Rating: 7.5 / 10
A bit long. Could have been a great article or post-series. But it gives some great points on learning. It feels like an amazing starting point for learning, but it could be built upon with other great strategies that one might discover for themselves.
Scott approaches learning with intensity and commitment. Being intense when it comes to learning seems to be able to lead to incredible growth.
The practice of directness is learning by doing. There's a huge difference between learning something new (passive learning) and practicing something new (active learning). They produce very different results: active learning creates skill. Passive learning creates knowledge.
Both are necessary. A great example given in the book is that you can read all about the bench press exercise, but you can't build strength without doing it. Another good example is sales. You can know all about every technique, but you won't get a single customer without making a sales call.
Language learning advice — Benny Lewis
Stardew Valley was made by one guy. Everything in it. The art, the game itself, the music... everything. He did it in five years through self-education.
Ultralearning is a self-directed and intense strategy for acquiring skills and knowledge.
You don't always have to apply it. It might be a (very) good choice, but it's not a must.
The self-directed part basically means that you control what you learn. You're not just mindlessly following along what a textbook tells you.
It's intense, and it's hard work. It's very much unlike just downloading Duolingo and having fun for 5 minutes a day. It's more like drilling 10.000 questions. It's not easy. But this can contribute to you going into the Flow state.
Taking on these Ultralearning projects can lead to incredible growth. For example, in your career. Learning new and very hard skills like this is amazing for your career (especially for entrepreneurs).
A trend that is occurring these days is that medium-skill positions are getting pushed out. Lower-skill positions are staying, and higher-skill positions are also staying. So if one is in the medium-skill position and isn't eager to learn (to go to a high-skill position), they'll fall to a lower-class position. This happens because most of the medium-skill positions are being outsourced to computers (or other countries). So if you want a high-skill(paying) position, you have to be eager to learn new, hard skills.
Information is more accessible now than ever. It's easier to get material to learn from than ever.
A lot of Ultralearners didn't do it for money. They did it because the project genuinely excited them. Those who blend the two are usually the best.
It's important to remember that not everyone will be able to put in the same effort as other Ultralearners and get the same results.
There are really three ways of implementing Ultralearning.
It's important to note that it's not a requirement that you spend the 60+ hours a week that other Ultralearners do. Yes, they're getting the results from spending that much more time, but you can still reap the benefits from implementing the commitment, intensity, and initiative.
Ultralearning projects all have one common denominator. They're all unique. Other than that, there's really no one-size-fits-all here.
Because of that, it's better to see it through the lens of principles, rather than some step-by-step protocol. The following 9 principles each have a chapter, but they're described here in short. They all underlie each ultralearning project and contributes to the success of it.
"If I have seen further it is by standing on the shoulders of giants" — Isaac Newton.
Meta comes from Greek and means "beyond". Typically used when something is "about" itself. So in this case, it's learning about learning.
Its learning how knowledge is structured and acquired in this subject. Learning how to learn it.
Almost like drawing a map of the relevant topics such that you can navigate (learn) it without getting lost.
There's two ways to draw this map. Over the short term, and over the long term.
When starting a project, and you're doing metalearning, it's useful to ask these three questions: "why", "what", and "how".
Why is about why you want to learn that skill. Make sure that learning the topic has the effect that you want before you get started learning. A good way to do research is to ask people who are where you want to go.
What is about what exactly you need to learn to be successful (concepts, facts, procedures). A good way to answer this question is to write down the three on a piece of paper, each in a column. Then just brainstorm what you need to learn — which you can revise later. Concepts are things that need to be understood. Facts needs to be memorized. Procedures need to be practiced. When you've done this, underline those which will be most challenging. This allows you to search for methods and resources to overcome these challenges. For example, if something requires a lot of memorization, then you'd probably want to overcome that by using some spaced-repetition system. For mathematics, a deep understanding of key topics is key, so maybe teaching them to others is key. Obviously this helps for finding ways to overcome challenges by adding tools to your kit. But it also helps you know what to avoid.
How is about which resources, environment, or methods you'll use when learning. There's two good methods to answer how you'll learn something. Benchmarking and Emphasize/Exclude Method.
Benchmarking: always start a learning project by finding the common ways that people learn the skill/subject. If there's a school that teaches it, find their recommended reading list, their course list, or their syllabus. This is an important step and can save you a lot of time later.
Emphasize/Exclude Method: when you've found a curriculum, you can make modifications to it. Sometimes it's possible to identify the importance of a subject on the list — and if it's not important, to just disregard it. But if you are trying to learn something more conceptual, or if you don't know why something is on the list, you'd probably be better off just keeping it there — at least until you know a bit more. As the name of the method suggests, it's mostly about emphasizing on the important parts of the subject/skill (those that you're interested in or need), and excluding those that might not be so important for your goals.
It's possible to plan too much, and to plan too little. If you find that you're using planning as procrastination for doing the actual learning, you should just get started. There's going to be uncertainty in your approach, so just get started and figure it out as you go.
A good rule of thumb is to use about 10 percent of your total expected learning time on research prior to starting. But this shouldn't scale (don't spend 100 hours researching a 1000 hour project). The important thing is that you don't just latch onto the first learning 'path', but you don't have to exhaust every possibility either.
Another good way is to research better methods as you go. Finding out what works and what doesn't, and then adapting to that.
Why do we procrastinate?
Some craving is driving you to do something else, there's an aversion to doing the task itself, or both.
Our excuses for procrastination are often silly when said out loud.
The first step to overcoming procrastination is to recognize that you are procrastinating.
Ask yourself why you are procrastinating. Is it because you want to do something else? Or is it because you really don't want to do what you should be doing?
When you have recognized that you are procrastinating, there's a few steps that you can take to overcome it. There's "crutches" / mental tools to help you.
If you are averse to the task, try working on it for just a few minutes. Then the aversion usually disappears. But it's all you have to do. Just a few minutes and then you can take a break.
If you find that you are taking breaks too often, try the pomodoro technique.
When it's no longer a problem, try to use a calendar where you schedule in work-blocks. If that doesn't work, start from the beginning (progressing through these steps again).
Eventually you can just do something without needing these crutches. But it's built up over time, and needs to be worked on constantly. Don't feel bad if you have to go back. It's natural.
Sustaining focus is key if you want to learn hard things.
First: which kind of focus is best to sustain?
Flow is ideal. The state right between boredom and frustration.
However, flow is hard to achieve during ultralearning. Scott's advice is to not worry about flow. Sometimes you'll achieve it, sometimes you won't.
For how long should you study?
Generally, you retain more of what you learn when your studying is broken into sessions, rather when it's done all in one session. Therefore it could also make sense to alternate between multiple subjects or aspects of the skill.
So if you have a long session ahead of you, try to do different things — but still inside the scope of your desired knowledge/skill. But if your study time becomes too fractured, it might be difficult to learn at all. Balance is key here. 50-60 minutes is good for most learning tasks.
Find what works best for you. Consider efficiency, retention, and your personal schedule.
Turn off the phone. Turn off the TV. Remove distracting noises/sounds (noise-canceling headphones or something). Be fully prepared to work — have everything ready. Multitasking doesn't work, so don't do it.
Use the tools (video/audio/reading) that works best for you. If you find that you're reading, but actually are somewhere else in your mind entirely, try to take notes while reading. This helps you process what it is you're reading, and makes it harder to become distracted. Explaining things out loud helps too.
Your mind can distract you, too. Sometimes feelings just pop up. If it's something that you don't have to (or should/can not) deal with right now, try to just acknowledge it and allow it to pass. This is, of course, a lot easier said than done. But persisting on the task builds long-term strength on your ability to persist on tasks, so you aren't wasting your time even if you feel that you aren't getting much out of your learning session.
How should you focus, when you are focusing?
Sometimes it helps to be incredibly focused. Other times (especially in creative tasks), you might want to remove your focus so that your 'unconsciousness' can help solve the problem (like getting ideas in the shower, etc) — meaning: take a break. Not 'stop working'.
Unrelated: when coming out of school, you only have the tools that the school gave you. But this doesn't always match with what the job market needs. So even if you did great in school, you might not be offered a job, simply because you don't have the knowledge or experience with the tools that they use. You can fix this by learning what exactly it is that you need to learn, and then learning it. This could be done directly by simply asking them what they use, or by figuring it out yourself from research. Then you'd want to develop experience and a portfolio using the knowledge and tools that they apply, such that when you apply, they don't see you as someone who needs a lot of training to begin.
The above does actually perfectly embody the idea of directness. Directness is the idea of learning being tied closely to the situation or context that you want to use it in.
We want to work on collaborative, professional programs but mostly code scripts in isolation. We want to become great speakers, so we buy a book on communication, rather than practice presenting.
The problem is that our learning is disconnected from what we want to use it for.
We might feel that directly learning the thing is uncomfortable, boring, or frustrating, so we settle for a book or something, hoping that it'll make us better at at the real thing. It wont.
This is also a problem with school. Students solve highly idealized problems that they will never see after graduation.
It might feel like you're making progress, but you aren't really.
Watching lectures is not where most of the learning happens. It happens when solving the problem sets. (in most cases — emphasized because it's applicable for me)
The easiest way to learn something is by doing it a lot. It's not always possible, but one should try to mimic the actual situations as much as possible.
Most of what we consider learning is passive. Active learning, however, is a lot better. Passive learning is simply reading or watching. Active learning is solving the problems. Putting theory to practice.
There's a few ways to apply direct learning.
Careful analysis & deliberate practice.
The strategy behind doing drills is identifying which action would help 'level up' your skill, and then doing that repeatedly. For example, learning a lot of words when learning a new language.
This isolation of a singular part of a skill allows you to progress much faster in your improvement, rather than practicing every aspect of it all at once.
A writer might practice writing by just writing. But you can also isolate the aspects of writing into
And so on. You simply figure out what matters, and practice that a lot.
Drills are good even if you don't think that there's one singular aspect that you might want to isolate and improve upon. They help simplify the skill such that you can focus your cognitive resources on just that aspect.
Since there is a 'conflict' between drilling and directness (one is isolating, the other is trying in the context in which your skill might actually be applied), you might use the Direct-Then-Drill Approach.
Practice the skill directly.
This initial connection and subsequent feedback loop ensure that the transfer problem won't occur (not being able to transfer knowledge to an applicable situation — from context to context)
Analyze the direct skill and try to isolate components that are important for your importance or subskills that you find difficult to improve upon because you also have to manage everything else happening. From here you develop drills and practice the components separately until you get better at them,
Direct practice again. Integrate what you've learned. This is important, otherwise, you might not be able to transfer the drilled components to actual practice.
To find out what you should drill, make a guess, try drilling it, and then incorporating it. Did it help? If not, try something else.
Drill 1: Time Slicing
Look for parts of the skill that you're learning that can be decomposed into specific moments of time that are more challenging or important.
Drill 2: Cognitive Components
For example. When learning a language, grammar, pronunciation, and vocabulary occur at all times. But you might want to only practice pronunciation. So you'd single that out (record yourself and get instant feedback).
Drill 3: The Copycat
Sometimes you can't practice without the full 'work'. So you copy the parts of the skill that you don't want to drill from either other's work, or from your own previous work. For example, if you are a programmer. Maybe only want to practice a smaller part of creating an app, so you'd use a boilerplate template for setting up an environment in which you can do so.
Then you don't have to do everything, and you can just focus on what matters.
Drill 4: The Magnifying Glass Method
Spend more time on one component of the skill than you would otherwise. Maybe this decreases your performance or increases your input time, but it'll allow you to practice the part that you want.
Scott improved his researching-for-articles abilities by spending a lot more time doing research for articles (10x). He still had to write the whole article, but he'd spend a lot more time researching than usual.
Drill 5: Prerequisite Chaining
Start with a skill that you don't have all the prerequisites for. When you (inevitably) do poorly, take a step back and learn one of the foundational topics, and then repeat.
"It pays better to wait and recollect by an effort from within, than to look at the book again" — William James, psychologist
A study found that testing yourself — trying to retrieve information without looking at the text — clearly outperformed all other conditions (re-reading, mind mapping etc).
Retrieval with the ability to look up the answers is the best. But retrieval itself is much better than conventional studying methods.
You want difficulty. You want to retrieve far enough form the learning such that it's hard to remember, but not so far that it's forgotten.
Retrieval can help you enhance what you've learned previously. But not only that. It can also help prepare you for learning new information.
But what should you actually try to remember? That's a hard question. But a way to answer it is by doing direct practice. So the things that you'll actually use, you'll be forced to retrieve, and therefore remember.
Sometimes we come across a potential solution to a problem that is better than the solution that we already know. But since we might not need the solution at that time, we won't use it, and we'll just forget it. The point is that its useful to have a certain breadth, which requires a certain amount of passive exposure, which benefits from retrieval practice.
Tactic 1: Flash Cards
Tactic 2: Free Recall
After reading a section from a book or sitting through a lecture, try to write down everything that you remember.
Tactic 3: The Question-Book Method
Instead of writing facts, write questions asking about the facts. So instead of 'the capital of Denmark is Copenhagen', you'd write 'what is the capital of Denmark called?'
Be careful that you don't only practice this method on useless facts. Important facts are things such as the big ideas behind a chapter. Don't just put a question mark on things that you've copied straight from the chapter. Do some actual thinking and boil down the ideas into a good question.
Scott recommends to only do one question per section of text.
Tactic 4: Self-Generated Challenges
The first three tactics work best with facts and such. But what about learning a skill? For example, you could challenge yourself to implement something that you've read about. This is great for programmers — especially if they're reading about algorithms and data structures.
Tactic 5: Closed-Book Learning
Don't allow yourself to look up the knowledge. Then you'll be forced to retrieve it. This works great with something like the Feynman method wherein you'd try to make something simple and understandable without actually looking at the book.
Ultralearning feedback has immediacy, accuracy, and intensity.
Feedback is essential. No feedback = stagnation.
Feedback on the performance of an individual is good. Feedback on the individual itself is bad.
"This answer was wrong because of x and you can do y and z to fix it" versus "Your answer is wrong because you ..."
Focus on what (part of the) feedback is useful, and disregard the rest. Feedback to either end of the extreme (too positive & too negative) can have negative consequences.
Harsh feedback can be hard to hear. Some people choose to avoid it. But it is an incredibly potent way to learn; to just dive in and take it. That's the best and fastest way to learn. For example, making videos for the public. You sidestep it because it's for everyone to see. It can be harsh. But it's the best and fastest way to learn.
There's three types of feedback.
Immediate feedback is best. Even better with delayed review (or further testing). In general, the faster the feedback, the better. But make sure that you've actually thought through the problems before wanting feedback.
Knowing something is useless if you can't recall it.
But why do we forget? Three theories:
Decay: forgetting with time
Interference: overwriting old memories with new ones
If memories are similar yet distinct, they can compete with each other.
There's two types:
Forgotten cues: a locked box with no key
How do we prevent it? Four kinds of systems:
Memory Mechanism 1 — Spacing: Repeat to Remember
Good studying advice: don't cram. Spread learning sessions over more intervals over longer periods of time.
Good software systems: SRS tools. Anki. Of course, not needed (but still amazing). You can do this with simple pen and paper.
Regular practice is good for skills that isn't just about memorization.
Don't let perfect become the enemy of good enough.
Refresher projects is also good.
Memory Mechanism 2 — Proceduralization: Automatic Will Endure
We can do certain things enough such that they become habitual. Procedures that we just do. Take driving as an example. When learning, you'd have to remember what the signs said, where to look and when, the speed limits, and so much more. But when you have experience, you can just do it without effort.
This can be applied to skills. Practice (some skills) enough, and they'll become 'automatic'.
Memory Mechanism 3 — Overlearning: Practice Beyond Perfect
Practicing beyond what is require to perform adequately can increase the length of time that memories are stored.
Memory Mechanism 4 — Mnemonics: A Picture Retains a Thousand Words
The more specific, the better. They are usually designed to remember very specific patterns of information. They usually involve translating abstract information into vivid pictures or spatial maps.
A common method is the keyword method: take a foreign-languaged word and converting it into something it sounds like in your native language. For example: s'il vous plaît in French (please) sounds like like silver plate. Then you combine the two words into a bizarre mental image. This method is especially good with language learning, but can be applied in most places.
A good book related to this topic: Moonwalking with Einstein by Joshua Foer.
Good methods for memory intensive subjects: active recall, spaced rehearsal, and an obsessive commitment to intense practice.
"Do not ask whether a statement is true until you know what it means." — Errett Bishop, mathematician.
When you think Intuition, think Prof. Feynman. There is no clearer example.
A principles-first way of thinking about things is clearly the most effective. But it's hard to develop. You must have enough experience with problem solving such that you can build up a deep mental model of how other problems work.
Intuition might seem magical, but it's really just incredible amounts of (organized) experience.
To truly understand this, take a chess grand master. They have astonishing amounts of boards/plays memorized. It's like a library of patterns. This is the same with intuition. People with great intuition (like Feynman), they have a huge library of patterns such that they always see something as 'another one of those'.
Spending lots of time studying something isn't enough to create a deep intuition. But here are some guiding rules:
Rule 1: Don't Give Up on Hard Problems Easily
Rule 2: Prove Things to Understand Them
Rule 3: Always Start with a Concrete Example - processing information (not just paying more attention to it) makes you remember it better. So creating your own examples will help you process something deeply.
Rule 4: Don't Fool Yourself — Feynman said: "The first principle is that you must not fool yourself and you are the easiest person to fool." Think of the Dunning-Kruger effect here; thinking that you know something which you actually do not. A good way to avoid fooling yourself is to ask a lot of questions. Not asking questions, in order to appear 'knowledgeable', isn't such a good idea.
Taking this straight from the book.
Write down the concept or problem you want to understand at the top of a piece of paper.
In the space below, explain the idea as if you had to teach it to someone else.
When you get stuck, meaning your understanding fails to provide a clear answer, go back to your book, notes, teacher, or reference material to find the answer.
This technique is great for things you don't understand at all, for problems you can't seem to solve, as well as just building a deeper intuition.
When trying to understand things that you don't understand at all, sometimes you might need to have the reference material by hand. That's OK. Just make sure to thoroughly and deeply process (and explain!) the concept.
When solving problems that you can't seem to solve, it's good to really go through step-by-step and explain — not just summarize. Go deep.
The third thing — just building a deeper intuition — is great for things that are really important to deeply grasp. A good idea is to use visual intuitions and examples here, as well. Analogies work too. Explain it as if you were explaining it to someone who knows nothing about it.
"Results? Why, I have gotten lots of results! I know several thousand things that won't work." — Thomas Edison
When starting to learn something, often it's enough to just follow the path of someone ahead of you. Learn from the experience of others.
But at some point, you're ahead of the curve, and your teachers (or materials) become few in number. So at some point, you'll have to experiment.
In the early phases, you learn a lot. But when getting even better, you'll have to unlearn things as well — things that aren't efficient. However, you also have to learn to solve problems that you couldn't before.
A novice and a master may be able to solve the same problems, but the difference is that the master will be able to solve it in a more efficient way, and in a way such that it won't cause headaches later (for programmers; fixing bugs, optimizing inefficient algorithms, etcetera).
So at some point, you'll have to go beyond your comfort zone and start experimenting. Trying new things.
It's important to realize that your learning styles are not fixed. You do not have immutable strengths or weaknesses. You can do it.
What topics are you going to learn? Approximate scope?
Start narrow — you can always expand
Which resources will you use?
Have everything ready before you begin.
Find a benchmark for how others have successfully learned this skill
You don't have to follow their footsteps. It's to make sure that you're not missing something important.
Which direct practice activities will you do?
Think about how you might use the skill. Starting to think about this early is good.
Backup materials and drills
Get some backup materials and possible drills.
Decide how much time you're willing to invest and when. This is also good so that you prioritize this project over other (less important) things. Be specific.
Figure out the length of the project. If it's too big, you can split it into smaller ones.
Ask yourself questions for each of the 9 principles to know if you've fulfilled it enough.
After your project; review! What went right? What went wrong? What should you do next time?
Knowledge decays without intervention. So what will you do?
For maintenance, it's enough to have regular practice. Maybe even integrate the use of the skills into your life.
For subjects that you learned that you won't use, you could just 'let them slide' and relearn them later if necessary. It's easier to re-learn than learn something for the first time.
For mastery, you can continue learning at a lighter pace or do another ultralearning project.
Ultralearning is a strategy. But you don't always have to learn everything with such intensity.
Alternative Strategy 1: Low-Intensity Habits
When you've learned something, you can keep learning by just building slowly. For example, reading books at a normal pace.
Alternative Strategy 2: Formal, Structured Education
There's lots of different, good reasons to undertake formal education. Ultralearning shouldn't be seen as a replacement, but as an alternative — or complimentary. You can use the strategies while undergoing a formal education.