Ask HN: How do you manage “knowledge”

64 points by PixelMath 8 years ago | 32 comments

I am not sure "knowledge" is the right word to put it but this is something I've been struggling with, how to manage all the things that are not in your mind yet

Things to Do, Things to Learn, Things to Consume (Watch/Read/Listen), Save good things to find later (Articles, Infographics, Quotes etc), Quick Note, Unsorted Web Links, Thoughts on various topics (Impact of various tech on specific domains, better team building, various design practices etc ), etc

I sometimes spend a couple of days or so restructuring all these and then slowly I drift away to reach the same point where I started from (usually takes less than 3 months) and when I reach that stage my productivity comes to a halt, and the cycle continues Every-time I try a different Method (Bookmarks, Google Keep, Plain Text notes, OneNote) but nothing seems to be "just right"

sp527 8 years ago |

I've found the opposite approach to be more useful: elimination, selectivity, and focus.

For example:

Read only the best books, take notes on them, and absorb them fully. Book lists are good, but they tend to grow unbounded and promote anxiety. It's not difficult to get to a book list of 300-500. Once you're there, there's little reason to update it often. The focus should instead become chewing through the list you've created and periodically retreading some of the best ground you've already covered. Most people who are reading are not truly learning, and even fewer are applying. When you find a book that's valuable, expending the additional time and effort to fully incorporate its content and lessons pays tremendous dividends. The best books should be given a chance to fundamentally change how you think. That doesn't happen on a first-pass read.

Aggressively cultivating a frontier of possibilities (I would call this "metaknowledge") is easy, often addictive, generally not useful, and usually detrimental to your emotional health. Most people are not short of possibilities. They're short of the focus, willpower, and grit needed to convert those possibilities into hard-won experience.

Jeff_Brown 8 years ago | |

Ah, now I realize this post is about how to process information, not just representing and storing it.

I once took it upon myself to map as much of my knowledge as I could. (I had heart surgery and wasn't sure I'd live a long time. Turns out I have.) I would collect lists of ideas, and then stare at them until I knew which categories they belonged in, and then split them up into the appropriate category. Long sentences I would compress into shorter ones, often two or three words (and maybe some symbols in an evolving language of punctuation). It's much easier to internalize a concept when it's written in few symbols.

I just looked through my Semantic Synchrony knowledge graph for an example. I found "?is thing = ?has logic". That's shorthand for "to wonder if a pattern is an entity is to wonder if it the pattern has a logic". Kind of abstruse, admittedly, but my point is the compressed version was much easier to internalize.

To create a knowledge base is to create a big collection of mandalas. Meditating on them is rewarding -- new ideas spring from them. There is a spectrum of cognitive processing states, from empty mind, to feeling an idea form, to having the words for it, to deciding it is worth writing down, to reading things (that you or someone else wrote) earlier deemed important. They are all valid and valuable.

Discovering writing was a big deal for humanity, because writing lets you collect a lot of good ideas, and then distill them into a more powerful form. We now have the capacity (with graphs) to process information nonlinearly. This lets us move faster, theoretically processing much more information than linear presentations like books permit.

I've written a little about the advantages of nonlinearity here[1].

[1] https://github.com/synchrony/smsn-why/blob/master/nonlinear-...

sudouser 8 years ago | |

this 1000x. accept you’ll never know everything so focus and trim and trim again on your interests. for technical stuff like setting up ubuntu machines a gist github work.

i don’t use bookmarks anymore it’s just too stressful trying to keep them updated, almost a full time job he! if I don’t read it in that moment it’s gone for good.

I use Evernote for temporary stuff but then it’s gone too after I’m finished with it or uploaded to a gist.

books in kindle are cleaned regularly too

nikhilrajendra 8 years ago | |

Really liked your answer, thank you!

dirktheman 8 years ago |

I use Evernote for exactly this. I organized it with a couple of notebooks: Filing cabinet, Family, Work, Inbox and Templates. I have about 20 tags that I can assign to each note and I rename all notes with the date in the name so a combination of tags, title and notebook alsways yields the result I want.

A couple of takeaways: - Evernote Premium is well wort hit, it converts all PDFs and images with OCR so that the text within it is searcheable

- I scan each incoming mail item that’s even remotely significant. All bills, letters etc. go into the filing cabinet, should I ever need it.

- The scannable app (by Evernote) outperforms almost any scanner below 500 dollars. And it’s free.

- The web clipper is awesome for storing websites for future reading

- I use Evernote not only as a filing cabinet, but also as a notebook for random notes/ideas, repository for my children’s drawings, scrapbook, wish list, to-do pad, grocery list, you name it.

We’re completely paperless (only some really important papers and drawings are kept) and it’s a huge improvement in our lives. As opposed to most people, I can tell you exactly how much our electric bill is, when the last PTA meeting was, what the VIN number of the car is, the exact dimensions of my living room and where/when we bought that nice IKEA vase. All within seconds. I find it liberating how much space in my head is cleared because I don’t have to remember all that stuff anymore.

If you're interested in going paperless: https://www.jamierubin.net/going-paperless/

contingencies 8 years ago |

Finer grained changes and git. When reading online or books, notes go in git. If you don't have time to type and just want to squiggle, try a wacom tablet and inkscape. If you have time, markdown all the things. Learn tools such as graphviz[0], mscgen[1], mermaidjs[2], syntrax[3] and wavedrom[4] to assist with clear communication.

[0] https://www.graphviz.org/

[1] http://www.mcternan.co.uk/mscgen/

[2] https://mermaidjs.github.io/

[3] https://kevinpt.github.io/syntrax/

[4] http://wavedrom.com/

z3t4 8 years ago |

If it's not interesting enough for me to want to learn it right away then it's not interesting enough. Also lets say you have a normal distribution of 1-10 scale, where 10 is the most interesting thing you ever learned about. And you find an 8 or better, just stop looking and learn that. Because it's very unlikely you will find any better in the near future. Also don't do todo-lists, make schedules instead! In order to add something to your time, you have to remove something else! So is this thing more important then what you currently spend your days doing !? Then set aside time for it.

jaccarmac 8 years ago |

Interesting that you mention those categories. I have an Org file organized into those headings. TODO, TOLEARN, TOCONSUME. Read/watch/listen subdivisions of each. Items are kept roughly sorted when I feel like it or when I update the progress of something (when I watch an episode I will note the next episode and move the show to the top of the category for example.)

On a day to day basis, I filter things through various channels (HN bookmarks, Reddit saves, emails, YT notifications, links shared to Keep from my phone) into that file. Time sensitive things get an email, etc. The system is certainly not perfect but it's good enough for me. It's made me more aware of the information I consume since there's no way I can possibly get to everything I want to.

mjmein 8 years ago |

I've been spending some time over the past few months developing a app for myself and family to help with this.

Basically, what I do is:

- make a list of books that I want to read

- over time, as I acquire the book, I have a tool to split up the book into separate chapters. These chapters are then loaded into the reading app.

- every day, I open the app and it chooses a chapter from a random book for me to read

- as I read, I highlight interesting, useful or difficult sections that I want to remember/study further

- instead of showing me a chapter, the app can also show a list of highlights from a previously read chapter. During this step it asks me to create little quizzes from the notes - either in the form of cloze replacements or simple q&a. Alternatively I can also create reflection questions or daily actions from the notes.

- in addition to the previous steps, the tool then also uses spaced repetition algorithm to review the little quizzes that I created from the books I've read

- every morning it chooses 10 of the actions that I identified and asks me how I am going to implement/practice them today

- every night it asks me how I practiced those actions

It does a bit more than this, but that's the main part. I've already seen great results from doing this, as I've always found the biggest problem with reading is that it tends to be too passive, and having the review and action steps helps alleviate that problem.

myfriendslinky 8 years ago | |

This sounds awesome. I’ve been using a spaced repetition app within my reading process for a while now. But your process looks intriguing! For for the curious here’s my current process: http://www.opensourced.me/2018/02/11/how-i-remember-what-i-r...

Can your app be made publicly? I’d love to experiment with your setup.

tmaly 8 years ago |

I have struggled with this same issue as I would keep things in a paper sketch pad or in something like a bookmark or note manager.

It just seemed too difficult to bring together all my ideas.

I recently picked up a copy of the 2001 GTD book and I am working through getting into the process of using an app called Things 3.

Getting it all out of my head and being able to partition it into projects even if they are someday type things has been very powerful.

Jeff_Brown 8 years ago |

Be like Google! Build a graph!

I use a free, open-source knowledge graph database[1] with an Emacs frontend. It resembles tree-based knowledge schemes like FreePlane or Org Mode, but things can have multiple parents, not just multiple children. It offers full-text search via Apache Lucene, which resembles regular expressions but also offers boolean operators. It also lets you selectively merge your knowledge base with other peoples' -- share what you want with who you want, without having to segment your experience of the data into separate files.

(I am one of the developers.)

[1] https://github.com/synchrony/smsn/wiki

guilhas 8 years ago | |

The intro video looks good. I will have to try this.

Currently I use Zim wiki.

godelmachine 8 years ago |

There's this habit I've been trying to inculcate since the past few months → Writing 10 ideas every day.

Whatever things I read about technology (most of them are research papers from Hardware Architecture / Emerging Tech section of ArXiv which I visit daily, The Morning Paper by Adrian Colyer, The Ryg Blog by Fabian Giesen et al) - I try to collate and correlate all these ideas and try to imagine something of my own.

Fun PS → James Altucher has promised we will get a seat in the Justice League Watch Tower for sure if we imbibe this habit of writing 10 ideas daily. Such is the power of this idea.

ssivark 8 years ago |

What started off as a few small lists have expanded (over ~5+ years) into a full-fledged "notebook" with pages on various topics.A lot of pages in my Zim notebook are lists of resources and tips, organized by theme (eg: speaking/presenting, things to buy, digital currency, sci-fi to read) i.e. hard to come by nuggets whose collective perspective seems valuable. I can look up things by theme whenever I seek, and also add to those pages whenever I find a new nugget. This structure also implicitly encourages spaced repetition when used for ideas/concepts [1].

One reason I like this approach is that it allows me to de-clutter my working memory, eg: number of open browser tabs [2]. It helps me avoid FOMO. As Scott Hanselman said: "It's not what you read, it's what you ignore". I keep adding to my lists over time, and this also helps me get a sense for which themes/resources/recommendations keep popping up repeatedly, which helps me prioritize what to spend time/money on.

I use the Zim desktop wiki, on linux [3]. Also, I download a PDF copy of any thought-provoking online content that I really like, since I don't want to risk the links not being around years later.

[1] https://en.wikipedia.org/wiki/Spaced_repetition

[2] I loved Firefox tab groups, and am on the lookout for alternatives compatible with the new Quantum FF.

[3] http://zim-wiki.org/

bewe42 8 years ago |

Teach others about what you want to acquire. That's why I believe that blogging is a very efficient exercise if you want to truly learn something.

Teaching requires you to absorb, eliminate and bring order into your thoughts. This is when learning happens.

Regarding what tool I'd recommend, I use Scrivener. As I mostly write in order to learn, and Scrivener is a writing tool, I've come to use Scrivener for almost everything.

schmidty 8 years ago |

OmniFocus for all actionable items Pinboard for tagging links Google doc listing quotes I like Instapaper for reading later YouTube for watch later Overcast + huffduffer for listen later

All of these work/sync cross my desktop laptop and iPhone.

In progress is a wiki type system for potentially typing it all together. Dropbox Paper has been great for the wiki part.

forapurpose 8 years ago |

I'm guessing you've considered it, but a wiki is a straightforward, highly flexible solution. Some wiki software supports structured data (alongside unstructured, of course), if you want that.

In my own experiments, I've found that the speed of the interface for input is very important. There's not enough time to input everything I would like; the longer input takes, the less knowledge I capture. Many wikis are weak in this regard; input requires a lot of reloads and scrolling, mainly because of the different modes for reading and editing:

1. Load the page on which you are inputting the data. Read it to decide where the new info belongs.

2. Click edit, wait for page reload.

3. In edit mode, scroll and read the page again to find the place you identified in step 1.

4. Add the new knowledge.

5. Click save, wait for reload.

6. Scroll and read to find the new knowledge on the updated page; verify that it's what you want.

7. If it's not what you want, return to step 2.

branweb 8 years ago |

Definitely agree with the synthesize, select, focus approach, though I find software can make this process cleaner / more enjoyable. You just have to guard against it (the software) becoming an end in itself. For suggestions, check out this HN discussion from a few years back: https://news.ycombinator.com/item?id=8270759 . Pay particular attention to Ixiaus' emacs org mode comment. That's pretty close to my ideal solution.

dyeje 8 years ago |

If I must do something it goes on the to do list. For everything else, if I don't remember then ah well I guess it wasn't that important.

lcall 8 years ago |

I use the tool I finally wrote for this (work in progress but very stable and I use daily; AGPL):

http://onemodel.org

josephernest 8 years ago |

I created http://bigpictu.re exactly for this. Maybe you ll find it useful.

tucaz 8 years ago | |

Mobile devices not supported (yet).

I really wanted to see what it’s about but I guess I won’t find out.

You could at least add some pictures and a simple description to entice people to go to a desktop to see it.

PixelMath 8 years ago | |

I once had stumbled upon this long back, been looking for it since then (for other purposes) cheers!

txsh 8 years ago |

Bookmarks aren’t enough. You lose old material when the URL changes or the site gets taken down. And you can’t search the content of bookmarked pages.

Text files don’t allow you to embed images, video. Structuring of data is too one-dimensional.

Other document files take too much time and energy to search and maintain.

Online services like Evernote cost money and have too many annoying quirks.

Mediawiki is too slow and bloated. It’s editor sucks and it takes too much work to upload and embed images.

BookStack is the way to go:

https://www.bookstackapp.com

It’s a wiki. It’s fast. It’s searchable. It’s got a great WYSIWYG/Markdown editor with code highlighting.

You can manually upload images or upload them by pasting then directly into the page.

It’s Book-Chapter-Page organization is easy and intuitive.

It’s also free, open source, and easy to self-host.

xstartup 8 years ago |

We have a bot which includes relevant discourse forum links in our GitHub issues as links.

We make it a point to summarize discussions from Slack in discourse after which the discourse topic is closed.

We also use Trello but not for extended discussions, just for prioritizing.