シャドーイング練習: They JUST Accidentally Revealed the Truth: The AI Bubble - 動画で英語スピーキングを学ぶ

練習モードを選ぶ

各モードは全画面の練習ルームで開きます。

389 文

文が短すぎたり長すぎる場合は、Editをタップして調整してください。

a seven -fold surge in profit.

The results boosted hopes that profits can be sustained as AI developers build data centers, which drives demand for Mary Mary.

Joining us, Bloomberg Tech.

The SK Hynix earnings were crazy.

Let's go through a breakdown of not just the raw numbers.

Let's go past that.

Let's go into what they're saying about the entire AI build out of the earnings call

and do a quick summary of all of it.

I'm really excited about this because I think this is absolutely wild.

So the first thing that you've got to know is they do these things called LTAs.

They're long -term agreements.

And the whole reason they do this is because back in 2023 and many times in many cycles before that, I think 2014 was another one.

The dot -com bubble was another one.

They're these cycles where memory basically goes through a shortage, prices of memory go sky high,

and then all the memory manufacturers are like, you know, what we should do.

We should build more.

And then all of a sudden the underlying desire to, you know, buy more memory chips fades or slows or the growth rate slows.

Now all of a sudden you have a glut and you have overproduction, overcapacity, prices fall off a cliff.

That's what everybody's worried about with memory stocks.

And honestly, I've been a little bit of a poopy doopy bear since this summer on memory.

I'm like, this is a meme movement at this point.

I mean, it's kind of going up like GameStop at 400 bucks and you know it's like

Leo Fold came in

and turned off the buy button you know what I'm saying like the momentum just evaporated

and so we were not very excited about these plays going even into the Leo Fold

because we saw this as sort of a short -term trade

and you better have a trailing stop many times called out have a trailing stop

now what do we hear now because what we're hearing now is actually really optimistic for these memory plays.

What we're hearing is they're doing more of these five -year contracts with their customers.

And right now, over 50 % of all of SK Hynix's capacity,

the production that they can generate for 2027 and 2028 is going to long -term agreements.

Now that's actually really impressive.

That's great because it gives them some stability.

But what happens when those long -term agreements expire?

Well, then you could be left holding the bag, right?

But that's actually where things are getting fascinating here.

Look at this.

Customers are requesting longer -term collaborations based on our product roadmap and supply capabilities,

indicating that long -term agreements are evolving to strengthen customer relationships and business continuity.

Today, we have long -term agreements with 14 customers, including pretty much all of the major cloud service providers,

and we're proactively in discussions about extending those contract periods.

Now, why would they want to extend contract periods?

Doesn't that create more risk for a company like Google or Microsoft or Amazon or Anthropic or whatever, or NVIDIA or AMD?

Sure.

But take a look at what they're also doing.

They are building now custom memory chips.

So we already know that Marvell and Broadcom work with companies like Anthropic and OpenAI to make custom application -specific integrated circuits.

Those are called ASICs.

Remember when those were all the rage for mining Bitcoin?

They do one thing really well.

That stands in contrast with an NVIDIA GPU, which is like a workhorse.

What do you want me to do today, boy?

I got you.

You want me to do that?

I got you.

Okay, the ASIC's like, man, I push this button all day long, bro.

I just push the button.

And so if needs change, those ASICs can lose value very rapidly.

It's one of the reasons people pay a premium for an NVIDIA GPU or some of the AMD products.

Custom high bandwidth memory is very interesting because this is now suggesting,

hey, how about these memory companies come in and actually customize what individual customers want for their product.

Now, what we're seeing a lot of is spec changes at companies like NVIDIA, where we're saying, hey, do we really want so much memory

or should we despec and go with less memory because we don't necessarily need all the memory?

What you don't want is stranded memory.

You don't want to oversize a chip with so much memory

and you can't even operate all of what you're trying to run through a chip because the bandwidth is your bottleneck.

All right, if you bottleneck one flow, what do you need all the RAM for?

What do you need all the high bandwidth memory for?

So that's where SK Hynix is like, look, we'll do even longer term deals if you want.

We'll even customize the chips for you.

These things are very smart.

And the bull case in investing in these companies is A, they've gotten cheap.

B, is they're returning a lot of money, capital to shareholders.

We just saw at the beginning here, 45 billion expected to be returned to shareholders soon.

And they promise to return even more in the future.

But what's really fascinating is that these companies are trying to kill the boom and bust cycle for memory.

And that's where reading this is either really good

if you're a bull or

if you're interested in what's going on with the overall artificial intelligence space and frankly what's happening with the AI buildup.

This is a key read.

So let's go through some of the highlights for you.

So we expect to increase capital expenditures throughout 2027 significantly compared

to 2026 to expand the share of leading edge processes and ramp up our first fab in Youngin.

Okay, generally, when investors see, wait, you guys are going to increase capacity even more?

You're going to spend more money on CapEx?

When people see that, they get nervous.

They get nervous that, oh, you're definitely going to lead to an oversupply.

So they actually got asked specifically about oversupply.

And what they said was this.

They said, thank you for your question.

It is true that a company is pursuing medium and long -term investments to accommodate strong demand,

but more investment and more fabs does not necessarily mean more supply.

With a rising share of high -performance products like high -bandwidth memory,

which, keep in mind, that takes about two to five times as many wafers as DRAM.

Okay, just to be clear here, like, RAM is like your overall umbrella, okay, and DRAM is really what everybody wants right now,

so just consider that as like almost the umbrella product.

And then you can divide that into like DDR5, which could be like a server product,

or you could stack DRAM and turn it into high bandwidth memory.

The point is, the more you're using these advanced chip packages, the fewer chips you're actually putting out.

100

But you're putting out potentially better quality or more capable chips.

101

And so that's what they're trying to say here is like, hey, just because we might be able to produce more at more facilities,

102

we might be creating more complex chips, outputting them with finer technology.

103

And therefore, that actually restrains our potential total bid output.

104

Now, that's obviously their bullish point of view.

105

It's their company trying to pitch their own products and their own stock.

106

Obviously, I generally see earnings calls as a sales pitch for their own stock.

107

But this is a really good comparison or argument for the bulls.

108

It's basically saying, hey, guys, like, just because we're going to manufacture more doesn't mean, you know, we're going to oversupply.

109

Obviously, at some point, if the AI trade slows down,

110

you know, it's all going to go poopy -doopy and you're probably going to be in a recession.

111

But, you know, until that point, so far, the AI boom seems like it is crushing it.

112

Now, I want to show you CapEx and what's going on over here because some of these numbers and a chart, really, really interesting.

113

A very brief reminder, though, Reinvest One, the super intelligence and one terminal coming out in this Q4,

114

does have a coupon code expiring on Friday, October 30th.

115

Join before the price goes up and make sure you lock in your lifetime access to the alpha membership, the portfolio transparency, trade alerts,

116

reinvest stocks terminal, the homes terminal, the reinvest research terminal, all nine courses, the course member live streams, the agentic AI modules.

117

You can learn all about it over at reinvestone .com.

118

We'd love to see you there.

119

Get that lifetime access before Friday.

120

We have a price increase.

121

Okay.

122

Let's take a look at this chart because this is fascinating.

123

So this is the cash machine of SK Hynix.

124

So basically, their CapEx is going up.

125

That's a little purple box here.

126

Their CapEx, how much they're spending and investing, going up bigly.

127

I mean, you're seeing a massive increase in CapEx from last year.

128

However, their free cash flow is also exploding.

129

The full bar chart here is their operating cash.

130

The green is their free cash flow.

131

So even after their CapEx, they still have massive free cash flow.

132

And this is where they're saying, hey, we want to give a lot of money back via dividends or whatever to shareholders.

133

Now, obviously, that's a sales pitch to get people to buy the stock.

134

Duh.

135

That's what earnings calls are.

136

There's nothing wrong with that.

137

It's just always worth taking things with a grain of salt.

138

So let's keep going through here.

139

So yeah, here they talk about returning 50 % of free cash flow to shareholders, but also ramping CapEx.

140

The chart I just showed helps a lot.

141

So what's going on with AI demand? because there's been talk about pacing the AI buildout.

142

Is this talk about pacing the AI buildout going to, you know, crush the economy?

143

Where do we stay?

144

Oh, my gosh, we're going to slow down AI investments.

145

This is the only thing holding up GDP, which, in fairness, should do a quick check on GDP.

146

GDP's been doing really freaking well, okay?

147

A lot of that, obviously,

148

driven by these expenditures and these investments that we're seeing by some of the richest companies in the world.

149

Atlanta Fed real GDP estimated to be at 3 .3 % right now.

150

And that's actually come down from as high as 6 % earlier in this quarter.

151

Those are the estimates for the third quarter.

152

They come in very noisy at the beginning of the quarter,

153

and then they kind of settle out a little bit towards the end of the quarter because they collect more data.

154

So it's kind of like a real -time response to what they're seeing with data.

155

But anyway, regarding the pacing of AI, there are two main reasons why they think artificial intelligence momentum will continue.

156

I think this is so fascinating that you could literally open up an earnings call

157

and see exactly what the industry insiders are telling you is happening.

158

Take a listen to this.

159

First, we are seeing customers' AI investments actually leading to greater adoption of AI services, and they are producing business results.

160

With major hyperscale customers reporting year -on -year growth of 30 % to 80 % of their cloud revenue,

161

so growth of 30 % to 80 % on cloud revenue,

162

a substantial portion of which SK Hynex believes is because of the expansion of their AI business and their AI products.

163

You know what I like to do sometimes is I'll sort of just see and Google has been so left behind

164

Okay, but sometimes I don't know if I can pull it up right now,

165

but sometimes I'll just go into Google admin and I like to go see like hey, what are they charging for?

166

You know upgrades or like well, you know what what can you add to?

167

Users or seats or whatever and

168

so what you could do is you could actually kind of go into the back end

169

and see how they're trying to

170

upsell people licenses right so

171

if I go jump in right now yeah AI expanded access here I don't think

172

that any of this is pretty proprietary here I'll just make sense

173

this is what I like to look at I like to keep an eye on this

174

so I look at this for example and I'll see okay

175

so we don't have any more Google work workspace enterprise licenses that's fine but then I look at this AI expanded access.

176

I could go add to this email address AI expanded access for 24 bucks a month.

177

And then what they'll do with this is they'll pitch you,

178

oh, hey, you know, now you could use a better version of the chat bot on,

179

you know, Google chat or Google Gmail or whatever, or we'll let you have more usage, blah, blah, blah, whatever.

180

And they're really trying to integrate AI throughout a bunch of different components of the business.

181

Like here's how Google does it, for example.

182

They'll come in and say, expanded access will get you more queries per month, build and edit spreadsheets.

183

You could do more of them, build more canvases, generate more videos, more avatars, whatever.

184

You could get more thinking responses.

185

You could get more argon responses.

186

Look at this, up to 200 of their argon.

187

Now that's Gemini 4.

188

So, like, you don't pay for that license, you don't get Argonne.

189

It's kind of like their way of saying, hey, you want to pay for Fable?

190

Oh, you're gonna have to add it to, you know, your workspace seats.

191

And it's so easy to check the box and add it.

192

And when an industry does it per seat, that's where companies like Google print freaking money.

193

It's because somebody goes in there and goes, you know what, we're gonna enable this for everyone.

194

It's like GitHub, you know?

195

Yeah, it's like, oh yeah, let's enable that security feature, because everybody's worried about cybersecurity great that's gonna cost

196

you 10 bucks more per month per seat you know

197

we got thousands of people it gets really expensive really fast

198

so anyway like I see that happening that's definitely happening have your loved one child been affected by social media

199

Morgan Morgan can fight to help you if you

200

or your child spent years growing up on social media

201

and experienced serious mental health struggles you should hear this recently a jury ordered meta

202

and YouTube to pay millions for their role in designing and promoting addictive platforms.

203

They knew their apps could contribute to anxiety, depression, and body image issues, but did not share those risks while prioritizing profits.

204

The recent historic verdict underscores the consequences of those decisions.

205

Now, Morgan & Morgan is stepping up to hold these platforms accountable for the harm they cost.

206

Click the link in my bio to take a short quiz, and you may be entitled to a potential recovery of over $1 ,000.

207

It's quite easy to do.

208

Go to the link in the description, fill out the form for yourself or a loved one.

209

It's quick, easy, and you could recover a potential recovery of over $1 ,000.

210

More long -term contracts for large -scale enterprise AI solutions as well, while other cloud order backlogs have grown two to five -fold

211

year on year cloud order backlogs growing two to five fold

212

year on year is incredible cash flow will only improve they say across ai like this is

213

so damn bullish ai it is insane it is it i almost want to be bearish

214

because of how bullish it is because i'm generally a contrarian i'm the kind of guy

215

That's gonna be at a party, okay?

216

We'll be at a party with a Diet Coke.

217

Hey, you wanna buy real estate?

218

And then 20 people will be like, real estate, F that, we hate real estate.

219

Fine, I'm gonna buy more real estate.

220

And then it's probably gonna be the same thing next time there's a stock market crash.

221

Everybody's gonna be like, ew, stocks suck.

222

And we'd be like, hey guys, we should buy stocks.

223

Ew, leave.

224

I don't know, I just always seem to want to do the opposite it.

225

And so I feel bad almost that I am aligned

226

and in agreement with this bullishness because that is not part of jaded Kevin.

227

But this makes sense.

228

This is what we're seeing.

229

And I believe there has been a massive inflection in the last six months and how good AI is.

230

And I think that's the contrarian piece is that most people don't even notice it yet.

231

How good it's gotten.

232

It's scary.

233

AI sales and cashflow will only improve and will ease the current market concerns on the investment burden.

234

In fact, we're going to see more multi -site development, smaller or mid -sized data centers to bring power and data centers available more rapidly.

235

This is basically saying, look, rather than have like the giant Kevin O 'Leary facility that everybody's going to go protest outside of,

236

what if we just did 10 smaller ones and then we'll have fewer protests?

237

And if two of them get shut down by city permitting issues, no problem, we got eight of them.

238

We actually do the same thing at reinvest with ADUs.

239

You know, we'll do six ADU applications in hopes of getting three of them because, you know, California is retarded when it comes to building.

240

And so you just, you just go in knowing that, you know, uh, and that's okay.

241

You know, you underwrite, you always underwrite as if you are not going to be able to build.

242

And as long as your deal makes sense and pencils out, as long as assuming you're not going to get what you want from the city,

243

if you get what you want from the city, it's pure bonus.

244

I love that.

245

But anyway, so a single company adjusting its investment in the past could directly result in overall demand volatility.

246

However, today, because AI is rapidly broadening to sovereign AI, on -prem demand, non -tech companies investing in their own AI,

247

all of a sudden you no longer have this single company problem.

248

So in other words, in the past, and I've changed, I changed this yesterday.

249

Remember how I always say there are these two canaries?

250

Okay, canary one is the labor market.

251

Unchanged, labor market's a big problem.

252

But the second canary, I've historically said is anthropic and open AI.

253

But yesterday what I did is I actually changed that.

254

I said, no, I think that's wrong now.

255

I don't think they matter as much as they used to because you have competition from Microsoft, from Gemini, from Amazon, from SpaceX,

256

from Grok, you know, whatever.

257

And what's actually happening is you are seeing more diverse AI expenditures

258

that insulate the entire AI boom from volatility from one of the companies slowing down

259

or one of the companies having problems with financing or whatever.

260

Like, it's not just

261

Oracle, it's everyone, right?

262

It's no longer just one company that moves the whole market.

263

So I actually agree with that.

264

Okay.

265

Then they mentioned here that they're sold out through the rest of the decade,

266

that DRAM and high bandwidth volume for the next three years through 2029 is effectively sold out,

267

that their goal is to improve longer term demand visibility that can lead to stable revenue and earnings growth.

268

Yep.

269

That's the whole point of these long -term agreements.

270

We talked about the overproduction fear.

271

We talked about the custom chips.

272

They do mention that prices for HBM4 are expected to increase, and that long -term high bandwidth memory demand outlook remains unchanged.

273

Demand remains so strong that customers' volume requirements are more than double what we can supply today.

274

That's freaking insane.

275

Then over here, this is where they talk about next -gen chips.

276

This is where things get a little bit geeky.

277

So I'm not going to go really crazy on this but basically right now they use a technology called a 6F.

278

Okay.

279

This has to do with, uh, the, the measurement of the chip technology that they do.

280

And the smaller they go, the more they can put memory cells on chips and potentially, uh, the, the more, you know,

281

you could charge for it.

282

But the problem with that is they say there's a risk to going to 4F right now, because you might have yield issues.

283

You could have reliability issues.

284

You could have competitiveness issues, like cost competitive initiatives.

285

So what we're going to do instead is we're going to focus on what we do really well, which is smart for a business.

286

We're going to focus on the 6F structure.

287

We're going to maximize our potential income from the 6F structure.

288

That's what we're going to focus on.

289

And we'll scale when the time is right.

290

So we're going to kick the can down the road a little bit on that.

291

Keep in mind that this company is both a DRAM play and a NAND play,

292

which is very interesting because I personally have this thesis that DRAM is going to explode in supply, which...

293

They counter -argue and say, yes, that might be true,

294

but we might net have fewer products out because we're adding more DRAM to high bandwidth memory or whatever.

295

We're increasing stack sizes potentially or whatever.

296

But they're also a NAND play.

297

21 % of their income comes from NAND.

298

Now, this gets confusing to people.

299

So NAND, the way to think about it is NAND is just, this is all storage.

300

When you hear NAND, when you hear SSD, when you hear NVMe, just think storage.

301

Okay, storage is different from RAM.

302

DRAM, high bandwidth memory, you're under the umbrella.

303

DDR5, right?

304

Those are all under the memory umbrella.

305

NAND, SSD, NVMe, one little category.

306

It's all basically storage.

307

NAND is the actual memory cell technology.

308

Then you've got SSD, which is the finished product.

309

That's what you buy.

310

The NVMe is what plugs in through your PCI port and the communication between those, blah, blah, blah.

311

The point is, this company actually does both.

312

I am bullish on NAND because I talked about this even before I read this.

313

I'm like, dude, I'm finding that these computers where I run agents on, if they don't have a good SSD,

314

they get bogged down, they shut down, they basically blow up.

315

If you're trying to push them to the limits too much.

316

Because I think what's happening with agents today is they're offloading more of this memory to your SSD

317

and not really telling you about it.

318

It's happening in the back end and it's crushing some of the weaker computers.

319

and it's actually making the case for like, oh yeah, I'm just going to go buy Apple Neos and run chatbots on them.

320

Not a good idea.

321

You need better computers.

322

You know, like for example, I got an M4 Max right here, a laptop, and I could run agents on this guy all day long.

323

But those other guys are just going to start smoking.

324

But what's happening here is they actually talk about how this is going to happen more and more,

325

where basically you're going to see NAND demand increase as AI inference demand increases.

326

So you could really store

327

some of that cash off of memory because you can't keep it in memory

328

because you're just going to saturate your memory and then you're going to be memory locked.

329

So this, I have to say, this is probably, and I know this was a lot,

330

and we're going to do a valuation on them in just a moment.

331

This was probably one of the most bullish earnings calls that I've ever read.

332

And I've read a lot of freaking earnings calls.

333

Okay, I put myself to sleep reading earnings calls sometimes just thinking about reading the early skull

334

But but yeah this this was incredible

335

so SK Hynix 10 out of 10 for giving us insight into this boom is nowhere near done now

336

What I do want to do is grab the SK Hynix evaluation now these are 80 hours

337

So I have to be a little careful with how we do these

338

But we're gonna pull the valuation here. So

339

uh let's see a SK HY okay okay

340

so SK Hynix is up and what we're looking at is their forecast earnings per share

341

and then their expected growth rate

342

so right now earnings per share are expected to come in and let's go ahead and change this all to US dollars.

343

Remember, this is a foreign company, so everything's going to be in Korean Hwan.

344

So we're looking for EPS adjusted of 270 Dallajas.

345

Now, let me just double check.

346

The ratio is 10 to 1.

347

So you need to divide EPS by 10.

348

This is where it gets complicated.

349

I need to fact check me on this.

350

But you need to divide EPS projections by 10.

351

So that means one to get to an ADR.

352

163 divided by, sorry, yeah, divided by 27, basically.

353

162 .73, that's the stock price right now, divided by 27 .1, we'll call it, six.

354

This is trading for a six PE ratio right now.

355

This is insane because people think these things are going to roll over and fall off a cliff.

356

And honestly, the analysts are pricing that in too there's a big negative year in 2029 that they're forecasting,

357

but then they go back to growth in 2030.

358

The analyst expectations are just looney tunes right now.

359

But anyway, 2027, we expect 35 % growth, 22 % growth minus 18 % growth, and then 28, or sorry, 38 % growth.

360

Let me do that math one more time to make sure I got it right.

361

3522 minus 18 plus 3895.

362

Yep.

363

Divided by four.

364

We're expecting a 19 % growth on average for EPS.

365

Honestly, I think they'll probably do better than that.

366

They're trading for a 6 %...

367

P .E ratio divided by 19.

368

Dude, this is trading for a point three peg.

369

This has been so discounted for cyclicality.

370

It's insane.

371

I want to be clear.

372

I don't own any memory or storage place, but I have to say your boy is salivating.

373

The only thing that gets me more excited about that is you checking out reinvestone .com

374

We did have a great options trade that we closed out on a sold put today,

375

which that was up about 40 % So we're really excited about that.

376

I'd love to see you there.

377

And once you join, you get that lifetime access.

378

If you like that video, check this one out.

379

I think you're going to love it.

380

Why not advertise these things that you told us here?

381

I feel like nobody else knows about this.

382

We'll try a little advertising and see how it goes.

383

Congratulations, man.

384

You have done so much.

385

People love you.

386

People look up to you.

387

Kevin Paffrath there, financial analyst and YouTuber.

388

Meet Kevin.

389

Always great to get your take.

アプリをダウンロード

流暢に話すために必要なすべて

AI発音採点文ごとに採点

IPA練習すべての音をマスター

単語学習自分の単語帳を作る

単語ゲーム遊びながら覚える

QRをスキャン

Download on theApp Store Get it onGoogle Play

FOR YOU

このレッスンの語彙とスピーキングのポイント

このスピーキングレッスンは、動画「They JUST Accidentally Revealed the Truth: The AI Bubble」を教材にしています。この動画には、シャドーイング用の文が389文、単語が4527語あります。音声の長さは25:59です。話す速さは1分あたり約174語で、日常会話に近い自然なペースです。単語の82%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
earnings 名詞		収入, 稼ぎ
bandwidth 名詞	/ˈbændwɪdθ/	帯域幅
divide 動詞	/dɪˈvaɪd/	分割する, 分ける
cloud 名詞	/ˈklaʊ̯d/	雲, 雲霞
expand 動詞	/ɪkˈspænd/	広げる, 拡大する
fold 動詞	/ˈfəʊld/	折る, 畳む
chart 名詞	/t͡ʃɑɹt/	チャート, 図表
custom 名詞	/ˈkʌstəm/	習慣, 風俗
pitch 名詞	/pɪt͡ʃ/	投球
insane 形容詞	/ɪnˈseɪn/	気が狂った, 狂しい
terminal 名詞	/ˈtɚmɪnəl/	ターミナル, エアターミナル
umbrella 名詞	/ʌmˈbɹɛl.ə/	傘, 蝙蝠傘
shareholder 名詞	/ˈʃɛəˌhəʊl.də(ɹ)/	株主
sudden 形容詞	/ˈsʌd.ən/	突然の
lifetime 名詞	/ˈlaɪf.taɪm/	生涯, 一生

動画に出てくる句動詞

単語	発音	意味
go up 動詞		上がる, 登る
slow down 動詞		遅らせる
shut down 動詞		シャットダウン
sold out 形容詞		売り切れ
blow up 動詞		爆破する, 爆発する
check out 動詞		試験する, 検査する
fill out 動詞	/fɪl ˈaʊt/	書き込む, 記入する
go ahead 動詞	/ˌɡoʊ əˈhɛd/	どうぞお先に

繰り返し練習したい文

動画に出てくる短く完結した文です。日常会話でそのまま使えます。

What do you want me to do today, boy?
You want me to do that?
What you don't want is stranded memory.
What do you need all the high bandwidth memory for?
Hey, you wanna buy real estate?

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
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	I've been · they've gotten · there's been
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	can be sustained · could be left · are called
関係詞節 who / which + 節 — 人や物について情報を加える	GPU, which is · well, which is · SSD, which is

注意したい発音

話し手はwe're, you're, they'reなど、短縮形や弱形を125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。

「th」の音: bandwidth /ˈbændwɪdθ/, mathematics /mæθ(.ə)ˈmæt.ɪks/, thesis /ˈθisɪs/, strengthen /ˈstɹɛŋ(k)θən/, anthropic /ænˈθɹɒp.ɪk/
「sh」と「zh」の音: shareholder /ˈʃɛəˌhəʊl.də(ɹ)/, potentially /pəˈtɛnʃ(ə)li/, ratio /ˈɹeɪ.ʃoʊ/, crush /kɹʌʃ/, artificial /ˌɑː.tɪˈfɪʃ.əl/
長い単語（アクセントの位置に注意）: potentially /pəˈtɛnʃ(ə)li/, expenditure /ɛkˈspɛndɪt͡ʃə/, necessarily /ˌnɛs.əˈsɛɹ.ə.li/, artificial /ˌɑː.tɪˈfɪʃ.əl/, momentum /ˌmə(ʊ)ˈmɛntəm/

日本語話者が苦手な音：

/r/ と /l/ の区別 — /l/ は舌先を歯茎につけ、/r/ はどこにもつけない: umbrella /ʌmˈbɹɛl.ə/, shareholder /ˈʃɛəˌhəʊl.də(ɹ)/, rapidly /ˈɹæpɪdli/, necessarily /ˌnɛs.əˈsɛɹ.ə.li/, trail /tɹeɪl/
/v/ — /b/ にならないように、上の歯を下唇に当てる: divide /dɪˈvaɪd/, invest /ɪnˈvɛst/, valuation /ˌvæ.ljuːˈeɪ.ʃən/, impressive /ɪmˈpɹɛsɪv/, server /ˈsɝvɚ/
/f/ — 「フ」ではなく、上の歯と下唇で出す: fold /ˈfəʊld/, lifetime /ˈlaɪf.taɪm/, fewer /ˈfjuː.ɚ/, artificial /ˌɑː.tɪˈfɪʃ.əl/, cliff /klɪf/

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。earnings, bandwidth, divide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: They JUST Accidentally Revealed the Truth: The AI Bubble - 動画で英語スピーキングを学ぶ

練習モードを選ぶ

アプリをダウンロード

おすすめ動画

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

Why Memory Prices Are Crazy High

The worst week in cybersecurity history... so far

The Investment Committee debates whether the AI trade is in trouble

We Let $200 GPT-Pro Think for 7 Days on Unsolved Math

China just dropped a new frontier model... and nobody can explain how

Why AI Tokens are so Expensive - Computerphile

How AI is helping brands identify consumer trends faster

Expensive RTX 5090 for LLMs? NO. Use This Instead. (Tesla V100 SXM2 + Z8 G4, #RACERRRZ)

I read Anthropic's 92-page risk factor document, so you don't have to | Explainer

The Only SaaS Moats Left in 2026

Why The AI Boom Is Reshuffling The Global Stock Market Hierarchy

このレッスンの語彙とスピーキングのポイント

この動画の重要語彙

動画に出てくる句動詞

繰り返し練習したい文

この動画の文法

注意したい発音

この動画での練習方法

シャドーイングとは？英語上達に効果的な理由