How I Learn Things Really Fast
By Tina Huang
Summary
## Key takeaways - **Define Clear Learning Goal**: The first step is to decide exactly what you want to learn and the specific goal like passing an exam, writing an essay, or building an application, because vague goals like 'learn AI' lead to scattered efforts. [01:44], [02:10] - **Gather Missing Resources**: After setting the goal, find all necessary resources by looking at courses, YouTube videos, social media, workshops, since most learning materials are incomplete like a puzzle missing half the pieces. [02:42], [03:41] - **Prime with Skimming**: Prime your brain by skimming content for titles and sections or doing quizzes first, which improves retention and learning speed by 10 to 20% and takes only 30 minutes. [05:10], [05:21] - **Layered Comprehension Approach**: Build understanding layer by layer starting with an overview at 2-3x speed, noting definitions, major concepts like the 50/30/20 budgeting rule, and full examples, avoiding getting stuck on one deep topic. [06:44], [07:32] - **AI for Resource Gathering**: Use Perplexity to search Reddit for learning paths and popular courses tailored to your needs, like focusing on listening and speaking Japanese, saving about 3 hours. [14:19], [14:40] - **Interleave Multiple Topics**: Learn multiple subjects simultaneously each day, like 1 hour Spanish, 1 hour personal finance, and 2 hours AI agents, to improve learning and stay motivated over focusing on one topic at a time. [23:06], [23:43]
Topics Covered
- Why does a clear goal accelerate learning?
- How does priming boost retention by 20%?
- Why learn layer by layer instead of diving deep?
- Can AI slash learning time from 30 to 10 hours?
- Does energy management outperform time management?
Full Transcript
How do I learn things really fast?
Just these weekly YouTube videos that I make alone involves me at least taking two to three courses which is like around 10 to 15 hours of content plus actually having to like do the projects and building and testing things out as well.
So luckily with the help of some learning frameworks and techniques plus the use of AI, I have managed to develop a learning system that cuts down the learning time by at least 60 to 70%.
It is a framework that is generalizable to learning anything as long as you have a goal in mind. So, in this video, I will explain to you exactly how I do this.
A portion of this video is sponsored by Warp.
First, I'm going to cover an overview of the learning frameworks and the techniques.
This is going to be like the foundation, which if you don't have a good grasp of, even if you use something like AI, it's like not actually going to be that useful.
It'll be like putting lipstick on a pig, you know?
I don't know if I got that analogy right, but you know what I mean.
Got to like actually know what you're doing first.
Then, I'm going to talk about where AI is able to come in and reduce the learning time significantly.
I'll also be covering like the exact AI tools that I personally use. And finally, I will end with some tips to make the process even faster and just things some things to keep in mind. At the end of this video, we will have a little assessment to help you retain all the information that you learn in this video.
All right, let's go.
The learning process is like doing a jigsaw puzzle.
It is composed of five different steps.
goal, research, priming comprehension and implementation.
Let me explain.
Say you woke up one day and went like, I want to do a jigsaw puzzle. So you're like, okay, oo, there happens to be a garage sale over there. Maybe they have some jigsaw puzzles.
So you go there and you see that they actually have a lot of different types of jigsaw puzzles.
So the first thing you got to do is pick the puzzle that you want to solve.
This is like when you want to learn a specific topic.
You first have to decide what actually it is that you want to learn and more specifically what is the actual goal you want to achieve.
Is it to complete a course and pass the exam?
Write an essay, build an application, do a report, do a presentation.
You see, actually knowing what is the end result is really important because you can say something like, "Oh, I want to learn about AI.
" But what does that even mean?
Like there's so many ways that you can be learning.
So you need to actually pick what is the final result that you're looking for. What is the actual picture of the puzzle that you want to solve?
After you figure out what your goal is, you can move on to the next step.
So, you got your puzzle box, right?
You're like, "Yay, what a beautiful puzzle.
It is about dinosaurs.
" So, you take your puzzle box and then you open up your puzzle box.
You look inside and your heart sinks because it is clearly missing some puzzle pieces here. The box says that it should have 500 pieces, but there's like literally half of that amount in here.
Alas, what to do?
You see, unfortunately, this is often the case when it comes to learning things as well.
Often times when you want to learn something, you don't actually have all the resources that you need to learn that specific thing and do your specific goal.
Hey, like sometimes you're lucky, all right?
Like your goal is to like pass this exam and you have the specific course that actually teaches you everything that you need for that exam.
But often times the case, this is actually not the case. Even if you are in school and then you have like a final exam, you will notice oftentimes the teacher doesn't even like teach the proper things, right?
Especially if you're in the workforce and you need to like write a report or like build an application or something like that.
You definitely are not just like given the resources to be able to learn the thing.
You often have to figure out what the resources are yourself and you got to like kind of piece that together, which is why it's like a jigsaw puzzle that only has like half the pieces.
You got to like find the other half of the pieces.
So, how do you find the rest of the pieces, the resources that you actually need?
And well, I actually spend a significant amount of time doing this, like looking at different courses, trying to pick the ones that I think are the best ones, watching YouTube videos, following some like decent people on X, attending workshops and conferences.
If you think about it, this step is actually really important and I feel like it's something that a lot of people neglect because like if you don't figure out what is the correct resources, how do you expect to actually learn the thing and be able to do the thing that you ultimately want to do, you know?
Anyways, now that you have all your resources gathered, you can move on to the next step, which is priming.
So great, somehow you have managed to find all the pieces of your puzzle by maybe like calling your friends and they somehow managed to have those pieces or something like that. And yay, now you have all your pieces in place and you also have the picture that you're trying to assemble.
The next proper thing to do is actually to start looking through your pieces and kind of like taking note of like, oh, like you know, these are the ones that have patches of color and you're kind of like arranging them in different ways.
You're like, okay, these ones are probably supposed to be together.
and you pick out all of the corner and the edge pieces so you know that oh these are supposed to be on the edges and the corners and just kind of like getting an idea like a really rough idea of how the pieces are supposed to be fitting together. Don't take this step lightly.
It's actually a very powerful subconscious process in which you're helping your brain orient itself to achieve the final result.
Similarly, when it comes to learning, you have to do priming by having a understanding, like a rough understanding of all of the information that you're trying to learn and how it's supposed to all fit together.
Priming is kind of like pre-learning.
You're actually just like getting a general sense of the information without actually understanding the information yet.
This is a step that is often skipped, but that's a bad idea because if you do priming, research shows that you can actually improve your retention and your speed of learning by up to 10 to 20%.
While priming itself doesn't take up that much time at all. It usually takes me like no more than 30 minutes to do, but it saves me like hours and hours.
So, how do you actually do priming?
It's actually really simple. The easiest way is just to simply skim through the content that you're trying to learn and just kind of like make a note of the different like titles and different sections that are there.
You don't actually need to like know anything.
Another way of doing priming is by doing quizzes and exams first.
Obviously, you're not going to do well in them, right?
Because you like haven't actually learned the information yet. But if you do them first, it sort of like ingrains into your brain the questions that you will be able to answer and you'll like subconsciously take more note when you come across that information when you are actually learning.
With projects, especially like coding projects, what I also like to do is I start the project and look at the like starter code for it if it's available uh and like not actually know what's happening, but just sort of like get an idea of what is the final product that should be built.
Okay, so we're ready to move on to the next step.
Now it's actually time to really go ahead and assemble the puzzle now.
And the way that you do this is that you start off with like the corner pieces and then start assembling the edges.
And you start assembling like the patches of pieces that have similar colors.
And you do that with all the different parts of the puzzle.
And you just keep doing that process like assembling the pieces here, assembling pieces there until they start forming a picture. And some of them actually like start joining together.
Now you're ready to like dive in and really understand that information.
And the way that you approach this is layer by layer like getting like a overview like an okay understanding of things overall and then diving deeper and deeper and deeper. This is in contrast to like a misconception that people have where sometimes they would like try to go really deep into a specific concept or a specific topic and then get like really stuck and just keep like bang their head against the wall like trying to figure out that one thing.
That's actually not a very productive way of of learning.
It's much more productive to just get like an overall understanding of the information and then going and like deepening your understanding layer by layer by layer. Let me give you a concrete example.
Like say you're learning the topic about personal finance and as part of this learning curriculum, you have like things like budgeting and saving and debt, investing and insurance.
These are kind of like the overall buckets that you're trying to learn, right? The way that I will approach this and actually I have approached this exact subject is that I would first take that learning material and I prefer like audio or video based stuff personally but it can also be like textbooks whatever it is and I would just go through it really really fast like I would usually listen to lectures at like two to three times the speed.
The goal is not to understand every single little detail. It's simply to understand like roughly what are the concepts that are in each module.
And when I'm doing this I'm also not taking like super detailed notes. I just let it play and I only jot down three specific things.
This first one is definitions.
Like what is the definition of budgeting?
What is the definition of debt?
What is the definition of saving?
What is the definition of investment?
The second type of thing I would make a note of is like major concepts.
For example, for budgeting, I would note down that people often times follow the 50/30 20 rule where 50% of your money is spent on things that you need, 30% of money is spent on things that you want, and 20% of money goes into savings.
Like stuff like that. And the third type of thing that I would note down is full examples of concepts. Stuff like, oh, like Bob has 100 monies and this is how he distributed these monies and then he invested in like a ETF like or something like that.
Full examples are really useful because they're able to like really illustrate how all the concepts are coming together in a very tangible way.
Anyways, I'm not going to go into too much more detail about this.
I've actually like made an entire video which I'll link over here uh where I talk about like how I take notes and stuff like that.
So you can go check it out if you want to. But the point that I'm trying to make here is that the first layer of learning I would do is just like grasping these like general concepts and then I would go and increase my understanding by then diving deeper into the details.
It's basically just like increasing the level of understanding deeper and deeper at every single level until you end up like getting a cohesive understanding of the things that you're learning.
There's a bunch of other little like tricks and techniques that you can use in this section to speed up your learning process, but I'll actually go into more detail about it when I talk about how to use AI to speed up the process because this section is what takes the most amount of time in the learning process.
And it's also where AI can actually save you the most amount of time, too.
So, it's like super exciting.
Now that you're learning how to learn things, one thing that I really think you should learn is how to use AI for coding and development.
And I promise you, if you try out Warp, it's going to blow your mind.
Warp is an agentic development environment built for developers.
I've been testing Warp's new AI agents and they are definitely game changers.
Instead of trying to oneshot a complex task, Warp's agents are supercharged in planning.
Like, you can literally see it write out detailed plans in a notebook before they start coding. And you work alongside it by editing a plan, steering the agent, and keeping track of it as it executes, which is huge if you're working across multiple files or repost.
You have so much control, which is a common complaint that developers have when using other tools. It's also built for real developer workflows with interactive commands and longrunning commands.
all the stuff that you actually deal with when you're actually in production.
Warp is already trusted by over 700,000 developers. And what I love the most is that it saves you time without taking you out of the loop.
On average, devs say that it saves them about an hour a day, which I concur.
You can now test out Warp AI's features for free.
Link is in the description.
Thank you so much, Warp for sponsoring this portion of the video. Now, back to the video.
The final step is the implementation step.
This is when you're like, you know, putting everything together.
now like connecting all of your clumps and putting together like your different patches like the patch of dinosaur head with the dinosaur body. You're like and then you have that patch of the sky which you're connecting with the bigger patch of the sky. Everything is just all coming together in a very satisfying way.
This last part of the learning framework is actually really interconnected with the previous step because as you're learning and understanding like the content for example let's go back to like the personal finance example. You would also have in mind the actual goal for your learning which is say to come up with your own personal finance strategy and actually implement that. You might also be like jotting things down and making a note like oh like this is where it's useful for me given my specific situation, right?
Like oh I'm interested in repaying my debts so I'm going to like focus a lot on this specific part.
Or if you see like equation for how to calculate wealth, you'll take a note of that and calculate your own wealth along the way too. So by the time that you're done learning, you probably also have like a good idea of what are the parts of your personal finance strategy you need to do. And then you would do like the calculations and finally implement the process by actually going into your bank and like setting up some automations and then also opening up like an investment account and setting up automations there as well.
What you will find is that if you implement this five-step framework in which you have the goal that you determined in the very beginning and you're like kind of having that goal in mind throughout this entire process, the final implementation step is going to save you so much more time if you didn't have that clear in your mind and you're just like either learning random things or if you're trying to implement something without the right resources. Anyways, so that is the five-step framework for how I learn things really quickly. In terms of amount of time that I spend on each of the steps, I would break it down roughly.
For goal, it would be around like 0% to 5% because I usually like already have a goal in mind, but if I don't, it's like not that hard to determine what your final goal is.
The resources section when you're finding the resources you need, that step can take anywhere from like again like zero to maybe 10% of the total amount of learning time.
If you're lucky and you just happen to have like all the material that's there for you to complete your final goal, amazing, perfect.
uh that's zero but often times it does take me like a significant amount of time to find the correct resources.
The priming section usually only takes around like 2 to 5% of the learning time.
It's doesn't take that long.
Now the understanding step usually takes up the bulk of the learning time anywhere from like maybe 40% to 60% of time.
And finally the implementation step also ranges depending on what it is that you're implementing. Like is it just like an essay that you're writing or are you trying to like build an entire application?
That varies as well.
But usually it takes maybe like say 20% to 40% of the time. Now let's talk about how AI can speed up this process.
For this five-step framework where AI can really help is three of these steps.
The first one is the resources part the finding of the resources. The second is the understanding process like learning the content and the third step is the implementation.
Now let me show you exactly what you can use AI for for each of these steps. So let's assume that you're going to learn a thing that would normally take you 30 hours, right?
If you followed this wonderful learning framework, which is already going to be like half the time it would take you if you were just doing like not following this framework.
But yeah, let's say like you know you already implemented all of learning strategies and you're sitting around like 30 hours to complete this entire learning process. I'm going to show you now how you can cut down that learning time to like 10 hours.
So you're cutting down like 2/3 of the time and then you can spend this amount of time like doing something fun instead like watch anime. By the way, if you're looking for good anime to watch, I just finished Tatami Galaxy and it's so good.
I'm now watching Sunny Boy, which is also so good. So, yeah, recommendations for your extra time. So, the first part of the framework goal, uh, honestly, it's not going to help you do that much, honestly.
Like, not really. You kind of still have to figure out what your goal is.
But the second part of the framework, gathering of the resources.
My favorite tool for gathering resources is actually perplexity. What I do is that I search up like a specific topic.
Like, say I'm learning a language like Japanese.
I would use Perplexity to search specifically on like social media like Reddit to see how other people have completed the process and compile the resources that they used. I would also use Perplexity to gather some of the most popular courses for learning Japanese and ask it to fit my specific case.
Like maybe I'm only interested in like listening and speaking Japanese and not like the reading and writing part of Japanese.
Deep research is also really useful.
I actually have a full video like covering how perplexity works which you can check out over here if you want to go into more detail about that.
So I would say that this is usually enough for most topics. In my specific case, when it comes to like learning AI topics, I do kind of go a step further because it's like literally my job.
I actually built an entire AI agent that searches for like specific courses and specific topics.
So, I'm able to compile like all the best resources about a topic and make sure I'm like not missing any potentially good resources.
But anyways, for most people, I don't think you need to actually go build an entire AI agent to do that. Using perplexity is probably enough.
Now, doing this with AI will save you like 3 hours of your time.
Now, next up for Prime Meade, a tool that I like to use is called Notebook LM.
It's a tool that specifically helps you understand and digest information.
One of the things they can do is that you can actually like take a course or like a video or whatever, like some type of resource, put it into Notebook LM, and then tell it to generate a study guide for you. And the guide will be able to surface up like all the different topics that you want to prime yourself about uh subconsciously.
And you can also ask it to include quizzes.
So, you can like take the quizzes prior to learning as part of the priming process as well. This is really helpful if you're covering resources that don't already have like embedded quizzes in them.
So with priming, if you use AI tools like notebook outlim, you'll save about an hour doing this. Next up with the framework is the understanding of the content.
This is where you're going to be saving the most amount of time by using AI.
Another thing that I like to use AI for is to actually convert the format of resources to my preferred formats.
For me, I'm a very like audio and video based learner. So whenever I have to learn something from a textbook or like written text, it is so painful for me.
like it takes me so much longer to learn that content if I have to like read it.
So what I like to do is actually take that information and transform it into a video form or transform it into a podcast form.
The way that I do this is I actually have a custom agentic workflow that I made that literally like takes the information and then transform it into audio form in which I would listen to it like two to three times the speed. And that's like how I digest the information really quickly.
Another way of doing this, even though it's a little bit more manual, upload it onto Google AI Studio and tell it to transform that information, whether that be like text, a report, or whatever, into a script for an audio podcast.
What I like to specifically prompt it is to transform the information that is contained within this resource into a single person podcast form and make it very concise, only containing information of definitions, concepts, and full examples.
Do not make any of your own commentary.
Then I would take this transcript and there's like a stream real time tab in Google AI studio and then just like put that in there and then ask it to make it into a podcast form and then download it.
>> The potato orelenum tubosum was first domesticated in the Indian highlands of South America between 8,000 and 5,000 BC.
Indigenous peoples in modern day Pu and Bolivia cultivated hundreds of native potato varieties relying on them as a staple. If you do have a script that is really long, you might have to do this in chunks and then you'll have to like download it and put it together which is you know a little bit annoying but still better in my opinion than having to like read a 30page report.
So, this is for if you're like an audio learner like me. If you happen to be like a video learner, Notebook LM does have the ability of transforming things into video format. Notebook also does have like a podcast format, like a audio format.
But I personally don't actually like to use this when it comes to just like learning information because I feel like it's a little bit too it's like a two-person conversation and is a little bit fluffy when I could just learn everything like really really fast, you know?
Instead, >> today we're diving head first into something well something many of us have tried or maybe are trying right now >> or thinking about trying.
Yeah, >> exactly.
>> I prefer like drinking from the fire hose kind of method, not like pleasantries.
If you're someone who's into text, you can actually do the other way around as well. You can give it like a video or like some sort of resource and tell it to transform it into text format, like a textbook format.
Uh you can really transform this information however it is that you like.
But the tool that I would recommend doing this is mostly going to be Notebook LM.
This saves me like four hours of time because if I learn the way that I prefer to learn, I just learn like so much faster than if I had to read a textbook.
Another learning technique that I like to use is by taking like resources and then only asking the AI to take out the parts that I'm interested in.
Often times when you have like a resource, you only have like, you know, one section that you actually care about, right?
But if you didn't have AI, you would have to like read through the entire thing until you identify that specific section.
But you can use AI like Notebook LM for example um to be able to extract just that section so you can learn just that section.
This will save you like another hour or so. One of my favorite AI features for this part of the learning process is Chatabit's audio mode feature.
What I like to do is actually just like talk to Chat to be and really like dive deep into whatever problem it is that I may be having. I would ask you to give me like examples of things and just like talk to me until I figure out like exactly what it is that I'm trying to learn.
So actually I'm not sure about the role of emotions in storytelling.
Like are all the emotions equivalent to each other or are some emotions more powerful than others?
>> Emotions tend to have a stronger impact on audiences depending on how they're used.
>> If you are a more visual person, you can of course also do this like using text and just like directly as a chatbot talking to it type stuff if you like.
You can also use it to generate diagrams too.
I just like am audio person so that's why I keep like default into audio options.
This would save me about like 3 hours I would say.
And finally, another tip that I have for using AI here is for organizing your notes.
I don't know about you, but the notes that I take are absolutely hideous.
Like, here's a sample. It's literally just like random text with a bunch of spelling errors and screenshots of things.
What I do is that I basically just take the entire thing and give it to Notebook LM and then tell it to process my notes and then take it to ChatBT, for example, and ask it to make it more concise, generate tables or even diagrams. If I really wanted to go overboard, I would actually take this and give it to Claude and have it generate an interactive dashboard.
These are some of my top tips. And in total, you're going to be saving around 10 hours.
Now, finally, for the implementation, the amount of time that you can save using AI for this step can really vary depending on what the final product is.
For example, if the final thing is that you want to have an essay or report, I would usually use Notebook LM as a tool to come up with the outline and then work with ChachiBT in order to actually write the report. If I wanted to create an interactive dashboard, I would usually use a combination of notebook LM plus like just putting data sets for anything that needs to be analyzed into Claude and then using it to generate an interactive dashboard.
If you want to know more details about exactly how to do this, don't want to make the video super long.
So, I'll actually like link the videos put up here or also in the description that you can check out how to use these tools uh to do these things. And finally, if I wanted to actually create a product like I wanted to code an entire application for example, I would use some sort of AI assisted coding war vibe coding tool.
Oh, and you know what? I'll just add in a bonus one. This one's really common, too.
If you want to generate a slide deck afterwards, I would usually use either Manis or Gamma to generate the slides.
On average, I would say it would save you about 6 hours. Wow.
If you actually add up all of these numbers together, assuming that it would have taken you 30 hours without using AI tools, you have now saved 20 hours of that time.
So, assuming that each episode of anime is around 30 minutes, that's like 40 episodes of anime or whatever more productive thing you would rather do with 20 hours, of course.
Anyways, those are some of my favorite ways in order to increase my study speed.
Now, before I end this video, I do want to give a few tips and tricks that I like to use in my learning process to speed up things even faster.
The first one is energy management over time management.
When we're usually setting up a learning schedule, we're usually looking in terms of time.
Like, oh, I can study for like 2 hours.
I have 2 hours free today and I have like 3 hours free tomorrow, like etc., etc. you try to like cram in your study sessions based upon the amount of time that you have.
But this is actually a mistake because usually when you end up having problems keeping the study schedule, it's not because you didn't have the time to do it. It's because you end up just being too tired and not having the energy to do it. That's why it's really important to pay attention to when it is that you're studying. Like if you are somebody who has a full-time job that is like super draining and then you expect yourself to come home after dinner at like 8:00 p.m. and then study for like 2 hours, that's probably not going to happen because you're going to be too tired and you're probably just going to like sit there and watch like trash TV or something.
Instead, if you try to get that 2 hours study time in before you start your work, that means your brain is completely fresh and you would have the energy to actually study.
So that's why whenever I plan out when I'm going to learn things and when I'm going to have study sessions, I think about where in my schedule I have the most amount of energy for me to have the greatest chance of having a productive study session.
The second pro tip I have is a concept called interle.
Interle means learning about multiple different concepts or topics simultaneously.
So, for example, if you're trying to learn about how to build an AI agent and you're also trying to learn Spanish as well as like personal finance, the research shows that the best way of learning these subjects is to not just focus on like one singular topic and then go to the next topic and then go to the next topic. It's actually better if you mix things up. For example, if you have a study day throughout the week, instead of saying like, oh, like today I'm only going to study personal finance and tomorrow I'm only going to study AI agents and the next day I'm only going to be studying Spanish.
It's actually better if you schedule your day like for each day you'll study like 1 hour of Spanish, 1 hour of personal finance, and then like 2 hours of AI agents.
This not only helps to actually improve the learning process itself, it also keeps you more motivated and excited because you're able to switch between different subjects.
Really, really powerful method.
All right, that is all I have for you guys for this video.
As promised, here is a little assessment.
Please answer the questions on screen in the comments below to help you retain the information or everything that you've just learned. I really hope this is a useful video for you.
Whatever it is that you're trying to learn these days, self-learning is really one of like my favorite things ever.
All right, thank you so much for watching until the end of this video and I will see you guys in the next video or live stream.
Loading video analysis...