waterfall.org

A Small Trick for Staying Organized With Folders

This essay is part of my Scale-Smart Moves series on tactics & operations that compound across your system. See the full index here.

Over the last few decades, folders quietly lost their charm. First, file names stopped being constrained. Then search improved dramatically. Now AI is going commonplace. The promise — don’t organize, just search — grows stronger by the year. Heck—kids growing up on tablets often don’t learn folders at all. Nowadays, everything is swiped, scrolled, typed, and retrieved. The mental act of “putting something somewhere” dissolves into “I’ll find it later,” and slowly, but surely, the file system turns invisible.

Maybe that’s efficient. Maybe even rational in a world of infinite indexing. But something important gets lost when you never decide where something belongs.

Searching is a plain act of retrieval from the 1st age of digital archiving. It’s primarily about access. Organizing, in contrast, is an act of interpretation. It’s about meaning. When you place a file into a folder, you’re making a small claim about reality. You’re saying: this belongs with that. This is that kind of thing. This connects to that domain of my life. Search can find your past. But only organizing with folders shapes how you understand it. And perhaps that’s why they feel outdated. They require commitment. And that has lost its charm as well.

But when placement becomes optional, orientation goes with it. Stop placing things, and you stop building a mental model of where things belong. Your folder structure, taken as a whole, is a map of your intellectual life. It reflects how your mind works. What you value. How your life has unfolded. Placement is not a neutral decision. It’s autobiographical. And in outsourcing it, you don’t just outsource labor — you outsource the quiet skill of knowing where you are.

Folders show a type of hardness. They act like digital stigmergy, leaving traces of coordination among your different versions. Your past self leaves a structure; your future self follows it. No traces, no paths. Just a growing pile and a search box.

But folders are not just about placement. They also involve maintenance. And maintenance has a cost. Moving a file down a deep hierarchy can take a few seconds. Do that fifty times in a row on an overloaded inbox, and you’ll feel the resistance in your body. That friction is real.

What usually breaks the system isn’t a big idea against folders. It’s the small resistances that build up over time. Eventually, the structure quietly collapses, and search fills the gap.

I won’t solve your entire organizing problem in this article — that’s what ±PEAKER is for. But I can zoom in on one small move I use constantly, one that makes relocating files noticeably easier. No, I am not talking about AI. Nor about automatic file-mover plugins. Those have their place,¹ but if you outsource the physical act of placement, you lose some of the mental value we just talked about. I’m actually talking about moving files manually — and making that just a little bit faster.

Let’s assume you already have a decent folder structure. Not perfect. But coherent. Few collisions in meaning. Most of the time, you can tell where something belongs. Now imagine yourself triaging an inbox on a busy day. Or walking through your root folder and spotting a stray file sitting there like a lost sock.

What do you do?

You just noticed a misplaced file. It would make sense to sort it, but it could take a few seconds — and you were on your way somewhere else. For cases like this, I use what I call waterfall organization. The idea is simple: if you don’t know the exact destination folder of a file or don’t want / can’t spend time on it right now, don’t force it. Do something smaller. Move the file one step closer. One drag-and-drop, one move command. That’s it. Think of it as the scouting rule of digital organization: leave every folder a little bit cleaner than you found it.

When I process an inbox and don’t immediately know the precise folder three layers down, I don’t stall. I don’t overthink. I move the file one step further into the correct org space (root-level folder). Later, when I’m already in that area, I move it one step further. Files trickle downward over time, like water flowing down a cascade. Each touch moves them closer to their natural resting place.

The key insight is this: organization doesn’t have to happen in one move. waterfall.org turns sorting into something incremental — a gradual process. No cleanup weekend. No grand restructuring. Just small corrections whenever you bump into something. Over weeks and months, entropy decreases. Not because you fought it once. But because you biased the flow.

Folders stop being a burden. They become channels. They stop feeling like a decision you have to get right. They become part of the flow of working. You adjust as you go.

If in 2026 you haven’t given up on folders, you can count yourself among the clergy of directories. That doesn’t mean doing grunt work for its own sake. Use AI where it makes sense. But when it comes to the main structure of your digital life, remember: where you put things is not a trivial decision. It carries weight. It carries value. It carries meaning.

1. In Obsidian, I use the Auto Note Mover plugin to move notes automatically based on tags. Any note tagged #isa/person gets routed to my people folder without any manual action. That’s a legitimate use case for automation: the destination is unambiguous, the rule is stable, and there’s nothing to interpret. No surprises.

Another narrow use case I recently tried: I had around 150 unread research papers sitting in a dump folder — years of feeding and ignoring (hello, SME). I hadn’t engaged with them yet, so there was little to lose by running an experiment. I tested Claude Cowork and delegated the task. The whole process took about twenty minutes and burned my entire free token limit for the day. But in the end, the papers were renamed in year_author_title format and sorted into folders — mostly while I did other things. I haven’t fully verified the results. Maybe it was worth it. Maybe not. But it was a useful dip into new tooling, and the tooling is clearly getting stronger.

Still, I would never let AI organize my whole vault. Asking AI to design your entire org structure is like asking a stranger to rearrange your kitchen without ever watching you cook. Even if that stranger were a genius who followed every best practice and never made a mistake, it still wouldn’t fit you. Because the point isn’t correctness. You are not building a public library. The point is that it’s yours. And it takes into account your uniqueness, your quirks, and your mental visibility.