쉐도잉 연습: You’re Not Behind (Yet): Learn AI Agents in 13 Minutes - YouTube로 영어 말하기 배우기
C1
Most people think they're using AI well when they get a decent answer from ChatGPT.
241 문장
문장이 너무 짧거나 길면 Edit를 눌러 조정하세요.
1
Most people think they're using AI well when they get a decent answer from ChatGPT.
2
That was enough six months ago.
3
It's not enough anymore.
4
The next shift is AI agents and the gap between people who understand them and people who don't,
5
it's about to get very expensive.
6
I've spent years in the boardrooms of billion-dollar companies
7
and the good news here is that agents are much simpler than most people think.
8
So in this video, I'll show you exactly how they work,
9
when to use them, and how to start before everyone else catches up.
10
An AI prompt and an AI agent are completely different.
11
But most of us are still stuck with old habits.
12
Let me give you the simplest way to think about this before we go any further.
13
This is our first framework right off the bat.
14
I call it ARR.
15
If a task is autonomous,
16
recurring, and reviewable, it's a strong candidate for an agent.
17
If it needs live judgment,
18
or it only happens once,
19
or can't be reviewed clearly, then use a prompt.
20
That one distinction alone will put you miles ahead of most people using AI today.
21
The internet gave us search, so we started googling.
22
AI gave us LLMs, or large language models.
23
But most people today still think of AI as a glorified search.
24
Now agents are here and we're still making the same mistake again.
25
We think of agents as just more capable chatbots.
26
They're not.
27
A chatbot waits for your next prompt.
28
An agent figures out its next move.
29
Prompting is like sitting next to a student driver.
30
You still have to guide them,
31
correct them, and stay very alert.
32
An agent, on the other hand is a hired driver you set the destination hand over the keys
33
and just sit in the back seat it handles the route the traffic
34
and all the step-by-step decisions that's the mental shift we have to make
35
so here's a prompt write me a linkedin post
36
and here's an agent watch my industry every monday find the
37
three most relevant stories study my previous posts draft the new post based on those stories in my voice,
38
revise against my style, and schedule it for Tuesday morning.
39
That's the power of AI agents.
40
And to wield that power,
41
you need to know what's actually running under the hood.
42
Everyone's talking about AI agents.
43
Almost nobody can tell you what's actually happening inside.
44
A chatbot predicts the next word.
45
An agent decides the next action.
46
Here's how a chatbot actually works.
47
It's a large language model.
48
When you type a question,
49
it's going to break that question into small units of words called tokens,
50
and it converts them into numbers.
51
And then it just finishes the sentence.
52
So if you said, Jack fell down and broke his crown,
53
now you know that, but LLM does not know that rhyme.
54
It will know words like bones and heart and crown,
55
and all of those words could make sense.
56
But based on its training,
57
it predicts that the most likely next word given that line is crown.
58
It's based on probabilities.
59
The agent has the same language model in the center,
60
but now there are four workers around it.
61
Analyst, planner, operator, auditor.
62
One finds the pattern, one decides the plan,
63
one does the work, one checks the results.
64
Let's make this real.
65
You can give an agent some instruction like every Monday at 7 a.m.,
66
review the past week's customer support tickets,
67
sales notes, and product feedback,
68
identify three biggest recurring issues,
69
summarize what changed, and email my leadership team a one page weekly brief.
70
That's it.
71
I mean, those are lots of steps,
72
but the agent will read the tickets,
73
notes, and feedback, and find the pattern.
74
Analyst.
75
It will then decide what matters most and what belongs in the brief.
76
Planner.
77
It will write and send the update.
78
Operator.
79
And then it'll check for weak logic,
80
missing context, or sloppy conclusions, and it'll refine it.
81
Auditor.
82
Now, by Monday morning, the brief is in your team's inbox.
83
You did not write the report.
84
You did not analyze it.
85
You just assigned the job of four people to one agent.
86
So here's your move.
87
Tonight, open ChatGPT agent mode and give it one recurring task,
88
but then watch what it does.
89
You'll see all four workers show up
90
in real time that's the anatomy of an ai agent now
91
that you understand the parts let's look at the entire loop the best thing about agents is
92
that they can adapt when things go wrong this is what makes agents genuinely different from everything
93
that came before it in the 1970s there was an air force colonel john boyd
94
and he studied a very intriguing puzzle from the korean war American pilots in their F-86 kept beating technically superior Soviet MiG.
95
Now the MiG was faster and it could climb higher.
96
It should have won, but it didn't.
97
And Boyd eventually found the difference.
98
The American pilots could see more from their cockpits and they could adapt faster.
99
So they got inside the enemy's decision cycle before the enemy could respond.
100
He called that loop, the OODA loop.
101
Observe, orient, decide, act.
102
And in the world of agents, it's the same thing.
103
That is the real test of an agent.
104
When the obvious path fails,
105
can it choose a better one?
106
Can it go through its own OODA loop?
107
So let me give you a concrete example.
108
You can build an automated workflow every Friday,
109
check this week's grocery prices,
110
build my shopping list, and place the order.
111
It works every Friday until one week your usual item is out of stock
112
and you have six friends coming for dinner on Saturday.
113
So that automated workflow is going to break.
114
Not because it's dumb, but because it's designed to be obedient.
115
It's designed to not think on its own.
116
An agent, on the other hand, does something very unique.
117
It sees the usual list.
118
It sees that that is not working, it finds substitutes.
119
It adjusts quantities for six people.
120
It checks your calendar, sees the dinner,
121
and rebuilds the entire order around it.
122
A workflow can follow the process.
123
An agent can reroute it completely.
124
That's the difference.
125
So when someone says they built an agent, ask one question.
126
When the first path breaks,
127
does the agent keep following the script or can it find a better path?
128
Can it find another way?
129
That's the agent adapting in real time.
130
So this begs the question, right?
131
If agents have autonomy, why do they still fail so often in real life?
132
That's next.
133
The most dangerous thing about AI agents is
134
that they will do wrong things faster and with more confidence than you ever could.
135
An agent is not magic, it's a multiplier.
136
I was working with a board and leadership team of a large consumer company.
137
I'm still working with them.
138
And they're profitable, well-run, great CEO.
139
And when I asked what was stopping them from using AI to drive customer acquisition,
140
for example, the CMO responded,
141
we have all the data,
142
but we'll still need to build a clean process so we can turn that into something useful, something insightful.
143
And I asked where the real challenge was.
144
And she said, you know,
145
we need the right people in the seats first.
146
That's the story everywhere.
147
Most AI problems are human problems in disguise.
148
An agent is just a mirror.
149
It reflects the quality of your thinking back at you.
150
It just amplifies it.
151
Give an agent vague goals,
152
sloppy directions, and no way to get feedback,
153
and it will drive the car straight into the tree,
154
faster and with more confidence than you ever could.
155
Here's the dangerous part.
156
An agent doesn't fix bad thinking, it formalizes it.
157
Usually, the agent fails because the human was vague,
158
not because the underlying model was bad or anything.
159
So, before you automate anything, run a GPS check.
160
Goal, proof, steps.
161
Goal.
162
Can I define the goal in one sentence very clearly?
163
Proof.
164
Can I tell what good looks like?
165
And how will I know if the agent got it right?
166
And steps.
167
Can I describe each and every step very clearly without a lot of hand-waving?
168
Unless you can do those three things very well,
169
your agent is not gonna make any difference.
170
I'll give you an example.
171
Here are two instructions for your agent.
172
First one, summarize my emails every morning.
173
It's good.
174
And here's the second one.
175
Every morning at 7 a.m.,
176
read my unread emails, categorize them by urgency,
177
draft replies to routine messages,
178
and flag anything from my top five customers.
179
So you see, there is a difference between those two instructions,
180
and that gap is exactly where the mess lives.
181
The winners who can wield the power of AI agents aren't just gonna be engineers.
182
They'll be the people who understand their work deeply enough to define it precisely.
183
Most companies want AI everywhere.
184
The ones actually winning are obsessively narrow.
185
If clarity is the bottleneck,
186
then the opportunity is not broad intelligence, it's narrow ownership.
187
I was visiting the customer conference of a construction software company
188
that I work with
189
and the product lead was on stage showing a demo of a single agent
190
that was focused on a very specific problem collecting field data
191
for a specific type of customers in a specific type of situation
192
and it was a beta launch the demo worked mostly with a few minor glitches here
193
and there but
194
when he showed the qr code at the end every hand
195
in the conference room went up with their phones everyone took a picture
196
because it solved a very specific
197
but very real pain they all had been living with for decades
198
that is where the real opportunity is in your career
199
or in your company narrow focus here's the test find a highly specific task people hate doing
200
but they have to do it repeatedly that's where the money is you know
201
we're entering an age where we'll have more agents than human beings on this planet.
202
On the business side, for every software company that exists today,
203
there will be an agent company trying to dethrone it.
204
The winners won't build the broadest agents first.
205
They'll build the one that understands one workflow,
206
one market, and one kind of user pain better than everyone else.
207
By the way, if you If you want to keep this conversation going,
208
I write a short newsletter once a week,
209
just useful ideas, tools, honest reflections.
210
You can subscribe below.
211
It's free.
212
AI will reshape almost every role,
213
but it won't replace what makes you irreplaceable.
214
You know, AI is a giant decoupling machine.
215
For most of our modern history,
216
your income was tied to your hours.
217
at the top, you're always trading time for decisions.
218
Agents are breaking that link for the first time.
219
Now they do the work and you scale your judgment in areas where it matters most.
220
And that changes what's valuable.
221
We're entering an era of infinite output,
222
content, code, and analysis all becoming super cheap.
223
When intelligence becomes that cheap,
224
judgment becomes even more expensive.
225
When output becomes infinite, taste becomes scarce.
226
Every time you define a task clearly enough for an agent to run it,
227
you're not just training the system,
228
you're clarifying your own standards.
229
You're learning what good actually looks like.
230
Sure, some roles will be reshaped.
231
The paralegal, the junior analyst, the coordinator.
232
But every disruption has always created new roles that never existed before.
233
Before the internet, nobody imagined the role of an online community manager.
234
The question is not whether the shift happens.
235
It is whether you shape it or get shaped by it.
236
The most valuable person is no longer the one who can think the fastest.
237
It's the one who can define good work,
238
spot bad work, and know when to trust an agent and when to trust a human.
239
That is where your value is going to move.
240
Ironically, AI will make human life less robotic.
241
Thank you and I love you.
앱 다운로드
당신이 말하는 모든 문장을 AI가 채점
TRENDING
인기 동영상
맥락 및 배경
이번 영상은 AI 에이전트의 활용 방법과 그 중요성에 대해 이야기합니다. 많은 사람들이 ChatGPT와 같은 AI 도구를 사용할 때 충분한 답변을 받았다고 생각하지만, 이제는 AI 에이전트를 이해하지 못하면 큰 손해를 볼 수 있습니다. 강연자는 수년 동안 대기업의 이사회에서 활동해온 경험을 바탕으로, AI 에이전트의 작동 원리와 이를 활용하는 방법을 쉽고 간단하게 설명하고 있습니다. 이 정보를 통해 AI 에이전트의 기능을 잘 이해하고, 다가오는 변화에 대비하는 것의 중요성을 알리고자 합니다.
일상적인 의사소통을 위한 5가지 주요 표현
- 자율적이고 반복적이며 검토 가능한 작업: AI 에이전트의 강력한 후보.
- 실시간 판단이 필요한 경우: 프롬프트를 사용해야 한다는 점.
- 채팅봇과 AI 에이전트의 차이: 후자는 다음 행동을 결정한다는 것.
- 프롬프트 예제: '링크드인 포스트 작성하기.'
- AI 에이전트의 예: 매주 업계를 모니터링하고, 관련된 이야기를 바탕으로 포스트 초안 작성하기.
단계별 쉐도우 스피킹 가이드
이 영상을 통해 영어 회화 연습을 더 효과적으로 할 수 있는 방법을 소개합니다. 쉐도우 스피킹은 원어민의 발음을 따라 하는 연습으로, AI 에이전트의 개념을 바탕으로 활용하면 좋습니다. 다음은 단계별 가이드입니다:
- 영상 시청: 처음에는 전체 영상을 보며 내용의 흐름을 파악합니다. 중요한 개념을 체크하십시오.
- 구간별 반복 시청: 각 섹션을 나누어 집중적으로 들어보세요. 이해가 어려운 부분은 여러 번 반복하세요.
- 쉐도우 스피킹 연습: 원어민의 발음을 따라 해 보세요. 'shadow speak'하여 발음을 교정합니다.
- 메모하기: 중요한 표현과 어휘를 노트에 기록해 두세요. 이는 유튜브 영어 공부에 도움이 됩니다.
- 주기적으로 복습: 일정한 간격으로 배운 내용을 복습하며 영어 발음이 개선되도록 합니다.
이러한 연습을 통해 영어 회화 능력을 향상시키고 AI 에이전트와 같은 기술적 개념을 쉽게 이해할 수 있습니다. 지속적인 연습은 여러분의 영어 실력을 한 단계 끌어올려 줄 것입니다.
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.
☕ 커피 한 잔 사주기
ShadowingEnglish는 여러분의 지원 덕분에 100% 무료로 운영됩니다. 서버 및 AI 유지 비용이 많이 듭니다. 여러분의 커피 한 잔이 큰 힘이 됩니다! ��