Thread by @fortelabs

1/ My latest thoughts on one of the most common questions I receive about building a Second Brain:

How and when do I *retrieve* all the knowledge I've saved in my second brain?

👇👇👇👇👇👇

2/ Retrieval is a trickier question than it seems on the surface, because it's not inherently desirable

If you never encounter a situation where a given note is relevant, than the ideal amount of retrieval is exactly zero

3/ Because retrieval takes work and energy, we want to minimize it unless we know it will be worth it

As with so many things in productivity, our motto should be "as much as necessary, as little as possible"

4/ Retrieval is really a matching problem:

We are trying to match a note captured sometime in the past, with a problem (or challenge, or opportunity, or constraint) that arises some time in the future

5/ We have a good answer for the first part: the best time to take a note is when you're consuming it

Especially with modern capture tools where it takes barely an extra moment to highlight, save, or bookmark a piece of content

6/ So we're back at the original problem: when to call up a note saved in the past

Because we don't know *when* that will happen, the critical factor is that retrieval methods have to take as absolutely little human attention as possible

7/ It might be worth tagging, linking, sorting, filing, graphing, etc. if I know a note will be relevant in the next 3 weeks

But if that time horizon is extended to 3 months or 3 years, and I have no way of knowing when, then even a tiny amount of effort becomes too expensive

8/ There's another important factor: there is no single perfect retrieval system

Different retrieval systems are needed in different situations

9/ For example, search is probably the single most powerful one in modern times

It takes no human effort besides creating the content in the first place, is almost instantaneous, perfectly thorough, completely random access, and searches can be iterated rapidly

10/ Yet studies have shown that search tops out at a certain percentage, no matter how fast and accurate it gets

This paper shows that some people some of the time prefer "orienteering," retrieving information in steps, like navigating a physical space

researchgate.net/publication/20…

11/ Other technology-based retrieval methods ppl use, all of which can be manual or automatic:

-Filing into folders
-Linking
-Tagging
-Visualizing (like in a graph)
-Sorting
-Special syntax for searches
-Annotating

All of these have their uses

12/ There's many other retrieval methods that aren't tech-centric, but probably the most powerful of all:

-Relying on one's social network
-Leaving physical artifacts in sight
-Making collages or arrangements
-Discussion and debate
-Setting up certain physical environments

13/ The problem is always how to enable or support as many of these retrieval methods as possible while spending as little energy as possible

I've settled on a few principles to do that in a way that is also platform-agnostic

14/ There are principles at each stage of my CODE framework – Capture, Organize, Distill, Express

We are taking each of these steps with ultimate retrievability in mind

15/ Capture:

Save things sparingly – only the best, most interesting, most relevant things that are least likely to be easily found through search, and that others are least likely to have access to

16/ Organize:

Take a single action to move (or tag) a note according to the most concrete, most short term way it's likely to be used (which is usually a project)

More than one action would take too much time/energy

17/ Distill:

Later, when preparing to put a note into action, identify the main points so that 1) You are familiar with their contents, 2) You can pick them up and move them easily, 3) You can minimize your switching costs from one idea to another

18/ Express:

Share your thoughts/ideas/knowledge as early, often, and widely as possible

The collective social brain is so much more powerful than even the smartest person, all you have to do is build your own Second Brain just enough to gain access to it

19/ The feedback, reactions, and contributions that come back to you from that shared knowledge gets captured back as fuel to start the process anew

Over time an increasing percentage of the knowledge you acquire comes through the collective hivemind, which is ideal

20/ So back to our original question:

The "how and when" of retrievability depends on the situation, so your goal should be to have as many retrieval methods at your disposal, not specialize in just one or a few

21/ Ppl naturally do this:

They start with a Google search, pivot to visual scanning of the search results, click one and do another search for a specific term, copy the link and send to a friend to discuss, share it on Twitter, save it to their notes & add annotations...

22/ I think ppl tend to think this approach is inefficient or wrong, and force themselves to follow some rigid retrieval routine

This might work in the short term, or for very specialized fields, but it's incredibly fragile and will break sooner or later

23/ This is why it's so important to have tech-agnostic methods for managing knowledge. Every time I've relied on a specific functionality I've been burned

Companies get acquired, features get discontinued, technology advances and replaces human effort with automation

24/ And most importantly of all, the specialized retrieval features work much better if you've done some pre-curation

They are complementary to each other

25/ Every year a task that I previously had to do manually gets automated

But since I'm always just reading and highlighting and creating what I would have anyway, I'm pleasantly surprised instead of devastated to see all my previous effort gone to waste

26/ So to answer the original question, imagine multiple retrieval methods like overlapping circles in a Venn diagram:

1. Search
2. Sorting (by title, date created, etc.)
3. Filing (usually by project)
4. Scanning (titles, images, headings)
5. Summarizing (within a note)

27/ Your job is to use as many of these as you need, pivot between them fluidly, and avoid or postpone as much heavy lifting as possible

You search your Second Brain much like you search Google search results, but now you know everything is curated by you

28/ To be concrete, you might start by looking through a project folder (or tag), you see an interesting term, pivot to searching for that term, move all the notes that contain it to a new page or folder, summarize all of them, extract the main points into a new note

29/ That process describes how I start creating anything, from a blog post to a new course to planning our wedding to buying a car to getting health insurance

It is an informal, iterative, improvisational, hacker way of assembling packets of knowledge just in time

30/ Ironically, it is also how I observe people working with information anyway. Finding shortcuts, going around problems, taking the path of least resistance

They think there's some other, more efficient, structured way of doing it, and there isn't

The natural way is the best