The brain alone is eager to make us feel good, eager to skip over inconsistencies.

By fixing arguments on paper, we can look at them with distance. If we occupy mental resources just keeping the argument in our head, we're happy we can follow its logic from beginning to end, but we're busying the very mental resources needed to properly analyze it, judge it, realize implications, or go on creative tangents that end up revealing a flaw.

Rely on internal reasoning? Poor idea. There are several reasons the unchecked brain sucks at knowledge work:

• makes things look as if they fit even if they don't
  • the effect of handwaving details with a "you know what I mean", even if on introspection we would find that we don't know what we mean (The slipbox confronts us with our lack of understanding)
• connects unrelated episodes into convincing narratives (faces in clouds)
• availability heuristic/Zeigarnik effect ([2023-02-07 Tue]: what? re-check Ahrens)
• edits memories every time we recall them
• …

Kahneman 2013: "The brain is a machine for jumping to conclusions." The WRONG machine for most knowledge work. Outsource what it's bad at. Many fields of science (psychology, neuroscience, …) have found the same conclusion: the brain needs external scaffolding of some kind in order to think.

Division of labor:

Slipbox

• Details
• References
• Unaltered long-term memory

Brain

• Get the gists, bigger pictures, deeper understanding
• Freedom to be creative

Just like it's possible to make bad flashcards, it's possible to make a #slipbox that doesn't converse with you, a "graveyard for notes".

• It must be trivial to follow links and jump to any note, with good hotkeys (Doom Emacs' <leader> n r f and Org-mode's C-c C-o may be dangerously clunky)
• You must have the backlinks view (in org-roam the org-roam-buffer) visible or easily togglable (and often in fact toggled)
  • And your color theme and fonts shouldn't undo its visual structure
• Don't attempt to recreate Wikipedia
  • Don't taxonomize topics like Wikipedia would. OK, sometimes you make a portal—a note that's an index over many others around a topic or purpose—but such indexes follow certain rules, see the book.
  • Though Sönke Ahrens doesn't say, try to title notes as active claims such as "Premature scaling can stunt system iteration" Matuschak: Evergreen notes should be concept-oriented. Jethro Kuan had a similar experience wrt "definition notes" vs "claim notes": Topic notes.
  • See where Matuschak uses proper noun titles: notes.andymatuschak.org/z6f6xgGG4NKjkA5NA1kDd46whJh2Gt5rAmfX "These note types are weakly evergreen. I may add to them over time, but because they aren’t concept-oriented, they’re not as useful to build on as an evergreen note… Non-trivial writing about proper nouns typically gets factored into separate evergreen notes which can be used in multiple places."
• Think of Luhmann's A6 cards. Sure, he had "card sequences", the equivalent of long notes, but try anyway to think of what might fit on an A6 card as it's possible you should be splitting your notes more (which ties back into the fact it shouldn't be a problem in the first place to juggle many small notes).
• Try to always rewrite instead of quote.
  • Usually you quote for an additional purpose, such as finding who originated a common aphorism and why. Expect to usually quote no more than one sentence and rarely a paragraph. If the quote is just too well-written to rewrite, perhaps come back after a day and then try to express the same idea in your words. The "mere ritual" of reproducing the insight is that good for making you learn.

When we have an "open task" or "open loop", it tends to stay atop our mind, causing our mind to involuntarily bring these up again and again. Zeigarnik effect.

Writing them down can convince our brain to let the matter rest. With all tasks written down (or loops closed) in a way that convinces us that they will be taken care of, we achieve David Allen's "mind like water". Get all the little stuff out of our short-term memory.

We can weaponize the Zeigarnik effect for knowledge work. Many mathematicians are familiar with the effect where if they think hard about a math problem and then do something else, the solution can come to them during the day.

As of [2022-08-09 Tue], the stuff atop my mind tends to be "there's something I have to do", which is not very useful to dwell on. What could I load it with that's an unsolved problem?

• Coding or other engineering challenge
• A current relationship question
• Something to do with work

Ahrens says we may divide notes into three kinds:

1. Fleeting notes
2. Book notes
3. Permanent notes

Fleeting notes can be taken on any medium, even a napkin, and should be processed within a day or two. When in doubt, everything's a fleeting note: even underlined sentences in books or copy-pasted book excerpts count as fleeting notes. Only if you write something in your own words can it be permanent.