Each with some key ideas:
1. Spaced Repetition for Effective Learning by Gwern
Free-recall questions are better than multiple-choice.
He estimates that you spend 30-40 seconds on a card in a 3-year period, and less than 2 minutes over 20 years, or 5 minutes to account for optimistic assumptions.
This leads to the fundamental metaphor of a purchase, where you trade the ability to recall an atomic fact for 5 minutes of your life (spread out over many years).
3. Andy Matuschak’s note Spaced repetition memory system (especially “Responses to common objections”)
People object that rote memorization is unimportant compared to conceptual understanding or creative production, but automating basic facts decreases the load on your working memory, which might be the limiting factor in understanding complex ideas.
People read entire books and remember only a few points, which is bad because you can’t have deep understanding without details.
“Knowledge work often requires search problems,” which you can do faster the more large chunks—or mental representations—you have in memory.
3. Augmenting Long-Term Memory (especially the section “Using Anki to thoroughly read a research paper in an unfamiliar field”) by Michael Nielsen
Describing his workflow through reading the AlphaGo paper, he starts by skimming the paper and identifying basic terminology and facts about Go (like the board size) and statistics (like the number of human game positions used for training) so that he can memorize them with Anki
After a few quick passes like this of increasing depth, he was able to analyze the paper thoroughly. (Compare this to the chunking process Matuschak describes.)
4. Using Spaced Repetition Systems to See Through a Piece of Mathematics by Michael Nielsen
This helps address the common complaint that math or computer science is too conceptual for flashcards to be useful.
He passes over a proof multiple times, starting with simple questions and building networks of basic implications.
This method facilitates the chunking that allows him to think through nonverbal abstractions afterward.
5. Twenty Rules of Formulating Knowledge by Piotr Wozniak
Each card should treat a single atomic fact.
Don’t worry too much about duplicating information from different angles.
Pay attention to interference between similar cards.