Anki as a supplement to Zk
Haven't been here awhile... always good to see the warm posts from @Will when I do!
Curious to hear your thoughts about "flashcards". I recently used them to get over a big hump.
I was trying to learn something specific in programming. It is a huge library, and although I am a fluent programmer (having programmed everyday for the last two decades), I was resisiting this.
What helped me over the hump was a combination of my Zk with Anki, a free flashcard system, which does spaced repetition. I think I now know the library I wanted to master really well: the size is what had been a hinderance and no longer seems to be. I made around 200 cards, and I know them inside out now.
Long term, of course, it is the zettelkasten that will help me with the big picture. But @Sascha or others: have you experimented with Anki and if so how has it worked out?
Howdy, Stranger!
Comments
There are some previous discussions about this. Here are just a few of the discussions that can be found by searching the forum for Anki OR "spaced repetition":
You may also like the articles on @alkhan's website, LeanAnki.com, for example:
@amahabal What are your goals?
As you found out already, memory helps
You use 20 rules of knowledge formulation already?
In ZK i personally see two use cases for SRS:
Solidifying in memory learned/encoded concepts/notes. You can write prompts (QnAs) right in the cards.
You may code a solution which will sync prompts with anki deck automatically. Logseq has
anki sync plugin that does this.
Pseudo-SRS - for keeping in memory some amount of unrelated ideas/concepts (which you can't connect to anything yet, or it just happened that they separate from everything by its nature) - you don't spend much effort here, you just reread/glance or briefly think about them.
Majority of those items are not meant to live in SRS for a long time - some for several week, some for several months, until you decide what to do with them or until you decide that it's stable enough in your mind
As for processing multiple zettels - better use incremental reading/writing
Thanks, @Andy and @Ydkd , for the helpful links.
Thanks for asking about my goals: I was going to knee-jerk respond with "two prongs", but there are three.
The first, and primary, is deep understanding of various aspects of cognition. I use my Zettlekasten for this (Archive), and tend to write very long structured notes with a huge amount of interlinking.
Second, I also program a lot, and for some tasks I want certain idiomatic patterns to be in my muscle memory, and here is where I use Anki. It has helped me be much faster at certain tasks. For example, in using, say, pgplot or pytorch, there are all sorts of magic incantations (such as how to get subplots, how to set colormaps, how to limit to a lower traingular tensor, etc.). Nothing deep whatsoever in each individually, and real no need to cross reference, and no real cost to not remembering except a bit of a slowdown. This I use Anki for, and it seems to work.
The third, which had not immediately come to mind, is writing. I feel guilty for not writing more, and I do not write publicly as much as I feel I should (https://amahabal.substack.com/). When I do, and also for talks and presentations, I am a big fan of using the method described in "The Clockwork Muse" by Eviatar Zerubavel, and then I draw a lot on my Zk, and mix that with Scrivener.
--Abhijit
I really like the combination of pulling information from my ZK and combining that with other information I've written in Scrivener, and then using Scrivener as a writing tool. I employ this method extensively for work projects, which involves a lot of review work. It's like drinking from a fire hose for 3 or 4 days, then taking a day to write a critique of everything that was presented.
I only wish I'd had all these tools when I was writing a PhD thesis in the late 1970's. Back then, it was pen (or pencil) and paper
A while ago I've read this piece Augmenting Long-term Memory by Michael Nielsen and has since put a more serious effort making Anki part of my regular learning sequence, and I think Anki is a very good complement to Zettelkasten-style note-taking. In my setup of Org Roam/Emacs, integrating Anki card creation into the workflow is relatively straightforward. I convert portion of my notes into Anki cards, when I think my muscle memory should remember it.
There are things that are fit for simple memorization. At the very least, it will help keeping them in your awareness until they have a chance to make a stronger connection with your mental model in some other context.
Anki also made me realize there are things that don't stick well no matter how much I repeat. It might stick temporarily, but it's so fragile that when I come back to it after a while, I still find it elusive. I take it as a sign that some knowledge really is not really important to me. If it doesn't stick short-term even with Anki, I'm trying to learn wrong stuff. (Or, perhaps I'm failing to make effective cards for Anki on the matter.)
I've read something about spaced repetition in the past, but I'm not convinced about the importance of the "spaced" term. The repetition in planned temporal slots.
I'se seen, in my experience, that the principles of Zettelkasten are enough, on their own, to make an "emergent repetition by resurfacing" dynamic over time.
More frictionless, more pleasant, and I feel, in my case, that it develops a more robust and elastic kind of learning than memorization.
Only my experience, anyway
The effectiveness of spaced repetition (and a few other learning techniques) appears very well established by now. I first learned about those from Make It Stick: The Science of Successful Learning by Brown, Roediger III, McDaniel, and people still talk about basically the same set of techniques. My most recent encounter is this podcast episode, Optimal Protocols for Studying & Learning by Andrew Huberman.
It's just how our brain works, and many do independently discover the technique, especially the kind who tends to excel in school. Anki is just one technique that makes use of the principle. So are all those cram schools for which East Asia is known.
But if you are at a life phase in which you aren't worried about exam-style assessment, then you don't have to optimize for the most effective method. I don't necessarily love doing Anki, but for stuff that I'd rather make part of me, I've started doing it because it's simple and effective.
Framing spaced repetition as a tool for exams is wrong.
SRS is not equal cramming and not analogue, people who say that srs==cramming are just using it wrong, furthermore it contradicts 20 rules of knowledge formulation.
Another perspective - optimizing this learning will free lots of time for other activities
@andang76 spaced term is one of the most important parts in this concept, it's a fundamental mechanism in our brains. FSRS is the most advanced algorithm now (aside supermemo), and it adapts to you.
"emergent repetition by resurfacing" is very unreliable mechanism if we speak about memory. Another strength of it is it targets your weak spots. Naturally items you know will have intervals over decades of years, some will be scheduled after hundred of years (which equals knowing it to the roots).
Those slots are calculated according specifically to your memory. Those intervals aim to give you strongest result per repetition, avoiding excess learning or wasting time on things you actually know.
Another point is microlearning - you can use little time slots, even 1 minute, for learning.
Memory is foundation of creativity. You can force your brain to avoid pruning information by force making connections with various techniques, which will decrease forgetting rate (which will equal to longer SRS intervals), but it's still about memory.