from Hacker News

Designing Tools for Scientific Thought

by harperlee on 5/25/25, 2:51 PM with 10 comments

• by chermi on 5/28/25, 10:14 PM

  "...a note should capture just one thing, and if possible, all of that thing.". This just doesn't seem feasible to me for an scientific workflow. I realize his scope is math, where I can picture it being a little easier, but still nearly impossible. How often can you break down a note into something self-contained? I'm probably just taking it too literally.

  Maybe this says more about my relatively disorganized thinking patterns. For me, notes are typically thoughts about the current thing I'm processing and it's relation to something else, often in an imprecise way, to be explored further. Maybe the point is that wouldn't qualify yet as a note, and only when you refine the thought into something more concrete with direct reference to precisely the relevant notes on the related thing(s). But where do I start?! How do I make the first atomic note? It seems I wouldn't be able to reference a book. Maybe I could reference a section of a book, hoping it's self-contained. But even then, is have to resolve the references needed for that section into atomic notes themselves. When do you ever have time to actually think?

  I hope enjoyed this unedited stream of consciousness that I should probably just delete.

• by jljljl on 5/28/25, 6:22 PM

  Is there a collection of note taking practices (such as they exist) among the great thinkers/scientists in history? Is there an assessment or study of their notebooks somewhere that distills best practices? Are techniques like ZK or spaced repetition based on those practices or studies? Have they been studied for desirable traits?

  The science of notetaking so far seems to have a lot of art and opinion, but little science

• by piombisallow on 5/28/25, 5:22 PM

  We must design better note taking software in order to continue improving our note taking software

• by lou1306 on 5/28/25, 3:34 PM

  > The existing tools for scientific thought can be divided into two main categories: interactive proof assistants and textual authoring and publishing tools (including LaTeX, as well as the Gerby software that runs the Stacks Project).

  This seems... reductive? At the very least, isn't data collection, storage and retrieval also essential to scientific thought? I really do not think those tools can fit into a "textual authoring and publishing tool"-shaped hole.