跟讀練習: We will all be generalists. - 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。

123 句

如果句子過短或過長，請點選 Edit 進行調整。

In the past, usually those developers who have deeply specialized in just one thing have been the ones paid the most.

So someone who was maybe into native Android performance or someone who specialized on Bluetooth low energy.

But now that output and research have gotten so much cheaper and faster with AI, there's ultimately more time for people to get better at more various things.

So how should we actually position ourselves in future?

First of all, what even makes someone an expert in a field?

Well for me at least it's someone who doesn't just have

the plain knowledge about a certain thing like Bluetooth low energy

but has implemented that in various different environments and gained real world experience that way.

So a BLE expert doesn't just know

which code has to be written in order to connect to a device for example

but they also know things like on which manufacturers there are which kind of quirks, what are edge cases to think of, which use case isn't it good for and why.

So I'm talking about those things you usually can't learn by simply watching a video or so.

Those things where you build something and then find out some environments or devices, it just doesn't work that well on.

Or you need some specific implementations to work around certain things.

And doing so not just for that one type of app, but for various.

So this kind of experience is what someone needs for me to at least consider them an expert.

And of course, if there's now a company having Bluetooth low energy problems, then they will be much better off having someone who spent a lot of time on that, to implement

that in various different environments than just taking let's call it

a general purpose android developer just like you would also prefer

going to a knee surgeon for a knee surgery rather than like a general practitioner

so can ai do

that can ai really implement such super specialized domain knowledge about

again let's take the example bluetooth low energy no ai in the end is trained on publicly available data

and such expert knowledge is of course hard to find anywhere in a structured format.

AI can build the happy path pretty well

but struggles at the edge cases especially when it's about such niche things like Bluetooth.

So AI cannot solely substitute the knowledge and the experience a real specialized expert has.

But still I think the level of expertise we as developers will have will become broader in future.

The thing is this while AI can't give you the specialized experience it makes gaining that significantly faster and easier.

You in the end still have to make this experience yourself

but this process of making it is just a lot faster

because first of all you can just get to a first working prototype a lot faster

so you can get this prototype on lots of devices faster

which makes you collect real world data on that prototype faster

and based on this real world data you can then again iterate on that prototype faster.

You can improve it fast.

You can again collect new real world data faster.

It just all got a lot faster nowadays to brainstorm how you could approach a problem

or a fix and all of that is what AI enables.

So what this ultimately means is, while you still have to do the same things to become really good at one niche thing, this process overall has strongly accelerated,

which means you have overall more time available to spend on other things, or in other words, more time to become an expert in other niches. And

that is why I do think the focus is shifting away

from being super duper great at one thing towards being super duper great at many things.

In times of AI, you're not special anymore, knowing a programming language

or knowing how to make a rest request that's simply not enough anymore to get a job

but something that any ai nowadays does very well in a minute

so our focus has to shift to a higher level and

that is knowing which tool to pick for which job

and this will be relevant across domains your main thing may be native android development

but just spending time on this one isolated domain is likely also not going to cut it anymore in future

but you will need surrounding knowledge so knowing okay

when would it be better to build the app non -native

for example how does the api design need to look like

to fit to the app what should your ci pipeline consist

of what is really the goal of the business behind the

app how we can improve this business purpose where does a more complex solution pay off

and where is it overkill maybe some kind of back -end

system design being able to design this whole thing not just the android app

but also the system that it connects to

so the bar to switch to related field

that is fundamentally different in implementation details

but very comparable on a higher level has just gotten

so much lower for example

if before this era of ai you are a native android developer

and you would feel well in this domain building android applications

in kotlin then you wouldn't just be able to open xcode

and start building an ios app some years ago

but now now it's possible you technically don't have to know

how to make a network call in swift anymore you don't

have to know how to set up a database table on iOS.

The AI does that for you

and as long as you know what a good network call in general looks like

and what a good database schema looks like you can also understand most of the code generated.

Just like you probably couldn't write a for loop in every

single programming language out there you would probably be able to understand a for loop in any language when seeing it.

It would likely be more difficult to build a stable backend as a mobile device

because it's a very different set of system design challenges

but an iOS app for example That comes with the same system design challenges we also have on Android.

Just the OS restrictions may be a little bit different here.

And I think this is really a big and important thing.

We as possibly native Android developers have to realize

that the exact look of low -level implementation details is becoming less and less important.

It won't matter

so much anymore how exactly a function is implemented as long as it does the job and is properly integrated, of course, into your architecture.

And as long as it fulfills the performance requirements you may have.

But whether a for loop is implemented or whether a while loop is used, how some variables are named inside, how it exactly works, that's not going to matter much anymore.

I really think we will ultimately all be software architects.

So we have to know how we build the house in a way

that we can put more and more floors on it, how we build it so it's not cold in winter.

But we don't have to know in detail anymore how we really put bricks on top of each other.

And that, by the way, is also a big question for me.

So what kind of content do I want to make in future?

And I think to some degree, I will always be the Android guy here.

And that's how most of you see me.

And it's also really something I love doing.

I love Android.

I love mobile development.

I love Kotlin.

But I also really see the value in gaining knowledge across domains.

100

Because it just has gotten so much faster, so much easier for us, mobile developers to maybe build an iOS app, a native one, or a website, for example, which is also just a client -based app.

101

And the ideas behind that are the same.

102

We have different frameworks to build the UI, different libraries, but in the end, it's a client application, it's making network calls, has maybe real -time communication, maybe has some kind of local storage.

103

Yes, it runs in a browser, that makes it a bit different, but all in all, it's very comparable.

104

So I'd actually love to cover a few topics here and there outside of the Android ecosystem, but really have no idea if that's something you would also want to do and which direction you find interesting.

105

For example, I always found backend interesting.

106

I found web development in general interesting, but also native iOS.

107

I kind of found everything interesting, but I also to some degree need to pick something.

108

So let me propose the following.

109

If you feel like, hey, that would be super cool to learn topic XYZ from Philip, please take a look in this video's comments if someone else already proposed that.

110

If that is the case, you simply give them an upvote so I can really see what most of you want to see.

111

If there is no such comment yet, then you just add a new comment so other people can upvote that.

112

That would really help.

113

This doesn't mean in any way that I will stop Android also.

114

Please don't think that.

115

I will keep doing primarily Android videos or mobile development content.

116

But I feel like I also need to adapt a little bit.

117

Because I'm not sure if in future it's enough to just be an Android guy.

118

So please let me know below what you would like to learn.

119

Thanks so much.

120

See you back in the next video.

121

Have an amazing rest of your week.

122

Bye bye.

123

We'll see you next time.

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

Wild times we live in

Daily ConversationWorking Adults

Why 95% of Android Developers Get Use Cases Wrong

Business EnglishUniversity StudentsWorking Adults

FOR YOU

本課的詞彙與口語要點

這段影片共有 123 個句子、1669 個單詞可供跟讀。講話部分時長為 7:57。說話人語速較快，每分鐘約 210 個詞，會出現較多連讀和弱讀。 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
android 名詞	/ˈænˌdɹɔɪd/	人形機器人 /人形機器人, 仿生人
domain 名詞	/ˌdəʊˈmeɪ̯n/	領域 /領域
implement 動詞	/ˈɪmpləmənt/	實施 /實施, 執行 /執行
loop 名詞	/luːp/	圈, 環 /環
prototype 名詞	/ˈpɹəʊtətaɪp/	原型, 雛形
implementation 名詞	/ˌɪmplɪmənˈteɪʃən/	履行
topic 名詞	/ˈtɑpɪk/	主題 /主題, 話題 /話題
connect 動詞	/kəˈnɛkt/	連線 /連線
collect 動詞	/ˈkɑlɪkt/	庋藏, 收藏
knee 名詞	/ni/	膝, 膝蓋 /膝蓋
database 名詞	/ˈdeɪtəˌbeɪs/	資料庫 /資料庫
nowadays 副詞	/ˈnaʊ.ə.deɪz/	現在 /現在, 現今 /現今
comparable 形容詞	/ˈkɒmpəɹəbəl/	比得上, 可比較的
propose 動詞	/pɹəˈpəʊz/	提議 /提議
stable 名詞	/ˈsteɪ.bəl/	馬廄, 馬棚 /馬棚

影片中出現的短語動詞

單詞	釋義
find out 動詞	找出, 查明

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	have been · have gotten · has implemented
被動語態 be + 過去分詞 — 強調發生了什麼，而不是誰做的	be written · is trained · is implemented
定語從句 who / which + 從句 — 補充說明人或事物	someone who was · someone who doesn · someone who spent

需要注意的發音

說話人用了 19 次縮略和弱讀形式，例如 doesn't, don't, can't。請按聽到的簡短形式來說。

“sh” 和 “zh” 音: implementation /ˌɪmplɪmənˈteɪʃən/, restriction /ɹɪˈstɹɪkʃən/, practitioner /pɹækˈtɪʃənəɹ/, specialize /ˈspɛʃəˌlaɪz/
長單詞——注意重音位置: developer /dɪˈvɛləpɚ/, implementation /ˌɪmplɪmənˈteɪʃən/, comparable /ˈkɒmpəɹəbəl/, primarily /ˈpɹaɪ.mə.ɹə.li/, architecture /ˈɑː.kɪˌtɛk.t͡ʃə/

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 android, domain, implement 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: We will all be generalists. - 通過影片學習英語口語

選擇練習模式

下載應用

同頻道

Wild times we live in

Why 95% of Android Developers Get Use Cases Wrong

推薦影片

Is ML Engineering Dead? (AI vs ML Engineer)

Self Educated Polymaths Will Win In The Digital Age

Why Coding Agents Keep Making Your Codebase Worse

Is Embedded Systems Still a Good Career in 2026?

LangChain in 5 Minutes (Explained Clearly)

How To Land Your First Junior Dev Role

Learn English with Podcast | Practice Smarter for Speaking, Shadowing & Real Conversation A2-B1

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

10 Skills That Become More Valuable As You Get Older

How to Become an AI Product Manager in 2026 | Ex-Google, Microsoft

Spec-Driven Development: AI Assisted Coding Explained