Ask HN: What's your learning strategy?

This might sound kind of bizarre, but I don't really try to learn. I think I have a minor learning disability where I have a very hard time remembering things that I deem (consciously or not) as unimportant in the long term. Instead, I try to focus on extracting the nature and patterns of relationships, instead of the specific details. An example might be: learning the purpose of a wristwatch and its power source, versus understanding how all of the gears work together. Unless I'm working on a specific problem involving how the gears work together, I won't remember it. Even then, I'll put it in my working memory just long enough to solve the problem and then forget it. But the patterns of relationships stay forever, and it's easy to associate with other patterns of relationships, once you forget the specifics.

In tech this can be an obstacle when communicating because many engineers have an extremely fine-grained memory, seemingly all the time. I will hand wave different concepts as a black box that we don't need to know right now, where they are happy to explain them at length, in detail. But if it's not relevant, I just don't care. They will also expect me to explain concepts that (to me) aren't relevant to the problem, and I have to say "I don't remember."

Go back and forth between theory and practice. If it's something practical like programing or music that means applying the techniques to make something. If it's something you can't really practice like history, that means explaining what I've learned to people. Also use a variety of material about the same subject (book + online articles + videos). Taking notes after having completed a piece of material is a form of practice. Taking notes while consuming material should be avoided, focus on the material and try to really understand it. Of course all of this is iterative.

I create the shortest possible time loop between consuming something (text, video, etc.) and applying it to a real project or problem. Then shortly after applying it, perhaps at the end of the day I'll write about the process as if I were explaining it to someone who doesn't know about the topic beforehand. If I can't write about it while clearly explaining it while having a complete and total understanding of every word then I go back and research anything on demand to fill in the gaps.

The idea is you consume and apply, then quickly turn a combination of theory + practice into your own words which is proving to yourself that you understand it. The key point there is "you", you don't need to publish the written component to anyone. Its purpose is to force you to come up with unique thoughts to combine this new information into something coherent which I'm sure has some benefit in making your brain remember it (but I have no scientific proof of this). It can also be used as a personalized detailed reference in case you ever forget the details which IMO is much better than scattered notes you tried to quickly write while you listened to or read something.

I don't know if this is a formal way of learning something. It's something I've done for most of my life because it felt intuitive at the time and still does.

As a working adult:

- don't take rigorous notes because it ends up replicating a lot of the material that I'm consuming anyway

- still write some notes if something interesting stands out; it's usually related to some core idea/explanation, not to technical details

- don't obsess over having "complete" notes or worry too much about organizing them

- try to write some relevant code alongside whatever I'm learning

- spaced repetition is for memorization; I'm not trying to memorize

This is different from how I used to learn for exams at school/university where I would repeatedly re-synthesize notes to a shorter and shorter format until I can reproduce the material from a cue.

I think the distinction between the two modes is important; it seems very weird to me to see knowledge workers putting so much effort to memorize stuff that they can look up anyway.

Learning is an exercise in creating space in your mind to represent a new map of interrelated symbols.

When you start something new, all the symbols are "too close together" and it takes time and repetition to increase the "resolution" on that map so that you can put the symbols into position with more precision and not have them confused with one another. You don't yet have a good intuition for the significance of each symbol and it won't come from being told, because sources have biases - frequency over time and space is a much stronger signal. It's the same with connections between symbols.

Doing, rather than just reading, is important; you need to try and find that Goldilocks zone of growth, where if things are easier you're not learning but if they're harder you lose hope. Try something just beyond what you know how to do, page in new information on demand when you get stuck (from docs, people, source code, whatever it is).

Writing notes helps me personally creating the space for symbols and reminding myself of what I already heard about but I don't know that it's necessary for everyone, and I especially think I've found it more useful as I've gotten older (and know more stuff already). Sometimes when I learn something "new", I pin it close to a concept I already know, which dulls its novelty and I'm probably more inclined to forget it, until I'm reminded of it again. And this is down to that "creating mental space" idea again.

Repetition increases symbol salience, particularly when you're surprised by the divergence between what you thought something meant and what it actually means. The risk of simply reading, and not doing, is that you're not surprised often enough. Doing stuff keeps you honest.

A research paper called "Improving Students’ Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology," which is backed up by experimental studies, can serve as a great introduction to develop a learning strategy [0].

The summary is that doing practice tests (e.g. creating flashcards in a question-and-answer format, or regularly doing practice problems) is highly effective, along with studying over a longer period of time instead of cramming. Switching between learning subjects ("interleaved practice") is also effective, along with creating explanations in your own words about a text, together with asking yourself questions about the text. In contrast, highlighting the text and rereading as a way to learn is relatively ineffective.

[0] Full text [PDF]: https://pcl.sitehost.iu.edu/rgoldsto/courses/dunloskyimprovi...

Giving an improvised lecture to an imagined audience in my head on a topic I am studying I find a very useful way of consolidating and systematising my understanding, and of finding gaps to be filled.

I recently wrote about my learning strategy under a post called “How Learning Works: Components, Loops, Systems”: https://romandesign.co/how-learning-works-components-systems...

> It's like Charlie Munger once said about mental models: “All this stuff is really quite obvious and yet most people don’t really know it in a way where they can use it.”

The long and short of it is that you go through a cycle of Knowing, Doing, Assessing. Each of those has sub sucked and pieces but they fit together really well and work generally across different types of things you’re trying to learn (intellectual, skill,ms, physical).

I'm a mapper [1], and as such, I tended to seek out information that matches some edge of what I know, and when I find it, I'm then able to incorporate that knowledge, and check for inconsistencies with the other fact I know. I just keep reading, and browsing until I find the right puzzle piece, and then I understand.

I like instructions that specify how things work, so I can route around problems, and understand what the goals and mechanisms of a system I'm using. I can tolerate a wide variety of discrepancies in those instructions, and work around small changes in procedures that might not be documented.

If I'm being tested on general knowledge, and how things work, it is stuffiest to read and understand the materials provided.

Practical Example - Right now I'm trying to understand the fine grained details of how quantum computers work, I've watched a lot of boring youtube with words like Hamiltonian, Eigenstate, etc... none of which has clicked, but all float around like loose puzzle pieces... eventually I'll find the detail I need, and a connection will be made, I'll then go back and find other pieces until I understand the landscape at the detail I want to know.

If I'm being tested on random facts (dates, buzzwords, etc) then it's time to break out repetition and wrote learning. This is also the mode when learning a new skill, practice, practice, practice.

---

If you're a packer, the opposite is true. You'll likely keep a lot of notes, and be good at memorizing facts.

I've worked with packers in the past, they like clear, precise, easy to follow instructions that very closely fit what they will see as they follow them. If the instructions don't match what they see, they will end up needing help, or get frustrated.

I'm not sure how packers best learn. I suspect it's repetition all the way.

---

1 - https://wiki.c2.com/?MappersVsPackers

One trick I have for taking lots of notes rapidly is the following:

Say somebody says a long string "The Cap Theorem is Consistency, Availability, and Partitioning," you write down the initial letters with space to come back and complete the words: T__ c_____ t_____ i_ c_______ a______ p_____. Once you have the initial letters down the hardest part is done. For very long sentences you want to remember this helps a great deal because you will still remember the last several words with the hint of the first letter when you do a second pass to fill the (rest of-the-word) in.

Diving into the deep end on projects, with a practical or interesting goal in mind. Along the way, pick up the needed skills, reference papers, tutorials, etc. Take an online class if it seems that would help. Ask for help when required.

I find that when taking classes, reading papers, articles etc, I learn far better if I can #1: Immediately see where the material is useful due to having encountered a problem it solves, and #2: Immediately put relevant information into code/CAD/notes etc.

I follow a very simple formula nowadays.

1. Always be reading.

2. Teach everything you know.

3. If you don't know, learn enough to start teaching.

4. Take notes when you feel compelled to.

5. Use sleep as a memory tool. One nap during the day, and before going to bed each night think about what you're learning or want to learn more about. You'll wake up with more clarity every time.

The best book on this topic is probably "Learn Like a Pro" which came out last year.

I think notes are generally a good idea to help you summarize the knowledge you've acquired while spaced repetition is the aspect that will help in retaining the knowledge you've learned. The important thing is that the notes are written in a way you understand and can be picked up and read-again with a limited amount of context. What I mean by 'limited context' is sometimes I forget big portions of what I've learned and writing notes as if I'm teaching someone who knows very little about the subject allows my future self to pick forgotten things up faster. I personally use one of the Anki apps to do this myself.

The best form of spaced repetition IMO is the application of the acquired knowledge in a big project over time. That way re-visiting the concepts comes naturally as you continue to develop something over time.

Can you do it `forwards', `backwards', `inside out'? What these mean depends on what you are trying to learn.

In programming, examples would be:

• writing code and reading and understanding other people's code.

• `executing' some code with pen and paper

• translating a formal statement of an algorithm into code

• translating some code into a formal statement of the algorithm

• translating code between two languages

• implementing the same functionality in as many different ways as you can think of

In learning scales on a musical instrument:

• ascend, then descend

• descend, then ascend

• start/end on notes other than the root

• all of the above but whilst playing the scale in intervals or arpeggios

In learning the times tables, examples:

• what are the factors of 24?

• recite the multiples of 7 backwards from 70

Two responses, as the context differs dramatically.

For coursework, I would take notes. At the end of the day, I would reread the notes and complete them. Completing them means correcting errors, filling in details from examples, and using assignments and course material embellish the notes. Embellishment means adding notes of personal interest from other course material, and also adding examples from problem sets. My notes were my source of truth, and I kept them in good order. For exams and future coursework, I would only use my notes since they were complete.

In my professional life, I consume a lot of material but do not take notes. I find that there are two important aspects to learning. The first is high exposure to material. Being able to put a name or provenance to an idea is helpful in finding my way back there again. The second, which is similar to what nickjj mentions in a comment above, is to close the loop. To wit, my engagement with material is to find aspects for which I can test it outside of the scope presented or examine in in contrast to another idea. By engaging with the idea, I am able to come to an understanding of its verisimilitude and utility to me.

In the later, I do not take notes while attending lectures and reading. However I maintain complete lab notes for my work. Lab notes include random thoughts, references, uri:'s of data sets, links to papers and wikipedia articles. The lab notes are written for me, and are mostly in chronological order. By the contain as thorough an appendix as I have the patience to maintain.

Rampant unmedicated ADHD.

Sounds like a joke response, but it isn't. My focus will settle on something because I want to do a specific thing in that area, and it just doesn't let go, even when I need it to. As a perfect handy example, my brain is currently full of a) chess, and b) compliant mechanisms, when it should be full of Rails/React interview prep.

I guess the transferrable principle here is to have something concrete you want to do with the knowledge you're gaining so that your brain has a tonne of hooks to hang the theory off.

My learning strategies are very personal, subtle, and hard to explain. I believe I build most of them a long time ago after reading Mind over Machine [1]. It's a critique of the, back then, glorious promises of AI, but it has a very thorough take on what learning is and is not.

[1] https://www.amazon.com/Mind-Over-Machine-Hubert-Dreyfus/dp/0...

I have been really obsessed with this one.

Somehow I failed to learn things when I was taught in "official ways", yet I managed to learn the language by being brutally forced into practice and environment that pushed me to make it happen. I moved countries when I was 12, they put me in school straight away without knowing how to speak (although I have been studying for a few weeks, at least the basic grammar).

The language group has been similar to my native tongue, however the alphabet was different (as I was used to cyrilic). Nonetheless, as a kid - I already saw some patterns and similar rules.

I managed not to fail my class as well as spoke and wrote really well after 2.5 years. After 5 years I was better than most of the native tongue speakers in terms of grammar and on par with their vocabulary. Obviously using the language every day and being pressed to do so contributed.

Years later I started learning German in high-school... was so bad it and never managed to even get to the basic level after 4 years. I moved to Berlin again years later but because English was already quite prevalent, it never pressed me to learn German and I never did.

Somehow the most effective learning curve was just doing the thing, making a lot of mistakes and I never took textbooks into account even when doing so.

The methodology was just deliberate practice and tacit knowledge in the end.

I take notes, as energetically as my interest allows. While I do read them subsequently, I do not really make too much use of them. For me, the act of taking notes strongly helps my mind connect to the information in ways that make me know it better, sooner.

I went through a bad patch as the nineties ended and I thought that computer note taking might be a good idea. It produced pretty good notes (I type well) but was much less conducive to learning. In fact, I felt it was distracting and counter-productive.

Handwritten notes. Definitely my first line of offense against ignorance.

After that, teaching it. If I can find an opportunity to make a presentation. That's great. I will always confirm understanding and level up when I explain what I know in a semi-formal way. Explaining to colleagues and friends is good. I didn't do study groups in college but found analogous social activities very helpful since.

After that, doing it. If it's a practical topic, putting the lessons into practice asap is very helpful. Like, I'm sure, many of you, I am a programmer. If I start using a new tool, technique or idea, I will soon have solid understanding.

Of course, I also keep going. If I do not keep seeking progress in a topic domain, then any knowledge I have will fade.

This is my learning strategy which originally posted on [1]:

I have created a learning framework for myself to learn literally anything. First of all It’s not going to be easy to get started learning something new because there are many holes in your mental model about the material and the aim of the learning is to fill those holes. We’re living in an information era. You can drawn yourself in surfing the web/docs/articles if you are some kind of perfectionism to convince yourself that you’re learning. But that’s the trap. You should spend about 30-40% of your time learning the material and 60-70% of your time applying it and then under the process you’ll find: “Oh I don’t know how or what is X” then you go to learn about X specifically. This works best. In other words as you will be applying what you have learned so far you start “learning on demand”. In other words filling the individual holes. This loop will force you to learn the topic.

[1]: https://news.ycombinator.com/item?id=29116026

In my opinion we are bad at deciding what knowledge is important and what ought to be forgotten. I leave this to my subconscious. If I forget something that means I wasn't using it anyway, why try to artificially retain a piece knowledge?

So my strategy is to work first and learn second. When I encounter a roadblock I try o learn something new, but in most cases I then quickly forget about it soon. This might happen several times over. Once I had to learn about unix file permissions, I knew all the commands and numbers and what they meant and it all made sense. But now I only remember `chmod +x` and that there is a command named `chown`. This satisfies for my day-to-day needs and is enough so that I can re-learn the thing if necessary.

And for the general context that are important to know in detail I found that teaching is a good way to remember and reinforce. I teach a course every year and every year I am surprised how much I've forgotten. But it quickly comes back after I start preparing materials for the class.

When I was young I used to think that you can learn a thing once and then remember. Now I know that, speaking strictly for myself, this is not the case.

I don't know if I have ever learnt anything or not and at this point I just know that no matter what I do, I am never enough. I have a knowledge base of concepts and mostly importantly where to find these concepts. I sometimes wished I had found SRS earlier so that I could have kept revising the things I learnt.

After that spoiler, here is how I learn things:

1. Get the difficult stuff and read through it as long as your brain can parse it.

2. Once you have the first overview, rest and visit the same thing next time.

3. On second attempt depending on the context: i. If it is problem solving: Attempt to solve problems (look at examples and try to establish relationship) ii. If it is theoretical: Note down the key assumptions, key steps, key proofs.

4. Until you are confident, loop 2 and 3.

5. After 1 month or so, you will forget what you have then, thus go back to step 1.

Yes, this is highly inefficient. I really want to break out of this loop. But the issue is I am so accustomed that my brain goes into this mode by default.

The result is I get my own version of understanding which may not be the actual true understanding. But it is as close as it can get (70-80%). So I am happy about it!

This is something I've thought about a lot; I ended up compiling tips on learning from a variety of different sources into a master document that I go back to every once in a while to see if there's any new things I can implement [0].

I only integrate one or two practices at a time, and I stopped short of implementing SRS or a Zettelkasten because I began to realize I was spending too much time on productivity to actually get anything done. I'll probably return to them later - I think SRS especially gets a bad rap from people who like to think about things conceptually as "just being rote memorization", but IMO the human brain is not a computer for Platonic objects. It thrives on hard data and real-world examples, and "memorizing" components or instances of a particular concept so they're at hand for effortless retrieval just makes connecting them to existing ideas easier (which is arguably all understanding things is).

[0] https://pastebin.com/sk6iNDpp

SRS works well enough for me. At least it's near the top in the ROI optimisation.

But remember that it's just a scheduling mechanism to save you time, not to help remember, necessarily. I like the Harry Lorayne book on memory for various tricks and techniques to actually remember. It's not groundbreaking but goes over a lot of the standard and proven memory hacks.

While I've generally been pretty good at picking up things by just doing them I was always a just ok student. Later in life when I wanted to start learning some things outside of work I took the coursera course "learning how to learn" and found it very helpful. It has the strategies above plus others. You can find links to the course as well as books by the instructor on her website - https://barbaraoakley.com/

Specifically though I use different methods for different materials. For some specific things I need to memorize verbatim I use hand written index cards/spaced repetition. For learning math, programming, etc there is no substitute for doing example problems with solutions you can check. For general reading I write my own summaries, notes.

The best way to learn is by doing.

I needed to learn javascript, So I created a web app https://text2db.com/ , which coverts simple text to database script. It took 3 months to make it. A long time but I learned alot.

How the brain handles memory is pretty complex, but specific recall of random factoids is not something the brain is set up for. If you don't work with the information day after day in different scenarios and there's no strong sensory or emotional tie, you won't have immediate recall. You can instead try tricks like Ken Jennings' links-in-a-chain (https://www.businessinsider.com/ken-jennings-jeopardy-how-to...).

You should probably just keep notes and refer back to them.

> Do you have a methodology for extracting common themes/patterns across topics?

Your brain does this naturally if you get a lot of sleep and don't think about it.

I do math, and most of what I know about learning is derived from my experience with math. YMMV.

Homework. So much homework. Do all the exercises in the book. Taking notes helps me focus on a lecture, but I might as well use a brush with water on slate -- I rarely ever actually look at them a second time. I do not learn by watching. I recall a few nuggets that my brain can reach for when doing the work. I learn through application. It's common in math that students do not understand the contents of a class until they're halfway through the next class that depends on it. That's where the integration (unavoidable pun acknowledged) happens.

1. I make mind maps and do inline annotation a la marginalia.

2. I haven't figured out a process for compilation but my plan is to condense big concepts into readable texts with the intent of sharing them with others.

3. Spaced repetition works, but it is labor-intensive. I use a combination of spaced repetition, mnemonic images (like Sketchy.com), memory palaces and reflective journaling to process my hands-on experiments.

4. I can combine my mind maps to see connections between texts. For example, maybe one book talks about scalability in passing, and another goes in-depth. If I compare those two maps I can see the gaps between each and it helps me see how the two texts are related.

I just note terms I dont know. I write one or two sentences about concepts. Sometimes, I will go back and rewrite/group common ideas if I notice commonalities. It all depends on the subject matter. When I studied neurology, there was a lot of memorizing- thats just how it is. When I studied philosophy, it was distilling big ideas down to concepts I could remember. When its a completely new subject, like UX design I'm studying now, I use immersion- I took an online course that was a broad overview, wrote down terms I didnt know, subscribed to rss feeds on the subject, found groups, now I'm taking a more detailed course online.

For me repetition is the way to go. Taking notes helps since I have to organize my thoughts to put them on paper, but in the end the repetitions are the things that counts. Things that I use most often will become almost like a muscle memory. Things that I've written down: I will not know how to do them right away, but I'll remember that I've encountered it somewhere and I know that it can be done.

0. My strategy is lots of unsucceses and result is close to nothing but i keep getting fun from reading any works which fits to my interest. 1. I do take notes when doing and do not while listening. But i write down some words to google and some professors jokes if funny. 2. Do not understand the thing. 3. Definite yes, our nature is to be forgetful to anything not repeating. 4. My old approach was just a desire to look for some special, but few days ago i have tried zettelcasten and it is surprisingly neat for that task imo.

The only reliable strategy i know is deep detailed understanding of why and what exactly i am learning.

Be it just passing stupid ielts test or getting a job - once I invested enough time to build a concrete picture of what outcomes do i need to get and why - everything else including motivation follows organically in the most efficient way I’m capable of.

I believe there is no single best strategy of learning for a single person. Since learning is as diverse as the subjects of learning might be.

Obsidian.

I copy/paste what I think are the "atoms" of the thing to learn. And i build a small graph of interconnected .md files in Obsidian.

To me, it is very important to break the medium that brings the information into a graph that suits you the most. (call that process my knowledge ETL, and the result .md files my personal knowledge graph :)

PS: the funny thing is that you indeed refactor your graph when you come back to it with more knowledge. You usually do not remove things, but reorganize it in a better way. Very interesting process...

1) no notes

2) n/a

3) yes, sort of. Learning something an hour per day over 10 days > learning something for 10 straight hours in one day

4) no method, but I tend to naturally go just below the surface on many different topics, so I think that helps.

A key thing for me is that I learn best by doing.

So maybe watching 5-10 hours of a structured "course", usually at 1.5x speed is about my limit before I have to get started on my own project applying what I'm learning.

After that, it's mostly blog posts, GitHub, and Stack Overflow to learn the more advanced aspects.

I use notion. I have one doc that's built out as a learning framework. I categorize the different areas I'm learning about. As I come across an insight (blog, tweet, book, news article, podcast), I put a link and some quick notes about it in there. Then, I take time on the weekend to go through and synthesize and pull out the signal. I'm also fortunate to be in a relationship with someone who's into learning just as much as I am. We talk about these things.

I’m a visual learner so when I take notes I use colors. If I need to be tested on the content after I finish the course I rewrite my notes in a more concise way and I organize them all in a quick reference / cheat sheet format. I may not need the notes but just rewriting, summarizing and reorganizing the content makes me remember the images of the pages better.

The next best thing is explaining the topic to someone who has no idea about it.

> Do you take notes?

I used to. 90 percent of them were about tools or other things to look up later. Now, I add those to the backlog, and the other 10 percent become Anki cards. These days the only things I write down are meeting related: action items, questions I want to find an answer to, and maybe 1 piece of key information I want to refer to later.

> Do you compile/rewrite notes?

Nope! I may adjust Anki cards if the phrasing turns out awkward, or if there's a term collision. And if a card is tagged 'leech' thats a sign to rework it but generally I'm not trying to memorize things word for word where this might help.

In practice the hardest cards, the ones that get marked leech, are the bill of rights. Especially the middle ones that are basically a paragraph joined into a single sentence. Like, I know a warrant needs to be supported by evidence but its not always immediately front of mind that this is a 4th amendment right (instead of say 5th or 6th), or the exact phrasing. Fortunately this is a hobby topic rather than a professional need -- and hopefully if it were I'd be better at it =)

> 3) Does spaced repetition work for you?

I think so. For papers I generally extract a few key topics or points. Today's was three sentences about Cox proportional hazards, after hearing about his most famous contribution in an obit thread.

4) Do you have a methodology for extracting common themes/patterns across topics?

No? My general approach is to spend a few minutes thinking about how I might apply it in my work, and the go looking to find out who did / what they found. So for proportional hazards, I ended up googling and found a few places where this was applied, like analyzing churn rate (https://towardsdatascience.com/strategies-for-customer-reten...)

But note that I'm not trying to merge two topics of study in order to unlock some deeper truth, or better learn a general pattern. I'm trying to advance the state of the art in a field I consider myself pretty good at by learning how other fields work and stealing their best ideas. Kind of the inverse of the Hamming / Feynman jumping fields every five years.

Carry a notebook around and frequently take notes for things to look back at. The wet concrete of the brain needs frequent imprinting for success in memory/recall. One notebook for career, one notebook for side business and projects. Have tried multiple notebooks, 2 seems to be the balance for me. Different colors. Keep colors consistent over the months/years of getting second and third ones.

1, 2) I always take notes, but then I discard them. It's just for mental correlation;

3) Never tried it, curious but not that curious;

4) Take long walks and think about nothing.

I don't really fell that I have the space to incorporate notes in my type of learning as I love reading about all sort of random stuff and getting to the peak of Mt. Stupid [0] and kinda leaving it there unless it's of real interest to me.

My biggest challenge is how to decide what to learn. And then how to keep at it without letting the FOMO in. It's an eternal struggle.

Just my 2 cents, this is what I find working for me:

- I found that I can retain much better if the course has readings. It can be lecture notes in textbook form (not presentation), or just a textbook. I found that without this, if all they provide are presentations, or blackboard work, then it is very hard to retain what I learnt long after the semester ended. (You could say I should create my own notes, but paying to learn is essentially to be more time efficient. I really don’t have time to create my own “textbook” from their blackboard work. I envy those that can transcribe in LaTeX in real time.) - creating cheat sheet (even if the exam don’t allow them and just for personal reference) is useful to keep the big picture, make you be successful in the semester, and can be useful in the long run. But especially if you are page limited due to exam requirement, the cheat sheet can becomes cryptic for your future self. I think this is a must for any course. - I lately also creates personal wikis to hold my knowledge. Basically Wikipedia is a very successful medium even for stuffs you know, you can look it up there to retain your old stuffs. Personal wiki is basically try to recreate the Wikipedia experience tailored for personal use, meaning only for relevant stuffs (say san history), or even links to various useful sections and other site such as math world. One tip is that in Chrome there’s an extension to force MathJax on Wikipedia, and with that you can click on a Math equation and copy and paste LaTeX in your personal wiki. (As it is personal, there’s no copyright issue, but if you share make sure the license is compatible and cite.) - 2nd brain: the last point about personal wiki may be similar to what some might call a 2nd brain. I still didn’t attain it (my “database” is fragmented in various places.) But I should learn more about that and may be one day I can build one.

What I said is more about being able to retain knowledge in the long run. Eg the real world doesn’t care if you make a search to find stuffs. So what I said is more about how not to let what you learnt be wasted because you can’t retrieve them.

Lastly, I find teaching very useful too. Being about to teach makes sure you truly understand a subject, and even reorganize it in a manner that’s easier to understand. But often you can’t control what you teach. You can’t say today I want to master Stochastic Differential Equation and let’s teach a course about that. But on the other hand, try to spend some time teaching some concept you find hard to grasp can be useful (audience and feedback like questions can provoke your mind to finally click.)

if it's a book, then I note down key points after every chapter. if it's a video course I just passively listen everything then take a practice test, then see where I am struggling then rewatch those videos (problem is that lot of times videos are NOT worth taking notes and you only realize after the end whether video was worth it or not, so it kills me to be taking notes continuously)

best combo I found was text to speech feature of online textbook, I'd listen to it and if I found anything even remotely important I'd stop the recitation and copy the text in notepad or something, and by the end id have compilation of important notes. The act of highlighting/copying/pasting was enough to form a good memory.

Getting high-level understanding of anything first and then looking and retaining details if required. If not required or not relevant, focusing on things that matter. If forget, revisiting it again (easy when you have high level understanding already).

Andrew Huberman has some really good podcasts describing how learning habits and how you can change you day to day to help learn more https://youtu.be/hx3U64IXFOY

No notes, just binging hands-on immersion iterating on short feedback loops.

This will make you look like a „genius“ to outsiders, but it only works on subjects that allow for short feedback loops.

(example where it works: coding, example where it wouldn‘t work: raising a child)

If you want to meta-learn, there are theories about how to learn out there; for example, https://doi.org/10.1038/nrn.2015.18

Gain familiarity and background (exposure to facts and ideas without struggling to understand them fully). Dig into particular paths of interest. Get my hands dirty--or discuss with someone.

In order to understand things I need to use what I learn. So I try it out and if it's useful I drill it until I remember a few ways to apply it.

If you find a problem that you are obsessed with, you will learn lots, whether you want to or not.

Use what you are trying to learn.

I like to use the phrase "practice makes permanent" - not "practice make perfect", cause every instructor disagrees, and not "perfect practice makes perfect" because perfection is subjective and a fool's errand. In a nutshell, whatever you do in practice will be what you do in application.

This leads to then what I consider "foundational knowledge", or refining your understanding on Bloom's Remember, Understand, and Apply categories. I think we try too hard to push students toward the Analyze, Evaluate, and Create categories too quickly, which results in poorer performance, especially in the classroom context where the next course assumes "mastery" (another term I dislike, I prefer proficient). Once a student is proficient, then we can talk theory all day.

So, in my classrooms, I mirror the methodology I commonly see in martial arts - warm-up, demonstrate technique, drill the technique, and (depending on the art) spar, or apply the technique. If you are struggling at one of these steps, then we can identify the issues and provide intervention practice at the appropriate level. Traditional classrooms don't really allow for extended spaced repetition, but I do believe it helps form autonomy.

More specific to your question since most of the above is just my general opinions on learning:

- Notes are good and writing notes is better than copy-pasting them [1]

- Highlighting your notes for important details is a lower level practice that is also beneficial [2]

- If you are given a technique/algorithm, attempt to do it on additional practice problems. If you can, through spaced repetition. For example, if you were learning the A* Search algorithm, change the starting and goal nodes in your current examples or make a new one with randomized numbers.

- Chi's ICAP Framework has a few additional points as well on which lower level practices are better than others [3]. However, as I'm arguing in my dissertation, better overall may not be better for the individual. If a student is struggling in a Constructive activity, I'm arguing it may be better to 'downgrade' them to a lower-level activity to reinforce understanding (again, my thoughts toward foundational knowledge).

[1] https://dl.acm.org/doi/pdf/10.1145/1240624.1240773

[2] https://www.sciencedirect.com/science/article/abs/pii/036147...

[3] https://files.eric.ed.gov/fulltext/EJ1044018.pdf

If I want to actually learn what I'm reading ... I basically follow Louis Frenzel's advice from his column[0] in Electronic Design years ago (HN post[1]):

--------------------

All learning is self-learning. Professors, trainers, and all teachers just organize and present the material to be learned. They don't teach it to you. You learn it. You're the one who actually absorbs, understands, and assimilates the knowledge by listening to the lectures, reading, thinking, solving problems, and other activities. Self-learning is a natural, human quality. While most of you have used this method in the past, you may want to do it on a more formal basis to speed up and fine-tune your methods. Here's a suggested approach (and trust me on this, you must write it down):

- Clearly identify what you want to learn. Write it out.

- Write some learning objectives for yourself. These statements clearly identify what you want to know and be able to do. For example, you should write something like "When I complete this learning assignment, I will be able to design and program an FIR DSP filter." The objectives should be expressed in "behavioral" terms, that is, using words that state some measurable outcome.

- Identify some initial resources. Start with books at the local bookstore or go to www.Amazon.com or Barnes & Noble at www.barnesandnoble.com. Most cities don't have good technical bookstores, and it's tough to find anything at regular bookstores. Consider yourself lucky if you have a good technical bookstore or a good college bookstore. Plan to get multiple books to give you greater breadth of coverage with multiple explanations, examples, and perspectives. Don't forget to look through your stack of magazine back issues.

- Check out online sources. Do one or more Web searches, or go to relevant company Web sites. You may run across an appropriate tutorial, white paper, or application note that will give you what you need. The large semiconductor and equipment manufacturers have tons of stuff on their Web sites, so start digging. Also check out the professional societies and other sources listed in the tables and sidebars.

- Watch out for any conferences or seminars on this topic. Usually, such events never occur when you need them, but you might get lucky. If you find one, attend because it will provide a big head start for your own learning.

- Organize your materials. Lay them out, mark them up, and then make an outline based on your objectives. See what you have and what you lack, and make an initial list of things to do.

- Dig in. Set aside an hour a day or whatever you can to go through the materials. Turn off the radio, CD player, and television. Make a habit of finding some quiet time to read and learn.

- Look for a human tutor. You could be working just down the hall from an expert on the very subject you're trying to learn. Pick his or her brain. Ask this person if he/she will help you understand and learn. Take this person to lunch or offer to pay for lessons. Most people will gladly share what they know, if you aren't too proud to ask. The best way to do this is to learn as much as you can on your own. Then, go for the professional, personal help with tough questions or when you get stuck.

- Include some hands-on. Is there any hardware you can buy or put together to help you learn it? Maybe there's some software that will help. Buy it or have your employer buy it.

- Write a paper or article or teach what you have learned. You have to know it to write it or teach it. There's no better way to learn for yourself than to have to explain it to others.

--------------------

[0] https://www.electronicdesign.com/technologies/dsps/article/2...

[1] https://news.ycombinator.com/item?id=30332471

Build

I think the only way to really learn is to do. So I'll skim through the content first without trying to retain knowledge

then I ask myself, is there anything interesting I can build involving this concept? or any useful activity I can do?

and I try to build it or do it, while using multiple sources of knowledge, but I never try to just remember stuff, the brain naturally remember the pieces of the knowledge that were actually needed the most while practicing. you don't know what's important to remember until you practiced, so in the end if you try to just retain knowledge you remember very badly a lot of things, but if you practice really early you remember very well what was actually important