€œNeural Zoo”.
50
200739
2919
그리고 이것은 β€œμ‹ κ²½ 동물원”이라고 λΆˆλ¦¬λŠ” μ‹œλ¦¬μ¦ˆκ°€ λ˜μ—ˆμŠ΅λ‹ˆλ‹€.
03:25
But when you look closely, what do you see?
51
205618
3629
ν•˜μ§€λ§Œ μžμ„Ένžˆ 보면 무엇이 λ³΄μ΄λ‚˜μš”?
03:29
There's no single creature in these images.
52
209289
3962
이 μ‚¬μ§„λ“€μ—λŠ” λ‹¨μΌν•œ 생물체가 ν•˜λ‚˜λ„ μ—†μŠ΅λ‹ˆλ‹€.
03:33
And AI augments my creative process
53
213710
3378
인곡 지λŠ₯은 μ œκ°€ μ§ˆκ°μ„ μΆ”μΆœν•˜κ³  μž¬μ‘°ν•©ν•  수 있게 ν•¨μœΌλ‘œμ¨
03:37
by allowing me to distill and recombine textures.
54
217130
4671
μ €μ˜ 창쑰적인 과정을 κ°•ν™”ν•©λ‹ˆλ‹€.
03:41
And that's something that would otherwise take me months to draw by hand.
55
221801
4588
λ§Œμ•½ 이것을 μ†μœΌλ‘œ κ·Έλ Έλ‹€λ©΄ λͺ‡ 달이 걸렸을 κ²ƒμž…λ‹ˆλ‹€.
03:47
Plus I'm actually terrible at drawing.
56
227015
1877
사싀 μ „ 그림에 λ³„λ‘œ μ†Œμ§ˆλ„ μ—†κ³ μš”.
03:49
So you could say, in a way, what I'm doing
57
229476
3169
κ·Έλž˜μ„œ μ–΄λ–»κ²Œ 보면, μ œκ°€ ν•˜κ³  μžˆλŠ” 것은
03:52
is a contemporary version of something
58
232687
1960
인간이 이미 μ˜€λž«λ™μ•ˆ ν•΄μ™”λ˜ κ²ƒμ˜ ν˜„λŒ€μ μΈ 버전이라고 ν•  수 μžˆμŠ΅λ‹ˆλ‹€.
03:54
that humans have already been doing for a long time,
59
234689
2753
03:57
even before cameras existed.
60
237484
2669
카메라가 μ‘΄μž¬ν•˜κΈ°λ„ μ „λΆ€ν„°μš”.
04:01
In medieval times,
61
241738
1793
쀑세 μ‹œλŒ€μ—
04:03
people went on expeditions,
62
243573
2211
μ‚¬λžŒλ“€μ€ νƒν—˜μ„ 떠났고,
04:05
and when they came back they would share about what they saw
63
245825
3504
λŒμ•„μ™”μ„ λ•Œ 그듀이 λ³Έ 것을 μ‚½ν™”κ°€μ—κ²Œ κ³΅μœ ν–ˆμ„ κ²ƒμž…λ‹ˆλ‹€.
04:09
to an illustrator.
64
249370
1377
04:10
And the illustrator, having never seen what was being described,
65
250789
4129
그리고 삽화가듀은 λ¬˜μ‚¬λœ 생물듀을 λ³Έ 적이 μ—†μœΌλ‹ˆ
04:14
would end up drawing
66
254959
1627
본인이 이전에 λ³΄μ•˜λ˜ 것을 λ°”νƒ•μœΌλ‘œ 그림을 그렸을 것이고,
04:16
based on the creatures that they had previously seen
67
256628
2460
04:19
and in the process creating hybrid animals of some sort.
68
259088
3170
κ·Έ κ³Όμ •μ—μ„œ μ–΄λ–€ μ’…λ₯˜μ˜ μž‘μ’… 동물이 λ§Œλ“€μ–΄μ‘Œκ² μ£ .
04:22
So an explorer might describe a beaver, but having never seen one,
69
262675
4338
κ·Έλž˜μ„œ νƒν—˜κ°€κ°€ 비버λ₯Ό λ¬˜μ‚¬ν–ˆλ‹€λ©΄
비버λ₯Ό λ³Έ 적이 μ—†λŠ” μ‚½ν™”κ°€λŠ” λΉ„λ²„μ—κ²Œ μ„€μΉ˜λ₯˜μ˜ 머리와 개의 λͺΈ,
04:27
the illustrator might give it the head of a rodent,
70
267055
2419
04:29
the body of a dog and a fish-like tail.
71
269474
2419
λ¬Όκ³ κΈ° 같은 꼬리λ₯Ό μ£Όμ—ˆμ„ κ²ƒμž…λ‹ˆλ‹€.
04:32
In the series β€œArtificial Natural History”,
72
272560
2670
β€œμΈκ³΅ μžμ—° μ—­μ‚¬β€œμ‹œλ¦¬μ¦ˆμ—μ„œ,
04:35
I took thousands of illustrations from a natural history archives,
73
275271
4588
μ €λŠ” μžμ—°μ‚¬ 기둝 λ³΄κ΄€μ†Œ μ•ˆμ˜ μ‚½ν™”λ₯Ό 수천 개 가지고 와
04:39
and I fed them to a neural network to generate new versions of them.
74
279901
5297
μƒˆλ‘œμš΄ 버전을 μƒμ„±ν•˜κΈ° μœ„ν•΄ 신경망에 그것듀을 κ³΅κΈ‰ν–ˆμŠ΅λ‹ˆλ‹€.
04:45
But up until now, all my work was done in 2D.
75
285990
4171
κ·ΈλŸ¬λ‚˜ μ•„μ§κΉŒμ§€ 제 λͺ¨λ“  μž‘μ—…μ€ 2D둜 μ΄λ£¨μ–΄μ‘ŒμŠ΅λ‹ˆλ‹€.
04:51
And with the help of my studio partner, Feileacan McCormick,
76
291246
3545
그리고 제 μž‘μ—…μ‹€ λ™λ£ŒμΈ ν•„λ ˆμ΄μΉΈ λ§₯μ½”λ―Ήμ˜ λ„μ›€μœΌλ‘œ,
04:54
we decided to train a neural network on a data set of 3D scanned beetles.
77
294833
5422
3D μŠ€μΊ”λœ λ”±μ •λ²Œλ ˆ μžλ£Œμ— λŒ€ν•œ 신경망을 ν›ˆλ ¨ν•˜κΈ°λ‘œ κ²°μ •ν–ˆμŠ΅λ‹ˆλ‹€.
05:00
But I must warn you that our first results were extremely blurry,
78
300839
4546
ν•˜μ§€λ§Œ μ €ν¬μ˜ 첫 번째 κ²°κ³ΌλŠ” 맀우 νλ¦Ών–ˆκ³ ,
05:05
and they looked like the blobs you see here.
79
305426
2253
μ΄λ ‡κ²Œ μ–Όλ£©μ²˜λŸΌ λ³΄μ˜€μŠ΅λ‹ˆλ‹€.
05:08
And this could be due to many reasons,
80
308346
1835
μ΄λ ‡κ²Œ 된 μ΄μœ λŠ” μ—¬λŸ¬ 가지가 μžˆκ² μ§€λ§Œ
05:10
but one of them being that there aren't really a lot
81
310181
2461
그쀑 ν•˜λ‚˜λŠ” 곡개적으둜 이용 κ°€λŠ₯ν•œ
05:12
of openly available data sets of 3D insects.
82
312642
4504
3D 곀좩의 데이터 μ„ΈνŠΈκ°€ λ§Žμ§€ μ•Šλ‹€λŠ” κ²ƒμž…λ‹ˆλ‹€.
05:17
And also we were repurposing
83
317188
2127
λ˜ν•œ μ €ν¬λŠ” 3Dλ₯Ό μƒμ„±ν•˜κΈ° μœ„ν•΄
05:19
a neural network that normally gets used to generate images to generate 3D.
84
319357
5214
이미지λ₯Ό μƒμ„±ν•˜λŠ”λ° 일반적으둜 μ‚¬μš©λ˜λŠ” 신경망을 μš©λ„μ— 맞게 κ³ μ³€μŠ΅λ‹ˆλ‹€.
05:24
So believe it or not, these are very exciting blobs to us.
85
324612
4380
λ―Ώκ±°λ‚˜ λ§κ±°λ‚˜, 이것은 μ €ν¬μ—κ²Œ 맀우 ν₯미둜운 λΆ€λΆ„μ΄μ—ˆμŠ΅λ‹ˆλ‹€.
05:29
But with time and some very hacky solutions
86
329784
3879
ν•˜μ§€λ§Œ μ‹œκ°„μ΄ μ§€λ‚˜λ©΄μ„œ 맀우 μ§„λΆ€ν•œ ν•΄κ²°μ±…μ˜ ν•˜λ‚˜λ‘œ
05:33
like data augmentation,
87
333705
2294
데이터 μ„ΈνŠΈλ₯Ό ν–₯μƒμ‹œν‚€κΈ° μœ„ν•΄
05:36
where we threw in ants and other beetle-like insects
88
336040
3754
κ°œλ―Έμ™€ λ”±μ •λ²Œλ ˆ 같은 곀좩듀을 집어넣어 데이터 μ„ΈνŠΈλ₯Ό κ°•ν™”ν–ˆκ³ 
05:39
to enhance the data set,
89
339794
2503
05:42
we ended up getting this,
90
342338
2628
λ§ˆμΉ¨λ‚΄ 이것을 얻을 수 μžˆμ—ˆμŠ΅λ‹ˆλ‹€.
05:44
which we've been told they look like grilled chicken.
91
344966
2544
κΌ­ ꡬ운 ν†΅λ‹­μ²˜λŸΌ 보인닀고듀 ν–ˆμ£ .
05:47
(Laughter)
92
347552
1752
(μ›ƒμŒ)
05:49
But hungry for more, we pushed our technique,
93
349345
4797
ν•˜μ§€λ§Œ 더 λ§Žμ€ 것에 κ΅Άμ£Όλ € μ €ν¬λŠ” μ €ν¬μ˜ κΈ°μˆ μ„ λ°€μ–΄λΆ™μ˜€κ³ ,
05:54
and eventually they ended up looking like this.
94
354183
3713
κ²°κ΅­ λ§ˆμ§€λ§‰μ—”
μ΄λ ‡κ²Œ κ°œμ„ λ˜μ—ˆμŠ΅λ‹ˆλ‹€.
05:58
We use something called 3D style transfer to map textures onto them,
95
358479
5047
μ €ν¬λŠ” 3D μŠ€νƒ€μΌ 전솑을 μ‚¬μš©ν•΄μ„œ 이 μœ„μ— μ§ˆκ°μ„ λ§€ν•‘ν•˜κ³ ,
06:03
and we also trained a natural language model
96
363568
3670
μžμ—°μ–΄ λͺ¨λΈμ„ ν›ˆλ ¨ν•˜μ—¬
06:07
to generate scientific-like names
97
367280
2544
과학적인 이름과 해뢀학적 μ„€λͺ…을 μƒμ„±ν–ˆμŠ΅λ‹ˆλ‹€.
06:09
and anatomical descriptions.
98
369866
1918
06:12
And eventually we even found a network architecture that could handle 3D meshes.
99
372869
4963
심지어 3D 그물망을 λ‹€λ£° 수 μžˆλŠ” 신경망 κ΅¬μ‘°κΉŒμ§€ μ°Ύμ•˜μŠ΅λ‹ˆλ‹€.
06:17
So they ended up looking like this.
100
377874
2377
κ²°κ΅­ 그듀은 이런 λͺ¨μŠ΅μ΄ λ˜μ—ˆμ£ .
06:21
And for us, this became a way of creating kind of a speculative study --
101
381836
4838
μ €ν¬μ—κ²Œ 이것은 μΌμ’…μ˜ μΆ”μΈ‘μ„± 연ꡬλ₯Ό λ§Œλ“œλŠ” 방법이 λ˜μ—ˆμŠ΅λ‹ˆλ‹€.
06:26
(Applause)
102
386716
2669
κ°μ‚¬ν•©λ‹ˆλ‹€. (λ°•μˆ˜)
06:29
A speculative study of creatures that never existed,
103
389427
3629
μ‘΄μž¬ν•˜μ§€ μ•ŠλŠ” 생물에 λŒ€ν•œ μΆ”μΈ‘ μ—°κ΅¬λ‘œ,
06:33
kind of like a speculative biology.
104
393097
2294
μΌμ’…μ˜ μΆ”μΈ‘ 생물학 같은 것이죠.
06:37
But I didn't want to talk about AI and its potential
105
397685
4421
ν•˜μ§€λ§Œ 인곡 지λŠ₯을 톡해 μ œκ°€ μ‹€μ œ 쒅에 더 κ°€κΉŒμ΄ μ ‘κ·Όν•  수 μ—†λ‹€λ©΄
06:42
unless it brought me closer to a real species.
106
402148
3629
인곡 지λŠ₯κ³Ό κ·Έκ²ƒμ˜ 잠재λ ₯에 λŒ€ν•΄ λ§ν•˜κ³  싢지 μ•ŠμŠ΅λ‹ˆλ‹€.
06:46
Which of these do you think is easier to find data about online?
107
406486
4880
이 λ‘˜ 쀑 μ–΄λ–€ 것이 μ˜¨λΌμΈμ—μ„œ μ°ΎκΈ°κ°€ 더 μ‰¬μšΈκΉŒμš”?
06:51
(Laughter)
108
411366
1918
(μ›ƒμŒ)
06:53
Yeah, well, as you guessed correctly, the red panda.
109
413326
4004
λ„€, μ—¬λŸ¬λΆ„μ΄ μ˜ˆμƒν•˜μ‹ λŒ€λ‘œ
뢉은 νŒ¬λ”μž…λ‹ˆλ‹€.
06:57
And this maybe could be due to many reasons,
110
417830
3420
κ·Έ μ΄μœ λŠ” μ—¬λŸ¬ 가지가 μžˆκ² μ§€λ§Œ
07:01
but one of them being how cute they are,
111
421292
4046
κ·Έ 쀑 ν•˜λ‚˜λŠ” 그듀이 μ–Όλ§ˆλ‚˜ κ·€μ—¬μš΄κ°€μž…λ‹ˆλ‹€.
07:05
which means we photograph and talk about them a lot,
112
425338
4588
μš°λ¦¬λŠ” 뢉은 νŒ¬λ” 사진을 찍고 그것에 λŒ€ν•œ 이야기도 많이 ν•˜μ§€λ§Œ
07:09
unlike the boreal felt lichen.
113
429926
2002
뢁방 펠트 이끼에 λŒ€ν•΄μ„  κ·Έλ ‡κ²Œ ν•˜μ§€ μ•Šμ£ .
07:12
But both of them are classified as endangered.
114
432679
2877
ν•˜μ§€λ§Œ λ‘˜ λ‹€ λ©Έμ’… μœ„κΈ°μ— μ²˜ν•œ μƒλ¬Όλ‘œ λΆ„λ₯˜λ©λ‹ˆλ‹€.
07:16
So I wanted to bring visibility to other endangered species
115
436349
4880
κ·Έλž˜μ„œ μ €λŠ” λ‹€λ₯Έ λ©Έμ’… μœ„κΈ° 쒅듀도 잘 λ“œλŸ¬λ‚˜λ„λ‘ κ°€μ‹œμ„±μ„ μ£Όκ³  μ‹Άμ—ˆμ£ .
07:21
that don't get the same amount of digital representation
116
441270
4755
κ·€μ—½κ³  λ³΄μ†‘λ³΄μ†‘ν•œ 뢉은 νŒ¬λ”μ™€ 같은
07:26
as a cute, fluffy red panda.
117
446025
2002
디지털 ν‘œν˜„μ„ 보여주고 μ‹Άμ—ˆμ–΄μš”.
07:28
And to do this,
118
448861
1710
그리고 이λ₯Ό μœ„ν•΄,
07:30
we trained an AI on millions of images of the natural world,
119
450613
5005
μ €ν¬λŠ” μžμ—°κ³„μ˜ 수백만 이미지에 λŒ€ν•΄ 인곡 지λŠ₯을 ν›ˆλ ¨μ‹œν‚€κ³  λ‚˜μ„œ
07:35
and then we prompted with text
120
455660
2085
μ΄λŸ¬ν•œ 생물듀 쀑 일뢀λ₯Ό μƒμ„±ν•˜λ„λ‘ 문자λ₯Ό μž…λ ₯ν–ˆμŠ΅λ‹ˆλ‹€.
07:37
to generate some of these creatures.
121
457787
2294
07:40
So when prompted with a text,
122
460123
3712
κ·Έλž˜μ„œ λ‹€μŒκ³Ό 같이
07:43
"an image of a critically endangered spider, the peacock tarantula"
123
463876
4713
β€œμœ„νƒœλ‘­κ²Œ λ©Έμ’… μœ„κΈ°μ— μ²˜ν•œ κ±°λ―Έ, κ³΅μž‘ νƒ€λž€νˆ΄λΌμ˜ 사진”
07:48
and its scientific name,
124
468589
1669
μ΄λ ‡κ²Œ μ“°κ³  κ·Έ ν•™λͺ…을 μž…λ ₯ν•˜μž
07:50
our model generated this.
125
470299
2378
우리 λͺ¨λΈμ€ 이것을 μƒμ„±ν–ˆμŠ΅λ‹ˆλ‹€.
07:55
And here's an image of the real peacock tarantula,
126
475972
3211
그리고 이것은 κ³΅μž‘ νƒ€λž€νˆ΄λΌμ˜ μ‹€μ œ μ‚¬μ§„μž…λ‹ˆλ‹€.
07:59
which is a wonderful spider endemic to India.
127
479225
3128
μ•„μ£Ό 멋진 인도 ν† μ’… κ±°λ―Έμ£ .
08:02
But when prompted with a text
128
482812
2878
ν•˜μ§€λ§Œ
β€œμœ„νƒœλ‘­κ²Œ λ©Έμ’… μœ„κΈ°μ— μ²˜ν•œ μƒˆ, 맹그둜브 ν•€μΉ˜μ˜ 사진”을 μž…λ ₯ν–ˆμ„ λ•Œ
08:05
"an image of a critically endangered bird, the mangrove finch,"
129
485732
4170
08:09
our model generated this.
130
489944
2669
우리 λͺ¨λΈμ€ 이것을 μƒμ„±ν–ˆμŠ΅λ‹ˆλ‹€.
08:14
And here's a photo of the real mangrove finch.
131
494532
2711
이것은 맹그둜브 ν•€μΉ˜μ˜ μ‹€μ œ μ‚¬μ§„μ΄κ³ μš”.
08:17
Both these creatures exist in the wild,
132
497702
2878
이 두 생물 λͺ¨λ‘ μ‹€μ œλ‘œ 야생에 μ‘΄μž¬ν•˜μ§€λ§Œ,
08:20
but the accuracy of each generated image is fully dependent on the data available.
133
500621
5965
μƒμ„±λœ 각 μ΄λ―Έμ§€μ˜ 정확성은 μ‚¬μš©κ°€λŠ₯ν•œ 데이터에 μ „μ μœΌλ‘œ μ˜μ‘΄ν•˜μ£ .
08:27
These chimeras of our everyday data
134
507462
2961
일상적인 λ°μ΄ν„°μ˜ μ΄λŸ¬ν•œ ν‚€λ©”λΌλŠ”
08:30
to me are a different way of how the future could be.
135
510423
3670
μ €μ—κ²Œ μžˆμ–΄ 미래λ₯Ό μ•Œ 수 μžˆλŠ” 색닀λ₯Έ λ°©λ²•μž…λ‹ˆλ‹€.
08:34
Not in a literal sense, perhaps,
136
514802
2836
문자 κ·ΈλŒ€λ‘œκ°€ μ•„λ‹ˆλΌ μ•„λ§ˆλ„ μ§„μ •ν•œ μ˜λ―Έμ—μ„œ
08:37
but in the sense that through practicing the expanding of our own imagination
137
517680
6423
μš°λ¦¬κ°€ μ†ν•œ μƒνƒœκ³„μ— λŒ€ν•œ 상상λ ₯을 ν™•μž₯ν•˜λŠ” μ—°μŠ΅μ„ ν•œλ‹€λ©΄
08:44
about the ecosystems we are a part of,
138
524145
3045
08:47
we might just be better equipped to recognize new opportunities
139
527231
3170
μƒˆλ‘œμš΄ κΈ°νšŒμ™€ κ°€λŠ₯성을 더 λΉ λ₯΄κ²Œ μ•Œμ•„λ‚Ό 수 μžˆμ„ κ²ƒμž…λ‹ˆλ‹€.
08:50
and potential.
140
530443
1335
08:52
Knowing that there's a boundary to our imagination
141
532236
3128
우리 상상λ ₯의 경계λ₯Ό μ•ˆλ‹€κ³  ν•΄μ„œ
08:55
doesn't have to feel limiting.
142
535406
2211
상상λ ₯이 μ œν•œλ˜λŠ” 것은 μ•„λ‹™λ‹ˆλ‹€.
08:58
On the contrary,
143
538159
1293
였히렀 그것을 μ•Œλ©΄
08:59
it can help motivate us to expand that boundary further
144
539452
3420
μš°λ¦¬κ°€ κ·Έ 경계λ₯Ό 더 ν™•μž₯ν•˜κ³ 
09:02
and to seek out colors and things we haven't yet seen
145
542914
3962
μ§€κΈˆκ» 보지 λͺ»ν•œ 색깔과 사물을 μ°ΎλŠ” 동기가 생길 것이고
09:06
and perhaps enrich our imagination as a result.
146
546918
3295
결과적으둜 우리의 상상λ ₯은 λ”μš± ν’λΆ€ν•΄μ§ˆ κ²ƒμž…λ‹ˆλ‹€.
09:10
So thank you.
147
550546
1168
κ°μ‚¬ν•©λ‹ˆλ‹€.
09:11
(Applause)
148
551714
4046
(λ°•μˆ˜)
이 μ›Ήμ‚¬μ΄νŠΈ 정보

이 μ‚¬μ΄νŠΈλŠ” μ˜μ–΄ ν•™μŠ΅μ— μœ μš©ν•œ YouTube λ™μ˜μƒμ„ μ†Œκ°œν•©λ‹ˆλ‹€. μ „ 세계 졜고의 μ„ μƒλ‹˜λ“€μ΄ κ°€λ₯΄μΉ˜λŠ” μ˜μ–΄ μˆ˜μ—…μ„ 보게 될 κ²ƒμž…λ‹ˆλ‹€. 각 λ™μ˜μƒ νŽ˜μ΄μ§€μ— ν‘œμ‹œλ˜λŠ” μ˜μ–΄ μžλ§‰μ„ 더블 ν΄λ¦­ν•˜λ©΄ κ·Έκ³³μ—μ„œ λ™μ˜μƒμ΄ μž¬μƒλ©λ‹ˆλ‹€. λΉ„λ””μ˜€ μž¬μƒμ— 맞좰 μžλ§‰μ΄ μŠ€ν¬λ‘€λ©λ‹ˆλ‹€. μ˜κ²¬μ΄λ‚˜ μš”μ²­μ΄ μžˆλŠ” 경우 이 문의 양식을 μ‚¬μš©ν•˜μ—¬ λ¬Έμ˜ν•˜μ‹­μ‹œμ˜€.

https://forms.gle/WvT1wiN1qDtmnspy7