Maybe Anki isn't a tool for you.
I personally don't use Anki because it bores me to death.
For me it needs to be interesting.
I think there are other people out there who even don't use Anki because it's a boring way to learn something.
Spaced repetition really clicked for me when I realized that it isn’t about learning, but about keeping memories fresh. If you don’t have an emotional attachment to each card in the deck, reviews become a terrible slog. Mine is a mix of random insights I’ve had, photos of places I’ve been, summaries of things I’ve learned, etc. The questions are simply there to force me to engage my memory, which drags up the whole web of mental associations— there’s no need to add every detail explicitly; that’s what notes are for.
Yes yes, I agree. The way I tackle the threat of detachment from my cards is to aggressively delete cards as soon as I see boredom creeping in. I switched from Anki to Supermemo, though. Much better for managing context and even keeping organized notes that may or may not be used for active recall at some later point.
Same. After about a year of using Anki the need to aggressively delete became apparent. With good culling and time to reflect, review can be almost a meditation.
How does Supermemo help you keep keep context?
> keeping organized notes that may or may not be used for active recall
I use a separate notes app (Bear) for this. Is there much advantage in your experience to integrating active recall notes with Evernote-style reference notes?
Can't reply to child for some reason but the reason that SM is so great for managing content is a feature called incremental reading (https://supermemo.guru/wiki/Incremental_reading). To sum it poorly, it lets you go through hundreds of articles in parallel and slowly convert them to active recall material over time. It sounds terribly complicated in practice, and learning it is indeed a pain, but once you know it there's really no way of learning more enjoyable. I can promise you that.
I don't understand how anki users are able to make cards for non-language things,I personally absolutely hate making cards while reading anything. It's hard to explain the exact process but incremental reading by contrast makes it much easier to take what's really important from an article and make cards from it.
My hunch is that Anki is mostly useful for stopping you from forgetting things you have already learned. The actual learning is better done outside of Anki.
Agreed - I've found that Anki really helps to recall underlying facts, but if I didn't grasp a concept well outside of Anki, it's not helping with A-HA moments.
Anki is "just" a tool to prompt you with flash cards. Designing them is up to you and extremely flexible. You're free to add text, images, audio, whatever you want.
Some ideas of how you could use Anki to learn the Greek alphabet:
- Front of the card has the upper- and lower-case glyphs and the back has their sounds, either English, IPA or an actual recording (Wikipedia has them for everything in IPA and the Greek Alphabet page has a mapping from letter to IPA). You should also do this in reverse.
- Front of the card has a Greek word with a missing space, back of the card has the complete word. You can use audio recordings here as well if it helps you (e.g. have the front read the entire word).
When I went to Greece I learned it in a few days just by trying to read all the signs I could see. I guess the trick is not to learn isolated letters but learn to read common words.
I think you'd be able to learn the Greek alphabet in at most an hour, just by writing it out and saying it repeatedly. Making it stick long term would be a matter of a couple of minutes a day for a week or so, then every few days. It's small and simple enough that flash cards aren't necessary.
For more involved facts, Anki seems great in principle, but in practice I've found it hard to create and maintain the habit of doing the cards.
Using it to learn an alphabet effectively probably requires more upfront work (setting up and improving the cards) than you’d expect. It’s not just 1 letter = 1 card.
You want to memorise the entire alphabet as a sequence not just as individual letters. So you need cards of little overlapping sequences, like ABC, BCD, etc. The supermemo essays cover this scenario. With that work in place you’d learn it quickly.
As just separate letters divorced of any meaning, not hooked up with any other neurons in your brain, you’d struggle to ever recall it quickly.
This is a detail that people overlook and then wonder why it’s not working. (Not picking on you for doing that: I did the same before eventually reading way too much about it)
Yeh, I hear ya. That’s why I used the word “formulated” rather than “discovered” —- because he really did the work to not just model it, but really implement a system with a very specific formula which he devised and which was the basis of the field we see today.
If you want to learn the Greek alphabet try leaving Modern or Ancient Greek on Duolingo. If there’s an Anki deck with modern Greek, translation and romanization that would almost certainly work better than trying to associate a sound to an arbitrary symbol with no scaffolding to help it stick.
I don't think you did it right. I also used anki to learn the greek alphabet, I added just one letter a day for less than a month, and now two years later I know it by heart with a review load averaging 2 notes a month..
That depends on what knowledge you're trying to retain. When I majored in Spanish in college I was extremely efficient compared to my classmates. Whereas most would memorize lists of words and sentences and dedicate tens of hours to it, I would average less than 5 hours of preparation time per exam.
I tried to use Anki to study for a tech certification.
In the end, I got a lot more out of doing the research to create the cards than I did from the actually looking at the cards. I might even go so far as to say that after creating the cards, they were practically worthless.
This would be less true if the exam was further from the creation of the cards, though. It was within a month.
Someone else's cards might have helped a little, but not nearly as much.
Really? I have a rule with Anki where the question is one line and the answer is no more than 3. No cards take longer than a minute to write, or more than 10 seconds to remember.
If it's anything longer, then it's probably not a subject you should be using anki to learn or you're trying to remember too large a bloc of information at once.
I have absolutely no idea how this could be a form of procrastination, or how I could more efficiently memorize information that must be memorized.
> I have absolutely no idea ... how I could more efficiently memorize information that must be memorized.
Outside of a schooling/testing context, memorization alone is rarely beneficial. You need to be able to practically apply the knowledge, which happens through experience; looking details up when you need them forms a natural kind of spaced repetition anyway that’s tuned to what you actually need to know.
Well, as a counterpoint, imagine how much time could be saved if you were able to recall from memory not just the most-used functions that you need for a problem, but the next level down of sometimes or rarely used ones. Or for patterns, or for other such things that can be simplified down into memorizable / recallable blocks.
Regularly looking things up is ok, but actively re-experiencing the thing you're trying to learn is a better way to make it stick.
In many cases less time that what I spend looking it up. I regularly look up things I haven't needed before, a few minutes of reading and I know it. Many of those things are something I expect to never need again in my lifetime. The few seconds to memorize all those things is greater than the time saved. Particularly since I don't know what I will need next week and so I'll be spending a lot of time learning things I turn out to never need.
For programming, memorisation doesn't matter that much. Not knowing the argument order for a function isn't a big deal. You use an IDE or you get a good offline docs viewer (Dash, DevDocs etc.) or you learn to Google efficiently.
Language learning is an obvious use case.
Also: law and medicine. Having knowledge mentally 'to hand' is pretty important if you've got a patient under general anaesthetic, or a judge asking you a very difficult question.
Memorizing word pairs can certainly be done with spaced repetition, but it’s unclear how much that translates to actual language ability. Second-language acquisition appears to be primarily dependent on reading (or listening to) the target language for content, and most words are learned via seeing them in context instead of being looked up in a dictionary.
I have no experience with law or medicine, but I expect the story is similar: practical knowledge is what you need to hand, and not book knowledge. Book knowledge is what gets you through the exams and into the practical part of your training.
Beyond the very basics necessary to extract some meaning from a second language, I’d be shocked if time invested in spaced-repetition drills of any design had better returns than reading the target language for pleasure.
Anki has two benefits: memorization and understanding. Memorization is what comes from going over the cards, which is what you allude to. Understanding comes from the process of creating the cards, where you think hard about the atoms of information in the text, and their interrelationships, and cast them into simple questions. Then memorization gives you a much deeper understanding than if you hadn't ankied the text.
I concur, what you describe is a very effective use of spaced repetition. It is different from what ‘ProstetnicJeltz described, where cards should take less than a minute to write — that describes memorization without first putting in the effort to understand.
I don’t agree. I was for a long time of the school that understanding is the only things that matters, and the details can always be looked up. Then during my post doc ( physics) I worked with a couple of Russian guys who had been forced to memorise a lot. What they gained was an enormous speed. They could try six different ideas in their mind while I still looked up the details necessary to try my first. So don’t underestimate root learning.
There’s a third category here, and that’s functional skill. It’s quite a separate thing from either understanding or rote learning. They all feed into each other, of course, but the point I tried (and apparently failed) to make is that it’s functional skill that’s useful in real-world contexts.
Sometimes the thing holding you back from improvement is a lack of facts, and sometimes it’s poor understanding of theory. In my experience, however, the fastest way to get better at doing something is almost always to practice doing that thing; most of the time a sufficient collection of facts and theoretical understanding will come along as a side-effect of that work.
How much of their prior training was memorizing specific facts, and how much of it was drilling the mechanics of solving typical problems? The former is what most people use spaced repetition for and what I believe is of limited utility; the latter is incredibly valuable and I never meant to imply otherwise.
It depends what you're doing. I mainly use it for learning vocabulary.
I'm learning German and Anki is nothing short of amazing. I'm easily able to learn ~20 words a day primarily just practicing on my phone during bathroom breaks.
On the other hand I'm also learning Russian and Anki just doesn't work. Words won't stick in my head the way they do with German.
I've heard very numerous accounts of it being great for Japanese.
>I've heard very numerous accounts of it being great for Japanese.
There's at least great pre-made Japanese vocab decks. I'm about two months into one. Maybe less so for kanji (I at least don't have one), so that's probably something you need to do outside.
>On the other hand I'm also learning Russian and Anki just doesn't work. Words won't stick in my head the way they do with German.
Because (for English speakers) Russian is harder than German. You need to reduce the number of words learned per day. Japanese is even harder to learn, but you have to be persistent. I spent more than an hour every single day for several years straight to learn the vocabulary.
Also there's a unique challenge with Russian is that nouns, verbs and adjectives are all conjugated, and the rules are quite complex. The noun cases would be completely foreign to non-Slavic language speaker. Even if you learn the vocabulary, using it in sentences is not trivial at all.
Just going to be a bit nitpicky here. Noun declension is a feature of various non-slavic languages, including Latin which English speakers have a decent chance of having been exposed to.
Modern Russian has 6 cases (with a couple more that pop up very rarely) with 3 genders and singular/plural endings. It's a total of 36 possible combinations, which is not that bad. Masculine and neuter also share most endings so there aren't actually 36 unique ones
Probably the most difficult thing about cases when I was learning Russian was remembering which verbs took a different case than I would expect, e.g. dative instead of accusative for what seems like a direct object.
Right: but one difference (which you sort of hint at) is that German is consistent in why something takes a particular case compared to Slavic languages (e.g. direction vs location), and at least the preposition is consistent. Is it "nad morze" or "nad morzem"? No way to know without the entire sentence. And "morzem" is the "tool case" -- the form you usually use to describe that something happens with the help of a tool even though the sea (morze) is not at all a tool in that sentence. Oh, and you only use "nad" with bodies of water by the way. Sorry!
Another incredibly tricky one for Slavic languages is imperfective vs perfective aspects. In most other languages, perfect vs imperfect is just a standard construction. In Polish (and Russian, though my examples here are Polish), you change the verb itself. How? Well, sometimes you put "po" in front of it, like rozmawiać / porozmawiać. Sometimes it's "z" (jeść / zjeść), sometimes "u", "na" or "wy". And sometimes you just give up, like oglądać (but obejrzeć in the perfective), widzieć (zobaczyć), mowić (powiedzieć).
(For this reason I have found it very hard to progress in Polish without conversation with native speakers who aren't too polite to correct me.)
Yes, perfective and imperfective aspects are very tricky, particularly how they interact with various other linguistic features: imperatives (and the negative imperative), the subjunctive, and verbs of motion come to mind, each of which modifies the use of the aspects in it's own way.
> Modern Russian has 6 cases (with a couple more that pop up very rarely) with 3 genders and singular/plural endings. It's a total of 36 possible combinations, which is not that bad.
36 isn't that bad but Russian also has 253 irregular verbs [0], each of which adds another set of combinations. A lot of them are just small tweaks and there are some that follow a set of patterns (e.g. идти with its various prefixes) but it still adds a lot of overhead.
That's a good idea, I'll give it a try with fewer words. I did notice that I'd just come back to a word I'd seen 30 seconds ago and not recognize it, maybe I need to reduce the number of switches in between so it has more opportunity to stick.
> The noun cases would be completely foreign to non-Slavic language speaker.
Picking those up was actually fairly straightforward. Cases may be completely foreign but the grammatical concepts behind them aren't. For example Russian has nominative, accusative and dative but English has subjects, direct objects and indirect objects and I learned all those in school.
I think somewhat ironically, the biggest grammatical problem I had with Russian was the tenses. English has 12 and Russian has ~5 so it was hard to figure out how to express a given English sentence in Russian.
But I think the biggest problem is phoenetics. On the surface it looks simple (unstressed o -> a) but there are so many extra rules if you want to speak properly (e.g. voiced consonant before a voiceless consonant becomes voiceless).
My time isn't so optimised that 15 minutes each day on Anki is otherwise going to be spent on deep work.
http://augmentingcognition.com/ltm.html
This writeup estimates his usage of Anki is 4-7 minutes of review time for each card over 20 years. So, if it's worth taking 10 minutes to remember a fact, it's worth putting in Anki.
(This writeup also explains it's not worth putting things you don't care about into Anki).
Anki seems well suited for domains where you need to quickly access facts from a wide domain, e.g. vocabulary for language learning.
It's harder to see as important for programming. If you're programming you'll be able to recall the things you have experience using.
I use it for programming. My general rule is anytime i need to keep looking up some command, i just stick it in anki. Here are two python cards I added this week:
What method do you use to format a datetime object in python:
strftime()
How do you create a new virtual env using python 3
python3 -m venv <NAME>
Those may seem stupid to you, but I was looking them up and now i do not need to. I also use them for other concepts when I want to know a language better. For example
What are the three prototypical methods of a js promise?
then, catch, finally
What does the constructor of a promise take?
and executor function
How many parameters does a executor take?
2
What are the two parameter of the executor function (constructor of a js promise)
resolve() and reject()
Things like this just help me as a programmer. I agree it's not for everyone, but I have the time an enjoying doing it.
Those are much closer to what I'd put in if I added Anki cards for programming than other things I've seen online. Nice and short.
Why not just `new Promise((res, rej) => ..)` instead of intermediate facts like "an executor takes 2 arguments"? I don't think I'd be able to recall what's meant by 'an executor' without first recalling that snippet.
I noticed from using Anki so much for everything that when I include answers in other questions titles I learn the material better. Your comment would obviously be faster because it is less cards to add, but I think my brain would be more likely to forget the concept that way. Personal preference I guess.
For math/programming, it's better to keep it short in my experience. But for other things, like knowledge from books, I use both methods, here are some examples:
----
What happen to slaves in Africa after the slave trade was abolished by the America/UK in the early + mid 1800s?
simply led to a redeployment of the slaves, who were now used within Africa rather than in the Americas
“So the abolition of the slave trade, rather than making slavery in Africa wither away, simply led to a redeployment of the slaves, who were now used within Africa rather than in the Americas. Moreover, many of the political institutions the slave trade had wrought in the previous two centuries were unaltered and patterns of behavior persisted. For example, in Nigeria in the 1820s and ’30s the once-great Oyo Kingdom collapsed. It was undermined by civil wars and the rise of the Yoruba city-states, such as Illorin and Ibadan, that were directly involved in the slave trade, to its south. In the 1830s, the capital of Oyo was sacked, and after that the Yoruba cities contested power with Dahomey for regional dominance. They fought an almost continuous series of wars in the first half of the century, which generated a massive supply of slaves. Along with this went the normal rounds of kidnapping and condemnation by oracles and smaller-scale raiding. Kidnapping was such a problem in some parts of Nigeria that parents would not let their children play outside for fear they would be taken and sold into slavery.”
What are two reasons why large scale wars no longer occur ?
1. Price of war has gone up because of Atomic weapons.
2. Weath is no longer is physical goods (gold,etc) but in the minds of the citizen (Silicon Valley).
Scholars have sought to explain this happy development in more books and articles than you would ever want to read yourself, and they have identified several contributing factors. First and foremost, the price of war has gone up dramatically. The Nobel Peace Prize to end all peace prizes should have been given to Robert Oppenheimer and his fellow architects of the atomic bomb. Nuclear weapons have turned war between superpowers into collective suicide, and made it impossible to seek world domination by force of arms.
Secondly, while the price of war soared, its profits declined. For most of history, polities could enrich themselves by looting or annexing enemy territories. Most wealth consisted of material things like fields, cattle, slaves and gold, so it was easy to loot it or occupy it. Today, wealth consists mainly of human capital and organizational know-how. Consequently it is difficult to carry it off or conquer it by military force.
Basically, I create cards with short answers but provide the context (usually the paragraph I found it in). This method works extremely well for remember everything you read. I could write a more technical blog post on this if people actually care and find this interesting.
I think that many people could think that fair assessment, but really it's just about what you value in the world.
Is becoming more knowledgable just "procrastination?" Is reading non-fiction outside your specialty procrastination?
I feel like for much of human history we've held people who have read an entire library's-worth of books in high esteem. Were they just procrastinating?
I still see where you're coming from. I used to think that people who meditated, or hiked the Appalachian Trail were "selfish," because they were doing something that benefitted only them. And I still find it hard to understand someone who decides to learn Mandarin at the age of 93. But people should be allowed to enlighten and actualize themselves however they wish, and it seems absurd to think that a planet full of more knowledgable people would be a bad thing.
Saves me a ton of time in medical school vs trying to brute force the facts in. Plus it keeps things I learned over a year ago fresh so it has been very helpful for preparing for USMLE Step 1. Medicine might be an ideal case for this kind of learning though.
- I have a huge Zettelkasten note archive (I manage it through Sublime-Zk, which is getting a bit slow to operate it). Can I somehow directly import all my Zettel notes?
- I tried to import my Anki cards, but it seems like while I could go through them, I couldn't assign them to any particular deck/reclassify them. They show up as part of a "Not Found" deck and I don't see any option/drag-and-drop method to move them anywhere else.
I'll definitely be giving Mochi a try. Thanks a lot for creating it!
> I have a huge Zettelkasten note archive (I manage it through Sublime-Zk, which is getting a bit slow to operate it). Can I somehow directly import all my Zettel notes?
You can import the markdown files as notes, but it won’t maintain the linking. If you send me a sample I can look into adding a Zettelkasten specific importer.
> I tried to import my Anki cards, but it seems like while I could go through them, I couldn't assign them to any particular deck/reclassify them. They show up as part of a "Not Found" deck and I don't see any option/drag-and-drop method to move them anywhere else.
Sounds like the importer derped a bit and couldn’t find the deck associated with those cards. If you send me the .apkg file I can take a look. Anki’s data model is a little weird and there’s probably some edge cases I haven’t accounted for.
Yeah you can export decks. The export is .mochi format, but it's just a zip file with some plain text data and attachment files (like images, audio, etc).
I'm interested in the longer form note style but it be thing that keeps me in anki-land is the Android app. Is there a way to export to Anki? I hope understand that that question is counter to you the app's model.
I created something similar which I no longer work on (but keep running on a maintenance-type basis). Feel free to be inspired/take features or just compare with another approach.
This is a joy-sparking product, as I've been turned off by Anki's UI/UX especially when it comes to code snippets, and I'm very accustomed to Markdown.
I had the problem where it felt too time consuming making the cards and I just wanted to start off learning a few cards at a time, so I made my own solution. Its free
https://www.flashcardmax.com/download
You might also try my website (and soon to be app) Seedlang: http://www.seedlang.com. It is a site for learning German, and later this year Spanish, French, and English. It uses video flashcards to create different learning experiences, and ties them all together with a Anki-like review deck where any word or sentence encountered on the site can be saved.
There is also SuperMemo 18 (https://super-memo.com/supermemo18.html) which is from the same people as supermemo.com except it's a desktop app with way more features but also a much higher learning curve.
If you use SM and you use anki both for a significant amount of time it's not hard to see that SM leaves you with way less reps. Anki is based on the very first version of the SuperMemo algorithm (from ~1990 or so which was the first SRS algorithm ever) known as SM2. Current version of SM uses SM18.
On linux you can use SM via vmware/wine and on os x it's quite usable with parallels. If you're using SRS just for language learning then being able to use on mobile can be helpful but I think for those cases supermemo.com can be a decent alternative. More generally, I think the platform specificness of SuperMemo is more than made up for with its incremental reading functionality (alongside better scheduling algorithm)
To add to this, there's a set of Winetricks recipes for running SM on Linux.[0] (Disclaimer: I haven't tried them, though I've been planning to, for a while.)
Is Anki a trademark? If someone wants to make a tool called AnkiSuper, is that allowed? I'm not a big fan of the UI/UX. So I am thinking of scratching an itch.
If you'd ask me it's a perfect example of how not to structure a database. It's not even in the first normal form.
I'd love to create an Anki clone that has better UI/UX and a better database format and I think that would remedy some of the biggest pain points about Anki. But since most Anki users (me included) use several plugins that are quite essential to their Anki experience, an Anki clone would have to support those plugins, too.
I've used spaced repetition to help learn stuff in general. It's one of the most rewarding things I've done. Some general thoughts:
1. A lot of spaced repetition apps don't optimize for fast card creation. I like to make a card as soon as I come across something interesting, but the delay to get Anki started, and to create the card, etc just makes it frustrating. I wound up building my own app for this.
2. Cloze deletion is surprisingly effective. For those who don't know, cloze deletion involves taking a sentence, blanking the interesting parts out e.g. "The moon landing happened on _______", and trying to recall the blank parts. This is effective because it's a quick way to make cards, thus solving the problem of slow card creation.
3. I fared better with lots of small, one-sentence cards. My rule of thumb is that a card should fit in a tweet.
4. Subjectively speaking, using spaced repetition didn't just help me recall stuff in my cards, it also helped me recall stuff in general.
For frictionless card-creation I have made an org-capture template for quickly adding cards to an org-mode file via a custom Emacs pop-up frame. This is accessible from everywhere with the keybinding (Super + c + a).
Exporting to Anki from Emacs is done via the anki-editor mode for Emacs.
I think org-drill is just better overall since it's all stored as plain-text at the end of the day and each card is really editable and accessible. Anki in my experience has a clunky, slow GUI. It can't beat the creation and editing speed from vim keybindings and macros.
This sounds like an issue with something else you have installed in Emacs. I suggest running the profiler to see what is actually taking time. Org-drill itself on an otherwise vanilla config is fast, so it could be that you have hooks that are slow or something.
The thing I miss with org-drill is being able to do it on the phone. I have considered trying to integrate it into Orgzly but it does seem like a big project, unfortunately.
One of the issues Anki has, I find, is that if you end up taking a long break for whatever reason, there's no way to "reset".
I used Anki to study Chinese for a period of about 10 years. At some point I decided that I wanted to memorize the poker "outs" (probabilities of filling out a hand based on what had currently been dealt). Then I went through a time where I was really busy and didn't study the poker deck for a month (but I made time for Chinese). When I came back to the poker deck, the "spaced repetition" system was completely broken: I had a massive long list of cards that had expired, most of which I'd completely forgotten; but it just kept showing them to me in one giant loop, rather than focusing on a few to actually teach me. And I didn't even have a clean way of telling it, "Just pretend I haven't seen any of these cards at all". I ended up just deleting the deck; that discouraged me from doing anything else I wasn't willing to commit to doing every single day.
I'm kind of new to Anki myself (started a couple of weeks ago) so I might be wrong, but I think you can limit the amount of cards you want to review on a per-deck-basis. The default-number is quite high (200 per day I think?).
If you make that number the same as the "new cards per day"-number, that might be a work around for your issue?
There are a couple ways to do it that I've found. None of them is perfect for every situation I've encountered. I think that part of the reason why there's no clean way to do it is that there are so many specific things someone might want to do:
- If you want to just push cards to the back of the queue, but remember timings and history, there's a built-in command to do that.
- If you want to change the current interval of a card, but keep your history, there's also a built-in command for that.
- If you need to do the above in bulk, the "Reset Card Scheduling" plugin can make this a lot more convenient.
- If you really do want to completely reset a card, forget all history, etc., and treat it as a new card, there's the "Remove Card History" plugin.
Finally, when returning to old decks, I found I usually get the most mileage by leaving all that stuff alone and just suspending the whole thing, and then un-suspending them at a steady pace. I neglected my Kanji deck for months, and I used that approach to get myself back up to speed by doing the catch-up review in the original (RTK) order I originally learned them rather than based on Anki's priority.
User plugins are a cool feature that let users do all sorts of things you couldn't forsee as the app developer nor want to encumber the UI with every possible feature.
But it feels so hacky when you need them for things that seem like rather elementary functionality like resetting a deck.
Calibre feels the same way like needing a plugin just to estimate page count. I feel lucky when a plugin actually works.
I don't see that there's any difference in responding "Again" to an expired card vs reclassing that card as new. When I'm catching back up on an old deck I just turn new card off and limit my reviews until I've worked through the backlog this way.
Suppose you have 50 unseen cards, and you have it set to show you 5 new cards every day, and energy for about 10 reviews.
Then on day 1 it will want to show you cards 1-5; day two it will show you 1-5 again and also 6-10; day three it will show you 6-10 again and also 7-15 (to simplify somewhat). So you're having some cards you know and some that are new.
If you have 50 cards you've forgotten, then on day 1 it will show you 1-10; on day 2 it will show you 11-20; day 3 it will show you 21-30, and so on -- all completely new. That makes it far more of a grind.
For what it's worth, I keep everything in a single deck. I never stop reviewing stuff and "interleaving" makes learning more effective than "blocking" (studying things in blocks). Source: Bjork et al 2013 (https://www.researchgate.net/publication/231610455_Self-Regu...)
I took a modified approach to this in university, though nowadays I mostly do the same.
For each new class, I would start a new deck, and after the course was over I would move it over to my main (huge) deck, and it's worked surprisingly well.
Typically (now that I'm out), I add small cards from time to time to my deck, but if there's a large quantity of things I am working on learning at any given time (say I'm going to add 50-100 new cards), I'll put them in their own deck for a little bit.
Ultimately, my repetition does end up being a wide array of subjects though.
This is why tagging is useful. Anything cards I added for United had the tag of the class it was from. You can then do a custom study session with just those tagged cards.
I don't know. I just tried to briefly check it out again. I'm sure I could be giving it a fairer chance.
Clearly, millions of people are benefiting from this app.
I'm just finding the features and interface very bombarding. And, the sheer number of screen taps just to study the flash cards in the manner that I'd like (and even then, not fully within my perceived control) is pretty repelling.
Thank you for pointing that out that feature. It might help eventually with conversion.
What's the use case for this dumb mode? Does Anki not have a "cram" mode that just goes through all flashcards with none of the fancy interval logic?
One of the wonderful things about software assisting with spaced repetition is that it automatically selects and remembers efficient intervals. It's crazy how little time is spent per flashcard with this automation.
I've used Anki for about ten years, and I'm going to tell you a huge secret...
There's absolutely no need to "reset". Ever.
If you have only ten minutes to devote to Anki, then only spend ten minutes. If at some later point, you have more time, then spend that time.
Set the maximum reviews per day to something you can do most days -- for me, that's 250 cards. If you're behind, turn off 'new' cards. And eventually, you will catch up.
What if you don't do Anki for a month or two? You still don't need to reset. Anything you remember after that month will have a much longer time until you next see it.
Only use Anki as much as you have time for it. Let it figure out which cards to show you. Resetting messes with the algorithm for which cards to show.
I'm also over ten years on Anki, and I've racked up 2/3 of a million review during that time. I've taken too-long breaks several times over that time period, on a variety of subjects (decks in Anki). Sometimes years.
Each time I've gone back to a deck, I've just slugged out the few days of heavy reviews and let Anki take care of the rest. The stuff that I've forgotten, Anki will nag me with. The stuff that I've retained for years, stays retained.
One thing that I do suggest is to limit the maximum interval to 365 days, and to remove the review limit. I also tend to use the "hard" answer on cards that I've retained in decks that I've ignored for some time.
Having used Anki for 10 years, reset is definitely the best option.you fly through the cards you remember at the start, and don’t get bogged down by cards your ‘supposed’ to remember.
I have some cool decks, but I am terrible about reviewing them frequently. It just becomes another chore. Maybe I should really embrace it. I spent a few months brushing up on biology and that was really amusing but I fell out of the habit :/
Using the reminder aka notifications feature of the android app got the habit to stick for me. Total pita to figure out how to get it turned on though :)
> 3. I fared better with lots of small, one-sentence cards. My rule of thumb is that a card should fit in a tweet.
Can someone tell me if this is a crazy good idea? Social network + spaced repetition. Why should you build all your own cards? Taking a class with classmates? Create a group & create cards during lecture, review/curate cards afterwards during study session, then rate which ones were most useful after the exam (or homework).
The actual act of creating the cards is also useful to the learning process. Even if you are copying someone else's cards, copying them by hand will lead to better retention than having a computer import them.
I've head a similar idea, except not in context of a group project but of a wiki. So there's a wider audience for any particular topic and more people to create and improve the exercises: cards, cloze deletion tests, etc.
I think it's important that exercises are tightly linked to the source material, like to a specific paragraph of an article, etc. So these materials should probably be added to the system as well.
* With the same starting point it should be easier to have a proper discussion.
* A new person can read the source and understand cards more easily with context.
* Later if they've forgotten the topic completely they can reread it and hopefully remember faster.
I also had some ideas about being able to discuss and alter every paragraph in the source. Allow it to evolve to be more clear as people come and discuss confusing points.
Since you've made you're own app, and have noticed that a good card is a good tweet: I want an app that puts my cards in my twitter feed. I check twitter way too much and don't like opening Anki to review my cards. Not sure exactly how it would work, maybe I see a few tweets, and then there's a 'card' tweet, and I interact with it as usual and then get more tweets. Just an idea.
For speeding up the card creation I've recently made a CLI application [1]. It generates cards from Markdown files.
I'd recommend having its own repository with Markdown files grouped by target domain using tags for every card. It'd help to search quickly relevant cards even in the same domain.
Nice! I made a similar thing for myself, but without nearly as much polish as your project [1].
I use a custom text file format to allow creation of cloze deletions and reversible cards as well as basic cards. I also annotate text files on export so that I can export the file again without creating duplicate cards.
The big shortcoming of my script is that it generates .tsv files and Anki only allows .tsv files to contain one note type. They also do not incorporate media.
I'd like to be able to sync edits made in Anki back to the text files and vice versa. It'd be really cool to integrate a text file parser and alternative card editing mode into Anki with a plugin.
Perhaps I'll send some PR's your way instead of duplicating effort.
I believe the mobile app is limited to just studying. Creating new cards and downloading shared decks must be done using either the desktop app or the web app.
It must be copyrighted or have some other broken dis-economy, because the most obvious product decisions that would make this huge have been rejected.
The whole point of this is to study in spare moments, like on public transit - using mobile. Uninstalled. That's so disappointing because it's so obvious, and I can't trust an ecosystem that dumb.
I didn't think it was possible to be as disappointed in an app as I am with this one.
I stand corrected. My recollection was that AnkiDroid (the Android app) didn't provide the capability of creating new cards, but I just checked and the current version of AnkiDroid does let you create new cards.
Anki (and spaced repetition) is awesome! Some things I've had used it for to great success:
1. Memorizing Japanese vocabulary, pitch-accent, and basic grammar rules. It would probably be useful, at least to some degree, for any language like this.
2. Any certification which required strict memorization. All the basic ComptTIA certs were like this, and the CCNA:R&S cert (unfortunately) required memorizing commands and their syntax.
3. Verses in the Bible (though this is pretty basic in comparison, just Address <-> Text).
In all honestly I think one of the best things you can do if you need to memorize something is make the flash cards yourself, whether with Anki, another app, or even just index cards. This forces you to think about what it is that you're trying to memorize and phrase the text of the cards in a way that you understand. After that, reviewing is just kind of "maintenance" in my opinion. There was definitely a marked difference in my retention when I was using pre-made Anki decks vs. creating them myself. It is also easier to create cards using the desktop app.
It's a side note, but I also don't agree with Anki's pricing model. The app is free on Android[1] but $25 on iOS[2]. I think I heard (I don't have a source) that the developer's justification was that they needed to make money from the all the time and effort they spent creating Anki, plus hosting costs, etc - so why not do a cheaper price on both Android/iOS, or do a free-to-download app with a subscription model? For what it's worth you can use the web version on iOS but the app is a better experience IMO.
> I also don't agree with Anki's pricing model. The app is free on Android[1] but $25 on iOS[2]. I think I heard (I don't have a source) that the developer's justification was that they needed to make money from the all the time and effort they spent creating Anki, plus hosting costs, etc - so why not do a cheaper price on both Android/iOS, or do a free-to-download app with a subscription model
People with money overwhelmingly buy iPhones. $25 is a pittance for anyone who can afford an iPhone. Anybody who uses Anki seriously gets far, far more value than $25 out of it. I used it for well over 200 hours before I stopped and I know I’ll go back to it again.
The Android version is not maintained by the developer of Anki. The iPhone app is. He chooses to charge for the iPhone app, which enables him to make a living making tens of thousands of people’s lives better.
If you don’t want to spend the price of two pizzas on an app that the modal user will use for over a hundred hours don’t.
> People with money overwhelmingly buy iPhones. $25 is a pittance for anyone who can afford an iPhone. Anybody who uses Anki seriously gets far, far more value than $25 out of it. I used it for well over 200 hours before I stopped and I know I’ll go back to it again.
I used Anki's web interface from my phone for about 6 months before I bought the iOS app. Then about 4 years after using the app almost every day, I sent them another $25 donation. I got way more than $50 worth of value out of it over the 10 years that I used it.
(I've now written my own study tool for Chinese which fixes some of the issues with using flashcards for language learning. Maybe at some point it will show up on "Show HackerNews"; but it's slow going when you've only got a few hours a week.)
Fwiw, two pizzas in US buy you up to 20 meals in a developing country. In a developing country, people often get cheaper iPhones (second hand or previous generation). The apps, however cost the same.
I agree with you - I definitely got more than $25 of value from the iOS app and was happy to pay the price. I guess my thing is that I wonder if the developer would be better off trying to solicit more users at a lower price (or a small, monthly/annual subscription) than a one-time purchase.
(co-author of the article) I wrote a longer list of categories [1] a while ago (some of which aren't mentioned in the article), here's the list:
* Basic information about countries e.g. population
* Ingredients and dishes from restaurant menus I didn't know
* Important people and places
* History facts (typically from Kindle highlights)
* Conversions between units (e.g. lbs to kg)
* Season for various vegetables and fruits
* Keyboard shortcuts for vim, readline, etc.
* Learning words and terms I don't know from Kindle/Instapaper highlights
* Useful statistics
Developing for iOS is a lot more expensive than Android - you need to pay for Apple hardware and there are ongoing fees for "notarizing" your app. It's quite normal that an iOS app would ultimately cost more.
Seems people are taking issue with my gripe about the pricing. To be clear:
I bought the iOS app, was happy to do so, and got a lot more than $25 of value from it. In my opinion, the developer might be better off with a subscription model rather than a one-time purchase.
For Japanese learners, I built two iOS apps that cater to the special needs of the language.
Manabi [0] is a flashcard app with the same algorithm as Anki but nicer UI.
Manabi Reader [1] collects a bunch of short-form reading materials, lets you tap words to look them up, and tap to add a flashcard. It tracks the individual words you read and charts your progress word by word and kanji by kanji. This app has gotten pretty popular so I have been improving it substantially.
I will check these out!
I use Anki for Japanese right now - mainly because I found Anki decks for each chapter of Genki 1 which I am working through right now.
Would love any feedback! The Reader app is really unique - no other app provides that kind of kanji-by-kanji reading progress functionality (probably because it took a tremendous amount of work under the hood to get it right).
It's a native iOS app so I would have to start from scratch on Android. I do plan to port it to macOS via Catalyst sometime after the upcoming WWDC which will help some Android users.
I'll note that this is a big pain point for me: although Android users don't pay as much for apps, they help boost word-of-mouth marketing substantially. I see some similar but (tbh) subpar apps get more word-of-mouth online even in recommendations to iOS users just because of the huge signal boost from Android users.
That looks really interesting, I'll definitely give this a try. How did you decide on the reading materials to use? There's obviously so much to choose from, I'm wondering what criteria you might use.
The app is a (very heavily) dressed-up RSS reader. I maintain a list of reading sources and add them so long as they're not going to have overly sensitive material (eg someone suggested an anonymous blog with short and colorful posts from contributors reflecting on their lives, but some of the posts talk about self-harm) or be too niche. I'm most interested in feeds that get regularly updated with new content, or where they have a trove of existing content. Also always looking out for content that's good for absolute beginners.
Some RSS feeds require additional work in the app to transform their content to make them work nicely with the reader mode, so I also take that into account.
Depends on the course, the exam as well as how long you used spaced repetition. The next time Anki will ask me what "å sløse" means is in over five years and I'm pretty sure I won't have forgotten it by then. For exams that rely less on memorization this might be different. I'm pretty sure that even if I remember all definitions from my measure theory course I still wouldn't be able to pass the exam in 10 years.
>>> It’s nothing crazy. Here’s an example from a very real non-work day:
Followed by a pretty long list.
How does he do ? There are 23 items on the list. None of which include daily tasks such as : dishes, sport, talking to family, to friends, eating, cooking, washing clothes, having a social life. All of these tasks take a lot of time, esp. if I consider the time to switch between them (which can be quite long since, hey, it's a non work day, so I'm slower...).
Do do the same thing this guy and do all the things you mentioned. I create cards on sundays only and i do anki every morning before i start working. It can be accomplished easily, sans children.
but that list is >8hrs of back to back chores (according to his estimates) for a non-work day without any of the ones your parent post mentions. I don't think that's realistic.
I imagine that the subject of the article is one of those genetic mutants who only requires 4-6 hours of sleep a day. They do exist, but are definitely statistical outliers.
Not needing 9 hours of sleep is the only real way to add time to your day.
Using Anki is less important than keeping relevant and digested notes. I use Anki daily but I don't really need to memorize everything. It's enough to be able to fast retrieve previous collected info. In the end I just don't want to repeat myself. For that I simply use one searchable text file (and a github public repository - to maintain some quality - for digested notes/cheatsheets).
A bit of a self-post, but I've compiled a list of Anki tips I've learned and found. My deck is about 14,000 cards over the last few years - so I've learned a bit from my own mistakes :)
Was just taking a break from my Anki session to check HN.
I'm a regular user, though I'm terrible at consistent, daily reviews. I somehow manage to keep on top of it, even after skipping weekends or entire weeks. You can tweak the settings to make it easier for less-than-perfect users.
Lifehack: Anki + exercise. Combining Anki with an elliptical or stationary bike improves both my review and exercise frequency and duration.
I've made my peace with the terrible UI/UX. Like many folks here I tried to roll my own SRS app at one point. But my review sessions would always get derailed with brainstorming new features or dealing with bugs. Personally, I'm not sure that the productivity gains from improving the UI/UX are worth the development time (especially when the whole point of using Anki is efficiency).
I find it difficult to do anything intellectual while exercising at a relatively high intensity. What intensity are your workout sessions? But perhaps review is different than trying to absorb a new concept?
Personal anecdote, but there's some optima for information retention at low intensity exercise. This is around moderate to vigorous walking pace or a very slow jog.
Increase intensity beyond that point and your brain's information processing ability drops below baseline.
I've found this consistently true. At lower intensity cardio for example, I can watch educational youtube videos or follow along a movie. At high intensity, I can't even watch an action movie and follow the plot. I have to switch to music and/or a sports action video (no meaningful dialog).
The more brainless the activity, the easier it is to do parallel flashcards. An elliptical or stationary bike session will be easier than an equal intensity run where you have to worry about your gait. Flashcards are little atomic chunks for you to process independent from one another. When you do flashcards the only thing you need to concern yourself with is the short prompt on that particular card in front of you. So even if you are working out pretty hard, you can still work your way through a deck, it's just going to be at a slower rate. This differs from reading a textbook or journal article, where you have to keep a certain amount of the previous passages in memory to understand the current line. You will hit a point where flashcards stop working but the ceiling is higher than with other types of learning activities. This will differ per exercise type, per intensity, and the subject you're learning.
With bluetooth headphones I've moved to doing everything I can while exercising. Even sitting down to read a book feels like precious time that could be spent combating a sedentary lifestyle like listening to the book instead.
One anti-feature makes me go away every time I return to give Anki another try: lack of a "dismiss once and forever" button. Many decks I'd like to study contain too many cards I already know. Yet I can only choose how easy it was so it would decide how often to repeat it, it won't let me choose "it's easier than saying my own name, don't repeat it ever".
It's been a while since I used Anki, but I'd suggest marking or suspending the notes you don't want to ever see again. Then you can keep them suspended, or if you marked them, you can delete them completely. Should do the trick?
Hey kids in school memorize baseball scores, league position, all the characters in Star Wars. It's very normal for people to remember/memorize things. A trick to do it more efficiently isn't going to drive anybody crazy.
They don't do it in a mechanical way attempting to maximize recall.
It's another thing to try to stuff absolutely every bit of information you've encountered in your head with lots of effort.
From reading HN I get a feeling (could be wrong feeling) that western society is always obsessed with something (or perhaps every society, but it's easier to notice this from the outside).
If it's money - you need to grab all the money you can get. If it's time - you are obliged to do something productive every second of your waking hours. And you must also find a way to sleep much less than 7-8 hours in order to maximize "productive" time.
Now you must also remember every bit of information you've seen.
It feels creepy and definitely doesn't feels sane thing to do.
What if I earned all the money in the world? What does it change?
What if I do something "productive" every second of my life? I would probably hate such life.
What's the point of trying to remember every detail from the book about telegraph history and ancient philosophy?..
It helps to read the article before commenting. The headline is overblowing it. He's not learning books by heart, just highlights some key facts he found valuable while reading and then reviews those and adds some of those go in a flashcard and zettelkasten system.
Impressive for doing it consistently but it's not learning entire books.
Spaced practice and retrieval practice, both of which are used in SRS, are two learning techniques for which there is ample evidence that they actually work [0].
Of course, You still need to decide what is worth remembering.
I just...can’t relate to living this way. Maybe you have to be a special type of person, but this whole lifestyle just seems exhausting and over-managed.
I mostly read for pleasure (and occasionally for work). Even for work, I don't tabulate what I read. I read a lot of paper books too, where uploading stuff is not practical without a lot of fuss (OCR?).
I love reading. And yes, I forget a lot of stuff. But if I had to go through all this whenever I pick up a book, I'd simply stop reading.
Ok, but I understood the person mentioned in the article does this with every book regardless of subject. And like the OP said, I just can't relate to living like that.
Maybe OP only reads books that he wants to remember, so doing that for every book is valid. On the other way around, maybe some people wouldn't be able to relate to reading a lot of books when you don't want to commit what you learned to memory
I feel exactly like you and when I read stuff like this one - I wonder whether those people understand they're doing stuff wrong and that writing about it won't magically make them better. It's as if they're trying to hack through their own shortcomings by throwing computer applications to the mix and writing an article about it makes it work.
The exact methodology in which he manages his life is probably less important than the fact that he's making a wide-sweeping effort in the first place, and maintaining consistency.
The opposite would be never writing anything down, never knowing what you're going to be doing on any given day, never estimating how long a task is going to take, reading books while making no effort to retain any information, etc.
We all dedicate energy toward transforming the natural chaos of life into order, even if it's just forcing ourselves to work a certain amount of hours per day, or re-reading paragraphs in books because we weren't paying enough attention the first time, or making grocery lists before we go to the store. These are things most of us do deliberately because we believe they make our lives better.
It's likely that there are certain things that he could be doing better, but I have no reason to believe that the structure he's set up isn't better than what he was doing before, and I think it's probably far more effective than the structure (or lack thereof) of most people's lives.
As a recently relapsed biochemistry student, with a wide variety of classes in both biology and chemistry, I use Anki exhaustively. I made at least 10k cards last semester, and that was a relatively light semester in terms of credits.
With that said, I can't imagine using such a powerful, and honestly demanding, tool for everyday knowledge and tangential facts (e.g. Greek mythology as a software engineer). It strikes me as far too much tool for the job.
It seems to me that, if you struggle with retaining things you read, some simple, lightweight note taking strategies, especially handwritten, would be sufficient. I believe that's actually a recommendation that his book recommendation, Where Good Ideas Come From, makes.
It was a light semester in terms of credits, but it still included classes in biochemistry, physical chemistry/quantum mechanics/spectroscopy, and molecular biology.
It's also important to make useful cards, and the most useful cards are almost always very simple and can be answered relatively quickly. So, for example, a complex, multi-step DNA repair mechanism (of which we needed to know quite a few, and they differ for eukaryotes and prokaryotes) for a molecular biology class becomes quite a few individual cards. In the end, there is just a lot of stuff that either just has to be memorized, or that benefits from improving recall speed for taking a 60-90 minute exam.
I kind of think it's in the same vein of the millennial trend of monetizing your hobbies. We've been told over-and-over that you have to maximize every experience in order to get ahead, so there's this compulsion to do so. I used to feel this way, but eventually I realized that I'm not in grad school, and it's ok to just enjoy things as I'm reading them.
I agree. In fact, what I find the most interesting part about reading is .... forgetting about it!
Basically concepts will sip thru, but learning the details is just a waste of brain cells, in my opinion.
Just leave the brain do it's job, don't read as it's a marathon, take breaks, think about that bits you've just read that was interesting, then promptly forget about it. You'll forget the details, but not the backbone of it.
Details don't matter in the end. Quite frankly the idea of knowing /by heart/ the name of the greek goddess blah blah blah he uses as an example would bore me solid. Worse, anyone knowing it and telling me about it would bore me solid :-)
Also notice how he mentions he remembers "trickle down economics" was a "very important idea" during Reagan's presidency. But there is no mention whatsoever about the idea itself, how it's considered today, whether it was a good or bad idea, its relation to neoliberalism, whether he understands how politically divisive the idea is, etc.
Someone remembering trickle down economics was something from the Reagan era tells me nothing about their understanding of that idea and their opinions about it. Precisely the bits I want to know!
The article not about 80's economics. He was just making a point, of course he knows the concepts he mentions with more depth than discussed in the post.
He is arguing for memorization. I'm saying memorization of this kind of "facts" seems unconvincing at best. Memorizing a "data card" about Reagan with this scribbled on it seems like something that would only be useful for a trivia quiz. "Mozart composed this or that", "Reagan's presidency had something to do with trickle down economics", "Washington had false teeth [1]", etc.
I know this isn't your main point, but I cannot resist to comment:
> The article not about 80's economics
It still pervades political discourse in many countries of Latin America, and many believe it's the cornerstone of neoliberalism, so...
----
[1] I wanted to write "wooden", basing my knowledge in LucasArts' Day of the Tentacle, but apparently this is false.
Indeed, a fact like "Shostakovich composed his 5th sympohny in 1937" may have little value on itself, but its value is compounded with related facts. For example, if you also knew that Shostakovich was Russian, you could deduce with reasonable certainty that he composed this piece while living in the USSR (knowing also that the USSR was in place in 1937).
This in turn might allow you to enjoy the piece more, because you can relate the music to the time period and better understand what it depicts and what the composers intention was.
Now, you could also simply read a brochure about the piece/composer before the concert and you would know the same or more, but this would cost you time you may not have at that moment.
---
Anyways, I don't think he is arguing for memorization, rather, he argues that you should first learn something (by reading) and then use some tools to remind you of what you have learned in order to not forget it. It turns out that the tool he chose for this is one that reminds him of some facts once in a while. The intention is not necessarily to know these facts, but to be reminded of the learned concept through these. Whether this is an effective method or not I do not know. One could end up becoming very focued on the facts, forgetting the deeper knowledge behind them (as happens in eductation, but for different reasons).
---
Maybe I went a bit too in depth with this :p
I dont understand your point about my point (the article not being about 80's economics). Are you saying that because it is important, the article should have elaborated on it?
Thanks for the reply. I see what you mean, but I'm not entirely convinced that knowing trivia is really helpful, except maybe to impress your friends (I tend to effortlessly remember all sorts of crap from Wikipedia, the more useless the better recalled, which makes for fun conversation... sometimes).
As for my nitpick, it was a tangential point I couldn't resist making: that trickle down economics is not exclusively an 80s thing, but (sadly, in my opinion) remains very relevant today.
> I dont understand your point about my point (the article not being about 80's economics). Are you saying that because it is important, the article should have elaborated on it?
Not the person you're replying to, but my feeling is that if you're writing an article about a system to use to usefully remember things, if you're going to then cite an example of something you've remembered, you should focus on what is actually important to remember about that fact. Saying "I remembered that 'trickle-down economics' was 'important in the Reagan administration'" does not do that, and could demonstrate that this memorization method is actually teaching you to remember the wrong things.
It doesn't matter if the post is or is not about 80s economics, but if you're going to use an example out of 80s economics to prove that your method is good, then show that your method actually helped you remember something important about 80s economics, not a piece of trivia that not only isn't useful, but is counter-productive to learn about if you don't remember the meat of the idea, and things like whether or not it was a good idea.
It's possible that the author did actually know and remember the useful parts of those facts, but did not actually demonstrate that his method helped him remember actual useful facts... which is kinda the point of his article.
I absolutely agree that memorizing facts help you understand the world. Famously, students find memorizing dates in history tedious and useless, but it's important. Any new event you hear of can be compared with many concurrent events. I don't know all that much about Shostakovich, but hearing that he composed a symphony in 1937 triggers a sequence of questions. What was his relationship to Stalinist purges in the 1930s? Did he have contacts with the Russian post-revolution emigre artists across Europe and America. What did he think of Stravinsky or Ravel? I'm listening to his fifth symphony now and I can't help but imagine the ominousness of the first movement having to do with all this.
> don't read as it's a marathon, take breaks, think about that bits you've just read that was interesting, then promptly forget about it.
That's basically the spaced repetition model. You need to forget a little bit, so that working memory isn't saving you, and the recall process takes effort. That's what builds long term memory. Spaced repetition systems just extend that process beyond the length of the book, in a time shorter than rereading a book.
> Quite frankly the idea of knowing /by heart/ the name of the greek goddess blah blah blah he uses as an example would bore me solid.
I agree, the names of greek gods is a bad example, unless you're a student of mythology. Spaced repetition is hard work, so probably only deploy it on things that matter. A better example might be Bayes Theorem. Incredibly powerful, but unintuitive and easily forgotten. Cards for Bayes' theorem might include the purpose: "Bayes' Theorem calculates how to adjust our prior beliefs given new evidence", as well as cards about the formula, or an intuitive visualization.
Or, maybe you want to learn more about the linux internals and spend a portion of your time memorizing the meanings of signals and errnos as a small part of a larger program of study. Sure, you can look these up in a man page, but part of the value in knowledge is knowing things exist. How many developers do you think know of SIGUSR2? Or... SIGBUS ;)
I don't see how cards would help me with mathematical ideas. You have to get them on a fundamental level, and then they're hard to forget. I'm more likely to forget the name of the theorem. But space repetition might be useful for formulas.
But I'm not too crazy about spaced repetition. I used it to learn all the capitals in the world, and after about a month I managed to do all of them without error. But after two years of barely using the knowledge, I only remember half.
Of course, I could keep reviewing it every few months to keep it fresh, but that's just not an efficient system, when you think of how many things I'm supposed to (and do) remember.
For me, at least, the best way to remember things is by tying them in as many associations and metaphors as I can, and that gives me a pretty reliable recall. It has some downsides: It's a bit more work than spaced repetition (requires creativity, for example), and it's not that good for unconnected data (but then, are random facts that useful anyway?). But I think that as a long-term method, it's much more solid.
I have had reasonable success using spaced repetition with math proofs (written in LaTeX). I create cards sparingly, often based on questions from problem sets which I shouldn't have gotten incorrect (i.e. the mistake stems from a fundamental misunderstanding of concepts, not a misstep in algebraic manipulation)
> I don't see how cards would help me with mathematical ideas. You have to get them on a fundamental level, and then they're hard to forget.
I find it too easy to trick myself into believing I understand a concept on a fundamental level. But often that "understanding" slips away, and six months later when faced with an example problem out of context I struggle to solve it.
I recently started reading "Clean Code" by "Uncle Bob" Martin and I've decided to incorporate adding notecards on what I've come across. I've created a few cards already.
But in terms of making cards on books I read for pleasure, I agree in that it seems like overkill. One of my favorite biographies is "Peter the Great" by Robert Massie, and as much as I love reading it attempting to memorize names and details and jam pack my brain with facts would fry my brain.
"At the end of the day people won't remember what you said or did, they will remember how you made them feel"
This is basically how I approach reading as well. I think it's fairly hopeless to have a huge organizing system that assists in trying to keep information sort of floating around in the conscious mind.
Human minds aren't really fact gathering machines so I think this is largely futile and the number of things to remember is too large anyway.
I think there is more sense in trying to absorb what you read, hope that it leaves an impression of some sort or gives you a cue when it becomes relevant and just go with it and using active time for creative things.
I used a lot of space repetition previously, but I've found a lot of issues with it
- Most of what I learned is not useful knowledge in anki. I made many variations of different decks, some for programming and some for other concepts
- Space repetition - the concept itself is built in many different places. For instance, google will remind me of photos that I took last year on my phone. I can check up instagram and facebook to see things my friends are doing and be reminded of things we've done. I can message people and be reminded of things too, or throw parties and likewise be hit with space-repetition if done frequently enough.
don't get me wrong though, space repetition is useful especially for learning a topic that requires a lot of vocabulary. The best way to learn something is just to be fully immersed into it, if you practice it daily you basically are applying the space-repetition algorithm in practice.
If you want to learn a new language, move to the country that speaks said language.
There's a lot of wasted time and effort in making cards and determining what is "Fringe value" - cards that expire and offer no values long term
If you want to remember things about people you care about, just message them pictures of things you;ve done together. No need to make a rolodex CRM like some of my friends do, its a waste of effort and time imo.
What anki really shines though again is language learning, or learning something with a lot of barrier-to-entry vocab. This is especically true for medical industry based applications
I practice what I use everyday and immerse myself in many programming communities and I've found that I've learned and retained information far more effectively when it's shared b/w many different perspectices and contextes
Zettelkasten has been making the rounds here this year. His process appears similar to the org-mode process using Jethro Kuan's org-roam. I've started using it and while it can be a bit tricky to set up and there's a couple mildly annoying bugs (with workarounds), for a solo project it filled the niche for me perfectly. I use it for all note taking and reading now and it is very satisfying.
Same here, org-roam just kind of "clicked" things into place for me.
I've modified a template for single-subject documents called "tags" so they're stored in ~/Org/tags/. I've also configured org-journal to store files in ~/Org/Journal/
One thing I'll be interested to discover, in the future, is how I archive or age out information. Most work stuff right now is related to a single project, so what happens when that project ends? Will I need to move things outside of ~/Org/ to remove clutter, or will I just get used to it?
I also did not find org-roam super easy to set up, but after a second try I have it running and am enjoying it.
Edit: I tried it a month ago - before installation was more than just (load)'ing the file. Now there are nicely documented use-package instructions and a layer for spacemacs
312 comments
[ 3.4 ms ] story [ 308 ms ] threadIs it possible I’m just not compatible with the way this works? Any tips on alternatives?
For the record, Anki actually works really well for me. But interest in the subject should come first.
How does Supermemo help you keep keep context?
> keeping organized notes that may or may not be used for active recall
I use a separate notes app (Bear) for this. Is there much advantage in your experience to integrating active recall notes with Evernote-style reference notes?
I don't understand how anki users are able to make cards for non-language things,I personally absolutely hate making cards while reading anything. It's hard to explain the exact process but incremental reading by contrast makes it much easier to take what's really important from an article and make cards from it.
Anki is "just" a tool to prompt you with flash cards. Designing them is up to you and extremely flexible. You're free to add text, images, audio, whatever you want.
Some ideas of how you could use Anki to learn the Greek alphabet:
- Front of the card has the upper- and lower-case glyphs and the back has their sounds, either English, IPA or an actual recording (Wikipedia has them for everything in IPA and the Greek Alphabet page has a mapping from letter to IPA). You should also do this in reverse.
- Front of the card has a Greek word with a missing space, back of the card has the complete word. You can use audio recordings here as well if it helps you (e.g. have the front read the entire word).
There are also a few pre-made decks that people have shared: https://ankiweb.net/shared/decks/greek%20alphabet
For more involved facts, Anki seems great in principle, but in practice I've found it hard to create and maintain the habit of doing the cards.
You want to memorise the entire alphabet as a sequence not just as individual letters. So you need cards of little overlapping sequences, like ABC, BCD, etc. The supermemo essays cover this scenario. With that work in place you’d learn it quickly.
As just separate letters divorced of any meaning, not hooked up with any other neurons in your brain, you’d struggle to ever recall it quickly.
This is a detail that people overlook and then wonder why it’s not working. (Not picking on you for doing that: I did the same before eventually reading way too much about it)
https://www.supermemo.com/en/archives1990-2015/articles/20ru...
...by the creator of supermemo who basically formulated the whole idea.
In Point 1 he explains you shouldn’t use spaced repetition to learn. Use it to remember.
In Point 10 he covers how to remember “enumerations” (with an alphabet example)
You have to go back to the classics when you want to do something properly.
Anki is better for inputting vocabulary with definitions and example sentences. That also puts the letters that you spent time writing into practice.
Did you use spaced rep?
In the end, I got a lot more out of doing the research to create the cards than I did from the actually looking at the cards. I might even go so far as to say that after creating the cards, they were practically worthless.
This would be less true if the exam was further from the creation of the cards, though. It was within a month.
Someone else's cards might have helped a little, but not nearly as much.
Anki is basically meant for longer term studying. The criteria for when a card matures in Anki is when the interval becomes 21+ days.
If it's anything longer, then it's probably not a subject you should be using anki to learn or you're trying to remember too large a bloc of information at once.
I have absolutely no idea how this could be a form of procrastination, or how I could more efficiently memorize information that must be memorized.
Outside of a schooling/testing context, memorization alone is rarely beneficial. You need to be able to practically apply the knowledge, which happens through experience; looking details up when you need them forms a natural kind of spaced repetition anyway that’s tuned to what you actually need to know.
Regularly looking things up is ok, but actively re-experiencing the thing you're trying to learn is a better way to make it stick.
On this we agree, but to me this means practical application in various contexts, rather than call-and-response memory drills.
Language learning is an obvious use case.
Also: law and medicine. Having knowledge mentally 'to hand' is pretty important if you've got a patient under general anaesthetic, or a judge asking you a very difficult question.
Memorizing word pairs can certainly be done with spaced repetition, but it’s unclear how much that translates to actual language ability. Second-language acquisition appears to be primarily dependent on reading (or listening to) the target language for content, and most words are learned via seeing them in context instead of being looked up in a dictionary.
I have no experience with law or medicine, but I expect the story is similar: practical knowledge is what you need to hand, and not book knowledge. Book knowledge is what gets you through the exams and into the practical part of your training.
Word pair is the easiest way to do it in anki, but it's not the only way to implement it.
Sometimes the thing holding you back from improvement is a lack of facts, and sometimes it’s poor understanding of theory. In my experience, however, the fastest way to get better at doing something is almost always to practice doing that thing; most of the time a sufficient collection of facts and theoretical understanding will come along as a side-effect of that work.
How much of their prior training was memorizing specific facts, and how much of it was drilling the mechanics of solving typical problems? The former is what most people use spaced repetition for and what I believe is of limited utility; the latter is incredibly valuable and I never meant to imply otherwise.
I'm learning German and Anki is nothing short of amazing. I'm easily able to learn ~20 words a day primarily just practicing on my phone during bathroom breaks.
On the other hand I'm also learning Russian and Anki just doesn't work. Words won't stick in my head the way they do with German.
I've heard very numerous accounts of it being great for Japanese.
There's at least great pre-made Japanese vocab decks. I'm about two months into one. Maybe less so for kanji (I at least don't have one), so that's probably something you need to do outside.
Because (for English speakers) Russian is harder than German. You need to reduce the number of words learned per day. Japanese is even harder to learn, but you have to be persistent. I spent more than an hour every single day for several years straight to learn the vocabulary.
Also there's a unique challenge with Russian is that nouns, verbs and adjectives are all conjugated, and the rules are quite complex. The noun cases would be completely foreign to non-Slavic language speaker. Even if you learn the vocabulary, using it in sentences is not trivial at all.
Modern Russian has 6 cases (with a couple more that pop up very rarely) with 3 genders and singular/plural endings. It's a total of 36 possible combinations, which is not that bad. Masculine and neuter also share most endings so there aren't actually 36 unique ones
Probably the most difficult thing about cases when I was learning Russian was remembering which verbs took a different case than I would expect, e.g. dative instead of accusative for what seems like a direct object.
Another incredibly tricky one for Slavic languages is imperfective vs perfective aspects. In most other languages, perfect vs imperfect is just a standard construction. In Polish (and Russian, though my examples here are Polish), you change the verb itself. How? Well, sometimes you put "po" in front of it, like rozmawiać / porozmawiać. Sometimes it's "z" (jeść / zjeść), sometimes "u", "na" or "wy". And sometimes you just give up, like oglądać (but obejrzeć in the perfective), widzieć (zobaczyć), mowić (powiedzieć).
(For this reason I have found it very hard to progress in Polish without conversation with native speakers who aren't too polite to correct me.)
36 isn't that bad but Russian also has 253 irregular verbs [0], each of which adds another set of combinations. A lot of them are just small tweaks and there are some that follow a set of patterns (e.g. идти with its various prefixes) but it still adds a lot of overhead.
[0]: https://en.wiktionary.org/wiki/Category:Russian_irregular_ve...
> The noun cases would be completely foreign to non-Slavic language speaker.
Picking those up was actually fairly straightforward. Cases may be completely foreign but the grammatical concepts behind them aren't. For example Russian has nominative, accusative and dative but English has subjects, direct objects and indirect objects and I learned all those in school.
I think somewhat ironically, the biggest grammatical problem I had with Russian was the tenses. English has 12 and Russian has ~5 so it was hard to figure out how to express a given English sentence in Russian.
But I think the biggest problem is phoenetics. On the surface it looks simple (unstressed o -> a) but there are so many extra rules if you want to speak properly (e.g. voiced consonant before a voiceless consonant becomes voiceless).
My time isn't so optimised that 15 minutes each day on Anki is otherwise going to be spent on deep work.
http://augmentingcognition.com/ltm.html This writeup estimates his usage of Anki is 4-7 minutes of review time for each card over 20 years. So, if it's worth taking 10 minutes to remember a fact, it's worth putting in Anki. (This writeup also explains it's not worth putting things you don't care about into Anki).
Anki seems well suited for domains where you need to quickly access facts from a wide domain, e.g. vocabulary for language learning.
It's harder to see as important for programming. If you're programming you'll be able to recall the things you have experience using.
What method do you use to format a datetime object in python:
strftime()
How do you create a new virtual env using python 3
python3 -m venv <NAME>
Those may seem stupid to you, but I was looking them up and now i do not need to. I also use them for other concepts when I want to know a language better. For example
What are the three prototypical methods of a js promise?
then, catch, finally
What does the constructor of a promise take?
and executor function
How many parameters does a executor take?
2
What are the two parameter of the executor function (constructor of a js promise)
resolve() and reject()
Things like this just help me as a programmer. I agree it's not for everyone, but I have the time an enjoying doing it.
Why not just `new Promise((res, rej) => ..)` instead of intermediate facts like "an executor takes 2 arguments"? I don't think I'd be able to recall what's meant by 'an executor' without first recalling that snippet.
---- What happen to slaves in Africa after the slave trade was abolished by the America/UK in the early + mid 1800s?
simply led to a redeployment of the slaves, who were now used within Africa rather than in the Americas
“So the abolition of the slave trade, rather than making slavery in Africa wither away, simply led to a redeployment of the slaves, who were now used within Africa rather than in the Americas. Moreover, many of the political institutions the slave trade had wrought in the previous two centuries were unaltered and patterns of behavior persisted. For example, in Nigeria in the 1820s and ’30s the once-great Oyo Kingdom collapsed. It was undermined by civil wars and the rise of the Yoruba city-states, such as Illorin and Ibadan, that were directly involved in the slave trade, to its south. In the 1830s, the capital of Oyo was sacked, and after that the Yoruba cities contested power with Dahomey for regional dominance. They fought an almost continuous series of wars in the first half of the century, which generated a massive supply of slaves. Along with this went the normal rounds of kidnapping and condemnation by oracles and smaller-scale raiding. Kidnapping was such a problem in some parts of Nigeria that parents would not let their children play outside for fear they would be taken and sold into slavery.”
Excerpt From: Daron Acemoglu. “Why Nations Fail.” iBooks.
----
What are two reasons why large scale wars no longer occur ?
1. Price of war has gone up because of Atomic weapons. 2. Weath is no longer is physical goods (gold,etc) but in the minds of the citizen (Silicon Valley).
Scholars have sought to explain this happy development in more books and articles than you would ever want to read yourself, and they have identified several contributing factors. First and foremost, the price of war has gone up dramatically. The Nobel Peace Prize to end all peace prizes should have been given to Robert Oppenheimer and his fellow architects of the atomic bomb. Nuclear weapons have turned war between superpowers into collective suicide, and made it impossible to seek world domination by force of arms.
Secondly, while the price of war soared, its profits declined. For most of history, polities could enrich themselves by looting or annexing enemy territories. Most wealth consisted of material things like fields, cattle, slaves and gold, so it was easy to loot it or occupy it. Today, wealth consists mainly of human capital and organizational know-how. Consequently it is difficult to carry it off or conquer it by military force.
Harari, Yuval Noah. Sapiens (p. 372). Harper. Kindle Edition.
-----
Basically, I create cards with short answers but provide the context (usually the paragraph I found it in). This method works extremely well for remember everything you read. I could write a more technical blog post on this if people actually care and find this interesting.
Is becoming more knowledgable just "procrastination?" Is reading non-fiction outside your specialty procrastination?
I feel like for much of human history we've held people who have read an entire library's-worth of books in high esteem. Were they just procrastinating?
I still see where you're coming from. I used to think that people who meditated, or hiked the Appalachian Trail were "selfish," because they were doing something that benefitted only them. And I still find it hard to understand someone who decides to learn Mandarin at the age of 93. But people should be allowed to enlighten and actualize themselves however they wish, and it seems absurd to think that a planet full of more knowledgable people would be a bad thing.
Incidentally it also supports Zettelkasten like note taking like described in the article.
- I have a huge Zettelkasten note archive (I manage it through Sublime-Zk, which is getting a bit slow to operate it). Can I somehow directly import all my Zettel notes?
- I tried to import my Anki cards, but it seems like while I could go through them, I couldn't assign them to any particular deck/reclassify them. They show up as part of a "Not Found" deck and I don't see any option/drag-and-drop method to move them anywhere else.
I'll definitely be giving Mochi a try. Thanks a lot for creating it!
You can import the markdown files as notes, but it won’t maintain the linking. If you send me a sample I can look into adding a Zettelkasten specific importer.
> I tried to import my Anki cards, but it seems like while I could go through them, I couldn't assign them to any particular deck/reclassify them. They show up as part of a "Not Found" deck and I don't see any option/drag-and-drop method to move them anywhere else.
Sounds like the importer derped a bit and couldn’t find the deck associated with those cards. If you send me the .apkg file I can take a look. Anki’s data model is a little weird and there’s probably some edge cases I haven’t accounted for.
Would love to see sample decks for languages. Like learning basic 1000 words & sentences in different languages.
I created something similar which I no longer work on (but keep running on a maintenance-type basis). Feel free to be inspired/take features or just compare with another approach.
Good luck dude!
https://about.vocabifyapp.com
- Memrise - https://www.memrise.com/ - Has a very extensive selection of decks for language learning
- Supermemo - https://www.supermemo.com/en - Has a (disputed) claim that its algorithm beats Anki, various pre-made decks
- Drops - https://languagedrops.com/ - Mobile-only, has a higher reliance on pictures and audio over written translations.
If you use SM and you use anki both for a significant amount of time it's not hard to see that SM leaves you with way less reps. Anki is based on the very first version of the SuperMemo algorithm (from ~1990 or so which was the first SRS algorithm ever) known as SM2. Current version of SM uses SM18.
To add to this, there's a set of Winetricks recipes for running SM on Linux.[0] (Disclaimer: I haven't tried them, though I've been planning to, for a while.)
[0] https://github.com/alessivs/supermemo-wine
You don’t want to get it wrong because you may corrupt people’s data
If you'd ask me it's a perfect example of how not to structure a database. It's not even in the first normal form.
I'd love to create an Anki clone that has better UI/UX and a better database format and I think that would remedy some of the biggest pain points about Anki. But since most Anki users (me included) use several plugins that are quite essential to their Anki experience, an Anki clone would have to support those plugins, too.
1. A lot of spaced repetition apps don't optimize for fast card creation. I like to make a card as soon as I come across something interesting, but the delay to get Anki started, and to create the card, etc just makes it frustrating. I wound up building my own app for this.
2. Cloze deletion is surprisingly effective. For those who don't know, cloze deletion involves taking a sentence, blanking the interesting parts out e.g. "The moon landing happened on _______", and trying to recall the blank parts. This is effective because it's a quick way to make cards, thus solving the problem of slow card creation.
3. I fared better with lots of small, one-sentence cards. My rule of thumb is that a card should fit in a tweet.
4. Subjectively speaking, using spaced repetition didn't just help me recall stuff in my cards, it also helped me recall stuff in general.
I have this URL in my bookmarks toolbar:
https://ankiuser.net/edit/
It enables me to directly add new cards (just don't forget to sync)
Exporting to Anki from Emacs is done via the anki-editor mode for Emacs.
https://github.com/louietan/anki-editor
Being able to add my slow-but-steady Lisp learning notes into Anki would be very cool.
I used Anki to study Chinese for a period of about 10 years. At some point I decided that I wanted to memorize the poker "outs" (probabilities of filling out a hand based on what had currently been dealt). Then I went through a time where I was really busy and didn't study the poker deck for a month (but I made time for Chinese). When I came back to the poker deck, the "spaced repetition" system was completely broken: I had a massive long list of cards that had expired, most of which I'd completely forgotten; but it just kept showing them to me in one giant loop, rather than focusing on a few to actually teach me. And I didn't even have a clean way of telling it, "Just pretend I haven't seen any of these cards at all". I ended up just deleting the deck; that discouraged me from doing anything else I wasn't willing to commit to doing every single day.
If you make that number the same as the "new cards per day"-number, that might be a work around for your issue?
- If you want to just push cards to the back of the queue, but remember timings and history, there's a built-in command to do that.
- If you want to change the current interval of a card, but keep your history, there's also a built-in command for that.
- If you need to do the above in bulk, the "Reset Card Scheduling" plugin can make this a lot more convenient.
- If you really do want to completely reset a card, forget all history, etc., and treat it as a new card, there's the "Remove Card History" plugin.
Finally, when returning to old decks, I found I usually get the most mileage by leaving all that stuff alone and just suspending the whole thing, and then un-suspending them at a steady pace. I neglected my Kanji deck for months, and I used that approach to get myself back up to speed by doing the catch-up review in the original (RTK) order I originally learned them rather than based on Anki's priority.
But it feels so hacky when you need them for things that seem like rather elementary functionality like resetting a deck.
Calibre feels the same way like needing a plugin just to estimate page count. I feel lucky when a plugin actually works.
Then on day 1 it will want to show you cards 1-5; day two it will show you 1-5 again and also 6-10; day three it will show you 6-10 again and also 7-15 (to simplify somewhat). So you're having some cards you know and some that are new.
If you have 50 cards you've forgotten, then on day 1 it will show you 1-10; on day 2 it will show you 11-20; day 3 it will show you 21-30, and so on -- all completely new. That makes it far more of a grind.
For each new class, I would start a new deck, and after the course was over I would move it over to my main (huge) deck, and it's worked surprisingly well.
Typically (now that I'm out), I add small cards from time to time to my deck, but if there's a large quantity of things I am working on learning at any given time (say I'm going to add 50-100 new cards), I'll put them in their own deck for a little bit.
Ultimately, my repetition does end up being a wide array of subjects though.
I just want a "dumb" flashcard application that doesn't try to apply "smart" techniques/heuristics.
I was just looking for a way to do the same paper card quizzing digitally without having to actually carry decks on my person.
Every app I've tried over the years powered under the hood by Anki had this issue.
It's been a few years since I last checked it out, but an option for a dumb mode would have had me throwing money at it just to even try it out.
There’s a “cram seen cards” feature under custom study that’ll just give you a temporary brand new deck you can run through in random order
Clearly, millions of people are benefiting from this app.
I'm just finding the features and interface very bombarding. And, the sheer number of screen taps just to study the flash cards in the manner that I'd like (and even then, not fully within my perceived control) is pretty repelling.
Thank you for pointing that out that feature. It might help eventually with conversion.
One of the wonderful things about software assisting with spaced repetition is that it automatically selects and remembers efficient intervals. It's crazy how little time is spent per flashcard with this automation.
I would prefer to iterate over small set of cards until I know them quite well.
There's absolutely no need to "reset". Ever.
If you have only ten minutes to devote to Anki, then only spend ten minutes. If at some later point, you have more time, then spend that time.
Set the maximum reviews per day to something you can do most days -- for me, that's 250 cards. If you're behind, turn off 'new' cards. And eventually, you will catch up.
What if you don't do Anki for a month or two? You still don't need to reset. Anything you remember after that month will have a much longer time until you next see it.
Only use Anki as much as you have time for it. Let it figure out which cards to show you. Resetting messes with the algorithm for which cards to show.
Each time I've gone back to a deck, I've just slugged out the few days of heavy reviews and let Anki take care of the rest. The stuff that I've forgotten, Anki will nag me with. The stuff that I've retained for years, stays retained.
One thing that I do suggest is to limit the maximum interval to 365 days, and to remove the review limit. I also tend to use the "hard" answer on cards that I've retained in decks that I've ignored for some time.
(SRS fails in this regards)
A big problem I see with pre-made decks is that they contain just too much information.
Can someone tell me if this is a crazy good idea? Social network + spaced repetition. Why should you build all your own cards? Taking a class with classmates? Create a group & create cards during lecture, review/curate cards afterwards during study session, then rate which ones were most useful after the exam (or homework).
but my cards are hard to read by other parties, when I optimize for my own learning
I think it's important that exercises are tightly linked to the source material, like to a specific paragraph of an article, etc. So these materials should probably be added to the system as well. * With the same starting point it should be easier to have a proper discussion. * A new person can read the source and understand cards more easily with context. * Later if they've forgotten the topic completely they can reread it and hopefully remember faster.
I also had some ideas about being able to discuss and alter every paragraph in the source. Allow it to evolve to be more clear as people come and discuss confusing points.
I'd recommend having its own repository with Markdown files grouped by target domain using tags for every card. It'd help to search quickly relevant cards even in the same domain.
[1] https://github.com/ashlinchak/mdanki
I use a custom text file format to allow creation of cloze deletions and reversible cards as well as basic cards. I also annotate text files on export so that I can export the file again without creating duplicate cards.
The big shortcoming of my script is that it generates .tsv files and Anki only allows .tsv files to contain one note type. They also do not incorporate media.
I'd like to be able to sync edits made in Anki back to the text files and vice versa. It'd be really cool to integrate a text file parser and alternative card editing mode into Anki with a plugin.
Perhaps I'll send some PR's your way instead of duplicating effort.
1. https://github.com/isaiahstjohn/flashbang
The whole point of this is to study in spare moments, like on public transit - using mobile. Uninstalled. That's so disappointing because it's so obvious, and I can't trust an ecosystem that dumb.
I didn't think it was possible to be as disappointed in an app as I am with this one.
1. Memorizing Japanese vocabulary, pitch-accent, and basic grammar rules. It would probably be useful, at least to some degree, for any language like this.
2. Any certification which required strict memorization. All the basic ComptTIA certs were like this, and the CCNA:R&S cert (unfortunately) required memorizing commands and their syntax.
3. Verses in the Bible (though this is pretty basic in comparison, just Address <-> Text).
In all honestly I think one of the best things you can do if you need to memorize something is make the flash cards yourself, whether with Anki, another app, or even just index cards. This forces you to think about what it is that you're trying to memorize and phrase the text of the cards in a way that you understand. After that, reviewing is just kind of "maintenance" in my opinion. There was definitely a marked difference in my retention when I was using pre-made Anki decks vs. creating them myself. It is also easier to create cards using the desktop app.
It's a side note, but I also don't agree with Anki's pricing model. The app is free on Android[1] but $25 on iOS[2]. I think I heard (I don't have a source) that the developer's justification was that they needed to make money from the all the time and effort they spent creating Anki, plus hosting costs, etc - so why not do a cheaper price on both Android/iOS, or do a free-to-download app with a subscription model? For what it's worth you can use the web version on iOS but the app is a better experience IMO.
[1] https://play.google.com/store/apps/details?id=com.ichi2.anki
[2] https://apps.apple.com/us/app/ankimobile-flashcards/id373493...
People with money overwhelmingly buy iPhones. $25 is a pittance for anyone who can afford an iPhone. Anybody who uses Anki seriously gets far, far more value than $25 out of it. I used it for well over 200 hours before I stopped and I know I’ll go back to it again.
The Android version is not maintained by the developer of Anki. The iPhone app is. He chooses to charge for the iPhone app, which enables him to make a living making tens of thousands of people’s lives better.
If you don’t want to spend the price of two pizzas on an app that the modal user will use for over a hundred hours don’t.
I used Anki's web interface from my phone for about 6 months before I bought the iOS app. Then about 4 years after using the app almost every day, I sent them another $25 donation. I got way more than $50 worth of value out of it over the 10 years that I used it.
(I've now written my own study tool for Chinese which fixes some of the issues with using flashcards for language learning. Maybe at some point it will show up on "Show HackerNews"; but it's slow going when you've only got a few hours a week.)
I bought the iOS app, was happy to do so, and got a lot more than $25 of value from it. In my opinion, the developer might be better off with a subscription model rather than a one-time purchase.
Manabi [0] is a flashcard app with the same algorithm as Anki but nicer UI.
Manabi Reader [1] collects a bunch of short-form reading materials, lets you tap words to look them up, and tap to add a flashcard. It tracks the individual words you read and charts your progress word by word and kanji by kanji. This app has gotten pretty popular so I have been improving it substantially.
[0] https://manabi.io
[1] https://reader.manabi.io/
I'll note that this is a big pain point for me: although Android users don't pay as much for apps, they help boost word-of-mouth marketing substantially. I see some similar but (tbh) subpar apps get more word-of-mouth online even in recommendations to iOS users just because of the huge signal boost from Android users.
The app is a (very heavily) dressed-up RSS reader. I maintain a list of reading sources and add them so long as they're not going to have overly sensitive material (eg someone suggested an anonymous blog with short and colorful posts from contributors reflecting on their lives, but some of the posts talk about self-harm) or be too niche. I'm most interested in feeds that get regularly updated with new content, or where they have a trove of existing content. Also always looking out for content that's good for absolute beginners.
Some RSS feeds require additional work in the app to transform their content to make them work nicely with the reader mode, so I also take that into account.
>>> It’s nothing crazy. Here’s an example from a very real non-work day:
Followed by a pretty long list.
How does he do ? There are 23 items on the list. None of which include daily tasks such as : dishes, sport, talking to family, to friends, eating, cooking, washing clothes, having a social life. All of these tasks take a lot of time, esp. if I consider the time to switch between them (which can be quite long since, hey, it's a non work day, so I'm slower...).
We're not equals.
Not needing 9 hours of sleep is the only real way to add time to your day.
A bit of a self-post, but I've compiled a list of Anki tips I've learned and found. My deck is about 14,000 cards over the last few years - so I've learned a bit from my own mistakes :)
I'm a regular user, though I'm terrible at consistent, daily reviews. I somehow manage to keep on top of it, even after skipping weekends or entire weeks. You can tweak the settings to make it easier for less-than-perfect users.
Lifehack: Anki + exercise. Combining Anki with an elliptical or stationary bike improves both my review and exercise frequency and duration.
I've made my peace with the terrible UI/UX. Like many folks here I tried to roll my own SRS app at one point. But my review sessions would always get derailed with brainstorming new features or dealing with bugs. Personally, I'm not sure that the productivity gains from improving the UI/UX are worth the development time (especially when the whole point of using Anki is efficiency).
Increase intensity beyond that point and your brain's information processing ability drops below baseline.
I've found this consistently true. At lower intensity cardio for example, I can watch educational youtube videos or follow along a movie. At high intensity, I can't even watch an action movie and follow the plot. I have to switch to music and/or a sports action video (no meaningful dialog).
I have a feeling that it's futile and could even be a bad thing for mental health and sanity.
It's another thing to try to stuff absolutely every bit of information you've encountered in your head with lots of effort.
From reading HN I get a feeling (could be wrong feeling) that western society is always obsessed with something (or perhaps every society, but it's easier to notice this from the outside).
If it's money - you need to grab all the money you can get. If it's time - you are obliged to do something productive every second of your waking hours. And you must also find a way to sleep much less than 7-8 hours in order to maximize "productive" time.
Now you must also remember every bit of information you've seen.
It feels creepy and definitely doesn't feels sane thing to do.
What if I earned all the money in the world? What does it change?
What if I do something "productive" every second of my life? I would probably hate such life.
What's the point of trying to remember every detail from the book about telegraph history and ancient philosophy?..
This obsession feels really weird to me.
Harnessing this for fun or profit is normal human behavior.
All those strawmen - I'll leave that for another time.
Impressive for doing it consistently but it's not learning entire books.
[1] https://news.ycombinator.com/item?id=22484060
Of course, You still need to decide what is worth remembering.
[0] https://www.learningscientists.org/
I mostly read for pleasure (and occasionally for work). Even for work, I don't tabulate what I read. I read a lot of paper books too, where uploading stuff is not practical without a lot of fuss (OCR?).
I love reading. And yes, I forget a lot of stuff. But if I had to go through all this whenever I pick up a book, I'd simply stop reading.
To me, this article is just more white noise.
The opposite would be never writing anything down, never knowing what you're going to be doing on any given day, never estimating how long a task is going to take, reading books while making no effort to retain any information, etc.
We all dedicate energy toward transforming the natural chaos of life into order, even if it's just forcing ourselves to work a certain amount of hours per day, or re-reading paragraphs in books because we weren't paying enough attention the first time, or making grocery lists before we go to the store. These are things most of us do deliberately because we believe they make our lives better.
It's likely that there are certain things that he could be doing better, but I have no reason to believe that the structure he's set up isn't better than what he was doing before, and I think it's probably far more effective than the structure (or lack thereof) of most people's lives.
With that said, I can't imagine using such a powerful, and honestly demanding, tool for everyday knowledge and tangential facts (e.g. Greek mythology as a software engineer). It strikes me as far too much tool for the job.
It seems to me that, if you struggle with retaining things you read, some simple, lightweight note taking strategies, especially handwritten, would be sufficient. I believe that's actually a recommendation that his book recommendation, Where Good Ideas Come From, makes.
It's also important to make useful cards, and the most useful cards are almost always very simple and can be answered relatively quickly. So, for example, a complex, multi-step DNA repair mechanism (of which we needed to know quite a few, and they differ for eukaryotes and prokaryotes) for a molecular biology class becomes quite a few individual cards. In the end, there is just a lot of stuff that either just has to be memorized, or that benefits from improving recall speed for taking a 60-90 minute exam.
Just leave the brain do it's job, don't read as it's a marathon, take breaks, think about that bits you've just read that was interesting, then promptly forget about it. You'll forget the details, but not the backbone of it.
Details don't matter in the end. Quite frankly the idea of knowing /by heart/ the name of the greek goddess blah blah blah he uses as an example would bore me solid. Worse, anyone knowing it and telling me about it would bore me solid :-)
Also notice how he mentions he remembers "trickle down economics" was a "very important idea" during Reagan's presidency. But there is no mention whatsoever about the idea itself, how it's considered today, whether it was a good or bad idea, its relation to neoliberalism, whether he understands how politically divisive the idea is, etc.
Someone remembering trickle down economics was something from the Reagan era tells me nothing about their understanding of that idea and their opinions about it. Precisely the bits I want to know!
He is arguing for memorization. I'm saying memorization of this kind of "facts" seems unconvincing at best. Memorizing a "data card" about Reagan with this scribbled on it seems like something that would only be useful for a trivia quiz. "Mozart composed this or that", "Reagan's presidency had something to do with trickle down economics", "Washington had false teeth [1]", etc.
I know this isn't your main point, but I cannot resist to comment:
> The article not about 80's economics
It still pervades political discourse in many countries of Latin America, and many believe it's the cornerstone of neoliberalism, so...
----
[1] I wanted to write "wooden", basing my knowledge in LucasArts' Day of the Tentacle, but apparently this is false.
This in turn might allow you to enjoy the piece more, because you can relate the music to the time period and better understand what it depicts and what the composers intention was.
Now, you could also simply read a brochure about the piece/composer before the concert and you would know the same or more, but this would cost you time you may not have at that moment.
---
Anyways, I don't think he is arguing for memorization, rather, he argues that you should first learn something (by reading) and then use some tools to remind you of what you have learned in order to not forget it. It turns out that the tool he chose for this is one that reminds him of some facts once in a while. The intention is not necessarily to know these facts, but to be reminded of the learned concept through these. Whether this is an effective method or not I do not know. One could end up becoming very focued on the facts, forgetting the deeper knowledge behind them (as happens in eductation, but for different reasons).
---
Maybe I went a bit too in depth with this :p
I dont understand your point about my point (the article not being about 80's economics). Are you saying that because it is important, the article should have elaborated on it?
As for my nitpick, it was a tangential point I couldn't resist making: that trickle down economics is not exclusively an 80s thing, but (sadly, in my opinion) remains very relevant today.
Not the person you're replying to, but my feeling is that if you're writing an article about a system to use to usefully remember things, if you're going to then cite an example of something you've remembered, you should focus on what is actually important to remember about that fact. Saying "I remembered that 'trickle-down economics' was 'important in the Reagan administration'" does not do that, and could demonstrate that this memorization method is actually teaching you to remember the wrong things.
It doesn't matter if the post is or is not about 80s economics, but if you're going to use an example out of 80s economics to prove that your method is good, then show that your method actually helped you remember something important about 80s economics, not a piece of trivia that not only isn't useful, but is counter-productive to learn about if you don't remember the meat of the idea, and things like whether or not it was a good idea.
It's possible that the author did actually know and remember the useful parts of those facts, but did not actually demonstrate that his method helped him remember actual useful facts... which is kinda the point of his article.
That's basically the spaced repetition model. You need to forget a little bit, so that working memory isn't saving you, and the recall process takes effort. That's what builds long term memory. Spaced repetition systems just extend that process beyond the length of the book, in a time shorter than rereading a book.
> Quite frankly the idea of knowing /by heart/ the name of the greek goddess blah blah blah he uses as an example would bore me solid.
I agree, the names of greek gods is a bad example, unless you're a student of mythology. Spaced repetition is hard work, so probably only deploy it on things that matter. A better example might be Bayes Theorem. Incredibly powerful, but unintuitive and easily forgotten. Cards for Bayes' theorem might include the purpose: "Bayes' Theorem calculates how to adjust our prior beliefs given new evidence", as well as cards about the formula, or an intuitive visualization.
Or, maybe you want to learn more about the linux internals and spend a portion of your time memorizing the meanings of signals and errnos as a small part of a larger program of study. Sure, you can look these up in a man page, but part of the value in knowledge is knowing things exist. How many developers do you think know of SIGUSR2? Or... SIGBUS ;)
But I'm not too crazy about spaced repetition. I used it to learn all the capitals in the world, and after about a month I managed to do all of them without error. But after two years of barely using the knowledge, I only remember half.
Of course, I could keep reviewing it every few months to keep it fresh, but that's just not an efficient system, when you think of how many things I'm supposed to (and do) remember.
For me, at least, the best way to remember things is by tying them in as many associations and metaphors as I can, and that gives me a pretty reliable recall. It has some downsides: It's a bit more work than spaced repetition (requires creativity, for example), and it's not that good for unconnected data (but then, are random facts that useful anyway?). But I think that as a long-term method, it's much more solid.
> I don't see how cards would help me with mathematical ideas. You have to get them on a fundamental level, and then they're hard to forget.
I find it too easy to trick myself into believing I understand a concept on a fundamental level. But often that "understanding" slips away, and six months later when faced with an example problem out of context I struggle to solve it.
But in terms of making cards on books I read for pleasure, I agree in that it seems like overkill. One of my favorite biographies is "Peter the Great" by Robert Massie, and as much as I love reading it attempting to memorize names and details and jam pack my brain with facts would fry my brain.
"At the end of the day people won't remember what you said or did, they will remember how you made them feel"
This is basically how I approach reading as well. I think it's fairly hopeless to have a huge organizing system that assists in trying to keep information sort of floating around in the conscious mind.
Human minds aren't really fact gathering machines so I think this is largely futile and the number of things to remember is too large anyway.
I think there is more sense in trying to absorb what you read, hope that it leaves an impression of some sort or gives you a cue when it becomes relevant and just go with it and using active time for creative things.
- Most of what I learned is not useful knowledge in anki. I made many variations of different decks, some for programming and some for other concepts
- Space repetition - the concept itself is built in many different places. For instance, google will remind me of photos that I took last year on my phone. I can check up instagram and facebook to see things my friends are doing and be reminded of things we've done. I can message people and be reminded of things too, or throw parties and likewise be hit with space-repetition if done frequently enough.
don't get me wrong though, space repetition is useful especially for learning a topic that requires a lot of vocabulary. The best way to learn something is just to be fully immersed into it, if you practice it daily you basically are applying the space-repetition algorithm in practice.
If you want to learn a new language, move to the country that speaks said language.
There's a lot of wasted time and effort in making cards and determining what is "Fringe value" - cards that expire and offer no values long term
If you want to remember things about people you care about, just message them pictures of things you;ve done together. No need to make a rolodex CRM like some of my friends do, its a waste of effort and time imo.
What anki really shines though again is language learning, or learning something with a lot of barrier-to-entry vocab. This is especically true for medical industry based applications
I practice what I use everyday and immerse myself in many programming communities and I've found that I've learned and retained information far more effectively when it's shared b/w many different perspectices and contextes
https://blog.jethro.dev/posts/how_to_take_smart_notes_org/
https://news.ycombinator.com/item?id=22337681
I've modified a template for single-subject documents called "tags" so they're stored in ~/Org/tags/. I've also configured org-journal to store files in ~/Org/Journal/
One thing I'll be interested to discover, in the future, is how I archive or age out information. Most work stuff right now is related to a single project, so what happens when that project ends? Will I need to move things outside of ~/Org/ to remove clutter, or will I just get used to it?
Edit: I tried it a month ago - before installation was more than just (load)'ing the file. Now there are nicely documented use-package instructions and a layer for spacemacs
[0]: https://github.com/Kungsgeten/org-brain