I run into this sentiment often and I believe it's fundamentally mistaken as to the purpose of notes.
Recording is not knowing, at best it is memorization of largely useless trivia. Sherlock has a good discussion on this (Sherlock solar system). To know something you have to have processed and at the very least achieved some compression of it.
Notes help with processing and thinking things through, aiding conscious and active understanding.
Unless recordings feed into one's process, they're a waste of space. Sturgeon's Law: The vast majority of what we encounter is junk, much of it even self-contradictory. The more you let into your corpus, the harder the problem of retrieval relevance.
My clippings are actively curated, carefully sized to fit in working memory while also independently comprehensible and guaranteed to have met my minimal standards on factuality. Because I have read them carefully before, I expand my mental canvas but am not slowed down too much by processing demands. The throttling on note growth means an easier time for information retrieval and much fewer irrelevant results to ignore.
Recording isn't knowing but archiving is better than losing.
As an exercise, look back at bookmarks that are 5+ years old. How many of them still load? Look back at your history of liked videos on YouTube. How many of them have been deleted?
That's what I use DevonThink for. Local macOS app with encrypted local filesystem. It just indexes basically everything from email to webpages to actual directories of plaintext files on my filesystem. And it can do automatic OCR of images/PDFs etc.
I don't do actual note-taking there, I use NotePlan for that, which is just a directory of plain text files on my local filesystem. Automatically indexed by DevonThink.
I just dump everything there, including a regular weekly cronjob to archive my email accounts from different IMAP servers including Gmail.
I use DevonThink regularly and it never fails to deliver (except syncing, which sucks, my databases are too large). I have 25+ GB of data there in all sorts of file formats, from PDF receipts to over 1 million+ text documents including emails, and DevonThink can still find stuff instantly (under 500ms) for any full text search.
Wouldn't that essentially be a system that tracked the entirety of your digital life in a sort of timeline, with metadata attached, making it easily accessible and searchable? I have thought a lot about this and curiously, it's also what many science fiction depictions of computer systems seem to aim for - a single, integrated, computing experience, across all devices and tools.
This is something I think about often when it comes to tools. The idea that Google had all of my location history for the past 8 years means I can see everywhere I've ever gone. I can lookup any email I've ever sent/received. I can look through my browser history for any particular day. With Spotify + Last.fm I can see stats about my listening history, and with Tract.tv I can see my TV/movie watching habits. Strava tracks exercise, and YNAB tracks budgeting/spending. There's an incredible amount of health data from my iPhone + Apple Watch.
It's really cool to think that all of this data exists, but I'm not really sure that it's useful. Occasionally I've gone back through my old search history for that one old post that I wanted to re-read, or go through my Spotify listening history to find a song I really enjoyed but forgot the name of, but 99% of that data goes to waste.
Quantified Self [0] is an interesting project to me since it seems to make use of data from all of these disparate sources, but I'm still not sure what the outcome of tracking all of this data is. It seems cool but not useful.
As of a couple of years ago, gmail search was a crap-shoot. It would consistently fail on some things, occasionally fail on others, and work just fine on cases that looked (to me) to be the same.
It's exciting that the data exists, but Google's (and every similar organization's) business model is not compatible with full data transparency. They want you locked into their platform, using their tools.
I think it's the same with a lot of other things, such as backups or proper testing - you wouldn't say backups were a waste because you never used them and you wouldn't say tests were a waste because they always passed!
It's also like all the wires I have in boxes and draws at my house. As soon as I throw them out I need them!
I've dumped my SMS logs into my IMAP mailbox since 2010. I've used the "History Trends" browser extension and religiously archive my browser history. I store client IP addresses and timestamps from my IMAP server as another data-point on my physical location in the world (i.e. what network my phone is connected to when it checks-in for email). My credit card transactions dump to email as well. EXIF from photos is helpful but I haven't automated dumping those into my email.
Most of my personal "note taking" is either time tracking entries for Customer billing or emails sent to myself with keywords in the subject line and the notes in the body.
I wish I could get at my phone's (iPhone) browser history, call database, and location services database easily to amalgamate into my "digital life panopticon" too.
I love being able to search over even this limited corpus of data. I can "remember" where I was and what I was doing on a given day pretty easily. I discovered that I can often get "organic" recollection when I prime my brain with a few context details gleaned from the digital corpus. ("I was at xxx site that day, I had lunch at yyy restaurant, and when I got back I looked at these websites... Ah, and now I remember what Bob and I talked about that afternoon!")
I'd love to have audio and low frame-rate video recording on my person during all my waking hours. If I had that, though, I'd keep it completely secret (and use parallel reconstruction if I need to divulge "memories" gleaned from it to others).
"There is no worrying about whether a document belongs in the Letters folder or the New Project folder. It simply takes its rightful place in the time-ordered stream." https://www.wired.com/1997/02/lifestreams/
> fundamentally I believe "taking notes" is the wrong way to think about it.
Hard disagree for a few reasons.
Even with history, sometimes I forget how to look up a paper I saw six months ago. A lot of stuff sits at the tip of my tongue, and I'm unable to adequately describe it well enough to find it again. Or if I can find it, it takes a great deal of effort. I'm pretty sure this problem isn't unique to me.
But even more importantly, writing helps your brain build connections. When you have to analyze, synthesize, and summarize, your brain is actively grappling with, contextualizing, and relating new concepts to the world you're already familiar with. It becomes an active part of your worldview.
Coupled with spaced repetition, note-taking is one of the best tools for learning.
we'll have to agree to disagree - everything you're saying is also true even if you take notes, except your notes are just a lossy form of reality itself. reality that could just be captured directly and indexed. computers can form the connections with human assistance. no notes required.
notes have a purpose, but most people that I talk to use notes to basically just record reality to remember details
I am building something like this. It's my most ambitious project to date and I know it will require a large team of people to support it's various connections.
Think automatically downloading all youtube videos you liked, all tiktok videos you heart, all tweets you favorite, make a crisp HTML backup of every bookmark you have. It will grab all the videos you "save" and automatically ffmpeg to the resolution of your choosing (in my case x265 720p).
It's a large effort, but one that I think it 100% worth it. People are losing their connection to the past on the whims of a few corporations. It's time to decentralize that.
A swarm of my apps all online, all uploading/downloads the content as people search for it and consume it. No taking it down since there is nothing to take down.
>I wish there were a tool that could just record everything [...] if you "know" everything, then notes serve no purpose.
I guess the completeness of your proposed "auto-memorize-everything" solution depends on what style of notes a person writes.
A lot of my notes are annotations or personal commentary that don't exist in the sources of the url text or video.
Here's an example of that type of notes I often take... E.g. I found some links that explain how one can grok "Docker" by going through the learning exercise of re-inventing it from basic os features with homemade scripts. But I wrote my own personal notes (in "##")about the limitations of the explanations in the urls:
## these tutorials explain the verb (runtime) but not noun (artifacts of image file)
https://github.com/p8952/bocker/blob/master/bocker
https://www.youtube.com/watch?v=8fi7uSYlOdc
Because content cannot meta-analyze itself and enumerate all the ideas and concepts that are not embedded in its own text, it means a hypothetical indexing engine that converted the 2 links above to searchable data still would not retrieve the extra meta information I wrote. A lot of good notes are generated from your own brain.
EDIT add to categorize different types of "notes" to aid discussion:
(1) notes as memorization aid such as copying facts: a user writes a note that says "1 inch equals 25.4 millimeters" after seeing it on a webpage or something. So the "universal-record-and-index-everything" would help make these type of notes obsolete because one could just skip writing that note down and just recall it later by searching for "inch to mm conversion" in the digital archive
(2) notes as synthesizing/interpreting/connecting/commenting the content. These type of idiosyncratic notes generated by the user's brain often don't exist on the internet so they cannot be replaced by a universal recording tool.
Of course the best option is doing BOTH things: indexing the context automatically via a modern personal search engine that accepts all kind of media and does vector
/ full text and relation indexing, plus the information wroten by the user
There have been a number of research prototypes in that direction including Susan Dumais' Stuff I've Seen [1], which was Windows desktop-oriented but I think had the right idea.
> much knowledge work involves finding and re-using previously seen information. We describe the design and evaluation of a system, called Stuff I’ve Seen (SIS), that facilitates information re-use. This is accomplished in two ways. First, the system provides a unified index of information that a person has seen, whether it was seen as email, web page, document, appointment, etc. Second, because the information has been seen before, rich contextual cues can be used in the search interface.
Might be worth revisiting their papers to see what they learned.
Maybe APSE is what you're looking for [1]. A while back the founder sent me a link after one of my blog posts hit HN. It's a tool that continuously records your desktop and offers text search of everything through OCR.
I personally found the idea interesting, but I was too afraid to ever try it out. The mere idea of a video record existing of everything that's going on on my computer, even if it's never supposed to leave the computer, was too scary for me.
> I believe "taking notes" is the wrong way to think about it
Recording only the inputs doesn't help me much, my notes are mostly always about the part of my brain it ticked off.
A lot of it is "I wonder if this will help me with X" connection points left off for me to follow up later rather than to go back to refer to the conversation in exact terms.
Good notes for me are a way of letting off my internal narrative pick up from there later and carry on paying attention to the conversation I'm actually having right now.
I tried a few different options for note taking, including Foam[1], Roam[2], and now Dendron[3]. Dendron has been the best fit for me because it's built on VS Code (I already know my way around and can be more productive in it), and has a great team & community supporting it.
Markdown in a git repo is so much better than anything else I've tried.
LogSeq is another interesting tool in this space. Its primary differentiator seems to be letting you attach arbitrary key-value attributes to pages and blocks and then query things in your workspace to create ad-hoc tables. https://docs.logseq.com/#/page/advanced%20queries
What I really like about LogSeq is the Journal feature that you can specify a template for, and the ability to query TODOs across all of your notes. I take meeting notes and add in action items and they all get collected up and organized using my queries that filter and order by due date. The way they’re presented also lets you easily jump back to the context where the TODO was created.
It’s a prolog based query language so it took a bit of time to learn but now it helps me stay organized and on top of all the stuff I need to do.
While we're shilling for Obsidian, I'd like to plug "Obsidian-Admonition"[0], which has been a gamechanger in the way I take notes. It's a simple extension of Markdown protocols with an enormous amount of flexibility for adding block-styled content. If you've ever wanted to glitz up your notes with some documentation-styled content blocks, this is for you!
I looked into Dendron after this thread, but the graphing is just so much worse than foam that I will pass. Creating a graph is super slow for some reason, zooming in and out is done in big increment instead of being granular, it doesn't show the direction of the links and as soon as you zoom out, it hides the name of the nodes.
I also just like the fact foam uses pure simple markdown and just uses the first heading as the node title in the graph.
First off, not everything is about "This is the problem and here is the solution", sometimes it's just about wanting to make incremental improvements in something.
That said, here are the incremental improvements listed from the post in the submission:
- I’m writing more
- I understand my productivity better
- I’m thinking more
- I’m managing my health better
- I’m managing my subscriptions better
- My spelling is better
- I’m tracking decisions
If it helps you, you could frame each section as "The problem is that I'm writing too little, so my second brain helps me to write more", "The problem is that I'm thinking too little, so my second brain helps me to think more" etc etc.
They're invested enough in it to use it, is my guess.
People tend to downplay the "neat" factor in productivity tools--and they can be total attractive nuisances where you play with them instead of doing stuff. But the truth is you do sometimes need something to keep you engaged in the short term while the long-term benefits add up. Wrestling with Notepad works against that.
FWIW, when I try to use Notepad for this, I end up with 150 open Notepad notes that just end up pickling/restoring with the app. I've never thought hard about it, but I suspect that happens because you naturally keep the note open all day, and often the next to refer back, but there's no obvious time to close it.
But for whatever reason it just tends to turn into a huge mess, whereas note-taking apps tend to either optimize to the idea of "everything all at once" or they manage a daily journal scroll for you somehow.
Author here... I had originally written something like, "I wrote my own tool but you might see these gains using any note tool". For whatever reason, I edited that out. Probably a mistake, given your confusion.
Ah, I didn't even realize this was a post/ad about a specific tool that the author was launching, I thought it was a general post about any "second brain tool". Sorry for that.
I use Obsidian to take notes. I like it. I keep a daily journal there and connect it out to notes I open on different projects I work on or books I read.
When I started using it I thought about it like a "Second brain" in the sense of it would somehow make me effectively smarter. I imagined referring back to stuff I'd written, building connections, and the notes letting me discover or recall insights I may have missed. Instead, like the author of this piece, I find that most of the benefits seem to come from inspiration and overcoming inertia and I almost never refer back to things I've written in the past.
The inspiration is usually simple stuff, like if someone told you a problem you might immediately react with a solution or question. Same kind of thing happens when you write out problems. For example, recently my wife was annoyed at me because I haven't done a chore I said I would. As I'm just jotting this down into my daily journal I have the obvious "Why don't you take out the boxes now?" question. I answer that now I don't want to do it because she scolded me and I don't want to reinforce her scolding me. As soon as I see those words, I realize that's childish and I go take out the boxes.
It's a realization I might not have confronted had I just existed in a non-reflective state. By typing out what's going I gave myself the chance to make conscious questions and observations which I can then react to. Sometimes they are useful.
Journaling also helps overcome inertia. Left to my own devices I might spend all day on hackernews, Twitter, YouTube, and taking care of kids. When I journal though I document my frustrations with living that way and that helps push me to work on my side projects more.
Further, just the act of writing things out helps me get started. If I have this nebulous idea that I want to work on "an idea" it's hard to actually start doing it. Decomposing a big idea into discrete chunks is inherently productive - it's drafting a plan and thinking through the system end to end. When I've written out a plan I have something specific and actionable to work on and my mind is already in the "Work on X" state.
Note taking has also helped with ideation. Last night I wrote that I wondered whether Spotify would ban Joe Rogan and challenged myself to predict whether or not they would. While I was writing a brief argumentative essay to myself making my case (low confidence they do ban this year) I came to hold an opinion on Spotify the company (long straddle). I was then able to queue up a buy order - which I wouldn't have done had I not been note taking. (Although, whether or not this idea was good for me remains to be seen)
Maybe second brain is a misnomer. I rarely check old notes. Main value to me is lower friction to start writing and I spend more time reflecting. Keeping a daily log (ironically) improves my memory of what's happening lately.
For me it's also a place to dump all those things that you want to say but you don't know who to say it to. So you write it out and realize 90% isn't not worth saying to anyone. But occasionally a tidbit here and there helps in a conversation or meeting and it all seems worth it.
Similar experiences here.
I started my analog bullet-journal system 5 years ago, and though it's evolved a lot, continue to use it today. In parallel, I'd spent a few years writing what I call "devnotes" in markdown files on local fs. But they were sporadic and poorly organized. Then in 2020 I tried Roam Research, loved it (backlinks ftw), spent just over a year using it daily, but eventually abandoned it, faced w/ a "choice" between abysmal performance [the offline UX depended on Chrome local storage, and as my graph grew it kept getting worse] and putting my private thoughts in the cloud. Enter Obsidian. Finally! Offline, local fs, markdown, backlinks, plugins galore (eg Excalidraw integration)... it's exactly what I wanted. I've been delighted with it for the last 6 months, and recommend it to anyone interested in "tools
for thought", PKM, note-taking, GTD, etc.
For what it's worth, Roam team has just recently added end-to-end encryption option for graphs.
I'm a Roam user, although I've tried Obsidian, Bear, Zettlr and vimwiki. I keep coming back to Roam because I find thinking in blocks suits me better than thinking in documents.
I have quite a very same feeling. But as a way to review things, I also write flashcards when I think there's something useful I'd like to remember, or when I'm studying. Once you start remembering a flashcard, it will appear less and less often.
It's quite simple in Obsidian with the extension "Spaced Repetition":
I built an entire app around the idea that every note participate sin the spaced repetition queue. For me it has made a lot of difference, as I have managed to internalize (as in put into a practice) a lot of the stuff that I put into my "second brain", for example insights from books I have read, videos I watched or blog posts, etc:
What I would like, as a productivity tool, is a coarse to do list + contexts. Contexts here means, essentially, snapshots of virtual machines with all of the tools open that I was using, along with some structure for shared ddata, probably managed by git.
You might recognize this as more or less the definition of an operating system, so I suppose I could just say that I want an OS which treats me as the CPU.
I would love to know more about it, I know about as little as it's possible to know about emacs while having also used it. If you have anything more guided than me just doing a google search, please share!
See in particular the passage "send text to a tmux session|window|pane".
In Emacs, `M-x desktop-save` and `M-x desktop-change-dir` will let you save your session state and restore from a saved state. Also handy is the command `process-send-string`. I use it in the below (from my .emacs config[1]) to mark one buffer as the "receiving GHCI buffer" and then send text from another buffer to be evaluated in that one.
I’d like to see these kind of second brain tools mergable with a global anonymized database. Where my notes merge with the collective, and I can traverse ideas that spawn from my database into the collective whole.
I agree this is something we should have and it's what I will be working on for the next year or so.
One aspect I would like to get out of it is something like Wikipedia but with no notability criteria or censorship, since each author can choose which edits to include.
Each editor can choose how they want to moderate incoming edits. So I expect most editors would delegate most pages a service who aggregates other editors, and focus on just their core interests.
Am I in a minority that has been looking for a tool like this ("2nd brain"), but just never knew the concept is called that way? This is what my current, best second brain, looks like :\
Haha I kinda do what you do (text editor + files). It works fine for me, and I keep it synced with Syncthing. I just repurposed my old installation of Sublime Text 3 and keep stuff separated by [topic].md.
Ha, this is what mine looks like too, but all in one tab and split up via line breaks. Sometimes I come back to it and realize I've completely lost the plot and clear it and start over. It's like having too many tabs open in a web browser.
"Notes", "personal knowledge base" and "(personal) knowledge graph" are terms I'd call both more informative and more popular than "second brain", but they all have history.
I do this with two notion pages. One is Idea Inbox for any ideas worth exploring later. Then a second page called Journal which is a running list of entries for each day. I pretty much work out of this page and use Idea Inbox to remember what to write about that day.
I’ve surprisingly seen health improvements as well.
For sure. I’m not exactly sure why but I make more consistent progress towards working out. Maybe I’m more aware.
For instance, one day I’ll write my neck hurts today. When I see that again, I think about why. Then start thinking about posture. I start doing a simple yoga routine. Find one area is super tight. Write it down. I learn my hip flexors need work. Now I’m on that simple yoga routine with some martial arts kick drills and I feel phenomenal!
Mental clarity is greatly improved as well.
The biggest benefit to me is not redundantly thinking about things. When I do, it’s obvious because I see it in writing and can take action.
I tried a bunch of complicated frameworks for organizing. I ended up with as simple as possible.
> Naming software experiments is a barrier to entry for me. Typically, I need to come up with a name before I can even create a project. Using numbers removes that burden and lets me start with a readme or a bit of code. This is the first of those numbered experiments.
(From the project readme on GitHub)
I have the exact same issue, not only for projects, but also servers, structures or variables. If the name doesn't clearly communicate the purpose, I can't go ahead. What helps me is settling for an unrelated naming scheme like greek gods or NASA missions (servers), going with concepts, like batch-importer or note-app (projects), or choosing the first generic name that occurs to me, eg. EncodingManager (code structures). The only important thing is a simple path to renaming stuff later on as its purpose becomes clear :)
The only points that the author brings up that I can relate to are "I'm writing more" and "I'm thinking more".
The interesting part is that my PIM is incomplete - I can put things into it very easily, I just can't get things out.
That didn't stop my random note-taking from rising significantly (from 1 note once every few days to 1-3/day on slow days and 30-50 on fast days) after I implemented the note capture bit. Although, even though I can't (easily) get things out of it again, my brain seems to think that I'll eventually find the stuff I write down again - perhaps that's what unlocked the flood of ideas?
I'm running on a Macbook that is not at all set up for PHP, so no "users.ini" file. I did this ugly hack.
<?php
class BasicAuth {
public function __construct($passwordFile)
{
}
public function auth() {
return true;
}
public function hash($password) {
return "hello";
}
}
I can create notes and search them, only possible name is "Untitled", found no controls for formatting (h1/h2/h3, italic, bold, code block, lists, etc). Nice start, could benefit from a better UE and more features.
Yeah, sorry about all that, this thing is very early in it's life...
The users.ini file is a simple ini with my list of users. I put that file in place. I put it outside of the root so that you can't read it. I thought I had explained this somewhere, but clearly I haven't. I need to fix this.
My plan with the name is to pull the first line of the text file, remove pounds and things, and use that as the "dynamic" name. I haven't needed to implement that for myself just yet, so it's on the todo list.
I just use wikilinks and markdown, edit it in VSC; it is helpful if you schedule yourself like a hour of study time for some topic which forces you to revisit the stuff you wrote previously, regularly. I'm going over some math / stat / probability at evenings for 1-3 hours and the wikilinking is definitely helping to learn the stuff.
But on topics where I am not doing revisits regularly even if to just to light a bulb in my head ("Oh that topic, yeah I remember it"), I don't see much benefit accumulating.
Joplin user here. I’ve installed it on every device I own, and sync my notes with my NextCloud code instance. A few synchronizing snags aside, I’m extremely happy with it.
I suggest all to pick up "Mircosoft OneNote" once again. Pick the older version, not the new Windows Apps version, its BS like every other metro style app.
After trying a bunch of tools, i found OneNote to be the least resistant. You can write and put anything anywhere on a page including text, images, files and even draw something. Every page is freeform like an infinite canvas and its all searchable.
I realize it's not hip, but I use OneNote. I used to keep notes + a work diary in markdown in a git repo, but it was difficult to integrate screenshots, and converting a note to a task was always a manual process of copying text into a separate system. OneNote's native integration with Outlook tasks was the killer feature for me.
Hey, post author and Nolific creator here. Nolific is very, very new, so I don’t think there will be a lot of other users, yet. For me there are a couple advantages. The main one is that it’s in the browser, where I spend a lot of my day, which makes it friction free. I mention a couple more in the post. I’m also super excited to add “actions” to Nolific. When I do, it will pass any text you’re working on to any API end-point. I’ll use it for things like, “Post to my blog”.
This is a killer feature that I've been wanting to create myself. When my thinking is divergent, I benefit from a single dialogue that allows me to write without distraction, but my outputs belong to many different contexts. When my thinking is convergent, I have difficulty recovering my collection of useful material.
My experience with a 'second brain' has mirrored yours. Nice post! Good luck!
Thanks. For me, it's important that it runs in the browser and is still minimal (just text). I spend much of my time in the browser. I spend a lot of time in VS Code too, but I found it awkward for notes for some reason. It may be that I've got a few VS Code projects opened and so I was confusing code with notes, I'm not really sure.
There are lots of note tools out there and I'm sure our preferences will each be different. I hacked this one together to work for my particular needs, which I'm not sure I even completely understand. But, the minimalist web based nature of it is making the work-flow work really well for me.
I have added a simple search feature recently. Calling it a tool for "most recent" notes is an accurate indication of how I'm using it. In fact, it defaults to your most recent note and then you can flip back and forth through recent stuff. The search feature does come in handy when I'm looking for something specific that I know I wrote in the past.
Second brain is a pretty common term for curating your knowledge in a structured way. While a text editor is a main component of it, it tends to allow us to:
1) connect notes/ideas using tags or links
2) automatically generate graphs/views (some even do sql)
3) make it easy to look up past information.
Some people can do it using a single txt file in notepad.exe and ctrl+f, some people add tags/headers/titles to make it easier, and some people go full graph DB with roam research. You might not enjoy that (and based on the comments, we enjoy dumping stuff to it more than using it), but no need to dismiss this.
fundamentally I believe "taking notes" is the wrong way to think about it. if you "know" everything, then notes serve no purpose.