For me, I always break down complex topics into smaller chunks and use the Pomodoro Technique. Also, teaching the concept to someone else (or even just pretending to) really solidifies my understanding.
- Focus on learning and retaining the fundamentals. Repeatedly, when possible. Why just the fundamentals? So I can keep a good mental model of how the systems work, in my head.
- Take notes[1] and document the fundamentals (for review), trivia, detritus, step-by-step procedures, and problems+resolutions for issues I encountered. I try not to keep this in my head, anymore. I have too many complex systems to manage at work. The days where I could keep everything in my head are long gone.
- When I'm taking notes, I try to write them out conversationally as if I were explaining it to a technically-adept person for the first time. Why? Because I'm technically-adept and often when I refer back to something I did a year ago, I want to remember and get back on track as quickly as possible. This approach will also help when I eventually point an LLM at them.
- I'm a visual person, so screenshots and photos and videos and whatever other visual information, as much as possible.
Small anecdote: speed reading is mostly bullshit but when I took a course it showed me how to approach books and new content in a different and more useful way. I usually just started from the beginning and started reading.
Now I look at the cover, back, chapter list, flip through the chapters a few times looking for high-level things that pop out, do it again, then look through the index for things that have lots of references. Then I can start reading the book, and I have a better idea of what I'm in for. Plus it's much easier to be engaged when you aren't so religious about the 'right' way to use a book.
That kind of thing has let me learn stuff I would otherwise feel too intimidated or unsure about.
I haven't taken a speed reading course but I've learned to skip read. I read the main idea of a paragraph and move forward. It's a good way to get the gest of a chapter so you have some idea of what the chapter is about as you read the detailed text.
For those curious, this kinda stuff is explicitly taught in Mortimer Adler's How to Read a Book, which is precisely about how to extract the maximum value from a book rather than speed reading.
One of the best classes I took in college was a "learn how to study class", I never had to study before college so I was woefully unprepared for engineering school. Probably the best thing I learned in the class was the Cornell Note Taking Method. I implemented the method the next semester of college and totally crushed it with limited studying, I probably should have continued using it but I was a knucklehead and didn't.
I glanced at it and it reminds me of my grandmother's system. She was born in 1918 and went to college in her mid-30s. Very bright.
Anyway, she took her lecture notes in shorthand. Shorthand is very fast, so she took down every word. She would go home and transcribe those by hand to regular English. Then she would type it out on her typewriter for a final copy.
She said by the time she did all that, she was done studying.
Make sure you understand before you move forward. I knew a guy that would read a text book cover to cover as fast as possible. His theory was that he could be ready for any quiz or test. The problem was that he did not take time to understand so the farther he got the less he learnt since he was not ready for the higher level ideas. All the reading was mostly a waste. Also, rereading does not equal learning. You need to stop and digest the info before moving further. So to make it easier take reading notes that you can review as you read new concepts.
I didn't know it at the time but that is what I used to get through college math.
I was always good at math as a kid so I never had to try, but I took time off between high school and college and it left me severely out of my weight class when college math started.
The Math 100 book started on chapter 6, as 1-5 was for remedial Math 99, so while I was struggling to cope with Math 100 stuff I was also working my way through the remedial chapters so I could catch up with my class. Took me 6 weeks of going to school and/or going to work, then going to Taco Bell and sitting down with a diet soda and doing math for 2-3 hours almost every day to catch up, but I managed to pull it off by doing every single math problem in the book.
Reminds me of my favorite dumb observation from grad school: the textbook is your friend.
I was doing ok in Abstract Algebra, but really wanted to bump up from a B to an A. So I spent the last few weeks slowly reading every chapter and doing almost every exercise. By the end of the semester I felt much more comfortable and got the top score on the final.
Only took me 20 years to figure it out, but reading the textbook and doing the exercises are effective ways to learn (duh).
Dive in, don't worry about perfect. Learn how to do each new thing you have to do. Incremental progress snowballs. Work on things you're bad at.
The hack is that you can log more hours per day than everyone else to go faster. There's not a super secret to learning things. Study, deliberate practice, and doing the thing.
Familiarity. Do something mindless over and over. Then do something slightly less mindless until it also becomes mindless muscle memory.
It’s not about memorization. It’s about repetition. Even things that cannot be memorized, like randomly generated puzzles, are solved by this. Through practice and life experience you learn what mindless things to practice and which to avoid.
Things that do not directly accomplish a goal but require intense learning are things best avoided. For example I have spent my career as a JavaScript developer avoiding the large stupid frameworks. This has harmed my career mobility but has allowed me to focus on skills most JavaScript developers cannot do: accessibility, performance, security, A/B testing, test automation, and really just about absolutely everything possible aside from some giant stupid framework API. There are pros/cons to not spending time with stupid. You are far less likely to be hired to do stupid beginner things, like put text on screen and pretend you are engineering, but once you get past that you are suddenly qualified for things that pay much more. This pattern applies to everything even unrelated lines of work.
I agree with your general point, but it is worth noting that chess is functionally a “randomly generated puzzle” and yet chess players do memorize patterns in order to perform better at these puzzles. Even if the activity isn’t entirely predictable, it’s still helpful to memorize patterns that you’re likely to encounter.
I have a somewhat controversial opinion about learning that I’ve been slowly forming over the last decade or so. I don’t buy it entirely yet, but I’m getting there. Here it is:
Reading traditional “narrative” books, articles, for information is extremely inefficient and almost entirely a waste of time. I have read hundreds of books in my life, and yet if you asked me to give a dozen facts from one I read a few months ago (for example, a biography of Rockefeller)…I can’t do it. I could tell you the rough outlines of his life, but I’ve forgotten the details. I put maybe twenty hours into reading this book and my return on that time seems to be minimal at best.
What’s the alternative? I’m a big believer in SRS/Anki and have found it to work almost flawlessly. And so I think a better alternative to reading for information may be to make a list of discrete facts from a book and put those into an SRS system.
Side note: I’m distinguishing reading for information from reading for enjoyment. I assume most people don’t read business nonfiction books for enjoyment and mostly just want to learn the information inside.
I can empathize with this opinion, but deep in me it feels (I am not saying it is) wrong. Maybe I love books too much. My fear is spending too much time optimizing life instead of living, so to say.
The value of Anki is clear, but I wonder how many hours it takes to create that list. Maybe that is time when I could be reading another book on the same topic, for example. Or I am rationalizing because I find it annoying to enter things in Anki (even a csv file) :)
The compromise I found is: if you own a lot of physical books, underlining (normally or with small vertical lines on the side) makes it easy to go back and remember the important facts. I also write my reaction to some sentences on the margin.
All of this helps me remember better, and it makes it easier to get back to it later to get the main points. It seems to work well enough for now.
I understand, and definitely love reading a lot, so it's not an easy realization for me either. But when I reflect on books I've read and find myself unable to recall most of the information inside, it does make me pause and reflect.
When I'm learning something there is a pretty simple trick I use which works for me.
1, I have the belief that I can understand the material. This is the most important part because if you do not believe you can understand the material then you will always fail to understand the material. If you do not have this, figure out why and address that first.
2, I will suck on a mint, since I was told as a child that sucking on a mint dilates the blood vessels in your head and gets more oxygen to your brain, more brain blood, more brain power, right? Honestly, though, it's more helpful for cram sessions and tests for some reason. If you're casually learning you can skip this.
3, Revisit the material a day or two later, especially if you had difficulty with it. This showed its hand in college maths a lot. I would fight for half an hour on a difficult problem at first, then come back two days later and try a similar problem and kill it in 5 minutes. Something about it being harder to carve a new line into a rock than to retrace it after the fact.
Doing this solidifies confidence in the new thing that you have learned.
4, After I have read and revisited the material, I will tell what I have learned to an imaginary child me as though I am teaching them what I have learned. If you can't teach it, you don't understand it.
Ever since I read this: https://cognitiveresearchjournal.springeropen.com/articles/1... in uni, i have followed the techniques mentioned to super great success. So much so that people are frequently surprised how I can know/do so much at work for someone so young.
24 comments
[ 3.5 ms ] story [ 71.8 ms ] thread- Take notes[1] and document the fundamentals (for review), trivia, detritus, step-by-step procedures, and problems+resolutions for issues I encountered. I try not to keep this in my head, anymore. I have too many complex systems to manage at work. The days where I could keep everything in my head are long gone.
- When I'm taking notes, I try to write them out conversationally as if I were explaining it to a technically-adept person for the first time. Why? Because I'm technically-adept and often when I refer back to something I did a year ago, I want to remember and get back on track as quickly as possible. This approach will also help when I eventually point an LLM at them.
- I'm a visual person, so screenshots and photos and videos and whatever other visual information, as much as possible.
1. https://obsidian.md with the Omnisearch plug-in.
Now I look at the cover, back, chapter list, flip through the chapters a few times looking for high-level things that pop out, do it again, then look through the index for things that have lots of references. Then I can start reading the book, and I have a better idea of what I'm in for. Plus it's much easier to be engaged when you aren't so religious about the 'right' way to use a book.
That kind of thing has let me learn stuff I would otherwise feel too intimidated or unsure about.
Anyway, she took her lecture notes in shorthand. Shorthand is very fast, so she took down every word. She would go home and transcribe those by hand to regular English. Then she would type it out on her typewriter for a final copy.
She said by the time she did all that, she was done studying.
While certainly not optimal from a speed perspective, it's no doubt extremely effective, and may even be efficient in the long-term.
I was always good at math as a kid so I never had to try, but I took time off between high school and college and it left me severely out of my weight class when college math started.
The Math 100 book started on chapter 6, as 1-5 was for remedial Math 99, so while I was struggling to cope with Math 100 stuff I was also working my way through the remedial chapters so I could catch up with my class. Took me 6 weeks of going to school and/or going to work, then going to Taco Bell and sitting down with a diet soda and doing math for 2-3 hours almost every day to catch up, but I managed to pull it off by doing every single math problem in the book.
I was doing ok in Abstract Algebra, but really wanted to bump up from a B to an A. So I spent the last few weeks slowly reading every chapter and doing almost every exercise. By the end of the semester I felt much more comfortable and got the top score on the final.
Only took me 20 years to figure it out, but reading the textbook and doing the exercises are effective ways to learn (duh).
The hack is that you can log more hours per day than everyone else to go faster. There's not a super secret to learning things. Study, deliberate practice, and doing the thing.
It’s not about memorization. It’s about repetition. Even things that cannot be memorized, like randomly generated puzzles, are solved by this. Through practice and life experience you learn what mindless things to practice and which to avoid.
Things that do not directly accomplish a goal but require intense learning are things best avoided. For example I have spent my career as a JavaScript developer avoiding the large stupid frameworks. This has harmed my career mobility but has allowed me to focus on skills most JavaScript developers cannot do: accessibility, performance, security, A/B testing, test automation, and really just about absolutely everything possible aside from some giant stupid framework API. There are pros/cons to not spending time with stupid. You are far less likely to be hired to do stupid beginner things, like put text on screen and pretend you are engineering, but once you get past that you are suddenly qualified for things that pay much more. This pattern applies to everything even unrelated lines of work.
Reading traditional “narrative” books, articles, for information is extremely inefficient and almost entirely a waste of time. I have read hundreds of books in my life, and yet if you asked me to give a dozen facts from one I read a few months ago (for example, a biography of Rockefeller)…I can’t do it. I could tell you the rough outlines of his life, but I’ve forgotten the details. I put maybe twenty hours into reading this book and my return on that time seems to be minimal at best.
What’s the alternative? I’m a big believer in SRS/Anki and have found it to work almost flawlessly. And so I think a better alternative to reading for information may be to make a list of discrete facts from a book and put those into an SRS system.
Side note: I’m distinguishing reading for information from reading for enjoyment. I assume most people don’t read business nonfiction books for enjoyment and mostly just want to learn the information inside.
The value of Anki is clear, but I wonder how many hours it takes to create that list. Maybe that is time when I could be reading another book on the same topic, for example. Or I am rationalizing because I find it annoying to enter things in Anki (even a csv file) :)
The compromise I found is: if you own a lot of physical books, underlining (normally or with small vertical lines on the side) makes it easy to go back and remember the important facts. I also write my reaction to some sentences on the margin.
All of this helps me remember better, and it makes it easier to get back to it later to get the main points. It seems to work well enough for now.
1, I have the belief that I can understand the material. This is the most important part because if you do not believe you can understand the material then you will always fail to understand the material. If you do not have this, figure out why and address that first.
2, I will suck on a mint, since I was told as a child that sucking on a mint dilates the blood vessels in your head and gets more oxygen to your brain, more brain blood, more brain power, right? Honestly, though, it's more helpful for cram sessions and tests for some reason. If you're casually learning you can skip this.
3, Revisit the material a day or two later, especially if you had difficulty with it. This showed its hand in college maths a lot. I would fight for half an hour on a difficult problem at first, then come back two days later and try a similar problem and kill it in 5 minutes. Something about it being harder to carve a new line into a rock than to retrace it after the fact.
Doing this solidifies confidence in the new thing that you have learned.
4, After I have read and revisited the material, I will tell what I have learned to an imaginary child me as though I am teaching them what I have learned. If you can't teach it, you don't understand it.
Seems to work pretty well.