"Ease of recording, ease of recall, low maintenance time. Pick two"

Okay, but why is this the case? And does it have anything to do with why scribes, clerks, and database administrators have been employed throughout history?

For the latter: yes, obviously. Let's break down the former in the context of modern personal knowledge management.

First off, PKM in the modern sense is pretty unprecedented; go back sixty years and describe what you're trying to do and you'll get a lot of raised eyebrows from pretty much anyone who isn't an academic or insane, and for good reason:

kit 🇱🇮 after dark's avatar

PKMSes don't solve the problem of realizing that you don't really need to remember all of that

PKM is downstream of a) information overload from modern life, and b) having too many hobbies. Once upon a time, a pocket diary and a rolodex could handle business for you, but now we have too much shit to track (and there isn't an excuse to not track it; the presence of a tool obligates one to use it); we add onto that by trying to know everything about everything, whether it be for its own sake, or improve our employment prospects, or because we feel like we should because everyone else is doing it, or something else.

As a result, we attempt to develop and maintain knowledgebases, something that even organizations for whom this is essentially their primary task have yet to perfect, solo. You're a pharaoh with no scribes, and thus turn to tools to act in their place, and those tools are subject to the limitations I described above. Let's take a look at each one:

  • Easy recording, easy recall: this is your zettelkasten, your personal wiki that you actually use, your longsite; you throw stuff in at the drop of a hat, and can find things in it instantly. You're the terror of Discord arguments and secretly hope no one finds out that your spend three hours a day dicking around with your ontology and search settings.

  • Easy recall, low maintenance time: this the German solution. All classification and metadata overhead is handled at insertion time, so every atom that enters with PKMS will automatically be indexed for fast retrieval. You haven't changed any settings since the day you checked your settings file into Git for the final time and documented it with your five-page submission form. Strangely, that was also the last time you actually filled that form out...

  • Easy recording, low maintenance time: this is the most common mode of behavior for anyone who needs to store data. You throw shit in, it piles up, and you spend half an hour digging through tags like recipes, recipes-old, grandma-recipes-2017 before realizing that it's a link in a screenshot from the phone at the bottom of your junk drawer.

There isn't a silver bullet here, with the possible exception of Nenexes and other AI-enabled knowledge storage and retrieval systems, aka "I hired some actual scribes, they just aren't flesh". If you're sitting on the edge of your chair waiting to jump in with "Anki!", then let me shove you back against said chair, because proper use of Anki trades point one off against two and three (unless you're in school).

Turns out it's hard to beat a librarian with a stack of paper or conflux of bytes, and those aren't as aesthetically pleasing, either.

Jibril, from No Game No Life