Second Brains and Why Most of Them Don't Work
The note is in there somewhere. I remember writing it — something about decision frameworks, maybe two years ago, in a folder I can't reconstruct from memory. I search. Three different keyword variations. Nine minutes. I give up and just rethink it from scratch.
That's the actual failure mode of personal knowledge management. Not forgetting to capture. Not poor organization. The failure is that you have the information and can't reach it in time, so you might as well not have had it at all.
A 2012 McKinsey study found that knowledge workers spent roughly 19% of their working week searching for and gathering information — not new information, but information they'd already seen and couldn't locate. A more recent figure is more striking: 83% of knowledge workers report having recreated a document that already existed somewhere in their systems. I'd wager the number is even higher for personal knowledge bases, because at least in an organization there's a help desk.
The strange thing is that "second brain" systems were supposed to fix exactly this. Tiago Forte's methodology — which produced the book, the online course, and approximately ten thousand productivity YouTube channels — correctly diagnoses the problem: information is scattered, you need one place, you need a system. The diagnosis is right. The prescription is where things go sideways.
The capture trap
Most PKM systems are designed around capture. Quick-capture modes, inbox workflows, browser extensions, Readwise integrations. The whole setup is optimized for getting things in. And this feels productive, because it is satisfying. Creating structure feels like doing work. Building a Notion database for your reading notes is a kind of creative labor — you go in, you organize, you tag, you feel a pleasant sense of order arriving.
And then you never open it again.
What second-brain systems mostly sell is a better filing cabinet. But a filing cabinet only works if you remember what's in it and know where to look. Most people don't. Notes compound, folders multiply, and the system becomes harder to navigate the longer you use it. After two years of faithful capture, you have more information that's harder to find than when you started.
The bottleneck was never capture. It was always retrieval.
What Luhmann was actually building
Niklas Luhmann was a German sociologist who published 70 books and more than 400 academic articles over a 40-year career using a physical slip-box — the Zettelkasten. The number gets cited constantly as proof that the system works. But the part that doesn't make it into most explainer posts is why the Zettelkasten worked.
Luhmann didn't design it as a filing system. He designed it as a thinking partner. Every note was explicitly cross-linked — not filed under a topic, but connected to whatever else it touched, across whatever conceptual distance. When he wrote something new, he wasn't asking "where does this go?" He was asking "what does this remind me of?" And when he retrieved notes, he regularly found things he hadn't been looking for. Unexpected connections. Things he'd written and half-forgotten.
He described the Zettelkasten as surprising him. That's a fundamentally different design goal than a folder hierarchy with good tags.
Modern note-taking tools mostly build the opposite. Obsidian has backlinks, but using them well requires sustained discipline that most people don't maintain. Notion has very little structural memory at all. Neither is really designed around the question: how does the right thing surface at the right moment, even when I'm not specifically looking for it?
The question you don't ask when setting up the system
When building a knowledge system, people ask: where does this go? Which folder? Which tag? That's the wrong question. The useful question is: when will I need this, and what will I be thinking about when I need it?
Sometimes the answer is clear. "I'll need this when I'm starting a project like this one" or "I'll need this when I'm writing about that topic." Sometimes the honest answer is "I have no idea" — which is also useful information. Notes you can't imagine a future use for probably belong in an archive, not a maintained knowledge base. Those are different things.
The other gap: retrieval isn't one problem. Looking up a specific fact you know exists is different from stumbling across a connection you'd made and forgotten. Full-text search handles the first. The second requires something more like a feed — something that surfaces older material alongside whatever you're working on now. Most tools don't attempt this.
A personal knowledge base isn't a library. Libraries are organized for retrieval when you know what you want. What makes a knowledge system actually useful is when it surfaces the right thing at the right moment, without requiring you to remember to go looking. Luhmann's slip-box did that through deliberate linking. A good colleague does it by paying attention to what you're working on.
The problem with most second brains is that they're very good at waiting to be opened.
Asgeir Albretsen is the founder of Harbor.