跟讀練習: Why AI art struggles with hands - 通過影片學習英語口語

選擇練習模式

每種模式都會開啟全屏練習室。

141 句

如果句子過短或過長，請點選 Edit 進行調整。

You're called to create.

A post -apocalyptic giraffe astronaut?

Generated.

Genghis Khan playing a guitar solo.

Pixel art?

Generated.

A man holding a delicious apple?

Ah, what's with his hands?

Why can't AI art make hands?

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.

Why is this so hard?

Seems easy enough, right?

We've got this weird situation where AI art can instantly make.

Abraham Lincoln dressed like glam David Bowie, but struggles with a woman holding a cell phone.

This isn't just a weird glitch.

The struggle of AI art with hands can actually teach you something bigger about how AI art works.

I mean, what is so hard about this?

I asked an artist who has taught thousands of people how to draw hands from imagination.

Before someone becomes or starts training to be an artist, like officially training, it's pattern recognition.

You just grow up seeing a whole bunch of hands and you start knowing what hands look like.

You learn how things look by living in the world and recognizing patterns.

An AI is similar but has key differences.

Imagine an AI is like you but trapped in a museum from birth.

All the machine has to learn from are the pictures and the little placards on the side.

Apple.

A red apple on a brown table.

That's like the images it sees from the web and the descriptions that go with them.

It's similar to how you learn but locked in that museum.

If you want to understand an apple, you can rotate it in your hand.

You can watch it whenever you want.

If AI wants to understand an apple, it has to find another picture of an apple in the museum.

Pattern recognition has allowed AI and people to draw decent apples.

But the processes differ.

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.

Artists in order to draw something complicated we tend to simplify things into basic forms and so

when you look at a hand you pretty much have the

big blocky part of the palm right you have the front you have the back

and then you have the thickness and you can pretty much just make

that into like a square with some thickness to it.

Then an artist can add all the style and texture and detail they want.

AI works differently.

Look at this hand.

The shapes are bizarre, but the AI has done a great job showing the light and texture here.

Remember, the AI knows how things look, but not how they work.

So these patterns in pixels are easy to understand.

It never learned, however, that fingers don't really bend like this.

It doesn't simplify to forms.

Remember, it's trapped in the museum.

So it is just trying to guess where hand -like pixels should be without knowing how hands work, like we do.

But listen, I find this kind of dissatisfying.

I mean, I'm basically just saying that AI can't draw hands because it's not a person.

But AI also doesn't know anything about construction, and it can still make a beautiful skyscraper in New York City.

So to understand this better, I spoke to two people who have worked with generative art models.

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.

Because of how popular these models have been in generative art, I've also been like reading a bit on that.

And I talked to Roy Shilkrat, who has a super varied resume, but has been teaching about generative art since 2018.

Good students that come in that are trying to break those models, take them to the next level.

Talking to them helped me figure out three big reasons.

Not every reason, but three big reasons that hands are tough for AI art models.

The data size and quality, the way hands act, and the low margin for error.

For the data size, let's go back to the museum idea.

The museum the robot hangs out in, it has a ton of rooms dedicated to faces, but not so many rooms for hands.

That means it has less to learn from.

Just as an example, available datasets like Flickr HQ has 70 ,000 faces.

70 ,000!

And this popular one annotates 200 ,000 pics of celebrity faces for lots of details, like eyeglasses or pointy noses.

There are There are a ton of great hand datasets that can really understand hands, like this one with 11 ,000 hands,

but these may not have been used to train the AI that makes art.

That data scarcity combines with the quality and complexity of the data.

Hand data in the art museum isn't yet annotated to show how they work, like the celebrity's pointy noses.

What they say is there's an image and there is a person in the image and the person is holding an umbrella.

You don't give the machine a lot of clues saying, this is a person holding the umbrella.

The thumb is going from one side of the handle and the fingers are curled.

And then the thumb is covering the index finger, but not the other ones.

All that is made worse because hands do lots of things compared to, say, faces.

So there's a pretty common portrait photo face.

There are a lot of these photos online.

And I think that everything's very well -centered, right?

Eyes are always around here.

There's always this order.

That's not true of hands which can do this, and this, and this.

I swear I'm sober right now.

Stan mentioned this too.

How many fingers do you see right now?

Like, two or three?

Like, it doesn't know there's five, because sometimes there's two, sometimes there's three, sometimes four, sometimes five.

You can see these problems with AI hands, but the jankiness is all over AI art.

Just look at horses.

You can often have, like, three legs, five legs, six legs.

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.

Okay, did you hear that last part he said?

Good, because it's really important.

It doesn't have as much bias as we do.

We care a lot about hands and need them to be perfect.

There is a low margin for error.

But because the model doesn't understand hands, hasn't seen many, and because hands act weird,

it makes pictures that are like hands it's seen in the museum, but not an exact hand.

That's good enough for a ton of stuff, but not hands.

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

下載應用

流利口語所需的一切

AI 發音評分逐句打分

音標練習掌握每個音

詞彙學習建立你的詞庫

詞彙遊戲邊玩邊學

掃碼

Download on theApp Store Get it onGoogle Play

同頻道

FOR YOU

本課的詞彙與口語要點

這段影片共有 140 個句子、1669 個單詞可供跟讀。講話部分時長為 9:47。說話人語速自然，每分鐘約 171 個詞，接近日常對話。 85% 的單詞屬於英語最常用的 3,000 詞，其餘的詞建議在練習前先學一下。

影片中的重點詞彙

影片中 15 個值得學習的單詞，附發音和釋義：

單詞	發音	釋義
finger 名詞	/ˈfɪŋɡɚ/	手指, 指頭 /指頭
umbrella 名詞	/ʌmˈbɹɛl.ə/	傘 /傘, 雨傘 /雨傘
pixel 名詞	/ˈpɪk.səl/	畫素
recognition 名詞	/ˌɹɛkəɡˈnɪʃ(ə)n/	認出 /認出
delicious 形容詞	/dɪˈlɪ.ʃəs/	好吃, 可口
trap 名詞	/tɹæp/	陷阱, 圈套
margin 名詞	/ˈmɑɹ.d͡ʒɪn/	邊界 /邊界
bias 名詞	/ˈbaɪ.əs/	偏見 /偏見
thumb 名詞	/ˈθʌm/	大拇指, 拇指
texture 名詞	/ˈtɛkst͡ʃə(ɹ)/	質地 /質地, 材質 /材質
thickness 名詞	/ˈθɪknəs/	厚度
stripe 名詞	/stɹaɪp/	條紋 /條紋
guitar 名詞	/ɡɪˈtɑɹ/	吉他, 六絃琴
complicated 形容詞	/ˈkɑm.plɪˌkeɪ.tɪd/	複雜 /複雜
compare 動詞	/kəmˈpɛɚ/	比, 比較 /比較

影片中出現的短語動詞

單詞	釋義
figure out 動詞	弄清楚
go away 動詞	離開 /離開, 走
grow up 動詞	長大 /長大

影片中的語法

說話人最常用的結構，並附上影片中的原話：

結構	影片中的用法
現在完成時 have/has + 過去分詞 — 過去發生但與現在仍有關聯的事	has taught · has allowed · has done
被動語態 be + 過去分詞 — 強調發生了什麼，而不是誰做的	are curled · is made · are generated
定語從句 who / which + 從句 — 補充說明人或事物	people who have · hands which can · appreciate, which is

需要注意的發音

說話人用了 32 次縮略和弱讀形式，例如 doesn't, don't, they're。請按聽到的簡短形式來說。

“th” 音: thumb /ˈθʌm/, thickness /ˈθɪknəs/
“sh” 和 “zh” 音: recognition /ˌɹɛkəɡˈnɪʃ(ə)n/, delicious /dɪˈlɪ.ʃəs/, imagination /ɪˌmæd͡ʒəˈneɪʃən/, variation /ˌvɛəɹiˈeɪʃn̩/
長單詞——注意重音位置: recognition /ˌɹɛkəɡˈnɪʃ(ə)n/, complicated /ˈkɑm.plɪˌkeɪ.tɪd/, diversity /daɪˈvɜː(ɹ)sɪti/, celebrity /səˈlɛb.ɹɪ.ɾi/, differently /ˈdɪf.ə.ɹənt.li/

如何用這段影片練習

先完整聽一遍影片，不要開口，記下不認識的單詞。
先用 0.75 倍速逐句跟讀，熟練之後再回到正常速度。
錄下自己的聲音並與原聲對比，特別注意 finger, umbrella, pixel 這類單詞。

什麼是跟讀法？

跟讀法 (Shadowing) 是一種有科學依據的語言學習技巧，最初開發用於專業口譯員的培訓，並由多語言者Alexander Arguelles博士普及。這個方法簡單而強大：您在聽英語母語原聲的同時立即大聲重複——就像是一個延遲1-2秒緊跟說話者的影子。與被動聽力或語法練習不同，跟讀法強迫您的大腦和口腔肌肉同時處理並模模擬實的講話模式。研究表明它能顯著提高發音準確性，語調，節奏，連讀，聽力理解和口語流利度——使其成為雅思口語備考和真實英語交流最有效的方法之一。

影子跟讀法: 閱讀完整分步指南 →

跟讀練習: Why AI art struggles with hands - 通過影片學習英語口語

選擇練習模式

下載應用

Why financial literacy education in the US sucks

The surprising pattern behind color names around the world

White, pink, and brown noise, explained #shorts

Five things you need to know for the 2026 midterms | America, Actually

Why are weddings so damn expensive?

Are headphones destroying our hearing?

How to be less neurotic | The Gray Area

We’re already using AI more than we realize

The Ticketmaster monopoly is over, now what?

The biggest peach myth in America

How to solve problems like a designer

The origin of the '80s aesthetic

Why “Buy one, get one free” isn’t a great deal

How F1 drivers survive their own cars

Why 350°F is the magic number for baking

How did pink become a girly color?

Why China's population is shrinking

The real reason Boeing's new plane crashed twice

Why is everything getting so expensive?

Proof of evolution that you can find on your body

watch this if you're an artist with no talent.

Disturbing Robot Thinks It's Human..

AI Animation Pipeline: How I Make Cartoons 100% With AI

AI Coding Is a Dead End

The world's most frustrating work of art

Generative AI explained in 2 minutes

The different levels of how Claude thinks

Every Generation Thinks This Time Is Different: The AI Job Myth

Dear Higgsfield, I reverse engineered you in 12 hours

Why Study Art? | TateShots