The traditionalist magistrates of productivity discussion spaces across the internet are holding some kind of line that some of us have already crossed, if not erased. Their wrists ache from carpal tunnel as they lower their bifocals down their noses and squint at their screens, all cluttered with community moderator tools that they may once have wielded toward more liberating, enlightened conversation.
“There shall be no discussion of vibe coded tools here, except perhaps in monthly self-promotion threads that command enough of our respect that we haven’t bothered to update their titles from six months ago.”
They are like the Thermidoreans of the French Revolution who turned their backs on enlightened ideals to take a more conservative posture. We can even accuse them (sort of like “J’accuse!”, if only because it’s French) of embracing the methods of the Ancien Régime that preceded and necessitated the revolution in the first place, though we must do so with some gratitude for a much, much lower body count.
I’ve always been a productivity software hobbyist. I didn’t expect vibe coding to make me a PKMS counter-counter-revolutionary, but I guess it does.
I read David Allen’s Getting Things Done when it came out in 2001, and actually got something out of the spiritual follow up, Making It All Work.
I was the first person I knew with a Palm Treo, and tried to get things done (and, when I had a Droid Whatever in 2009, tried to make it all work).
I loved Nozbe.
Todoist entered the conversation and, if I remember correctly, didn’t even do recurring tasks in its early days. (I might be wrong about that, but it was the kind of thing that happened.)
And, Evernote was a true sandbox. The muse that danced above its developers was one of helping you work rather than deciding how you should work.
Hell. I almost bought the Evernote messenger bag. I look back on it like my grandfather looked back on some Chevy Whatever for which he pined in the 1950s. It was cool and all Jetsons-y at the time, but let’s be honest. It’s objectively and timelessly ugly.
After the Evernote bubble burst, I spent the next 15 years being dissatisfied with powerful and capable apps that I loved, each for a hot minute before feeling like I was being boxed into a corner.
And don’t even get me started on Obsidian. After jumping in early and toying with it for years, I finally realized that I needed to step away from the on-again-off-again dysfunctional relationship I was having with Mr. Markdown. No matter how many plugins I installed or even wrote, I just couldn’t get Obsidian to change for me.
(Oh, how I lied to myself! How much time I wasted!)
I also have to admit that I had a fling with Emacs. I’m not proud of it, but I’ve learned that I can’t hide from my past.
After all of that, I even tried to use a (gasp!) Traveler’s Notebook with an actual pen full of actual ink. I moved to notebooks from Japan that had whisper-thin paper. I came out of this analog phase with a favorite pen. It’s a satisfyingly heavy and fidgety copper Karas Kustoms monster that now sits largely unused because I’m back to digital.
But when I met Claude Code, when I first laid eyes on him, I was immediately smitten.
Like, I’ve-never-felt-this-way-about-any-app-before smitten.
No longer a productivity gigolo, I’m now in deep partnership with AI to vibe code everything I need. It started with simple scripts and plugins, evolved into strange and clunky Electron apps, and turned toward much more robust and comprehensive platforms, like a fully block-based PKM web app that conjures images of Obsidian and Capacities having an illegitimate FOSS baby (with Ollama serving as the midwife) and a fiction-writing platform that categorizes your characters into archetypes, analyzes your writing style against a growing list of mentor texts, and produces a fully-formatted manuscript for submission.
Oh, and, yeah, you can use it to actually map and write that unfinished opus that you started for NaNoWriMo, and which has been nagging at you ever since that whole movement collapsed.
An aside: there’s no upsell in my sharing those links. The tools are open source and yours to fork/install/whatever. But, I learned the hard way that “gift” or “seeking feedback” means “self-promotion” to those PKMS Thermidoreans.
Disagreeing proved … pointless.
Whatever anyone thinks of the tools I’ve made, they’re mine. I’m not renting them. I’m not subscribing to them. I’m not buying into someone else’s vision of how 21st-century knowledge work should happen.
I’m at the helm.
So, vibe coding. “Vive la révolution!”
But, my project and your project and everyone else’s projects … it could all be slop, unless we can make something of it all together. And here’s where I have to have some sympathy for the PKMS Thermidoreans. All of those Reddit mods are drowning under everyone’s vibe coded solutions. And, without some interoperability, all of that sharing might still just mean signing on to someone else’s solution. There are just a whole hell of a lot more solutions.
And, that’s where the futurist in me sees something of the future, even if just as a hazy vision for half a second.
What if we actually had productivity tool standards?
What if there was a standard way, for example, of describing and handling a task? And, I don’t mean GFM or todo.txt. I don’t mean just an export target. I mean a fully annotated and governed JSON or whatever structure for “task,” “event,” “note,” etc.
Like we have for a GIF. Apps know how to open and work with a GIF because “GIF” is a standard.
What if every app knew how to open and manipulate a “task” in the same way?
Folks have tried. About ten years ago, the Linux desktop project KDE did a thing and then stopped doing it. There are some white papers out there, each a false start. And, we do have some good elements in standardization in iCalendar, but not flexibly enough to say “recurrence of a task must be flexible enough to happen along all different understandings of what ‘every week’ could mean, and when to mark the opening of the next instance window (like, from when you actually did it or from last occurrence).”
They all failed to achieve what I’m suggesting, and we really ought to dig into why. I suspect it may be because most of them traveled on the backs of some app or ecosystem that their authors wanted other people to use, rather than being an agnostic standard. And, for some of them, monetization or monopolization were the real motivations.
I think about HTML, and how the first web browser was more of a proof of concept than an attempt to corner the market. In fact, when folks have tried to corner the browser market, purists actually bucked those trends. The purists were those who fought against Internet Explorer, and who still turn up their noses at Chrome rather than the people saying “There shall be no discussion of alternatives because it’s slop.”
I should note, here, that my own projects are guilty of fragmenting rather than standardizing the effort. But, I’d rapidly rewrite them to embrace standards.
In, like, 15 minutes.
Because I can.
Because vibe coding.
And, does it have to be Claude? Am I suggesting that we just substitute one rental for another? One subscription for another?
Not if we, as a community, embrace and support open weight models. They’re getting more and more capable every day.
So, common building blocks. Productivity Legos. So much different from dominant apps. More capable even than MCP, which has made a lot of impact in my own work but really only standardizes accessing the tool rather than the actual stuff that the tool manipulates.
Widely-adopted standards for the actual things rather than just the transport of things would turn PKMS community conversations from “slop versus not slop” toward the realization of a true sandbox in which we’re able to get things to work together in really productive ways.
And, imagine this.
Imagine a PKM-centered operating system using these standards. Imagine a layer in the depths of your laptop that knows those standards, and makes everything that touches the tasks, events, notes, and everything else interoperable, or as interoperable as you want it all to be.
When Siri or Clippy or whatever does your bidding and adds a task to your ever-growing list of things to do, every app you deem helpful could know about and manage that task. Did you finish that spreadsheet? Tell your todo list right from Excel. Or, better yet, let Clippy or Fido or Talking Moose or whatever knock that off your list for you.
Any app you allow could know your agenda for the day. Did you take notes at that meeting? Maybe your calendar should know about that, and also be able to show you the tasks that rose from that meeting. And I mean any calendar app you use. Not just the one that locks you into a notes/tasks/events ecosystem.
And, as far as notes go, the sky’s the limit. Imagine if your browser knew more about what you were searching for from the context offered by that PKMS layer, or if that video conferencing platform could read the relationships between your notes about all of those other Zoom-heads in that meeting?
Granted, the line between “empowered and interconnected applications” and “privacy hellscape” is awfully thin under our current conception of how information works and travels through apps and services. I’ll be the first to admit that I’m skating over a big concern in that. But that’s where we are in the process of thinking about this. We’ve got to bookmark the real challenges and work them into the framework, not elevate them to the level of showstoppers.
We also have to recognize that privacy is a concern because the only answer to interoperability right now is to give an inordinate amount of access to our deepest thoughts and information, because every app and service needs to figure out how to connect it to other things, if it even can.
That job could, instead, just live on your machine or your personal server. Imagine if the only permissions you needed to manage were on your equipment rather than across the many services with which you engage.
The privacy concern could therefore be a good indicator of the revolution that’s really in front of us. It’s not about evolving into a new, more capable set of tools laying over a context we already know. It’s really about the opportunity to start over. Right now, our computers reflect paradigms of what work looked like when it was confined to paper and file cabinets.
Think about it.
Folders, for God’s sake. Folders!
We have the chance to build systems on the foundation of what modern knowledge work—even what the biology of thinking—actually looks like. And, that’s all about interoperability, connections, and the expanding rather than reducing promises of tags, categories, types, and more.
One machine’s slop is another’s grease.
When everything looks like slop, it might be time to change the machine.
