跟讀練習: Should You Become an AI Engineer? - 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。

162 句

如果句子過短或過長，請點選 Edit 進行調整。

I get a lot of messages from people who want to become AI engineers, and honestly, I get it.

The field is everywhere right now.

The salaries look incredible, and the work is genuinely meaningful.

But at the same time, you're hearing people say that AI is just another tech bubble.

So which is it?

Is AI engineering the dream job everyone makes it out to be?

Or is it just hype?

Today, I want to give you an honest look at what becoming an AI engineer actually involves,

so you can make a real decision based on what's actually true.

If you're new here, my name is Jean Lee.

I've spent 20 years in tech, including as a 19th engineer at WhatsApp before its $19 billion acquisition by Meta, and as a software engineering manager.

Today, I help people navigate tech careers in the age of AI.

And thanks to Coursera for sponsoring this video.

More on them later.

Before we get into the challenges, let's answer a much more basic question.

What is AI engineering?

And how is it different from machine learning?

AI engineering exploded in popularity for two simple reasons.

AI got dramatically better, and it got dramatically easier to use.

Ten years ago, building AI systems typically required a team of machine learning engineers and specialists who spent months collecting data,

fine -tuning, and training models.

Today, that foundation already exists.

Companies like OpenAI, Andrope, Google, and Meta have spent billions training powerful models.

So AI engineers don't meet to build those foundations from scratch anymore.

They built products and systems on top of them.

An AI engineer might use GPT, Claude, Gemini, or other open source models like Lama to create a product to solve a problem.

They're not necessarily reinventing the models.

They're choosing the right model, connecting it to data and tools, and tuning them into something that solves real business problems for them.

That's why AI engineers can move much faster than traditional machine learning teams.

So it's less like building the engine itself, but moving on to building the cars.

And perhaps the biggest misconception is that AI engineering is generally not as math intensive as traditional machine learning.

This role is much closer to software engineering than most people realize.

If you want a deeper breakdown of the differences, I do have an entire video on that.

You can watch that later.

And before we get into the five real challenges, let's clear up two misconceptions.

I want to first talk about the math.

Most AI engineers are not doing heavy math every day.

The job is mostly coding, building systems, working with APIs, and turning AI into useful products.

But if you're going into machine learning engineering specifically, there is math involved like linear algebra, statistics, probability calculus.

So if math is really not your thing, machine learning may not be the right fit for you.

But that doesn't automatically rule out engineering necessarily.

On the money side, people love to say, don't go into tech for money.

But I grew up prioritizing financial stability, and that's a real legitimate reason to choose a career for a majority of us. And realistically,

like many of the people who say don't do it for

the money have probably never had to seriously worry about it

that said the reason these jobs pay well is

because they're really hard ai engineering should not be seen as

a shortcut to a big paycheck the salary reflects the skill level required the pace of change

and the difficulty of building systems that actually work

while i do think it's okay to care about the money

and prioritize money but there are also downsides that you need to understand before going all into this career.

So let's go through the challenges.

Number one is that AI is genuinely hard.

And that might sound really obvious, but I think a lot of people underestimate it.

I've never met anyone in AI who says this is the easiest thing I've ever done.

Companies are not paying six figures because the work is so easy.

They're paying for skills that are really difficult to develop.

One of the most common questions I get is, is it easy to get an AI engineering job if I just follow a roadmap?

The answer is no. And honestly, that's the wrong question to be asking in the first place.

You won't ask if it's easy to become a brain surgeon.

You're learning coding, AI systems, APIs, prompting, evaluation, debugging, and product banking.

And then even after having all those skills, you're competing against thousands of other people with strong qualifications.

That doesn't mean you can't do it.

It just means you should go in with a realistic expectation.

If you're looking for shortcuts, I don't have one.

This path is for people who are willing to learn hard things and keep going.

And if you are trying to learn AI, one of the hardest parts isn't just the learning itself, it's having the right discipline and the structure.

to follow through.

Luckily, Coursera has built learning paths designed to solve this problem exactly.

If you want a beginner -friendly introduction, AI for Everyone covers core AI concepts,

real world use cases, and business applications without requiring a technical background.

If you want something more practical, the Google AI Professional Certificate focuses on mastering prompting to automate data analysis,

streamline communication, and build web applications from scratch.

And if you want to go deeper into both AI and machine learning,

Microsoft AI and machine learning is a much more structured path toward building and deploying real systems.

I'll link all these in the description along with the 40 % discount code.

Now back to the video.

Okay, So reason number one was that AI is hard.

But unfortunately, it's not just hard.

It's hard while thousands of other people are trying to do the exact same thing you're doing.

Think about California Gold Rush.

Some people struck gold, but most people didn't.

The same pattern has played out with the internet, mobile apps, and now AI.

That doesn't mean the opportunity isn't there or is hype.

And I'm living proof.

I benefited from one of those waves

when I joined WhatsApp as the 19th engineer

and I was lucky enough to be there when Meta acquired the company for $19 billion.

That experience helped me reach financial independence that I experienced today, and these opportunities can be life -changing.

But most people won't hit that outcome, not because they're not working hard or they're not smart,

but because whenever a lot of money concentrates in one area, a lot of people show up.

That's the supply and demand problem,

and right now the number of people who want AI jobs is growing much faster than the number of AI jobs available,

especially at the entry level.

Everyone wants to be an AI engineer, but very few have developed the right skills companies are actually willing to pay for.

And that gap creates intense competition.

And these high -paying jobs usually come with high expectations.

That's not unique to AI, but every high -paying role I've seen in 20 years of working in tech came with similar pressure.

Just look at what happened at OpenAI.

100

Sam Altman was pushed out of the company.

101

He helped build, then fought his way back in within days.

102

The board that tried to remove him ended up getting removed themselves.

103

And it was fast, intense, and showed exactly how high the stakes are at the top of this industry.

104

Do you think work -life balance is near the top of Sam Altman's? or executives at OpenAI's priority list right now.

105

And it's true, yes, most of us are not running AI companies, but the culture at the top usually trickles down and teams are moving fast,

106

expectations are high, companies are racing to ship and go IPO before the competitors do.

107

Hiring happens quickly, layoffs happen really quickly too.

108

And if building a balanced, sustainable lifestyle is your top priority, this is probably the most completely opposite of what you can go for.

109

So just think about the trade -offs.

110

You can pursue money or you can pursue work -life balance.

111

They're both fine options, but make a choice.

112

The fourth reason is that everything in AI moves really fast.

113

Remember when prompt engineering was the hottest skill for a while, and then MCP came in, then context engineering.

114

The other day, I tried to look for prompt engineering jobs

115

and I could not find any and now the job market is flooded with forward deploy engineering.

116

The conversation is shifting every few months and the tools you spent months learning might be replaced tomorrow.

117

The people who thrive in AI tend to kind of enjoy this environment.

118

They like learning new things, experimenting, and pivoting.

119

Uncertainly may feel exciting to these people.

120

Every week brings new challenge and that's fun.

121

But it won't be fun for everyone.

122

Not everyone operates that way because some people just do their best work

123

when things are stable and clear and expectations are laid out for you.

124

And if that is you, there's nothing wrong with it.

125

But AI engineering right now can be exhausting for that type of person.

126

The job isn't learning AI once and work for 20 years.

127

The job is continuously relearning AI tools and whatever comes next.

128

So know which type of person you are before making a commitment.

129

Now this last one might be the most important one on my list

130

because breaking into AI engineering at junior level is genuinely difficult.

131

The challenge isn't just competition from other new grads, you're often competing against experienced people.

132

A lot of them might have been working as a software engineer

133

or other fields for a long time and just switching into AI.

134

Or maybe they have masters

135

or phd degrees someone with 10 years of engineering experience only needs to learn the ai part

136

while new graduates usually would need to learn both

137

And that's a really tough spot to be in.

138

Most AI teams want people who can contribute quickly because they're competing with other companies too.

139

They often don't have the time or the resources to train someone from scratch.

140

And that's why internships matter so much.

141

The more real -world experience you have before graduation, the better chances you have.

142

It's also why many people are better off starting in software engineering, data engineering or some other technical role before going into ai

143

and of course if you can land an ai engineering role today right away go for it

144

but if you're having a hard time just understand

145

that there are other roles in the field of ai

146

that pay just as well and have more room for people who are starting out

147

so after all that why would anyone choose this path well

148

if you do make it the upside is significant you won't just have a job you'll be working on technology

149

that is reshaping entire industries healthcare education finance science manufacturing ai is transforming every field

150

that touches everyone's lives you'll be helping building tools

151

that millions of people use

152

and you'll often see where technology is heading long before anyone else does for the right person that's genuinely interesting

153

and exciting but notice i said for the right person right don't become an ai engineer

154

because everyone else is talking about it and doing it don't do it

155

because your mom or your dad

156

or your professor told you to do it the people who are really going to make it

157

and thrive in this field are people who are going to

158

be genuinely curious who enjoy learning who enjoy figuring things out

159

and don't fall apart when everything around them is changing constantly

160

if that sounds like you

161

and you've decided this is the path you want to pursue

162

you want to watch this video where i talk through exactly how to learn it i'll see you there

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

FOR YOU

本課的詞彙與口語要點

這節口語課以影片“Should You Become an AI Engineer?”為素材。影片中反覆出現的詞有：engineering, learning, engineer, machine。這段影片共有 162 個句子、2016 個單詞可供跟讀。講話部分時長為 12:01。說話人語速自然，每分鐘約 168 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
skill 名詞	/skɪl/	技巧, 技能
mathematics 名詞	/mæθ(.ə)ˈmæt.ɪks/	數學 /數學, 算學 /算學
expectation 名詞	/ˌɛk.spɛkˈteɪ.ʃən/	等待
solve 動詞	/sɒlv/	解決 /解決
compete 動詞	/kəmˈpiːt/	競爭 /競爭, 高下
pursue 動詞	/pɚˈsu/	追求
scratch 動詞	/skɹæt͡ʃ/	抓, 劃
necessarily 副詞	/ˌnɛs.əˈsɛɹ.ə.li/	必然地, 必定地
priority 名詞	/pɹaɪˈɒɹɪti/	優先權 /優先權, 更重要的事
salary 名詞	/ˈsæləɹi/	薪金, 薪水
tune 名詞	/tjuːn/	曲調 /曲調, 曲子
hype 名詞	/haɪp/	炒作
thrive 動詞	/θɹajv/	繁榮 /繁榮
deploy 動詞	/dɪˈplɔɪ/	部署
practical 形容詞	/ˈpɹæk.tɪ.kəl/	使用的, 實用 /實用

影片中出現的短語動詞

單詞	釋義
end up 動詞	到達
go through 動詞	通過 /通過
grow up 動詞	長大 /長大
rule out 動詞	排除
show up 動詞	露面

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	I've spent · have spent · have probably never had
定語從句 who / which + 從句 — 補充說明人或事物	people who want · people who say · people who are
被動語態 be + 過去分詞 — 強調發生了什麼，而不是誰做的	should not be seen · was pushed · is flooded

需要注意的發音

說話人用了 49 次縮略和弱讀形式，例如 you're, don't, they're。請按聽到的簡短形式來說。

“th” 音: mathematics /mæθ(.ə)ˈmæt.ɪks/, thrive /θɹajv/
“sh” 和 “zh” 音: expectation /ˌɛk.spɛkˈteɪ.ʃən/, specialist /ˈspɛʃəlɪst/, acquisition /ˌæk.wɪˈzɪʃ.ən/, evaluation /ɪˌvæljuˈeɪʃn̩/, graduation /ˌɡɹæd͡ʒuˈeɪʃən/
長單詞——注意重音位置: mathematics /mæθ(.ə)ˈmæt.ɪks/, genuinely /ˈd͡ʒɛn.ju.ɪn.li/, expectation /ˌɛk.spɛkˈteɪ.ʃən/, necessarily /ˌnɛs.əˈsɛɹ.ə.li/, priority /pɹaɪˈɒɹɪti/

中文母語者容易讀錯的音：

/v/ — 不要讀成 /w/，上齒輕觸下唇: solve /sɒlv/, thrive /θɹajv/, involve /ɪnˈvɑlv/, evaluation /ɪˌvæljuˈeɪʃn̩/, intensive /ɪnˈtɛnsɪv/
詞尾子音 — 要讀清楚，後面不要加母音: mathematics /mæθ(.ə)ˈmæt.ɪks/, prompt /pɹɑmpt/, solve /sɒlv/, compete /kəmˈpiːt/, scratch /skɹæt͡ʃ/
/r/ 和 /l/ 的區別: necessarily /ˌnɛs.əˈsɛɹ.ə.li/, salary /ˈsæləɹi/, practical /ˈpɹæk.tɪ.kəl/, reflect /ɹɪˈflɛkt/, popularity /ˌpɑ.pjəˈlæɹ.ə.ti/

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 skill, mathematics, expectation 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: Should You Become an AI Engineer? - 通過影片學習英語口語

選擇練習模式

下載應用

推薦影片

Fastest way to become an AI Engineer in 2026

Do Elite Degrees Still Matter In The AI Era? | Jamie Beaton from Crimson Education | EP 410

Advantages and Disadvantages of AI

Is ML Engineering Dead? (AI vs ML Engineer)

How Modern Data Teams Work (Engineer, Analyst, Scientist, Architect, ML)

Why the Arab World never united

Every Level of Engineering in 15 Minutes

No, Really — Is A.I. a Bubble? | The Ezra Klein Show

Can This Vibe Coder Beat a Senior Developer?

Jack Dorsey’s New Dessa AI Voice Sounds Scary Human

3 Ways AI Will Completely Change Your Life by 2029

AI can do your homework. Now what?