You can now teach yourself anything completely for free. That's the power of AI. You can get dangerously self-educated and build your own personal university right about now.

I've been a monk in training, an MIT grad, a tech CEO, and I've seen this one pattern.

The people who unlock real intelligence are almost always relentlessly curious and self-driven. So, today I'm giving you that precise learning system that nobody has shared so far.

I call it the university in a box. Genius name, [music] right? But, it's a way to use AI to become a self-taught expert at almost anything.

You'll be able to learn how to build your own personal curriculum from A to Z.

So, let's get going. Uh a few months ago, I was in Los Angeles, and I had to give a keynote to a class of 120 incoming interns at one of the largest media companies.

And they were all from top universities and schools, very sharp, very hungry, [music] very young.

And I opened my speech with a simple question.

I said, "How many of you use AI for at least 30 minutes a day?" And I was expecting [music] each and every hand to go up, but I was thoroughly surprised.

Less than maybe 30 hands went up. And then I asked, "How many use Instagram for more than 30 minutes a day?

" Every single hand went up. Think about the tragedy hidden in this trend.

Here's another surprising finding. Many of the world-class universities, including Harvard and MIT, Stanford, IIT, Oxford, LSE, many [music] others, they have put many of their entire courses online, the entire semester worth of classes online for free.

And 88% of all users who start those courses don't finish them.

There's almost nine out of 10 of us. And it gets even better now because with AI today, you're holding a multi-million-dollar university in the same pocket with Instagram.

The most powerful intelligence engine [music] in human history is waiting for you.

And you don't even have to pay $250,000 for that education. And that's the system we want to build today.

And I don't mean some new Notion template or anything. Instead, we'll go through five radical roles that you can give AI.

One easy way to remember these five roles is to arrange them by their first letters.

It spells alter. Which kind of fits, right? Because these roles alter the way you think, the way you learn, the way you behave, and the way you [music] graduate.

So, let's build them one at a time. First things cool, you need an advisor. That's where we begin.

To understand that concept a little better, let's go to Liverpool. In the '60s, there was a city full of shipyards and docks.

And there were three teenagers who started a band called Johnny and the Moondogs. They didn't know then that they would change the world of music forever.

John Lennon, Paul McCartney, George Harrison eventually went on to become the Beatles.

But they were lost for a while. They got turned down by every label in London until they found their creative advisor, the fifth Beatle, George Martin.

He became the music producer for the band, and he was able to hear what the four of them couldn't.

[music] And the band went on to create some of the most loved music ever made. And this story keeps repeating since the dawn of humanity.

If you don't have Aristotle, you don't get Alexander the Great. Every great Michael Jordan needs a coach like Phil Jackson.

Because good advisers study you, and they build a personalized path for you. If today all the universities started offering customized course wear for each and every student, they would all go bankrupt.

If you go to a gym and every person walking in gets a personal trainer, they'll never make money.

But here's why things are changing dramatically for all of us. Today's AI can become your perfect academic advisor when you're ready to learn anything.

And how do you make AI a great advisor? Ask it for five decisions. First, destination.

[music] What do you want to know or do at the end when you're done learning it? Second, baseline.

Where are you on this topic right now? Basic level, intermediate, nothing? Third, sequencing. What order do you learn this topic in? That's really important.

[music] And fourth, cut list. What should you ignore for now? And fifth and final, milestones. What output from you will show that you mastered the topic and you're ready to move on.

Let's make this concrete. Let's say if you wanted to learn business finance.

Obviously, you're not going to give a prompt saying, "Teach me corporate finance." Instead, type, "Act as my elite academic advisor.

We want to build a 6-week custom course. I want to learn about how finance works in business, so I can understand how to use those skills to build my own business.

First, interview me to find my weaknesses, my baseline.

I want to define the destination, build the sequence, list what I should ignore for now, and also give me weekly milestones, so I can prove that I'm ready to move on.

Ask me up to five questions for each of these five steps, one question at a time. Now, the exact prompt is not that important.

In fact, you can ask AI to make you an even better prompt than the one I just gave you. The goal is to leverage AI to create the right personal curriculum for you.

This doesn't have to be just business or finance or anything.

What if you wanted to master something grounded in human judgment, in art, in taste, in style? What if you wanted to master design theory?

You can use the same structure. In fact, you can use Gemini because it has the multimodal capabilities.

So, it's almost like giving a pair of eyes to your advisor. In all such self-education adventures, AI no longer remains as your lazy text generator.

>> [music] >> It is an intelligence engine that becomes your path builder. But, even the best paths will break if the sources you learn from are garbage.

That's where we go next. Once you build your curriculum, the next battle is sticking to it.

Stanford researchers tested more than 7,800 students on how well they judged information online. The result was embarrassing.

82% of middle schoolers could not tell the difference between sponsored content and real news. Now, these are not some ancient villagers discovering the internet for the first time, right?

These were digital natives. They knew how to use the internet. They just didn't know how easily internet could abuse their trust.

The algorithm and the feed is not interested in teaching you anything. It only cares about keeping you there.

That's why the AI librarian matters. Their job is to defend the curriculum from distraction. It's a triaging function that tells you which 10 sources deserve your attention, >> [music] >> your time, and which thousands you can ignore.

It helps you figure out this belongs, this doesn't. This is credible, this is too advanced. This is signal, this is noise.

So, let's see the librarian in action. Let's use Notebook LM. First, [music] build the shelf. Find the three or four sources that actually matter.

Uh textbook, uh YouTube course, or a bunch of videos, a podcast, um an academic paper, a bunch of newspaper articles, whatever it is that you can find.

What if you don't know what sources to look for?

Well, you can use deep research on Gemini or Claude or ChatGPT to find those sources. Then, load these documents to Google's Notebook LM.

Nowadays, you can create a new notebook from Gemini itself. Now, your AI is not wandering homelessly across the open internet, it [music] is anchored to the ground truth you just gave it.

But now, how do you make the most out of these documents?

This is where Notebook LM changes the game. You can turn anything into a podcast or a video or a slideshow.

Let's say you want to educate yourself on building great money habits. Okay, so let's do it live. I love this book called The Psychology of Money.

It's by Morgan Housel and I've read this book. I grabbed a bunch of very good videos from YouTube and interviews that the author himself has done and I put them into Notebook LM.

And now I click on studio and then it's going to start a podcast >> [music] >> and you can listen to it during your walk, during your commute and you can even call in.

>> Welcome to today's deep dive. We have a uh a really fascinating mission ahead of us today, entirely custom tailored for you listening right now.

>> Yeah, it's a great one. We're looking at a huge stack of sources today.

>> And they all point to one specific subject. So we are going to be summarizing the core lessons from Morgan Housel's incredibly popular book, The Psychology of Money.

>> Now you can even call in. >> Read honestly. >> Oh, it really is. And the mission of this deep dive is to prove a premise that kind of Oh, hey, I think our listeners got something to say.

>> Yes, I love your podcast. I just had a quick question.

What's the one-line summary of the book? What's the key message? >> That is a brilliant question. We love it when you jump in like that.

>> Seriously, it is the perfect way to kick things off. >> If we had to boil it down to one key message, is this.

>> Doing well with money has little to do with how smart you are. >> Exactly, it has everything to do with how you behave.

>> Financial success is a soft skill, not a hard science. >> And the ultimate goal of that >> This is all AI. This is not some futuristic thing.

It's all here, available right now. But a good source alone can't guarantee understanding. At some point the material will stop making sense to you and this is exactly when you need the tutor.

A teacher and a tutor serve two different roles.

The teacher explains it, the tutor [music] diagnoses your specific confusion one-on-one. A good tutor makes it harder for you to fake understanding.

Tutoring has always been the luxury good of education. [music] Um, psychologist named Benjamin Bloom wrote about this in the '80s and it became known as the two-sigma problem.

And he found [music] that students with one-on-one tutoring performed way better than the rest of the students who sat in conventional classrooms.

The average student who got tutored performed above 98% of the students in the control class.

That was one of the most important findings in education. And also one of the most frustrating because everybody got to know that tutoring works, but not everyone can afford a private tutor until now, that is.

AI completely breaks the economics of human intelligence.

A world-class tutor can now sit with you at midnight, ask you uncomfortable questions, refine your logic, and refuse to leave until the concept finally clicks.

I use AI both as a teacher and a tutor. So, here's the tutor in action. I usually use either ChatGPT voice or Gemini voice because tutoring should feel like the office hour.

So, how do you actually do this? Just Let's do it live.

I have Gemini voice mode on my iPhone right now. I'm going to hit voice mode and say the following.

Be my tutor on the idea of enough from the book The Psychology of Money. Ask me one question at a time. Don't lecture me. Find the gap in my understanding.

Be precise. Let's go. >> How does constantly comparing yourself to others make it difficult to define your own personal version of enough?

>> Okay, you get it. So, that was a complete improvisation because I didn't know what the AI tutor was going to ask, but that's what would happen in real life as well.

Sometimes, you will get questions you don't know the answers to, and you'll have to work with the tutor to learn about it.

That's the whole point. The two best commands in AI are teach me and test me.

I love both of them. But, there is still one problem. Learning all the skills is great, but when you have to deliver something, you need honest feedback.

You need refinements. That's why the next rule is so important, the editor.

That's where we go next. A tutor helps you understand. An editor helps you deliver.

A typical Formula 1 race is about 90 minutes, but the difference between winning and losing can be just a few seconds.

That's why a large team of engineers for every race car company monitor thousands of data points in real time, constantly feeding tiny corrections back to the driver.

Because if you can shave off just a fraction of a second at every turn, you can change the outcome.

Real-time feedback is everything in that game because excellence lives in feedback. That's why authors have editors.

Filmmakers have editors. Journalists have editors, too. Even CEOs have boards. They're editors. Your AI can be a great editor.

When I write anything nowadays that's important. I usually take time to review it with chat GPT or Gemini or Claude.

And I ask AI to challenge my thinking, my logic. Find the weakness in my arguments. Cut repetition, tighten the structure, make the language more precise or concrete.

AI can be your powerful ally when it comes to editing your work. But you still have to fly the plane.

[한글자막] AI로 혼자서도 압도적인 실력을 쌓는 독학 방법

Scripts