Pratique du Shadowing: Why AI art struggles with hands - Apprendre l'anglais à l'oral avec la vidéo
Création de la leçon...
1
You're called to create.
2
A post -apocalyptic giraffe astronaut?
3
Generated.
4
Genghis Khan playing a guitar solo.
5
Pixel art?
6
Generated.
7
A man holding a delicious apple?
8
Ah, what's with his hands?
9
Why can't AI art make hands?
10
It doesn't matter what AI art model you use, if you have a man holding a delicious apple, his hands will look weird holding it.
11
Why is this so hard?
12
Seems easy enough, right?
13
We've got this weird situation where AI art can instantly make.
14
Abraham Lincoln dressed like glam David Bowie, but struggles with a woman holding a cell phone.
15
This isn't just a weird glitch.
16
The struggle of AI art with hands can actually teach you something bigger about how AI art works.
17
I mean, what is so hard about this?
18
I asked an artist who has taught thousands of people how to draw hands from imagination.
19
Before someone becomes or starts training to be an artist, like officially training, it's pattern recognition.
20
You just grow up seeing a whole bunch of hands and you start knowing what hands look like.
21
You learn how things look by living in the world and recognizing patterns.
22
An AI is similar but has key differences.
23
Imagine an AI is like you but trapped in a museum from birth.
24
All the machine has to learn from are the pictures and the little placards on the side.
25
Apple.
26
A red apple on a brown table.
27
That's like the images it sees from the web and the descriptions that go with them.
28
It's similar to how you learn but locked in that museum.
29
If you want to understand an apple, you can rotate it in your hand.
30
You can watch it whenever you want.
31
If AI wants to understand an apple, it has to find another picture of an apple in the museum.
32
Pattern recognition has allowed AI and people to draw decent apples.
33
But the processes differ.
34
You start training to become an artist and now you're like, okay now I have to learn the rules and that's where it becomes very different from how AI is learning.
35
Artists in order to draw something complicated we tend to simplify things into basic forms and so
36
when you look at a hand you pretty much have the
37
big blocky part of the palm right you have the front you have the back
38
and then you have the thickness and you can pretty much just make
39
that into like a square with some thickness to it.
40
Then an artist can add all the style and texture and detail they want.
41
AI works differently.
42
Look at this hand.
43
The shapes are bizarre, but the AI has done a great job showing the light and texture here.
44
Remember, the AI knows how things look, but not how they work.
45
So these patterns in pixels are easy to understand.
46
It never learned, however, that fingers don't really bend like this.
47
It doesn't simplify to forms.
48
Remember, it's trapped in the museum.
49
So it is just trying to guess where hand -like pixels should be without knowing how hands work, like we do.
50
But listen, I find this kind of dissatisfying.
51
I mean, I'm basically just saying that AI can't draw hands because it's not a person.
52
But AI also doesn't know anything about construction, and it can still make a beautiful skyscraper in New York City.
53
So to understand this better, I spoke to two people who have worked with generative art models.
54
Ilun Du is a grad student whose heart is in robotics, but, you know, AI art is like a big deal now, so he got pulled into it.
55
Because of how popular these models have been in generative art, I've also been like reading a bit on that.
56
And I talked to Roy Shilkrat, who has a super varied resume, but has been teaching about generative art since 2018.
57
Good students that come in that are trying to break those models, take them to the next level.
58
Talking to them helped me figure out three big reasons.
59
Not every reason, but three big reasons that hands are tough for AI art models.
60
The data size and quality, the way hands act, and the low margin for error.
61
For the data size, let's go back to the museum idea.
62
The museum the robot hangs out in, it has a ton of rooms dedicated to faces, but not so many rooms for hands.
63
That means it has less to learn from.
64
Just as an example, available datasets like Flickr HQ has 70 ,000 faces.
65
70 ,000!
66
And this popular one annotates 200 ,000 pics of celebrity faces for lots of details, like eyeglasses or pointy noses.
67
There are There are a ton of great hand datasets that can really understand hands, like this one with 11 ,000 hands,
68
but these may not have been used to train the AI that makes art.
69
That data scarcity combines with the quality and complexity of the data.
70
Hand data in the art museum isn't yet annotated to show how they work, like the celebrity's pointy noses.
71
What they say is there's an image and there is a person in the image and the person is holding an umbrella.
72
You don't give the machine a lot of clues saying, this is a person holding the umbrella.
73
The thumb is going from one side of the handle and the fingers are curled.
74
And then the thumb is covering the index finger, but not the other ones.
75
All that is made worse because hands do lots of things compared to, say, faces.
76
So there's a pretty common portrait photo face.
77
There are a lot of these photos online.
78
And I think that everything's very well -centered, right?
79
Eyes are always around here.
80
There's always this order.
81
That's not true of hands which can do this, and this, and this.
82
I swear I'm sober right now.
83
Stan mentioned this too.
84
How many fingers do you see right now?
85
Like, two or three?
86
Like, it doesn't know there's five, because sometimes there's two, sometimes there's three, sometimes four, sometimes five.
87
You can see these problems with AI hands, but the jankiness is all over AI art.
88
Just look at horses.
89
You can often have, like, three legs, five legs, six legs.
90
The model does not learn to explain this because there's too much diversity, and it doesn't have as much bias as we do.
91
Okay, did you hear that last part he said?
92
Good, because it's really important.
93
It doesn't have as much bias as we do.
94
We care a lot about hands and need them to be perfect.
95
There is a low margin for error.
96
But because the model doesn't understand hands, hasn't seen many, and because hands act weird,
97
it makes pictures that are like hands it's seen in the museum, but not an exact hand.
98
That's good enough for a ton of stuff, but not hands.
99
Here, let me give you some examples.
100
Come over here.
101
So I typed, make me a person with exactly five freckles.
102
So this one's from Dolly 2, this one is from Stable Diffusion, and this one is from Mid -Journey.
103
So it's like, you know, great job.
104
You've got, you know, a red -haired person, they're more likely to have freckles, but there are not exactly five freckles here.
105
Here that doesn't really matter, because we see a freckly face, but hands require higher standards.
106
Look at our apple holding man again.
107
I made three other variations.
108
The hands are all weird, but don't look at them right now.
109
It changed the shirt stripes, the buttons, the apple style.
110
None of that matters because it's stripe -like and button -like and apple -like, but hand -like isn't good enough.
111
I came away from this thinking a couple of things.
112
AI art is basically bad at art.
113
We're just able to see it with hands.
114
And B, it's never going to get any better.
115
But both of those things are a bit wrong.
116
I will say
117
that the newest AI art generator to come out at the time of this video is Mid -Journey version 5.
118
And they made some progress with hands, for sure.
119
But it's not totally fixed yet.
120
Don't tell the AI to hold an umbrella.
121
I think they're, like, spending lots of time on, like, some things that you appreciate, which is why you like the images, and a lot of stuff that you don't actually even notice.
122
I think, like, for a lot of natural scenery or something like that, I feel like the model might be getting at that people.
123
And they are working on two things.
124
First, they have the AI look at a ton more pictures, which requires more computing power.
125
They're trying to solve that on a big scale, because if you want to train on more than a handful of images
126
if you want to train
127
or more than 100 images this would take tremendous resources from
128
you to retrain the model itself the other solution might be
129
to invite more people into the museum there's an interesting analog
130
so like have you heard of like chat gbt the big difference was
131
that it basically used human feedback so like they generated many many sentences
132
and asked people to rate which ones are good and
133
which ones are not good they basically fine -tune the model
134
so that it would generate sentences
135
that are convincing to people I guess it would require a lot of engineering to get people to label
136
so much data but i think
137
if we could just get like people to rank how good
138
the images are generated by these models then like a lot of these issues will go away actually
139
because they're just training the models to do what people like it's not just the hand teeth
140
and abs anything where there's like a pattern a large amount
141
of something it doesn't know the rule of there are this many because it's trained on different amounts
📺 Même chaîne
B21
How bite marks made one man a murder suspect | Part 1
IELTS PrepTOEFL PrepUniversity Students
B22
How Vladimir Putin won Republicans' approval
IELTS PrepTOEFL PrepUniversity Students
B21
We’re already using AI more than we realize
Business EnglishDaily ConversationUniversity Students
B28
How to solve problems like a designer
Business EnglishUniversity StudentsWorking Adults
B2Why Latino voters have turned on Trump | America, Actually
Business EnglishDaily ConversationUniversity Students
B26
Five things you need to know for the 2026 midterms | America, Actually
Business EnglishDaily ConversationUniversity Students
B24
Why are weddings so damn expensive?
Business EnglishDaily ConversationUniversity Students
B21
Are headphones destroying our hearing?
Daily ConversationUniversity StudentsWorking Adults
C11
How to be less neurotic | The Gray Area
Daily ConversationUniversity StudentsWorking Adults
B22
The origin of the '80s aesthetic
Daily ConversationUniversity StudentsWorking Adults
B217
Why “Buy one, get one free” isn’t a great deal
Business EnglishWorking Adults
B22
How F1 drivers survive their own cars
Business EnglishUniversity StudentsWorking Adults
B210
Why 350°F is the magic number for baking
Daily ConversationUniversity StudentsWorking Adults
B26
How did pink become a girly color?
Daily ConversationUniversity StudentsWorking Adults
B22
The real reason Boeing's new plane crashed twice
Business EnglishDaily ConversationUniversity Students
B21
Food waste is the world's dumbest problem
Business EnglishDaily ConversationUniversity Students
B22
How the rich avoid paying taxes
Business EnglishDaily ConversationUniversity Students
B2What's safe to eat now?
Business EnglishDaily ConversationUniversity Students
B25
Zohran Mamdani on his first priorities as NYC mayor (Full video interview)
Business EnglishDaily ConversationUniversity Students
B210
Why we say “OK”
University StudentsWorking Adults
✨ Vidéo recommandée
B21
104 phrases
We Need To Stay Human
Daily ConversationUniversity StudentsWorking Adults
B21
138 phrases
Starting art "late" is actually a HUGE advantage | YTartschool
Business EnglishDaily ConversationUniversity Students
B22
32 phrases
The world's most frustrating work of art
British AccentDaily ConversationUniversity StudentsWorking Adults
B2154 phrases
Are We Getting Too Comfortable With Claude?
Business EnglishDaily ConversationUniversity Students
B21
298 phrases
Godfather of AI WARNS: "You Have No Idea What's Coming"
Business EnglishDaily ConversationUniversity Students
B21
421 phrases
ChatGPT Religion: AI is Being WORSHIPPED now
Daily ConversationUniversity StudentsWorking Adults
B21
195 phrases
What Happens to Language When AI Can Speak for Us?
British AccentDaily ConversationUniversity StudentsWorking Adults
B22
32 phrases
AI challenges human identity | Vietnam Today
Daily ConversationUniversity StudentsWorking Adults
C11
254 phrases
AI Will Create New Wealth, But Not Where You Think | Carnegie Mellon University, Po-Shen Loh
Business EnglishDaily ConversationUniversity Students
B11
6 phrases
am i an artist?
Business EnglishDaily ConversationUniversity Students
B27
32 phrases
Why We Should Draw More (and Photograph Less)
British AccentDaily ConversationUniversity StudentsWorking Adults
Vocabulaire et conseils d’expression pour cette leçon
Cette vidéo contient 140 phrases et 1669 mots à répéter en shadowing. La partie parlée dure 9:47. Le locuteur parle à un rythme naturel d’environ 171 mots par minute, proche d’une conversation courante. 85 % des mots font partie des 3 000 mots les plus courants en anglais ; le reste mérite d’être vu avant de commencer.
Vocabulaire clé de cette vidéo
15 mots de la vidéo à retenir, avec leur prononciation et leur sens :
| Mot | Prononciation | Sens |
|---|---|---|
| generate verbe | /ˈd͡ʒɛn.ə.ɹeɪt/ | générer |
| finger nom | /ˈfɪŋɡɚ/ | doigt |
| umbrella nom | /ʌmˈbɹɛl.ə/ | parapluie, parasol |
| pixel nom | /ˈpɪk.səl/ | pixel |
| recognition nom | /ˌɹɛkəɡˈnɪʃ(ə)n/ | reconnaissance |
| delicious adjectif | /dɪˈlɪ.ʃəs/ | délicieux, savoureux |
| trap nom | /tɹæp/ | piège, collet |
| margin nom | /ˈmɑɹ.d͡ʒɪn/ | marge |
| bias nom | /ˈbaɪ.əs/ | préjugé, parti-pris |
| thumb nom | /ˈθʌm/ | pouce |
| texture nom | /ˈtɛkst͡ʃə(ɹ)/ | texture |
| thickness nom | /ˈθɪknəs/ | épaisseur, grosseur |
| stripe nom | /stɹaɪp/ | rayure |
| guitar nom | /ɡɪˈtɑɹ/ | guitare |
| complicated adjectif | /ˈkɑm.plɪˌkeɪ.tɪd/ | compliqué, alambiqué |
Les verbes à particule que vous entendrez
| Mot | Sens |
|---|---|
| come over verbe | venir |
| figure out verbe | cerner |
| go away verbe | partir, s'en aller |
| grow up verbe | grandir |
La grammaire de cette vidéo
Les structures que le locuteur utilise le plus, avec les mots exacts de la vidéo :
| Structure | Dans la vidéo |
|---|---|
| Present perfect have/has + participe passé — une action passée qui compte encore maintenant | has taught · has allowed · has done |
| Voix passive be + participe passé — l’accent est mis sur ce qui arrive, pas sur qui le fait | are curled · is made · are generated |
| Propositions relatives who / which + proposition — une précision sur une personne ou une chose | people who have · hands which can · appreciate, which is |
Prononciation à surveiller
Le locuteur utilise 32 contractions et formes réduites, comme doesn't, don't, they're. Prononcez-les sous leur forme courte, telles que vous les entendez.
- Les sons « th »: thumb /ˈθʌm/, thickness /ˈθɪknəs/
- Les sons « sh » et « zh »: recognition /ˌɹɛkəɡˈnɪʃ(ə)n/, delicious /dɪˈlɪ.ʃəs/, imagination /ɪˌmæd͡ʒəˈneɪʃən/, variation /ˌvɛəɹiˈeɪʃn̩/
- Mots longs — placez bien l’accent: recognition /ˌɹɛkəɡˈnɪʃ(ə)n/, complicated /ˈkɑm.plɪˌkeɪ.tɪd/, diversity /daɪˈvɜː(ɹ)sɪti/, celebrity /səˈlɛb.ɹɪ.ɾi/, differently /ˈdɪf.ə.ɹənt.li/
Comment s’entraîner avec cette vidéo
- Écoutez la vidéo en entier une fois sans parler et notez les mots que vous ne connaissez pas.
- Commencez à la vitesse 0,75×, répétez phrase par phrase, puis revenez à la vitesse normale quand cela devient facile.
- Enregistrez-vous et comparez avec l’original, en faisant attention à des mots comme generate, finger, umbrella.
Qu'est-ce que la technique du Shadowing ?
Le Shadowing est une technique d'apprentissage des langues fondée sur la science, développée à l'origine pour la formation des interprètes professionnels. Le principe est simple mais puissant : vous écoutez de l'anglais natif et le répétez immédiatement à voix haute — comme une ombre suivant le locuteur avec un décalage de 1 à 2 secondes. Les recherches montrent une amélioration significative de la précision de la prononciation, de l'intonation, du rythme, des liaisons, de la compréhension orale et de la fluidité.