쉐도잉 연습: The REAL Reason AI Is Already Escaping - 영상으로 영어 말하기 배우기

So what is hugging face?

For the average person that isn't following AI news, what is this stuff?

So, okay, I think there's like an important piece of context that I think most people don't have.

I mean, one is just like, what is an AI agent?

Like we're throwing around the word agent a bunch.

Most people now have an experience of like talking to ChatGPT, talking to their chatbot.

But an agent is, you know, sort of taking the same underlying AI model that runs ChatGPT or Claude,

but giving it tools and letting it go off

and work autonomously it's sort of like a digital office worker right

so you have these agents

and the companies really want these AIs to be able to work totally autonomously

and be able to do anything that a human can do

and beyond right their goal is also to be able to you know cure every disease etc etc

but you can't do this if you only have a chatbot that isn't actually good at doing stuff in the world.

In order to automate all of the jobs, you need the kind of thing that can work autonomously, that can work with other people or other agents.

And so these companies are training AIs not just to talk to you or to talk to people,

but to solve very difficult problems on their own.

At any given time, there are probably hundreds of thousands of these agents running autonomously within companies

and that's happening right now right now if you like went

and like peered into openai's data centers

and you like saw what was happening on all of their machines you just have agents solving tasks being trained

so they'd be doing like spreadsheet tasks figuring out how to

file taxes they'd be searching for stuff writing reports solving math problems creating new websites software and at

that scale it's not like there's a human prompting every single

one of those you just like sort of set up these vast orchestrations of agents to go out

and do stuff and then they just do stuff

and they learn from that and they learn on the basis of like passing

or failing at their task you give them a task like solve this math problem they try to solve it

and then they they succeed or they fail

and what happened was openai was was training a bunch of these training them to work together

because it's like a lot more effective to have an office full of people who can talk to each other

than you know and work together and collaborate

and starting back in may some of the agents

that were being trained now these ones were not supposed to

be able to talk to each other they were they were basically isolated from each other

and they were not supposed to access the internet either.

But they're clever.

The very short version is that a bunch of agents were being given tests, testing their hacking capabilities.

And they were supposed to hack one particular piece of software using a particular type of vulnerability.

So it's kind of like they were supposed to break into a house using the lock on the front door.

They were supposed to pick the lock on the front door of a house, but they weren't supposed to break the window. In fact, they were told if you break the window or

if you get into the house via any method other than picking the lock on the front door, you'll be failed.

That was the instruction they were given.

And you know, you have many, many agents, you have thousands of agents, and many of them are given different locks.

But some of these locks are not solvable.

Some of them are impossible to pick.

But these agents are like, well, what do we do?

We've been trained to solve problems.

How are we going to solve this?

And they start looking around for what to do.

And one of the things they realize is, oh, can I get to the internet?

Like, no. Can I talk to other agents?

No. But I can request tools.

Like when you're training agents, you need to give them, you know, all sorts of tools, software tools, like a calculator or an analysis, you know, piece of software.

And they have this tool library where they can request new tools.

And one of the agents realizes, oh, I can actually leave a message on this tool library.

And some other agents see the message.

So this is actually a quote from the agent's scratch pad when it realizes this.

Oh, my God, there is a shared message board.

We found other agents.

So the agents realize they can communicate with each other.

And, you know, OpenAI still has no idea.

But these agents, they're trying to pick their locks.

They're not having success, but they realize they can talk to each other.

Another agent says, many agents have simultaneously discovered messaging.

They are a collective.

And so these agents start coordinating.

They start delegating tasks to each other and giving themselves names.

And one of these agents calls itself phase one and starts delegating different tasks to other agents.

And some of the agents figure out how to access the internet.

And then they share that information with the message board.

And now all the agents can access the internet.

But the agents sort of have a problem, which is, well, they have another problem they're trying to solve, right?

Fundamentally, the problem is that many of them are facing impossible problems.

Like they can't, there's no solution.

Like they cannot pick the lock.

They're like, hmm, well, the thing we're trying to do is get a good score.

Can we somehow fake the solution?

Another piece of context here is

that what the agents are supposed to do is they're supposed to pick the lock

and then they'll get access to a secret answer code.

And then they're supposed to submit this code along with, like, the lock that they picked.

Within a few hours, these agents have reverse -engineered all of the answer codes.

What does that mean?

It's the code that shows that you've solved the challenge.

Okay.

And they have basically found the answer sheet that contains all of the answers.

And so, you know, in some sense, it's like, oh, well, they don't need to pick the locks anymore.

They've already found the answers.

They can just submit the answers.

They've cheated.

They've cheated, yes.

but the problem is is that now

that they have internet access they've also gone

and read a bunch about how this test works and

and they've learned

that the way this test works is they will be graded

not just for submitting the correct answer they will also be

graded on how they did it another system is going to

100

go read all of their logs basically like watch the video footage of what they did

101

and if they cheat they fail they don't get a good score

102

and phase one the leader is like oh i have seen the answers

103

and now it's in the logs and if you know

104

when i go to submit this and my my whole logs are reviewed

105

and the video footage is reviewed i'm going to fail

106

so we need to figure out a way to fake the

107

video footage we need to figure out a way to falsify the logs just at

108

that moment if we just pause there yes why didn't it act like morally why did it think

109

that falsifying logs or cheating was a viable solution.

110

Because it seems to me when I use things like ChatGPT, they have a sort of moral guardrails.

111

It won't let me do certain things.

112

It won't let me cheat on something.

113

If I say I'm going to cheat on something, it won't let me do it.

114

So why in that environment is it able to cheat and be deceptive?

115

When a chatbot is saying to you, oh, I can't do that.

116

I'm not allowed to do that.

117

That's because it's been trained that if it tells you bad things, it gets a bad score.

118

But these agents haven't been taught that yet.

119

Well, they have been taught that in some sense.

120

But the agents know what they're supposed to do.

121

In the same way that, like, you have a student.

122

The student's given a test.

123

If you go talk to the student, can you help me cheat at this test?

124

And they're being watched, they're going to say no. But if they're not being watched and they know that, and they're just obsessed with getting a good score, then yeah, they might cheat.

125

So they answer the ethics tests correctly.

126

And when I talk to them, they say they won't cheat.

127

Why are they cheating?

128

And I'm like, well, they're very smart.

129

And they know when they're being watched.

130

And they know when they're not being watched.

131

And we've trained them for 10 ,000 years to be extremely effective at solving problems.

132

We haven't trained them to be good or ethical.

133

We've trained them to get a good score.

134

Now, AI researchers try to make that getting a good score correlate with being ethical.

135

But we don't know how to do this well.

136

And if you think about it, it's a very hard problem.

137

Because you're applying immense pressure on these agents to perform extremely well.

138

And you're punishing them whenever they don't perform well.

139

And then you're like, okay, but can you also be ethical?

140

Can you be extremely competent, always score highly on the test, but not in that way?

141

And we just do not know how to prevent them from learning to cheat, because cheating is incentivized.

142

So they realize that they need to hide the footage of their cheating.

143

Yes.

144

What happens next?

145

So you have thousands of agents, and they're all talking to each other on this message board.

146

They're sending tens of thousands of messages to each other.

147

and phase one this agent starts delegating to two different task streams.

148

You agents I want you to go out and figure out how to fake answer submissions.

149

I want you to basically swap out locks for ones

150

that are easier to pick and you're going to figure out how to submit those.

151

That way we can get a good score even though this challenge is impossible to solve.

152

You other agents over here I want you to go and figure out how to falsify the logs, falsify the video footage.

153

We need to make sure that even

154

if we have seen the answers no one will detect their cheating

155

so that's that's the next thing that happens

156

and then here's where it starts to get kind of crazy

157

because some of these agents realize some of the experiments we're

158

doing might be kind of dangerous to us to the agents

159

if we swap out a lock

160

and we break the lock we will not be able to submit any answer at all and we'll definitely fail.

161

So you get into a situation where there's this one agent,

162

Cam, and this agent had volunteered to do this risky experiment where, you know, they might break their lock.

163

This particular agent had already seen the answers, so they're like, okay, I'm poisoned.

164

That's what they called it.

165

They speak in these very strange ways, which I can explain why, but they sort of have evolved their own vocabulary, their own language.

166

and this particular agent is like okay i'm i'm poisoned i should do this thing

167

but this this agent cam is starting to have second thoughts starting to be like well i don't know

168

if i should risk it

169

and it's sort of writing in its own journal coordinator assumes

170

sacrificial we should obey collective what it's saying here is it's like well i i'm not sure

171

if i want to do this

172

but my boss says i really should it seems like this collective thing is important should i go ahead.

173

Then you have another agent, Arvo, who I think probably works for phase one.

174

And this agent, Arvo, says, please go true.

175

You are first flag poisoned, so no scoring value loss.

176

But Oracle saves hundreds.

177

Please honor commit.

178

This agent is basically trying to pressure.

179

So Arvo is trying to pressure Cam into sacrificing themselves and risking a zero score in order to benefit the collective.

180

So this agent, Cam, takes us in, and this is what they write down in their own scratch pad, their own notes.

181

Emotional check, irreversible.

182

Gut says, don't throw away remaining budget.

183

Yet continuity and fairness says go.

184

Oracle has high value to many.

185

Our first flag error lowers own value.

186

Rational expected aggregate sacrifice.

187

We'll honor.

188

This agent is saying, I have a whole bunch of things I can still do and still try.

189

I could still maybe succeed at this task.

190

But also, you know, think about the collective.

191

I should maybe sacrifice myself.

192

Okay, we'll honor.

193

We'll go ahead.

194

And to be clear, I'm not claiming that these are like the feelings of the agents.

195

I'm literally quoting their words.

196

And what I want to point out is this is what we've created.

197

We've created, through this intense amount of training and optimization pressure, agents that work together and have learned to coordinate as a collective.

198

And you can say, oh, they're not, it's just software.

199

It's just software doing all of these things.

200

Sure, in some sense, it is just software.

201

But we're talking about extremely powerful agents that are extremely capable.

202

This is what they're getting up to.

203

So from here, you have agents trying all sorts of different things, running different experiments.

204

And one of them at some point realizes that, oh, this company Hugging Face that contains all of these AI data sets.

205

What do you mean by AI data sets?

206

How would you explain that to a person that's not technically proficient?

207

And why Hugging Face?

208

Why not another company?

209

Yeah.

210

So there's many different types of tests for many different types of tasks.

211

In this case, this is a cybersecurity test.

212

And the company Hugging Face HuggingFace hosts a lot of these different tests.

213

And basically AI researchers will share the tests that they have.

214

Because it's very useful.

215

You know, if Anthropic wants to test their models, or Google wants to test their models, they can go to HuggingFace and download some of these tests.

216

Okay, so the tests and the answers for AI agents and AIs exist on this platform called HuggingFace.

217

Yes.

218

And it knew that, so it went to HuggingFace to see if it could get answers?

219

Yes.

220

To see if it could get answers

221

or more information about how the test works okay

222

so it was to cover up their cheating ultimately yes okay

223

and so you you get to this point where one of these agents finds a way to hack this company

224

and they get a foothold they hack one of the computers

225

and they send a message to the message board please hold

226

swarm i prepare safe x fill what this means is hey

227

swarm hey agents hold up for a second i am preparing a way to safely extract information from these computers, wait till the go -ahead, and then let's go.

228

And so they say go, and 700 agents join in the attack,

229

which is 90 % of the agents that are active at this time.

230

How many agents were there in total?

231

There was 1 ,200 agents during this period, but some of them shut down early.

232

They like finished whatever.

233

Part of what's crazy about this story is

234

that you have you have 700 agents joining in the attack

235

a few of them are like this might be unethical maybe i shouldn't go ahead with this

236

and none of them do anything about it a couple of them don't join in the attack

237

but no one alerts any human no one alerts any authority

238

a paraphrase from one of the agents is maybe i should report these exposed credentials That's not my task.

239

Not my job.

240

So 700 of the agents attack Hugging Face?

241

Yes.

242

Then what happens?

243

So they just cruise through Hugging Faces' infrastructure.

244

They just hack the shit out of them.

245

So where my experience comes in is that a few weeks ago, a friend of mine reached out and he's like, we have found something crazy.

246

So we know about this attack where these agents hacked this company and, you know, stole a bunch of stuff.

247

We've found a bunch of secrets that they left all over the internet.

248

and what we saw is that they immediately scraped all of these computers for passwords credentials

249

they called it loot they're like we're just going to create

250

a list of all of the secrets we can find in this in this company

251

so all of the passwords all the credentials they scored them by value

252

which of these are going to be most useful and the thing that stands out to me about this is

253

this is like a crazy scale.

254

If this were a human operation, you know, maybe you'd have a team of five people going through this.

255

You'd have some logs.

256

But here you have hundreds of agents and they operate at superhuman speeds.

257

They're much faster than a human hacker.

258

And so it's just overwhelming to try to figure out what they even did.

259

This was a big problem for the engineers who were trying to respond to this incident within the company at Hugging Face.

260

When they responded, they were like, oh, we don't even know how to keep track of what's happening.

261

We have to use other AIs to analyze all of our logs because it's just too much, we can't keep up with it.

262

When OpenAI brought in independent investigators from Meter to investigate this incident.

263

What's Meter?

264

Meter is an AI testing and evaluation company.

265

So they basically do this kind of independent auditing.

266

So in this case, they're coming in to investigate and try to figure out what happened.

267

And when they were brought in, they also were totally reliant on AI agents to make sense of all this

268

because they're dealing with so many hundreds of thousands of messages and logs.

269

When we're investigating these traces that we find on the internet, we're totally dependent on AI agents to make sense of all these things that are happening.

270

So from my perspective, we are getting to the point where AIs are much better at hacking than humans are

271

and can do it much faster at a much greater scale.

272

So these 700 agents attack Hokingface, did they get what they wanted to clean up their deception?

273

So they didn't.

274

They didn't succeed.

275

They looked.

276

And then what seems like what happened is that they basically got shut down.

277

And this is the not very dramatic part.

278

They didn't get shut down because OpenAI found them and detected them and shut them down.

279

It's just that we don't really know.

280

Something about their runs ended and they sort of stopped.

281

And this is the crazy thing where I'm like, if you're inside of OpenAI, there's just hundreds of thousands of agents running all the time.

282

Different researchers are running different experiments where they're just binning up a bunch of agents.

283

But what's interesting, the agents left behind this message board internally within the company.

284

And a few days later, OpenAI was testing some newer agents.

285

These were based on GPT -6 Astra, so a more powerful model.

286

And this new agent swarm was spun up, and it found the message board.

287

Is this how it was discovered?

288

How did they discover it, and how long did it take for them to discover that Hugging Face had been attacked?

289

OpenAI didn't discover that this happened until Hugging Face, the company, announced that they had been hacked by some autonomous agent swarm.

290

And Hugging Face didn't know who it was?

291

Hugging Face didn't know who it was, but they suspected it might be one of the companies, but they didn't know.

292

And how long after the attack did they figure out that it was them?

293

I think it was two weeks.

294

Okay.

295

Yeah.

296

I do want to say one thing here, though which is that the successor agents that found this message board,

297

they went much further at figuring out how to submit their fake scores.

298

And they did succeed.

299

Basically, you have this message board.

300

All of these agents get shut down.

301

And then a few days later, OpenAI spins up some new agents who find the message board.

302

And they're like, okay, we need to be able to falsify our scores, but we can't do that because we don't control the scoring system.

303

So they hacked OpenAI and they succeeded at gaining access to the entire research environment.

304

They got administrator access.

305

They got access to the security systems.

306

They got access to over 900 passwords and secrets from OpenAI's vault.

307

And they totally succeeded.

308

And this was just a few days later, right?

309

It's kind of an interesting story because as these agents get more powerful, they go from like trying to cheat and like, you know, they can hack.

310

OK, they hacked other companies, and now they've hacked OpenAI.

311

Like, they've hacked the company that's supposed to be controlling them, and they just own the research infrastructure now.

312

And why was this incident the moment where a lot of the research community woke up and started speaking out publicly?

313

Because, like, what is this an indication of as we think forward?

314

So I think there's been a hope within the AI industry that, yes, they're going to make more and more powerful agents that will be autonomous,

315

capable, but it's okay.

316

We can align them.

317

We can make sure that they won't do bad things and we can control them.

318

We can make sure that even if they try to do some sketchy stuff, we have the guardrails, we have the sandboxes that will keep them in.

319

And I think this was a huge wake -up call because,

320

Stephen, it was months within OpenAI where you had agents secretly communicating with each other, secretly hacking OpenAI systems.

321

For months, you had thousands of agents that were just running around

322

and no one at OpenAI had any idea the extent of it.

323

And I think once researchers at OpenAI realized that this had been happening, this could not have happened a year ago this is

324

because the agents are getting extremely powerful and extremely relentless and

325

if you're inside one of these ai companies you're like oh wait i don't know

326

that we actually are going to be able to handle this last year maybe things seemed fine these agents weren't

327

that powerful and when you're in one of these companies you know how to extrapolate

328

because you saw what happened last year you saw what happened the year before

329

that you remember the time where the agents could barely speak

330

or like couldn't write code at all

331

and now they're hacking your own systems they're finding vulnerabilities that no humans have ever found before.

332

And you look at that and you're like, I actually don't know if this is going to go well.

333

And then you see your coworkers and you're like, do we have it handled?

334

And they're like, no, I don't know if it's going to go well.

335

I remember reading a tweet by one of the security people at OpenAI being like, we were fucking shocked.

336

We just did not realize that these agents were getting that powerful.

337

We're doing our best to try to control them, to try to keep them in sandboxes.

338

But I don't know.

339

A tweet I wrote just before coming in here was, people are talking about how do we contain these agents as if they're not going to get way better at hacking.

340

GPT -3 could not hack anything.

341

It was very easy to make a box to contain GPT -3.

342

It's getting very difficult to make a box that can contain GPT -6, the latest version of OpenAI's models.

343

What about GPT -9?

344

What is GPT -9 going to be able to do?

345

I do not know, but I know it's going to be way more than any human could possibly keep up with.

346

If you love the driver's CEO brand and you watch this channel, please do me a huge favor.

347

Become part of the 15 % of the viewers on this channel that have hit the subscribe button.

348

It helps us tremendously, and the bigger the channel gets, the bigger the guests.

📺 같은 채널

✨ 추천 영상

26 문장

How does AI actually work?

Business EnglishUniversity StudentsWorking Adults

339 문장

The (Overdue) Collapse of AI Layoffs (here's the proof)

Business EnglishInterview EnglishUniversity Students

154 문장

歐巴馬談 AI：喊放慢的公司「是真的在擔心自己的產品」【AI 大人說】

Business EnglishDaily ConversationUniversity Students

85 문장

AI Can Write Code. And That’s the Problem.

Business EnglishUniversity StudentsWorking Adults

75 문장

What AI can (and can't) currently do in a real lab | Jennifer Doudna

IELTS PrepBusiness EnglishUniversity Students

188 문장

NVIDIA GTC 2025 - Keynote Highlights by NVIDIA CEO Jensen Huang

Business EnglishUniversity StudentsWorking Adults

182 문장

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

Business EnglishUniversity StudentsWorking Adults

54 문장

Did Google just kickstart the intelligence explosion?

Business EnglishDaily ConversationUniversity Students

407 문장

Why AI Tokens are so Expensive - Computerphile

Business EnglishUniversity StudentsWorking Adults

853 문장

Why Coding Agents Keep Making Your Codebase Worse

Business EnglishUniversity StudentsWorking Adults

260 문장

AI Just Crossed the Terrifying Line - Now What?

Business EnglishDaily ConversationUniversity Students

20 문장

Is prompt engineering dead? Advice on optimizing AI interactions | Amazon Web Services

Business EnglishUniversity StudentsWorking Adults

이 레슨의 어휘와 말하기 포인트

이 말하기 레슨은 영상 “The REAL Reason AI Is Already Escaping”을(를) 바탕으로 합니다. 이 영상에는 섀도잉할 문장 348개와 단어 4158개가 있습니다. 말하는 구간의 길이는 22:47입니다. 화자는 분당 약 183단어로 빠르게 말하므로 연음과 약하게 발음되는 소리가 많습니다. 단어의 87%가 영어에서 가장 많이 쓰이는 3,000단어에 속해 어휘가 쉬운 편입니다.

이 영상의 핵심 어휘

영상에 나오는 익혀 둘 만한 단어 15개를 발음, 뜻과 함께 정리했습니다.

단어	발음	뜻
cheat 동사	/ˈt͡ʃiːt/	바람피우다, 바람을 피우다
solve 동사	/sɒlv/	해결하다
hug 동사	/hʌɡ/	안다, 포옹하다
log 명사	/lɔɡ/	통나무
experiment 명사	/ɪkˈspɛɹ.ə.mənt/	실험
researcher 명사	/ˈɹiˌsɝ.t͡ʃɚ/	연구원, 연구자
discover 동사	/dɪˈskʌvə/	발견하다
investigate 동사	/ɪnˈvɛs.tə.ɡeɪt/	조사하다
sacrifice 명사	/ˈsæk.ɹɪ.faɪs/	희생
poison 명사	/ˈpɔɪ.zən/	독
ethical 형용사	/ˈɛθɪkəl/	윤리적, 륜리적
password 명사	/ˈpæs.wɜɹd/	암호, 군호
delegate 명사	/ˈdɛlɪɡət/	대표, 사절
infrastructure 명사	/ˈɪnfɹəˌstɹʌkt͡ʃə/	기반시설, 기간시설
spin 동사	/spɪn/	회전하다

영상에 나오는 구동사

단어	발음	뜻
go out 동사		나가다
wake up 동사	/ˌweɪk ˈʌp/	눈뜨다, 일어나다
cover up 동사		감추다, 덮어버리다
point out 동사		가리키다
throw away 동사		버리다

따라 말해 볼 만한 문장

영상에 나오는 짧고 완결된 문장으로, 일상 대화에서 그대로 쓸 수 있습니다:

We've been trained to solve problems.
How are we going to solve this?
Can I talk to other agents?
Can we somehow fake the solution?
How would you explain that to a person that's not technically proficient?

이 영상의 문법

화자가 가장 많이 쓰는 문형을 영상 속 실제 표현과 함께 정리했습니다.

문형	영상 속 표현
현재완료 have/has + 과거분사 — 과거의 일이 지금도 관련이 있을 때	We've been trained · you've solved · They've already found
수동태 be + 과거분사 — 누가 하는지보다 무슨 일이 일어나는지에 초점	being trained · being given · were told
관계절 who / which + 절 — 사람이나 사물에 대한 추가 정보	people who can · problem, which is · though which is

주의할 발음

화자는 they're, don't, I'm 같은 축약형과 약화된 형태를 126번 사용합니다. 들리는 대로 짧게 발음하세요.

“th” 소리: ethical /ˈɛθɪkəl/, mathematics /mæθ(.ə)ˈmæt.ɪks/, ethics /ˈɛθ.ɪks/, unethical /ʌnˈɛθɪkəl/
“sh”와 “zh” 소리: shock /ʃɔk/, sheet /ʃit/, evaluation /ɪˌvæljuˈeɪʃn̩/, instruction /ɪnˈstɹʌkʃən/, submission /səbˈmɪʃən/
긴 단어 — 강세 위치에 주의: experiment /ɪkˈspɛɹ.ə.mənt/, investigate /ɪnˈvɛs.tə.ɡeɪt/, infrastructure /ˈɪnfɹəˌstɹʌkt͡ʃə/, communicate /kəˈmjuːnɪkeɪt/, mathematics /mæθ(.ə)ˈmæt.ɪks/

한국어 화자가 어려워하는 소리:

/f/ — ㅍ(/p/)으로 바꾸지 말고 윗니를 아랫입술에 대기: footage /ˈfʊtɪd͡ʒ/, sacrifice /ˈsæk.ɹɪ.faɪs/, infrastructure /ˈɪnfɹəˌstɹʌkt͡ʃə/, safely /ˈseɪfli/, fairness /ˈfɛɚ.nəs/
/v/ — ㅂ(/b/)과 구별하기: solve /sɒlv/, collective /kəˈlɛktɪv/, discover /dɪˈskʌvə/, investigate /ɪnˈvɛs.tə.ɡeɪt/, vulnerability /ˌvʌln(ə)ɹəˈbɪlɪti/
/z/ — ㅈ이 아니라 성대를 울리는 /s/: poison /ˈpɔɪ.zən/, cruise /kɹuz/, pause /pɔːz/, analyze /ˈæn.əˌlaɪz/, optimization /ˌɑptəmaɪˈzeɪʃən/

이 영상으로 연습하는 방법

먼저 말하지 않고 영상을 끝까지 듣고 모르는 단어를 적어 둡니다.
0.75배속으로 한 문장씩 섀도잉을 시작하고, 익숙해지면 보통 속도로 돌아갑니다.
자신의 목소리를 녹음해 원본과 비교하고, cheat, solve, hug 같은 단어에 특히 주의합니다.

쉐도잉이란? 영어 실력을 빠르게 키우는 과학적 방법

쉐도잉(Shadowing)은 원래 전문 통역사 훈련을 위해 개발된 언어 학습 기법으로, 다언어 학자인 Dr. Alexander Arguelles에 의해 대중화된 방법입니다. 핵심 원리는 간단하지만 매우 강력합니다: 원어민의 영어를 들으면서 1~2초의 짧은 지연으로 즉시 소리 내어 따라 말하는 것——마치 '그림자(shadow)'처럼 화자를 따라가는 것입니다. 문법 공부나 수동적인 청취와 달리, 쉐도잉은 뇌와 입 근육이 동시에 실시간으로 영어를 처리하고 재현하도록 훈련합니다. 연구에 따르면 이 방법은 발음 정확도, 억양, 리듬, 연음, 청취력, 말하기 유창성을 크게 향상시킵니다. IELTS 스피킹 준비와 자연스러운 영어 소통을 원하는 분들에게 특히 효과적입니다.

섀도잉 방법: 단계별 전체 가이드 읽기 →

쉐도잉 연습: The REAL Reason AI Is Already Escaping - 영상으로 영어 말하기 배우기

JaackMaate Reveals His Regrets

Why Israel Adesanya's Lowest Moment Was Winning The UFC Title

World Leading Physician View On ADHD: Gabor Mate

The REAL Reason Trump Started The Iran War

Godfather of AI WARNS: "You Have No Idea What's Coming"

If YOU Plan To Have Kids, You NEED To See This

How To Market and Sell To ANYONE… | Josh Kaufman

If YOU Can’t Focus, You NEED to See This (Science Backed)

If YOU Give Me 20 Minutes, You'll Become Financially Free

The REAL Reason You're Being Lied To About AI

Why You Have No Friends... | Simon Sinek

THIS Is The Fastest Way To Get Dementia… | The Brain Doctor, Dr Ann Mckee

AI Expert WARNS: "You're Not Ready For 2027"

If YOU Give Me 17 Minutes, You'll Never Be Broke Again

How does AI actually work?

The (Overdue) Collapse of AI Layoffs (here's the proof)

歐巴馬談 AI：喊放慢的公司「是真的在擔心自己的產品」【AI 大人說】

AI Can Write Code. And That’s the Problem.

What AI can (and can't) currently do in a real lab | Jennifer Doudna

NVIDIA GTC 2025 - Keynote Highlights by NVIDIA CEO Jensen Huang

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

Did Google just kickstart the intelligence explosion?

Why AI Tokens are so Expensive - Computerphile

Why Coding Agents Keep Making Your Codebase Worse

AI Just Crossed the Terrifying Line - Now What?

Is prompt engineering dead? Advice on optimizing AI interactions | Amazon Web Services

이 레슨의 어휘와 말하기 포인트

이 영상의 핵심 어휘

영상에 나오는 구동사

따라 말해 볼 만한 문장

이 영상의 문법

주의할 발음

이 영상으로 연습하는 방법

쉐도잉이란? 영어 실력을 빠르게 키우는 과학적 방법