| > Do you take notes? I used to. 90 percent of them were about tools or other things to look up later. Now, I add those to the backlog, and the other 10 percent become Anki cards. These days the only things I write down are meeting related: action items, questions I want to find an answer to, and maybe 1 piece of key information I want to refer to later. > Do you compile/rewrite notes? Nope! I may adjust Anki cards if the phrasing turns out awkward, or if there's a term collision. And if a card is tagged 'leech' thats a sign to rework it but generally I'm not trying to memorize things word for word where this might help. In practice the hardest cards, the ones that get marked leech, are the bill of rights. Especially the middle ones that are basically a paragraph joined into a single sentence. Like, I know a warrant needs to be supported by evidence but its not always immediately front of mind that this is a 4th amendment right (instead of say 5th or 6th), or the exact phrasing. Fortunately this is a hobby topic rather than a professional need -- and hopefully if it were I'd be better at it =) > 3) Does spaced repetition work for you? I think so. For papers I generally extract a few key topics or points. Today's was three sentences about Cox proportional hazards, after hearing about his most famous contribution in an obit thread. 4) Do you have a methodology for extracting common themes/patterns across topics? No? My general approach is to spend a few minutes thinking about how I might apply it in my work, and the go looking to find out who did / what they found. So for proportional hazards, I ended up googling and found a few places where this was applied, like analyzing churn rate (https://towardsdatascience.com/strategies-for-customer-reten...) But note that I'm not trying to merge two topics of study in order to unlock some deeper truth, or better learn a general pattern. I'm trying to advance the state of the art in a field I consider myself pretty good at by learning how other fields work and stealing their best ideas. Kind of the inverse of the Hamming / Feynman jumping fields every five years. |