シャドーイング練習: We will all be generalists. - 動画で英語スピーキングを学ぶ

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.

📺 同じチャンネル

Wild times we live in

Business EnglishUniversity StudentsWorking Adults

Why 95% of Android Developers Get Use Cases Wrong

Business EnglishUniversity StudentsWorking Adults

✨ おすすめ動画

55 文

このレッスンの語彙とスピーキングのポイント

この動画には、シャドーイング用の文が123文、単語が1669語あります。音声の長さは7:57です。話す速さは速く、1分あたり約210語です。音のつながりや弱く発音される音が多くなります。単語の85%は英語の頻出3,000語に含まれます。残りは練習の前に確認しておきましょう。

この動画の重要語彙

動画に出てくる覚えておきたい単語15語を、発音と意味つきで紹介します。

単語	発音	意味
android 名詞	/ˈænˌdɹɔɪd/	人造人間, アンドロイド
implement 動詞	/ˈɪmpləmənt/	実行する, 実現する
loop 名詞	/luːp/	ループ
prototype 名詞	/ˈpɹəʊtətaɪp/	プロトタイプ, 原型
implementation 名詞	/ˌɪmplɪmənˈteɪʃən/	実施, 実用化
niche 名詞	/nɪt͡ʃ/	壁龕
topic 名詞	/ˈtɑpɪk/	話題, テーマ
connect 動詞	/kəˈnɛkt/	繋げる
collect 動詞	/ˈkɑlɪkt/	集める, 収集する
knee 名詞	/ni/	膝
database 名詞	/ˈdeɪtəˌbeɪs/	データベース
nowadays 副詞	/ˈnaʊ.ə.deɪz/	現在, 現代では
comparable 形容詞	/ˈkɒmpəɹəbəl/	比較できる
propose 動詞	/pɹəˈpəʊz/	申し込む
stable 名詞	/ˈsteɪ.bəl/	馬小屋, 厩舎

動画に出てくる句動詞

単語	意味
find out 動詞	見出す, 調べる
pay off 動詞	報われる

この動画の文法

話し手がよく使っている文型を、動画の実際の表現とともに紹介します。

文型	動画での表現
条件文 if + 節, will/would + 動詞 — 条件とその結果	if that's something you would also · If you feel like, hey, that would be
現在完了形 have/has + 過去分詞 — 過去の出来事が今も関係している	have been · have gotten · has implemented
受動態 be + 過去分詞 — 誰がするかより、何が起きるかに焦点を当てる	be written · is trained · is implemented
関係詞節 who / which + 節 — 人や物について情報を加える	someone who was · someone who doesn · someone who spent

注意したい発音

話し手はdoesn't, don't, can'tなど、短縮形や弱形を19回使っています。聞こえたとおりの短い形で発音しましょう。

「sh」と「zh」の音: implementation /ˌɪmplɪmənˈteɪʃən/, restriction /ɹɪˈstɹɪkʃən/, practitioner /pɹækˈtɪʃənəɹ/, specialize /ˈspɛʃəˌlaɪz/
長い単語（アクセントの位置に注意）: developer /dɪˈvɛləpɚ/, implementation /ˌɪmplɪmənˈteɪʃən/, comparable /ˈkɒmpəɹəbəl/, primarily /ˈpɹaɪ.mə.ɹə.li/, architecture /ˈɑː.kɪˌtɛk.t͡ʃə/

この動画での練習方法

まず声を出さずに動画を最後まで聞き、知らない単語をメモします。
まず0.75倍速で一文ずつシャドーイングし、慣れてきたら通常の速度に戻します。
自分の声を録音して元の音声と比べます。android, implement, loopなどの単語に特に注意しましょう。

シャドーイングとは？英語上達に効果的な理由

シャドーイング（Shadowing）は、もともとプロの通訳者養成プログラムで開発された言語学習法で、多言語習得者として知られるDr. Alexander Arguelles によって広く普及されました。方法はシンプルですが非常に効果的：ネイティブスピーカーの英語を聞きながら、1〜2秒の遅延で声に出してすぐに繰り返す——まるで「影（shadow）」のように話者を追いかけます。文法ドリルや受動的なリスニングと異なり、シャドーイングは脳と口の筋肉が同時にリアルタイムで英語を処理・再現することを強制します。研究により、発音精度、抑揚、リズム、連音、リスニング力、そして会話の流暢さが大幅に向上することが確認されています。IELTSスピーキング対策や自然な英語コミュニケーションを目指す方に特におすすめです。

シャドーイングのやり方: ステップ別の完全ガイドを読む →

シャドーイング練習: We will all be generalists. - 動画で英語スピーキングを学ぶ

Wild times we live in

Why 95% of Android Developers Get Use Cases Wrong

Cloud Providers are changing RAPIDLY [REUPLOAD]

How to Manage Multiple Interests

Google Interview Tips That Actually Get You Hired

Is ML Engineering Dead? (AI vs ML Engineer)

#001 Well, hello there engineers

How do elite performers automate their habits? | Wendy Wood

Can you learn English with AI?

24 Hours Inside a YC AI Startup

How I’d Learn Coding in 2026 (Without Wasting Years)

Self Educated Polymaths Will Win In The Digital Age

Are we heading to an era of stagnation or innovation?

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