Personal Knowledge Management Is Bullshit | Hacker News Reader

Personal Knowledge Management Is Bullshit

otherlife.co

125 pointsleephillips98 comments

This is mostly a set of surface-level arguments about superficial aspects of personal knowledge management.

Yes, influencers pitching specific methodologies are often full of shit. Yes, network graphs are often boring and gratuitous. Yes, being overly focused on the tools used to complete tasks often gets in the way of completing tasks.

None of that is super interesting.

Particularly disappointing are the references to obesity and hoarding. Taking notes isn't hoarding. Storing notes isn't obesity. Physical and mental health issues should not be fodder for bad analogies. The act of writing notes and being able to search and read notes is useful for anyone, and is essential for those with memory issues.

The idea that ought to be central to this post, cutting a line through the knowledge graph, could have been really useful and compelling if not surrounded by trivial complaints. It would be fascinating to explore how knowledge management tools can help people do the job of cutting a line through a graph, though that's not really explored in this post.

Maybe with this post the author got a little lost in the knowledge graph in his head rather than cutting a line.

His business Indie Thinkers seems to share a market with many personal knowledge management tools. Of course it does.

I agree, this blog post is too long for what he was actually trying to say, he could of gotten his point through in a few paragraphs rather than an entire page, I'm surprised that I had the patience to read all of it today.

But he _forcefully_ and _linearly_ made his prolonged point.

Can you elaborate on the last point? From what I can tell it seems to be a private community, not a tool for knowledge management.

I use the Johnny Decimal system (with Zim Wiki). In 30 years of trying, it's the best personal knowledge management I've ever used.

All these note taking tools missed out on the straightforward categorization that the Johnny Decimal provides. And it's a big one.

Never heard of the Johnny Decimal system. Are there good any writeups or videos on it you recommend?

johnnydecimal.com but have a look at the HN search there was a heated discussion back in the day.

I wrote an article about it a few days ago: https://dsebastien.net/blog/2022-04-29-johnny-decimal

I take notes relentlessly with Obsidian and I'll honestly never go back. I would have been more stubborn about this earlier on in my career. Now I realize that I've probably forgotten more over time than I ever would have imagined so I realized having good notes is critical for me. I end up looking up answers on stack overlflow and then realize I wrote them.

To each their own.

Can you recommend a workflow for obsidian? Or some pointers for resources. I am struggling to "get" how to take notes with it and I am feeling a lot of friction

Second this. I even have a bookmarklet in the browser that adds text selection or a page automatically to Obsidian. And yet, when it comes to saving something I feel is important, the maximum I can do is just create another note and put the text/link and forget. I feel like I need to find proper place for this note in the folder hierarchy or think how to categorize it ("should it go to the Economics->Behavioral or Sociology?"), and quickly becomes a burden. In the end I feel like heavily underusing it, but perhaps I just don't get the right workflow..

The best workflow is to consciously underthink it. If you find yourself spending more than 10 seconds thinking of where to put something then you're doing it wrong. Force yourself to go with the first place you were thinking. You'll most likely find it later using search tools.

Sometimes you don't even need to take notes. Just the act of pasting in a link or a snippet from an article you read is enough to reference it in the future.

If you insist on making things clean, then you can use some sort of version control and review all the notes that you've taken once a week and organize them better. The most important thing is to take the note.

Right, agree with <10 sec rule. But, again, the main promise of second brain tools is to keep your "second memory" layer organized. Apple Notes are good enough for copypasting links and excerpts you want to make searchable later, right? So I assumed that Obsidian's markdown/hyperlink system should bring some benefits into organization.

For me, the value is somewhere in between. Priority number one is getting the data into the second brain. Afterwards try to do a little organizing but don't fuss around too much with it because I can always make improvements later (especially if/when I access it for a second time). That's kind of how the real brain works.

Browsing through Linking Your Thinking's YouTube channel[1] might be a good first-start. Despite appearances here, I am not a notes fetishist, so I find his videos rather approachable especially his "Start Here" video[2].

Jamie Rubin's blog series[4] is pretty good, too.

I personally just create atomic (single-topic) notes, with tags in the YAML front-matter, and store them in a folder structure along the lines of the Johnny Decimal System[3]. I store checklists, procedures, howtos, concept summaries, cheatsheets, and a daily log stored in a YYYY-MM.markdown file.

I don't have any comment on the blog post in the article, because this system works really well for me. Obsidian is extensible at hell, but I'm not spending all day looking for new tweaks, I'm just writing notes. Also, since notes are just files stored in directories, I have all sorts of shell/Python-scripted automations and shell aliases to quickly do stuff.

1. https://www.youtube.com/channel/UC85D7ERwhke7wVqskV_DZUA/vid...

2. https://www.youtube.com/watch?v=QgbLb6QCK88

3. https://johnnydecimal.com/

4. https://jamierubin.net/blog-series/practically-paperless-wit...

> I personally just create atomic (single-topic) notes, with tags in the YAML front-matter, and store them in a folder structure along the lines of the Johnny Decimal System[3]. I store checklists, procedures, howtos, concept summaries, cheatsheets, and a daily log stored in a YYYY-MM.markdown file.

You must be a very organised person, any sort of Zettelkasten or knowledge-based linking system for me is too complex, I'd rather just take a simple note, maybe with the date in the filename, to each their own I guess.

I rarely write TODO lists.

I’m not organized at all, but this system takes me very little effort.

As a quick example, I used to do daily notes, one markdown note per day but that was too much mental overhead, so I switched to one markdown note per month with an <HR> tag between each day.

But, what works for me won’t work for everyone else.

Zettelkasten seems like inhuman, nightmarish garbage to me.

If simple notes are working for you, then stick with that!

Interesting, so how is this Johnny decimal thing different to Zettelkasten?

With Obsidian you really don't have to have any system, other than linking shared concepts between notes. Optionally, limiting each note file to a single atomic idea is also helpful. Obsidian takes care of the rest, organizing, displaying, linking.

Its probably not for you then, people makes Zettelkasten look like a magical thing that will help you get more organised in life.

You would've "gotten" how to create a workflow that fits you by now if this is what you really wanted.

i would focus on just taking notes on related topics, without much structure for a bit. try to link them together if they are on related ideas a bit. once you have a network of notes that feels a little less manageable (~100), i create a "map of content" where i link to all notes related to that concept, and in it i essentially tell a higher level story about how those notes relate to each other.

This is if he wants to take notes in the first place.

I started up a job a few months ago and threw myself into Obsidian, and it has definitely been paying off. Here's how I work, ymmv of course:

- I create a "Work" folder. Inside it are folders for Weekly Notes, Tasks and Information.

- At the beginning of each week I make a new Weekly Notes page. I have a subheading for each day, and a subheading in each day for a todo list. After the todo list, I put down a heading for each of my meetings/activities that I want to take notes for and fill them in accordingly.

- When I begin work on a non-trivial ticket or project, I make it a note in my Tasks folder named with it's Jira tag. That makes it easy to link to from my weekly notes, and gives me a loose idea of when I was working on which tickets through backlinking. I then use these notes to plan out the problems, store personal research and organize code snippets.

- The Information folder serves as a kind of "Misc." folder. I have sections for travel, guidelines, best practices, and I even store some PDFs relating to my job in there.

A few other general tips for Obsidian:

- Create a group in your graph view for "Work" and it will highlight all of the notes in your work folder with a specified color. Pretty handy.

- Make heavy use of linking, both internal and external. Tags are also nice, but don't be afraid to break out your notes when things get too heavy; that's what it's made for.

- Liberal use of the keyboard shortcuts is a gamechanger. Ctrl+O to open notes, Ctrl+P to open the command palette (which I mostly use to split panes), and hold Alt+Arrows to quick-switch to adjacent panes.

- Plugins are nice, but it's easy to get carried away. Mind Map, Tasks, and Admonition are the only ones I keep enabled.

Hope you get as much out of it as I do. It sounds like a lot, but it feels like second-nature to me now. Feel free to adapt this to your own needs or skip it all together, this is just my personal workflow and what made me fall in love with Obsidian.

When sharing/saving links to Obsidian, can it automatically add the title of the linked page (and date)?

It just saves the URL for me.

I'm not sure, I normally link it with Ctrl+K and then fill in the fields manually.

There seem to be a plug-in https://github.com/zolrath/obsidian-auto-link-title

But I wrote a raycast script that automatically appends the current page with the title and time into my scratch page

I don’t overcomplicate it. Just create a new note and write. Often times I’ll create something for each day and just jot whatever down throughout the day. A lot of developers keep a random text file open all the time anyway for that type of thing, so just think of it that way.

I’ll add a ## header if needed to separate a topic here and there too.

If I remember an earlier note that was similar I’ll [[ to create an internal link.

Copying and pasting PDFs, images or other reference material is easy since by Obsidian can just copy those files into a local attachments folder.

If there’s a good reference video on YouTube or something, you can just embed it.

If my sidebar of notes gets longer than the screen I’ll take a few minutes and sort them into folders.

Honestly, there’s no fancy or organized workflow here. I just keep it open and write as it makes sense, then clean up every now and then when needed. Whatever is most comfortable for you is the right way. It’s your notes.

I understand the friction you're feeling. There are several ways to do the same thing in Obsidian, which can make it confusing to know whether you're doing it the right way. The way I managed to grok it was by reading this excellent post about this user's Logseq workflow [0]. Logseq is an open source copy of Obsidian. The advice applies to Obsidian as Logseq works the same way.

https://discuss.logseq.com/t/three-choices-new-users-need-to...

Shameless plug: Check out tht Obsidian starter kit: https://developassion.gumroad.com/l/obsidian-starter-kit

Obsidian is an amazing piece of software.

Perhaps PKM as a cottage industry is overdone, and the author's point that everybody's system is highly personal and should not be replicated is valid. That said, Obsidian has allowed me to organize all of my reading, learning, writing, and thinking into a format that allows me to recall knowledge and make connections that I wouldn't have otherwise found. Seeing in what other contexts you have read about a concept or a historical figure or an essay, etc. is so valuable for life-long learning that I can't imagine going back to disconnected or separated note-taking.

Obsidian is new to me, thanks for the recommendation. I see the files are local, does the app need to phone home for anything? I'd like something that will still work even if the company goes away at some point.

Obsidian being _just_ a UI on top of Markdown files is one of it's selling points.

That sounds exactly like Org-Mode inside Emacs.

have you heard about logseq? It seems to me that it is what you are looking for

Thanks! That looks great too. And it seems like I could actually use both at the same time on top of the same files in a complimentary fashion, though I'd have to test that out to make sure, and see if the use cases made it worthwhile.

I use todoist for basic task management, but it's ability to view history is virtually non existent so I don't bother attaching documents or anything related to a project there. But I use the API to compile task history, so I could integrate them as markdown files into this so I don't have to start from scratch in building things up.

Nope. Works on local files. Open source.

It’s one of the main reasons I went with it.

Wait. Is Obsidian open source? Any links to their code?

Definitely thought it was open source in my brain…maybe I’m wrong though.

Now I don’t actually see the code…so my mistake. At least with markdown files you don’t have to worry about accessing your notes if anything ever happened to them though.

> The sex appeal of the Knowledge Graph derives from the fantasy of not having to decide what’s most important

I write way more notes because that decision is taken away from me.

> Not only will the graph tell you what’s most important

While I’m sure some people do, I never thought to use the graph for finding what’s important. It’s always been about the connection between topics rather than the number of connections. The act of consciously connecting the topic to another topic is a great way to internalise it.

The notes also do help me write blog posts when I want to. But their structures do not get determined by the graph.

The thing I like the most about having a graph is just doing a “wiki surfing” in your own wiki. It’s quite fun and sometimes I make new connections with an old topic.

I do agree that showing off how many nodes/connection your graph has is bullshit.

Despite the disagreements, I still quite liked the post :)

This was a weird article for me - I agree with much of what he says but I also use Roam (and have tried many of the others) and it is very useful for me. I’ve also tried it with my business partner and it was a complete failure for a couple of reasons:

* We have different knowledge consumption habits. I consume a lot of snippets on the internet during the day and then consume books via an eReader, which allows me to dump the highlights straight to Roam. He prefers purely physical books and takes his own notes instead of capturing direct quotes.

* He is visio-spatial and I am verbal. He needs to draw, move, and position things (whiteboards everywhere!). Roam is useless for this.

* I tend to do more convergent thinking - I take in a lot of info from a lot of sources which, when thinking about the problem, tends to snap into view for me as something new. This works well in Roam because I can reference/quote things from across many areas rapidly.

> The concept of the Knowledge Graph deserves the classification of bullshit because its allure derives primarily from the false impression that it can mechanistically deliver—or substitute for—the brute, linear willfulness that defines all non-trivial writing.

This particular critique stood out as a bit strange to me, and maybe it is simply a difference and how my brain and the author’s brain works, but writing is linear (on a page) but the content gathering and referencing is not, which is where I find it most helpful. I brute force the narrative/story, but fill in with the details more slowly and with more references.

I completely agree with your points and they also relate with me a bit, Zettelkasten systems don't work for me at all, nor do TODO lists, it depends on the type of person, it doesn't work for everyone; I think what the author is trying to say is that Zettelkasten or other knowledge-based linking systems shouldn't be advertised to everyone as a "way to organise your thoughts", mainly because of the fact that it depends on the type of person, like an example is that as you mentioned, it works amazingly for you, but for me it doesn't really make sense with how I do things, I think that is what the author is trying to say.

Anyone sharing a picture of their graph, or who even pretends to navigate/look at the graph as a graph in any useful way, is definitely a bullshitter. In my experience, the graph is just a consequence that is sometimes pretty to look at, and that's it. But what is it a consequence of? I think there is a ton of value in taking notes (probably more value than in having notes), so what I get out of Obsidian, my PKM tool of choice, is probably no more than vim keybindings on a text editor, but that text editor can autocomplete references to other files without getting in the way of bulleted lists or prose.

Pretty much agreed for the most part, but if these chic new PKM tools get people who need to take notes but kind of suck at it (like myself) to take notes, then that's a good thing. Definitely not as miraculous as some claim. The cult around Roam in particular is well beyond laughable at this point.

Edit: The other thing I find very pragmatically useful, and is also a relatively simple thing (but a thing nonetheless!) is querying your notes. So I can take meeting notes, add some followup items as checkboxes, and then have a note that aggregates all of those to one place while linking back to their original context.

> Anyone sharing a picture of their graph, or who even pretends to navigate/look at the graph as a graph in any useful way, is definitely a bullshitter.

Yep. Graphs are not "scalable" unless the visual elements are meaningful in themselves; they are mostly for "feeling-good-about-oneself" --- "look at all the notes I took and linked!".

When I created my own Roam clone from scratch, I completely left out the "knowledge graph". I agree it's a failed idea.

My clone is optimized for terminal usage, since the clunky javascript UI is the other flaw of Roam.

Yeah, the three things I get out of Obsidian compared to my previous approach of a generic text editor + folder of markdown are:

1. As you mention, I can just shove a todo in a note and then I have a note to query all of these so I can make sure I'm following up on things I should be. For me todo systems came and went but it took integrating at the point I'm keeping records when most "todo" items come in for one to stick.

2. Embedding can be kind of handy. I think there's more places I can use it, but I do it so I can e.g. embed my discussion of a ticket in both my 1:1 notes when reporting to more senior leaders, and also my personal log of what I did that day. It's not like a lifechanging moment but it makes me appear more professional and prepared when I'm not going "Hold on a moment, let me pull that up waits for JIRA to load", or having to search for files to open them up as I'd closed them down to avoid tab overload.

3. Templates. I have some e.g. for standup, for 1:1s with specific people, for a daily/weekly note (which includes some recurring todos in the template, to link back to point 1)

---

For the graphing features, yeah, I take a lot of notes, and a lot of them are leaf nodes apart from a few highly connected ones, and those tend to be stuff like "Weekly XYZ meeting agenda", rather than anything enlightening. Having a leaf node each for each ticket I'm tracking in my system or each day's standup is going to make any diagram an absolute mess.

However, the other thing I do use Obisidian for is my tabletop game sessions. These tend to have much more "stable" entities (Player character X, location Y, Maguffin Z). While I don't personally use the graph there, it at least looks much more sensible there to the extent that I could maybe imagine it makes sense for people using the tool in a creative writing/research context. People find mind maps valuable too, even if that never did it for me, so I can imagine people inclined in that way would appreciate basically hyperlinked mind maps.

---

For the point on whether such a structured note taking system is useful at all, I think it really depends on "Does the number of distinct items you need to keep track of exceed your mental organisational capacity to keep track of them?". When I was just a mid level developer, I never took personal notes. For the "now" period, if I had 4 topics on my mind it was a lot, and that was well within my capacity to keep track of. For longer term keeping, either it was useful technical information for _other_ developers as well and I shoved it in a shared wiki, or it was information about how the feature we're working on is evolving as we work on it, in which case it made sense to keep it recorded in a ticket or document. At that point, there was very little left that was valuable to write down but didn't go into some other shared document bucket.

But once my role changed to one where I am now also responsible for the work of other people, that's when things changed and where I needed a note taking system. It might be a very simple topic and we might have talked about it yesterday, but I have n things I'm working on, plus x team member's and their individual n things, plus things that are upcoming that need to be ready for the team to work on.

This reminds me story: during my first (failed) PhD attempt, I had a supervisor working on a knowledge graph system for education. Great! I was doing something very similar, except I had an algorithm for presenting it in a linear way. So I shared my concern that students will not be able to understand and make use of a graph representation. Like almost everything I said in the following years, it was immediately dismissed and I quickly stopped trying to work on this with that guy.

Fast forward 2-3 years later, I ask about a student that was somewhat working on that project with this supervisor, he replied "oh, I stopped working on that, it doesn't work, student are confused by the graph!" In the meantime, I've also read a paper in another field that said users don't understand graph representation.

I still think that graphs can be useful for knowledge management, but it has to stay an implementation details not shown to the user.

This is very interesting! Is there any public content available for further reading on your (or similar) approaches? I would be keen to try it in my Obsidian KG if so

PKM is not bullshit; it is extremely helpful to a lot of people. This article seems a bit pointless. Yes, the graph is used to advertise a lot of these systems, but the point of the graph isn't the graph itself; it's just showing the power of backlinks. I use Obsidian and haven't looked at my graph since I started. But I use backlinks regularly. The graph is merely advertising the power of backlinks.

And it is extremely unhelpful to a lot of people aswell. It doesn't work for everyone.

That a bullshit does not make.

Roam user here. My experience is that it's not about the graph as a whole, it's about the neighborhoods. I never look at the graph view and I don't meaningfully use the global structure of the graph for anything, but I find that projects develop local neighborhood graphs that are definitely useful.

The thing Roam-style note taking software accomplish is letting you be productive by organizing bottom-up rather than being forced to decide what notebook/tab/folder any given new thing needs to go in. It's very pragmatic and low-bullshit in that sense. Being able to make references to other nodes (rather than only to other pages) is also very pragmatically useful in my experience (e.g. refer to a TODO I wrote in my weekly TODO list when writing my daily TODO list, so when I check off a daily TODO it's also checked off in the weekly list), and I never saw anything like this in pre-Roam tools like OneNote.

I get that other people operate differently and are obsessively "gardening" their graphs and searching around for connections to make. I don't do any of that. Doesn't seem valuable for me in particular, though maybe if I were a professional writer, I could imagine it being useful.

Indeed, I don't use the graph either. It's all about [[pages]] for me. Its such a low friction way to take notes. Just create the page tag and note away. Next time the tag you created will show an auto complete option when you start typing, making things even faster.

I agree with the general idea here but there are some opinions here I can't get behind.

> Individuals blessed with high degrees of industriousness and orderliness will build sophisticated media diets, note-taking systems, and automated archiving pipelines much more effectively than those less blessed with these traits.

I can tell you by nature I am not a smart organized person but implementing certain systems in my life has made this easier. I don't think what I'm saying necessarily contradicts this quote but this quote seems like it is reducing the argument to "the haves and have nots."

I think what the quote is saying is that naturally organized people are the ones who build all these management systems. For them, they work because they are naturally organized; pretty much any system will work. Because they lack this insight, they think they found the One True Way, but for people less naturally organized, every system will be a burden, including this One True Way.

I think everybody has a different sense of organization, and the prevalence of organization schemes and apps is in hopes that enough people can relate to a particular app-developer's preferences and patterns. At the end of the day though, I think productivity is such a chaotic and fragmented genre of tools because everybody is different and there are as many productivity methods as there are people. I've certainly never found one that I've been comfortable depending on, and it just so happens that I've had the idea (and the skills) to create The Best Todo for years, details of which are lurking on some note pages in this pile over here (and a few over there).

Naturally organized people built VisiCalc, and yet us mere mortals were still able to make use of electronic spreadsheets. The blog author is needlessly pigeonholing people because he thinks the knowledge graph is silly.

There are exceptions, and it doesn't invalidate the article's premise. Spreadsheets are a huge deal, agreed. At this point they are like a general purpose visual programming language, rather than an "organize my thoughts" tool.

This rant ignores the fact that our technology has failed to provide the average user with a common, ubiquitous, low-effort way to recall all the useful information that we can now dig up.

These billion dollar tech cos either think that I'll just be able to search google again when I really need it (what a joke!) or that I'll spend more time using tags to be able to find it again, but that requires a list of tags way to long to be easily fluent with and way too much librarian work on my part.

I have endless lists of bookmarks organized into ~50 categories that are still too broad to be useful (mostly because folder in folder bookmarks are a bad UX in Firefox).

I started using Notion because I like to see URL previews, but that list gets unwieldy fast and it is more of a pain to populate than saving a bookmark.

And now, I am down to a system where I keep 10 browser windows open with up to 40 tabs in each and if a tab gets too stale, I dump it in a bookmark folder and hope I'll find it when I need it. Each window is given a name based on its topic with the Window Titler add-on and uses non-treed ('cause trees are a stupid waste of space) side tabs ('cause top tabs are unreadable and STUPID) with the Sidebery Add-On. Each window with up to 16 pinned tabs that are major sources for that window's topical links. This system actually works for me and requires a low-effort flow to manage information for someone with a huge variety of work, interests and pursuits. I have to use the Total Suspender Add-On to keep Firefox from grinding my machine to a halt due to up to 400 tabs open at once.

What I really want is a browser that has all that out of the box so that it works seamlessly (e.g. Sidebery doesn't recognize Window Titler titles so moving tabs between windows is a huge pain), isn't a pain to setup on a new machine (I hate Mozilla for making it so hard to get rid of the top tabs!!!) and doesn't require me to trust several nearly anonymous Add-On makers.

Google knows that search solves pretty much all of these knowledge management issues. But there are deep conflicts of interest, and curious corporate blind spots to solving this individually.

So why don't all my web browser keep a complete copy of everything I have viewed (ok maybe not those sites) - so I an do a text search for "that thing about female doctor who was first ever".

Partly because if I had that on my local disk I would not need to go to the google home page. More fairly that particular search has been followed by X thousand previous people and Voogle knows how many ended up finding something useful.

But this does not solve for Outlook. The second biggest search opportunity surely is corporate emails. Gmail must have shown the outlook product managers how to do search ... 15 years ago. Yet it is still clunky unintuitive slow and frequently wrong.

> So why don't all my web browser keep a complete copy of everything I have viewed (ok maybe not those sites) - so I an do a text search for "that thing about female doctor who was first ever".

This existed for a bit, it was a Chrome extension called Memex. They pivoted to social bookmarks and it absolutely crushed me losing it. I had become very used to only needing to remember a few keywords from something I’d read in the past and then being able to rapidly finhd it.

Google has one of the most severe knowledge management issues I've experienced in my career and search is not very helpful in taming it. I'm not convinced that's a good example.

I've written pretty much your same comment a few times on HN in the past.

The web browser could be a provider of rich personal context. Firefox is probably in the best position to take advantage of this. But they, too, are dependent on Google (>80% of revenue is Google). Apple doesn't care enough about search or the web to pursue this. Which leaves... open source? Some fringe startup? I'm not hopeful anyone will tackle this problem soon.

We've known bookmarks are not the solution for over two decades now.

> The second biggest search opportunity surely is corporate emails.

This has largely been solved in a roundabout way by Slack. They didn't solve email or search. They just made email completely irrelevant at orgs that use Slack and their search interface, while not perfect, does an okay job. They only succeeded because email is so bad in contrast.

It's great to have a note system where you can get general information about a subject. However, such a system will not encourage you to read and recollect the information on a regular basis.

It seems to me that the only good personal knowledge management system is a self-testing tool, like Anki. When going through your daily tests, completed questions can have links to the source of the information. If you find yourself interested in studying the material again, you can follow the link, or retest using decks that you can build on the fly based on the tags of the question that caught your interest.

Personally, I just link to the original material that I found, instead of a personal note system. When I write notes, I write them by hand and regularly throw away binders of material. The handwriting was the important part, fully reinforced by constant self-testing.

The reason that this is so difficult, though, for technology, is the rate of change and how much technological bifurcation exists. Each company that I might go to will have their own stack of similar technologies with the same goals. How many languages do I actually want to know all the details of? Probably one. The ability to think as a programmer (or server administrator, network administrator, container system engineer, etc.) is the important part. Otherwise, I can learn the details of any utility on the fly as needed. It will stick with practice, and doesn't require the same level of personal knowledge management.

Learning a language, or things that don't change much (STEM subjects, for example), can benefit from the use of such a system.

    The Knowledge Graph says you don’t have to use your conviction and will to execute risky and time-consuming, linear projects (which will fail if your conviction is wrong). All you have to do is take cute little notes all the time, and the hard work is magically done for you!

I don't think the knowledge graph says this in so many words, but the author sure hears it.

It seems not to have occurred to him that for any topic, the loudest people on the internet are not necessarily worth listening to. Nor are they even representative of the larger group (see Twitter).

PKM has its share of grifters. I mean grift in a gentle sense, like someone who becomes successful by writing a book on how to be successful. I wish them inspiration and success.

They're selling people solutions for problems, but the problem has near infinite solution space because the problem is "how do I process information effectively?"

So what we get is individuals marketing their take on this ultimately rather personal question. It's marketing that makes them highly visible to people like the author. They're among the first personalities we encounter when one start searching on PKM (is this a good thing? I'm not sure).

I would tell the author the good news: the users without vlogs are doing just fine. You haven't heard from them because they aren't trying to grow an audience, but they're seeing added value in their lives as a result of PKM.

It's obvious that the author has no use for PKM in his understanding of it. I just don't think he's bothered to understand it as anything beyond a self-help topic.

Wow, does this ring true for me. I had been using Evernote for at least 7 years, then switched to Notion a few years ago, completely "reinventing my digital life" and creating my "Second Brain" with the P.A.R.A method by Tiago Forte.

I loved that for the first few months, then slowly but surely the sheer amount of interconnected nodes became... oberwhelming, to the point of me almost completely stopping to enter in new data.

I recently realized that I kind of hated my Second Brain, and did a painful migration away from it, to the stupidly simple Notes.app (the built-in Apple one).

Feels much better now: Less worrying about bloat, I deleted a bunch of stuff and do less procrastinating and fiddling about with the structure of it all.

Some stuff must be digital of course but I always find paper notebooks superior in the end. Something about physically writing it down makes me remember it better. And even though they can become disorganized I like occasionally flipping through an old one - it often triggers memories associated with the time I wrote the note or people I worked with that I would not get with digital files.

I think the other thing is it forces you to be more space conscious - you do not have unlimited real estate. Large notebook for some things, smaller handbook for portable, and notes that are compressed or adjusted for each.

A lot of technical people think to themselves "if I master this tool, something wonderful will happen," when in reality you usually have to figure out the wonderful thing you want to happen, not just trust a tool to manifest it for you.

Exactly, you have to know what you wanted to do before mastering the tool in the first place, this is true for a lot of things not just note-taking systems.

I would revise that to say "The Knowledge Graph View is mostly bullshit". That seems to be the core of the argument anyway.

It's not about "node flexing" and neither is it about letting the system do the work for you. You still have to do the work of turning information into understanding.

I want this article to be correct because it would validate my failed attempts at knowledge management. But I can't say that it isn't just some random person stridently arguing against the positions of a bunch of other random people. (And the author has the disadvantage that they totally misused the "bull in a china shop" metaphor...)

I certainly find it believable that people will fetishize knowledge graphing, because it's something I would do: spend lots of time to produce something sophisticated and fancy and utterly useless for real-world impact. But the abuses of people like me don't disprove the value. And the argument that it won't magically do some things, like derive the importance of specific bits of data, doesn't prove that it won't do other things.

Anything we produce or decide is based on a substrate of information, sort of a soil in which ideas and conclusions can grow. It doesn't seem ridiculous that it could be beneficial to take some intentional action to enrich that soil. Just don't expect that it's going to produce amazing results all by itself; I agree with the author that willful linearization and filtering is critical when you're trying to make something out of it.

I’ve tried a bunch of these sorts of tools. And nothing stuck over the years.

But two things did after all, recently:

1. Using Bear.app. I realized that I just hated slow, sloggy note apps. Bear is native and very fast! Faster than Apple’s own Notes.app!

2. Not worrying about organizing too much. Especially upfront. I just start dumping everything. Then _maybe_ organize later. But more often not. Just leave it as is. I can usually find it fast with a simple keyword search anyways.

> I’ve tried a bunch of these sorts of tools. And nothing stuck over the years.

Yeah same with me, I've tried many tools like Org-Roam inside Emacs which is yet another Zettelkasten type of system, doesn't really make sense for me.

So I just stuck to using plain text notes with whatever structure I want them to have, no standard structure, just whatever structure I want on a given day, and that is how I can actually get myself to write shit down, rather than wondering about structure and how this note links to that etc.

> The problem with the Knowledge Graph ideology is that everything in your mind is already situated in a graph structure. That is precisely the problem. It’s big and complicated, with way too many connections everywhere. There’s no good in replicating that web in digital form.

I completely agree with this, I'm not an organised person, I rarely write TODO lists, yet in Emacs with Org-Roam I thought I would be highly productive because I would start writing notes and linking them together or whatever magic Org-Mode has inside Emacs (if I can even be bothered to write a TODO list in the first place);

But I don't think Org-Roam is for me, I rarely wrote TODO lists even before using Org-Roam, and I'd have never heard of Zettelkasten or whatever until trying out Org-Roam, you could say that I'm not an organised person, and that is true, but I don't care.

Using a Zettelkasten system for people like me isn't going to work, it will never feel right to be honest, in the first place before trying it out I didn't want any of this stuff anyway, I don't need a graph linking notes together, I'm a simple man, just write a few notes sometimes.

The problem I see is, as you say, not the structure of personal knowledge, but recall.

Once I can recall something, that usually leads me to where I need to go. It's sometimes that initial failure to recall what I wanted to remember that can prevent me from acting until much later.

This is why I don't like any structure to my personal data. I throw every single thing I jot down into Apple Notes and either I'll come back to it when I see it above the fold or if I do a search query.

I like Apple Notes, but it really sucks in a lot of ways. The search function is crappy as crap gets. I don't know how indexing could be so poor. There's no way to tag or label notes other than by placing notes in folders. Notes does nothing to try and help you extract meaning from your unstructured text.

What I want more than anything that tries to impose the job of organization on to the user is something that lets me treat it like a dumping ground, does its best to extract any implicit structure or meaning and keywords from my notes, lets me add relationships optionally, and then gets out of my way. I don't need silos or folders or schemas or graphs or any such nonsense.

> Once I can recall something, that usually leads me to where I need to go. It's sometimes that initial failure to recall what I wanted to remember that can prevent me from acting until much later.

Exactly my frame of thinking, if I start worrying about structure, I forget the thing I wanted to write about in the first place.

> I like Apple Notes, but it really sucks in a lot of ways. The search function is crappy as crap gets. I don't know how indexing could be so poor. There's no way to tag or label notes other than by placing notes in folders. Notes does nothing to try and help you extract meaning from your unstructured text.

It lets you get the job done quickly, you just remembered you had to write something down, you can just write it down as quick as possible without worrying about other nonsense.

> What I want more than anything that tries to impose the job of organization on to the user is something that lets me treat it like a dumping ground, does its best to extract any implicit structure or meaning and keywords from my notes, lets me add relationships optionally, and then gets out of my way. I don't need silos or folders or schemas or graphs or any such nonsense.

Exactly, this is what I am thinking too, organization shouldn't be worried about when you just want to write something down NOW, as apposed to later after you've worried about organization and you've forgotten the thing you wanted to write about in the first place.

The actual thing that you are using the write notes on should do the organization itself, make tags automatically and link notes automatically, without you having to worry about anything, that for me, is the best way.

And you should be able to search through notes quickly and find keywords etc.. all just by looking at what is inside the notes that have any sort of structure.

> lets me add relationships optionally

Totally agree, relationships should be worried about after the note has been written down, not before.

Apple Notes got tags with iOS 15 / macOS Monterey. You can combine them with Smart Folders for organization.

I hope this isn't true, hah. I don't really invest in any existing PKMs but i nevertheless spend a fair bit of time writing the foundation of my own.

Regardless of the implementation, i want something to archive and/or retain (in the working memory sense) my conclusions. As without repetition i feel like a majority of what i read, and any conclusions therein, evaporate. Maybe the author would argue that if i want to retain it i should study it! .. but i don't, it's just some article i read.

Are my only choices to invest heavily or become increasing narrow minded? The things i actually invest in i know quite well.. but what does that matter? I'd like to be generally, reasonably informed. I'd like some minor effort spent reading broad topics to not be in vain. Besides, even if i invested heavily - there's only so many hours in the day; only so many things i can invest in to properly retain.

It seems reasonable to me to seek software to aid in this. But it feels like the author is attacking the idea of PKM just as much as the implementations. I can imagine many implementations are akin to snake oil.. but the goal seems reasonable. No?

> and any conclusions therein, evaporate.

I'm in a similar but different boat: my brain remembers the conclusion but discards all of the "useless" details used to arrive there. This works fine on a personal level, but when someone challenges your position, you find yourself unable to defend it because all of the supporting details have long since evaporated.

So PKM might be useful to archive away the details after you reach a conclusion.

I highly, highly recommend having your own private wiki. It is the 80/20 sweet spot of PKM.

> I'm in a similar but different boat: my brain remembers the conclusion but discards all of the "useless" details used to arrive there. This works fine on a personal level, but when someone challenges your position, you find yourself unable to defend it because all of the supporting details have long since evaporated.

> So PKM might be useful to archive away the details after you reach a conclusion.

100%. I hold the same mindset!

I've been playing with an app idea that helps people communicate these sorts of conclusions to each other to whatever level of detail they require. You can drill all the way down to the premise/axiom level if you want but you don't have to. The graphical animation is nice, but like you say, a private wiki can get you most of the way there.

I see knowledge as being a dag where the edges are directional and where the edges actually mean something, meaning multiple types of edges. Generic backlinks are nowhere near sufficient and lead to the sort of combinatorial explosion the author talks about. But when you have "statement" and "statement" necessitates "conclusion", (the edge means "necessitates") then reviewing the graph over time can actually lead toward you paring down the graph and distilling it to the knowledge that is most useful to you.

The more I deal with the topic of PKM the more I'm inclined to agree. My life tends to compartmentalize into projects with related themes. The themes are the only thing I have to keep in my head, the rest is a matter of orbiting a specific approach and taking notes towards that. Arbitrarily sucking up notes from all my interactions isn't project-focused and raises the noise floor. I could be spending that time drilling and internalizing the stuff I definitely will use, so that it's as near at hand as possible.

A related aspect of this is that there's a kind of implicit scope creep to technologizing your information management, where the end point is to have "Personalized Wikipedia" or something. Well, I'm not building Wikipedia. I am compressing knowledge that is, for the most part, already in Wikipedia. Why would I do the same work twice?

》After only two weeks, my knowledge graph was utterly unintelligible and distressing. It's called combinatorial explosion.

I've built a few node design tools and this is one of the main problems I've had to deal with, how to keep users from building this forest of wires that won't let them see the trees, I mean, nodes. The only way around is limiting the "depth" of nodes the user can work on, requiring them to organize those in "scenes" and hide everything else that is distracting. Then we get from "graphs are beautiful, behold their complexity" to "graphs are useful".

I'm currently building a concept web for a new search engine, and I want users to explore it without getting lost in the forest. The problem is there once again.

I never look at the knowledge graph on Logseq; it is pretty but mostly decorative than useful.

Linked notes on the other hand have been a game changer, as have little features like Excalidraw integration. I can believe that we've known about the concept of linking information for 25 years (surfing the WWW of course) and set up linked cells and functions in Excel numerous times, yet it's only now that software exists to allow that for our personal journals and notes.

Yes, I realize Org-mode probably has had this for years but most people aren't Emacs users and need something a bit more user friendly.

I read a similar piece a few years ago complaining about the proliferation of "everything bucket" software.

Turns out, one size doesn't fit all. Both articles read to me like someone finding a given mode of data management isn't useful for them, and then assuming theirs is the universal truth.

https://www.al3x.net/blog/2009/01/31/the-case-against-everyt...

I'm possibly one of those people who aren't genetically inclined to these kinds of applications...all I know is, if an app doesn't work seamlessly across MacOS, iOS/iPadOS and Windows, it's no good for me. Obsidian is great, but totally failed for me on the iPhone. I couldn't navigate at all. I don't know if that's my fault or Obsidian's, but that meant it was broken, and I moved on. (OneNote is great these days.)

PKM is not bullshit, but many of the people creating content about PKM do seem to be professional bullshit artists, for e.g Tiago Forte and his “build a second brain” method, a $700 workshop that can be summarized as a distinction between “projects” and “areas”.

I use Windows Explorer. Simple folder names like work, with subfolders for employers. Another folder for car, identity, one for medical, another for my son with subfolders for reports, pictures he drew, etc.

I can find anything very quickly. When search fails me because I forgot a vendor name or something, I just navigate the folders.

No software required other than what comes with Windows.

Exactly!

My GF's grandmother was very confused when we didn't know what database (sic!) she has on her computer, she said every computer has it... and then she has shown us her perfect categorization of documents into folders... We (both programmers) felt very dumb that day :-)

I liked the first part of the article, up to the knowledge graph discussion.

The author is lumping all KGs together. The term also applies to KGs that help run Google, Meta, etc. Good tech for using disparate data sources.

Personal KGs? Then I agree more with the author. Perhaps the problem is companies like Roam reusing existing terms?

The best system for PKM is called sub-consciousness. When one learns to facilitate it right and all there is is at a thoughts distance. Writing stuff down is a way to reinforce the subconscious stress of memory, also sleeping well, walking and taking time to be with oneself.

>...is genetic. Individuals blessed with high degrees of industriousness and orderliness...namely that most of the variance between individuals is genetic and relatively invulnerable to intervention.

This is a very strange thing to attribute to genetics, certainly not without some reference that supports it-- I could not find any. It seems much more likely to be a product of what you observed in your parents growing up, and perhaps more importantly how you were trained early on in your career.

That statement immediately put me on the defensive to be skeptical of the rest of what the author wrote, which was perhaps justified: the rest of the article was basically the author complaining about organization methods they didn't like and didn't think we're effective even though other people seem to get some value from them. And also that some people seem to brag about their system, which is entirely irrelevant to assessing actual utility. All of this while also coining their own buzz phrases, e.g., "brute linear willfulness".

TLDR: just another person claiming a tool is useless because it's not a tool they like.

By “bull in a china shop” does he really mean “the elephant in the corner”?

I am using Logseq and it’s an important tooling in my work/life now.

>Sure, you could say that I just didn’t learn to use it correctly (and link me to a paid course, no doubt).

Nailed it on the head. Productivity is ripe full of gurus and experts looking to make a buck.

I feel productivity is one of those areas that people get into and spend a lot of time thinking and writing about, to the detriment of their productivity.

Procrastinating so much that their relief became their passion.

My life dramatically improved in more ways than I can count as a direct result of using software to help manage my knowledge and tasks.

It’s a shame to discover that I’ve been wrong all this time. Back to my old, worse self I guess, because last thing I want is to look like a bullshitter to some random ranter on the internet.

Reply on news.ycombinator.com