Too bad Piotr Wozniak (inventor of SuperMemo) is such a hermit and uncompromising of his lifestyle choices, that most of his life's work was eventually superseded by open-source solutions (Anki + FSRS).
I had tons of material in SuperMemo for years. Gave up and fully switched to Anki.
At least I'm thankful for his spaced-repetition algorithms. Also, his articles restored my love for learning and helped me confirm that school was an insane waste of time and resources.
Another key point: its dependencies on native but very outdated Windows libraries.
No plugins, no way to modify or improve anything. Fully closed source. Feature request? Wait months until your email shows up in Dr. Wozniak's queue for incremental-reading of his emails. And one year until maybe the next version of SuperMemo. Or write the request on his wiki, to be debated by community members. It's so impractical.
Yeah I use it to read HN threads, articles from my RSS reader, Wikipedia pages, code that I want to review etc. Also, material that I need to make into flashcards eventually. The ability to schedule a chunk of text and then turn it into flashcards in the same application is really valuable to me. I typically open a thread and import it a few days later after most of the comments are done. Right now I have ~40-50 tabs I will import in a few days.
This video [1] is the second step of the process, the first would be making that extract from an article or larger chunk of text imported. Often, I read a little bit of the text and just delete the whole entry if it turns out not to be good/interesting/useful to me.
Yes, I do, and I'd say the opposite. I started using incremental reading after using Anki for 11 years, and now, five years later, there is no way I would go back to the Anki style.
You can check my YouTube channel, 'Pleasurable Learning', for a course and many videos on how to get started with incremental reading.
Exactly. Incremental reading was touted as some ultimate productivity/pleasure hack but it's impractical for the real world were humans have to synchronize with each other's calendar. It may work for Piotr because he doesn't have a schedule to follow, at all.
When I need to read and learn, it needs to happen within a timeframe. Not "some day" when it shows up in my queue again.
That's usually not a problem. You can always schedule the reading of an element manually. And if you do that for the first repetition, it will not have any negative impact on the algorithm.
You can also put stuff in something called "pending queue" and then just pull it from there once you're ready for it.
Only for the subsequent repetitions the algorithm will take over. This is like reviewing notes basically.
Also, you can manually advance repetitions if needed, for example before an exam.
It works very well for me and I constantly have to (and want to) learn lots of stuff.
If I hadn't stubbled upon this essay and Anki, I think by way of a Wired article, I'm not sure I would have gotten into or passed medical school. They were eye opening and turned me from an average smart student into a leader of my class. I had planned to keep doing my cards after moving on to fellowship but alas, life gets in the way.
You deserve some credit, though. Setting this up properly takes quite a bit of discipline and persistence. I only did it for certain memorization-heavy classes and I still found it to be kind of a chore. Mostly in college I just tried to understand things deeply and hoped that this kind of understanding/memory would be reasonably durable. That got me good enough grades for what I was trying to do, but it probably would not have been sufficient for a very competitive professional school like medicine.
That's how I did undergrad. Wasn't even much of a note taker, usually just paid attention during lectures and did the reading. That's not enough for the insane volume of information you have to learn in med school. There are certain subjects like physiology that are very conceptual but even there you need to know quite a bit of detail, and then of course the memorization heavy topics like anatomy, microbiology and others.
There's a reason Anki is used so heavily by med students these days.
If you've used AI to create flashcards then you've probably encountered many annoyances. For example, AI-created flashcards tend to be too wordy. Share this article with them and ask them to follow the rules. It makes for better flashcards.
It feels circular. How is learning different from understanding? How is understanding different from knowledge? I’m supposed to understand before I learn. How do I understand if not through learning? I’m supposed to understand to gain knowledge. Isn’t knowledge understanding?
There is a lot of fetishization of this terminology in the spaced repetition community. It's really best to ignore it, it's not based in much.
I think their mental metaphor is that cards allow you to memorize nodes, and understanding is having a feel for the entire graph. But cards also help you to memorize links between nodes, subgraphs, overviews, principles, etc...
I also think it's mostly a ready made array of excuses to read off to somebody who is having a crisis of faith about whether a Anki is helping them or not: you're holding it wrong. You haven't put in the work. Are you making your own decks, you can't use other people decks because making your own decks = understanding (for mysterious reasons, do you really understand something you can't remember?) Are your facts atomic enough? Basically direct or indirect paraphrases of the Supermemo wiki.
Supermemo didn't discover anything, he computerized something that desperately needed to be computerized, and at that point wasn't restricted to the algorithms that could be executed by shuffling around physical cards, such as Leitner boxes (which are awesome, still, by the way.) His analysis is great to read and often insightful, but is no more profound than many others and often far less scientifically grounded. People just are addicted to parasocial relationships with self-improvement gurus.
Not sure if this will help (or if it's even correct), but Wilfrid Sellars in his essay "Philosophy and the Scientific Image of Man" talks about the difference between "knowing that" to ride a bike one needs to put pressure on the pedals to rotate the tire and move the bicycle and rider forward and "knowing how" to ride a bike. To me the latter is indicative of understanding and the former is information or knowledge.
Lack of understanding would mean you haven’t decided meaning from the words on the page. He gives the example of memorizing a German textbook without speaking German.
Memorization without learning means that you don’t know the relationships between individual facts that give them meaning and relative importance that lets you make wise decisions about what to memorize and what not to for your purposes.
So: make sure you understand what the words and phrases you’re reading mean (understanding). Look up terms and definitions. Identify the main points of the sentence, paragraph, page and chapter and why they matter (learning). Then memorize those main points, starting with the most important basics.
Mortimer Adler's How To Read A Book talks about "reading for understanding" a lot. By understanding, he means getting a holistic feel for the content. Grokking, if you will. An analogy is reading a newspaper article on what happened in Gaza yesterday versus reading a 500-page book on the history of Palestine and writing a report on it. In the first example you are reading for facts (knowledge), in the second you are reading for understanding
How does one walk? To walk one needs to move their foot forward, to move their foot they have to lift it up, to lift a foot up they first have to place it down, and that requires moving it forward. It feels circular!
Learning is circular. You do it step by step, one bite at a time: you learn a fact, you understand its connection to other facts you know, you gain a little knowledge, you repeat.
Understanding is prior to learning. Learning is required when prior understanding fails to extend/generalize/transfer. Understanding is a residue of learning.
You are correct that those concepts are interrelated. It works well not to get caught up in precise definitions. Instead, reflect on your current level and take the next best step.
I always thought that point 1 is obvious (don't try to memorize Goethe in the original if you don't understand any German) and that point 2 is where it's at, and what most people underemphasize (learn the material before you try to memorize it). Granted, some types of learning and memorization go hand in hand; but for me the key point is to not try to use SRS to learn the material. Writing and rewriting notes; explaining topics out loud to myself and others; and using information to create something of my own — those are the ways I learn best. And at that point I've naturally memorized a lot already because I've "internalized" it; the spaced recall system becomes more of a repeating task list to remind me to practice recalling what I already learned, right before I would forget it. In that way it's similar to my OmniFocus lists of repeating maintenance tasks and chores, except that the repetition scheme varies with my forgetting curve instead of on a plain daily, weekly, monthly, or yearly schedule.
> Writing and rewriting notes; explaining topics out loud to myself and others; and using information to create something of my own — those are the ways I learn best. And at that point I've naturally memorized a lot already because I've "internalized" it.
Yes precisely. That's generally where i end it. If I've understood the topic well enough to reason about it, rephrase in a compact form, or explain it to someone else, i consider the process done. Doing so during university always made me pass reliably.
Flash cards & repeated practice always felt like a cheat to memorize parts i had not sufficiently understood or learnt. Memorization techniques are a great way to pass exams and get grades, but terrible way to learn in the longrun. Any details that would truly require memorization techniques (numbers or large lists of terms) are things i would want to look up to be sure of in the real-world anyway, so why try memorize them.
Whenever i look up "study advice" i often see memorization techniques, but very rarely see plain old; "read the book slowly page-by-page, Write down any questions, summarize whenever you feel you understood the section, go back to previous chapters when confused, cross check with other sources or explanations if not sufficient, try practice problems to check if you've understood correctly"
This article leans in that direction as-well; 1&2 are mentioned briefly, but encompass the vast majority learning process. All the rest of the points are memorization aids that i would consider footnotes to the learning process at best.
Too bad Piotr Wozniak (inventor of SuperMemo) is such a hermit and uncompromising of his lifestyle choices, that most of his life's work was eventually superseded by open-source solutions (Anki + FSRS).
I had tons of material in SuperMemo for years. Gave up and fully switched to Anki.
At least I'm thankful for his spaced-repetition algorithms. Also, his articles restored my love for learning and helped me confirm that school was an insane waste of time and resources.
Last time I checked, Anki was nowhere near SuperMemo in regards to incremental reading features. Especially not with FSRS.
SuperMemo is actually almost the only reason I still keep a Windows VM.
> "the only reason I still keep a Windows VM."
Another key point: its dependencies on native but very outdated Windows libraries.
No plugins, no way to modify or improve anything. Fully closed source. Feature request? Wait months until your email shows up in Dr. Wozniak's queue for incremental-reading of his emails. And one year until maybe the next version of SuperMemo. Or write the request on his wiki, to be debated by community members. It's so impractical.
These are all fair points, and yet SuperMemo meets my needs better than Anki.
Which is a thing I experience quite often. I love the idea of open source, but often the proprietary products are „better“ for what I need them.
Do you use Incremental Reading? I've tried many times but I don't see the point in it, probably never got it.
Anki way makes more sense.
Yeah I use it to read HN threads, articles from my RSS reader, Wikipedia pages, code that I want to review etc. Also, material that I need to make into flashcards eventually. The ability to schedule a chunk of text and then turn it into flashcards in the same application is really valuable to me. I typically open a thread and import it a few days later after most of the comments are done. Right now I have ~40-50 tabs I will import in a few days.
This video [1] is the second step of the process, the first would be making that extract from an article or larger chunk of text imported. Often, I read a little bit of the text and just delete the whole entry if it turns out not to be good/interesting/useful to me.
https://www.youtube.com/watch?v=h8oyvBmevn8
Yes, I do, and I'd say the opposite. I started using incremental reading after using Anki for 11 years, and now, five years later, there is no way I would go back to the Anki style.
You can check my YouTube channel, 'Pleasurable Learning', for a course and many videos on how to get started with incremental reading.
Incremental Reading in SuperMemo is essentially my default mode of reading for anything I expect to become part of "knowledge".
I don't use it for things like e-mail, news or fiction.
Exactly. Incremental reading was touted as some ultimate productivity/pleasure hack but it's impractical for the real world were humans have to synchronize with each other's calendar. It may work for Piotr because he doesn't have a schedule to follow, at all.
When I need to read and learn, it needs to happen within a timeframe. Not "some day" when it shows up in my queue again.
That's usually not a problem. You can always schedule the reading of an element manually. And if you do that for the first repetition, it will not have any negative impact on the algorithm.
You can also put stuff in something called "pending queue" and then just pull it from there once you're ready for it.
Only for the subsequent repetitions the algorithm will take over. This is like reviewing notes basically.
Also, you can manually advance repetitions if needed, for example before an exam.
It works very well for me and I constantly have to (and want to) learn lots of stuff.
What you describe fits the 'Plan' feature, not incremental reading. You don't sync with anyone other than your future self...
If you have deadlines for certain knowledge, then this can still be managed; it just requires some manual overruling via 'advance' or subset reviews.
But the whole Anki community is surrounded by his idea! Thanks for his effort; he makes learning possible for an uncountable number of people.
If I hadn't stubbled upon this essay and Anki, I think by way of a Wired article, I'm not sure I would have gotten into or passed medical school. They were eye opening and turned me from an average smart student into a leader of my class. I had planned to keep doing my cards after moving on to fellowship but alas, life gets in the way.
You deserve some credit, though. Setting this up properly takes quite a bit of discipline and persistence. I only did it for certain memorization-heavy classes and I still found it to be kind of a chore. Mostly in college I just tried to understand things deeply and hoped that this kind of understanding/memory would be reasonably durable. That got me good enough grades for what I was trying to do, but it probably would not have been sufficient for a very competitive professional school like medicine.
That's how I did undergrad. Wasn't even much of a note taker, usually just paid attention during lectures and did the reading. That's not enough for the insane volume of information you have to learn in med school. There are certain subjects like physiology that are very conceptual but even there you need to know quite a bit of detail, and then of course the memorization heavy topics like anatomy, microbiology and others.
There's a reason Anki is used so heavily by med students these days.
That's impressive! Would you mind sharing how you used Anki for your studies?
https://www.reddit.com/r/medicalschoolanki/wiki/index
Related:
Rules of formulating knowledge in learning (1999) - https://news.ycombinator.com/item?id=22524122 - March 2020 (2 comments)
Effective learning: Twenty rules of formulating knowledge (1999) - https://news.ycombinator.com/item?id=18404150 - Nov 2018 (17 comments)
Effective learning: Rules of formulating knowledge (1999) - https://news.ycombinator.com/item?id=13047576 - Nov 2016 (35 comments)
Effective learning: Twenty rules of formulating knowledge (1999) - https://news.ycombinator.com/item?id=10785221 - Dec 2015 (1 comment)
If you wanna go deeper, I’d recommend checking out https://supermemo.guru/wiki/SuperMemo_Guru
It’s a bit of a treasure trove for us spaced repetition nerds!
Sadly, it seems to (as of 6 hours after parent post) intermittently produce "max connections exceeded" messages :( Poor old hug of death!
If you've used AI to create flashcards then you've probably encountered many annoyances. For example, AI-created flashcards tend to be too wordy. Share this article with them and ask them to follow the rules. It makes for better flashcards.
It feels circular. How is learning different from understanding? How is understanding different from knowledge? I’m supposed to understand before I learn. How do I understand if not through learning? I’m supposed to understand to gain knowledge. Isn’t knowledge understanding?
Memorization is part of learning. Memorized knowledge is knowledge. But memorized knowledge isnt understanding.
You can fully understand something without being able to recall it perfectly later.
There is a lot of fetishization of this terminology in the spaced repetition community. It's really best to ignore it, it's not based in much.
I think their mental metaphor is that cards allow you to memorize nodes, and understanding is having a feel for the entire graph. But cards also help you to memorize links between nodes, subgraphs, overviews, principles, etc...
I also think it's mostly a ready made array of excuses to read off to somebody who is having a crisis of faith about whether a Anki is helping them or not: you're holding it wrong. You haven't put in the work. Are you making your own decks, you can't use other people decks because making your own decks = understanding (for mysterious reasons, do you really understand something you can't remember?) Are your facts atomic enough? Basically direct or indirect paraphrases of the Supermemo wiki.
Supermemo didn't discover anything, he computerized something that desperately needed to be computerized, and at that point wasn't restricted to the algorithms that could be executed by shuffling around physical cards, such as Leitner boxes (which are awesome, still, by the way.) His analysis is great to read and often insightful, but is no more profound than many others and often far less scientifically grounded. People just are addicted to parasocial relationships with self-improvement gurus.
Not sure if this will help (or if it's even correct), but Wilfrid Sellars in his essay "Philosophy and the Scientific Image of Man" talks about the difference between "knowing that" to ride a bike one needs to put pressure on the pedals to rotate the tire and move the bicycle and rider forward and "knowing how" to ride a bike. To me the latter is indicative of understanding and the former is information or knowledge.
Lack of understanding would mean you haven’t decided meaning from the words on the page. He gives the example of memorizing a German textbook without speaking German.
Memorization without learning means that you don’t know the relationships between individual facts that give them meaning and relative importance that lets you make wise decisions about what to memorize and what not to for your purposes.
So: make sure you understand what the words and phrases you’re reading mean (understanding). Look up terms and definitions. Identify the main points of the sentence, paragraph, page and chapter and why they matter (learning). Then memorize those main points, starting with the most important basics.
Mortimer Adler's How To Read A Book talks about "reading for understanding" a lot. By understanding, he means getting a holistic feel for the content. Grokking, if you will. An analogy is reading a newspaper article on what happened in Gaza yesterday versus reading a 500-page book on the history of Palestine and writing a report on it. In the first example you are reading for facts (knowledge), in the second you are reading for understanding
How does one walk? To walk one needs to move their foot forward, to move their foot they have to lift it up, to lift a foot up they first have to place it down, and that requires moving it forward. It feels circular!
Learning is circular. You do it step by step, one bite at a time: you learn a fact, you understand its connection to other facts you know, you gain a little knowledge, you repeat.
Understanding is prior to learning. Learning is required when prior understanding fails to extend/generalize/transfer. Understanding is a residue of learning.
You are correct that those concepts are interrelated. It works well not to get caught up in precise definitions. Instead, reflect on your current level and take the next best step.
Enter: knowledge. It's a messy thing. Before you know it, it 'clicks'. That's the only thing worth chasing.
I have notes with me for anything of importance. rule 2 and beyond is of no concern to me.
I got through uni on entirely point 1, and only relied on accidental memorization from the process of understanding.
I find alot of study advice under-emphasise point 1, and over-emphasise memorization techniques.
I always thought that point 1 is obvious (don't try to memorize Goethe in the original if you don't understand any German) and that point 2 is where it's at, and what most people underemphasize (learn the material before you try to memorize it). Granted, some types of learning and memorization go hand in hand; but for me the key point is to not try to use SRS to learn the material. Writing and rewriting notes; explaining topics out loud to myself and others; and using information to create something of my own — those are the ways I learn best. And at that point I've naturally memorized a lot already because I've "internalized" it; the spaced recall system becomes more of a repeating task list to remind me to practice recalling what I already learned, right before I would forget it. In that way it's similar to my OmniFocus lists of repeating maintenance tasks and chores, except that the repetition scheme varies with my forgetting curve instead of on a plain daily, weekly, monthly, or yearly schedule.
> Writing and rewriting notes; explaining topics out loud to myself and others; and using information to create something of my own — those are the ways I learn best. And at that point I've naturally memorized a lot already because I've "internalized" it.
Yes precisely. That's generally where i end it. If I've understood the topic well enough to reason about it, rephrase in a compact form, or explain it to someone else, i consider the process done. Doing so during university always made me pass reliably.
Flash cards & repeated practice always felt like a cheat to memorize parts i had not sufficiently understood or learnt. Memorization techniques are a great way to pass exams and get grades, but terrible way to learn in the longrun. Any details that would truly require memorization techniques (numbers or large lists of terms) are things i would want to look up to be sure of in the real-world anyway, so why try memorize them.
Whenever i look up "study advice" i often see memorization techniques, but very rarely see plain old; "read the book slowly page-by-page, Write down any questions, summarize whenever you feel you understood the section, go back to previous chapters when confused, cross check with other sources or explanations if not sufficient, try practice problems to check if you've understood correctly"
This article leans in that direction as-well; 1&2 are mentioned briefly, but encompass the vast majority learning process. All the rest of the points are memorization aids that i would consider footnotes to the learning process at best.
Reimplementing it in 'proper' way with LLM assistance in development and functionality looks like a great project. Hold your breath...