Unbundling Tools for Thought

I have a blog post that I've been sketching out in my head about this whole area of Tools for Thought, future of computing, cybernetics, etc. Basically I find that the thoughtleaders of this space seem to always claim that the true manifestation of their ideas is just out of reach (Project Xanadu, memex, object oriented programming, etc.), but then never deliver this true manifestation.

Indeed it's rather remarkable how many of these figures like Vannevar Bush, Alan Kay, Douglas Engelbart, etc., never actually shipped much. Now, you could argue that they were ahead of their time and couldn't ship, and for some of these people you would be right. But c'mon now, Alan Kay is still kicking, and still talking about how programming should be reinvented. Douglas Engelbart lived into the 2010's. It's more that these fantastical futuristic ideas, if they were released, would probably not live up to their sales pitches. They're all chasing the high of Engelbart's famous demo. And their followers, who are perpetually waiting, who attempt interpretations of these thoughtleaders ideas and get dismissed as flawed manifestations, well they are essentially the parish. They're waiting for the second coming.

This isn't to say that none of these people have accomplished stuff. They have accomplished an extraordinary amount. But they are fundamentally salesmen, salesmen for the future. And salesmen only have a job as long as they have something to sell.

For sure. I get the appeal of Having Everything Organized. It's conceptually compelling. But I think the right way to look at it is in terms of minimizing total cost of retrieval and filing.

As an example, take physical paper. Receipts, bills, anything that ends up in one's mailbox and doesn't immediately get recycled. I used to oscillate between two approaches: over-elaborate filing systems and just ignoring the problem and letting the mail pile up in snowdrifts.

Eventually I realized that my love of elaborate systems was a giant fucking problem for my actual life. I thought about it like I was designing a production system. I very rarely needed to retrieve old documents; most of it was for "just in case" conditions. I needed to frequently file things, and if the cost of filing was too high, I wouldn't pay it. So I bought 8 filing boxes, each 3 or 4 inches high and big enough to comfortably hold legal-size paper. Each one is marked with a year, and almost everything for that year just gets tossed on top. A few exceptional kinds of paper then have their own separate file folders (e.g., tax documents, my current landlord, key retirement paperwork, key medical stuff). Once a year I throw out the contents of the oldest box and relabel it.

This works great. It turns out I almost never need anything from an old box. When I do, it's a quick rummage in one spot. With infrequent, hard-to-predict retrieval, storage-optimized organization is the best organization.

I spent quite a bit of time in this particular rabbit hole. Zettelkasten, Roam, Org-Roam, LogSeq, GTD, Anki.... yet I don't think I've stuck with any habit for more than a week or two. I have the same sinking feeling with these tools as I do with programming language development.

I'm sure many others here have followed note-taking/PL ideas for years, waiting for the one that will come along and give us superpowers. It feels like we're "so close"... all these tools/languages are almost the perfect blend of concise, expressive, interactive, etc.

It's occurred to me recently that we may already be among the last programming languages and note-taking systems to be built. We've tried for decades to design notes and code in ways that suit storage access patterns of our brains. I used to think that meant a breakthrough was inevitable, but now I'm starting to see the efforts having diminishing returns.

ChatGPT and its brethren are almost certainly the way forward for most knowledge storage (and the knowledge work that goes with it). The entire class of organization problems melt away when you can communicate in plain language what you'd like to store and retrieve. As chat assistants start to become more integrated in our work, the storage phase will become completely passive. The assistant will automatically accumulate the context it'll need when we come back to it for retrieval.

Why would we design new languages and note systems when, quite literally, a general-purpose second brain already exists?

I don't mean to discount the creativity and effort of all language and note system designers today. All exploration is valuable, and there are certainly better designs out there to be found. But the best designs don't always win, and we might just be at the point where the imperfect designs that have critical mass might be the ones that stick around forever.

I'm someone who feels very happy about my sprawling obsidian vault, combining journal, project notes, wiki, flashcards, zettelkasten, drafts, finished writing. I use pretty much no plugin, but I do have different methodologies that have emerged in multiple waves.

I started with just notes in bear, crosslinking when i saw fit, until that turned into 300 notes where I couldn't find anything. Tags and folders never stuck, and I abandoned those pretty quickly.

That's when I switched to obsidian, with a few very broad top level folders (projects, logs, zettelkasten, writing). Over time, this filled up too, and the zettelkasten got split into wiki and zettelkasten. The zettelkasten filled up, so I added index notes and structure notes and a numbering scheme.

The only plugins I use are some templates and the graph analysis plugin to help me find things I haven't linked yet. As said, most of the logic is in my workflow / way of tagging things. I don't care about being organized as much as I care about regularly using the vault.

Now that I'm relying more on research and external reading, I am integrating readwise and developing a workflow around that.

I'm sure though that if I were to document my setup, it would seem overengineered and impossible to use and over ambitious, but the fact is that it's very organic for me.

While there is some truth to the feeling that many people over-engineer and get overwhelmed with plugins and methodologies, I know many of us who just silently enjoy their vaults and plow away and are perfectly productive. If anything, I would indulge in the recurring urge to get to know better systems, but the foremost thing is having fun. I really enjoy my sunday mornings where I just write and throw things in there. I've had a 2 months pause, where I mostly kept to daily logs and drafts, but I'm about to go back to "proper" content management.

Furthermore, I can easily publish the vault and have a site that is not a chronological blog with edited posts. My writing is messy, iterative, raw, sprawling, multimedia, and pressing publish still gives me a clear signal "you finished something."

My vault is here: https://publish.obsidian.md/manuel

I've found the same, the catharsis of writing the "Daily Notes Pages" [0] is probably the main benefit of these systems.

In my case, the immediacy of handwriting has been a better fit than typing for this purpose. I also use the reMarkable 2 with a linked pdf planner [1] that I built, and with some custom collections I find it hits about 90% of what I'd use a "proper" TfT for. Obsidian and Logseq still look pretty seductive, but I know I'd spend most of my time in the weeds configuring plugins etc.

I'm hoping the rM2's OCR and export capabilities improve over time so that we could combine the benefits of quick, effortless capture on that device (via handwriting) with automated categorization, linking and search (on my laptop, maybe via Obsidian/Logseq/etc). There's a lot of potential if someone can effectively bridge that gap! It's something I hope to explore some more this year.

[0] https://maggieappleton.com/daily-notes [1] https://hyperpaper.me/ (as seen recently on HN :)