Decision Debt: When Capture Is Free, Judgement Is the Bottleneck

I had more than 300 things sitting in my task list under a single heading: decide. Build it, park it, or kill it. Not one of them was there on purpose. Each had slipped past the point where I should have made a call, because writing it down felt like progress and calling it didn’t.

Capture stopped costing anything

Writing an idea down properly used to take enough effort that you didn’t bother unless it mattered. That friction did real work. It forced a small verdict at the door: worth keeping, or not.

That friction is gone. A voice note on the way to a meeting, a line typed into a tool between tasks, and the idea is captured, structured and queued, ready for someone to build it the moment there’s a slot. Capture went from a judgement call to a reflex. A reflex doesn’t decide anything. It just defers.

An undecided pile is not a parked one

There’s a version of this that costs nothing and is entirely fine. You look at something, weigh it, and choose to wait. That’s a decision, and it’s honest about what it is. The thing still has somewhere to go.

Most of mine weren’t that. Undecided means you never looked hard enough to choose. “Decide” wasn’t a path, it was a waiting room with no door out of it. Every item I’d never ruled on was a small debt against a future version of me who’d apparently deal with it. The interest doesn’t show up as a number. It shows up in the one thing you needed to see clearly, buried under three hundred things nobody ever called.

The friction was doing the triage

This isn’t just my list. When capture and building both cost something, the cost did the triage for you — you only started what had already survived a rough mental sum. Take that away and put nothing in its place, and nothing dies at the door. Everything gets in, and it all waits on the one thing that never got cheaper: someone looking at it and calling it.

That’s a personal task list. It’s also a team’s roadmap, or a department’s list of things it could digitise given the budget. The tools got faster at producing candidates. Nobody got faster at judging them.

So what actually fixed it?

Judgement, still. Nothing I’ve built makes a call for me and I wouldn’t want it to. What changed is that judgement now has one place to happen.

It’s the same task list everywhere. Phone, laptop, wherever I am. I spend nothing deciding where a thing goes, because there’s only one place it can go — no notes app, no half-used kanban board, no service I signed up to last year and forgot.

It deduplicates itself. I capture the same idea three times — I know I do, and I won’t stop — and the system sorts that out rather than me policing it.

The payoff isn’t tidiness. It’s that the idea is out of my head. I’m hypervigilant by default, and anything unlogged keeps running in the background. That’s what eats focus. Once it’s in, it’s in an action path, even if that path is paused on its workstream. Paused is fine. Waiting in the queue to get on the roller coaster is part of the ride.

So the debt was never the things I hadn’t done. It was the things with no path at all. Cheap tools were always going to hand us more ideas than anyone could build in a lifetime; that was never the hard part. The hard part is the bit that hasn’t got cheaper: deciding what’s worth your time, and having one place you trust for everything else.

Companion essay to Capture outran decision, and everything got rebuilt — the AI Sabbatical, post 4.