What's Left on the List

I've been maintaining lists for twenty years. To-do apps, GTD, outliners, sticky notes. In recent years I came back to handwritten notebook and pen, Bullet Journal, interstitial journaling between tasks, some Apple Notes and Reminders, a shared family calendar. Fewer tools, more analog, less overhead.

They all worked. For about six weeks.

The problem was never the system. The problem was that I couldn't sustain any of them. GTD collapses if you skip a weekly review — and I always skipped the weekly review. Bullet Journal is a better design. The daily migration forces you to reflect on yesterday and choose what actually matters today. The cognitive work is the value, not overhead. I sustained a version of it longer than anything else — but the energy waned with each new notebook. And even at its best, I was still staring at a list where half the items weren't things I could act on. They were things I needed to research, coordinate, or decide. The migration helped me prioritize. It couldn't shrink the list.

That's the generous reading — I lacked the discipline. The less generous one: making the plan was the point. The execution was always optional. A new system feels like progress. It isn't.

Oliver Burkeman landed on the most elegant solution I've found. In Four Thousand Weeks, he proposes two lists: an open list and a closed list. The open list holds everything — every task, project, and someday-maybe. You never look at it during the day. The closed list holds at most ten items. That's your work. Nothing new gets on the closed list until something comes off. It's a forcing function for focus, and it works because it accepts a truth most systems deny: you will never do everything, so stop pretending you will.

I lived on that system for a while. But even Burkeman's method requires you to decide what moves from open to closed, to notice when something's stuck, to keep the list honest. You're still the one managing the list.

The no-list people

Bezos structures his day around a small number of high-quality decisions. Not tasks. Decisions. He putters in the morning, doesn't schedule early meetings, and focuses his energy on the two or three things that actually need his judgment. Everything else is delegated or systematized.

Andreessen went further — he stopped keeping a to-do list entirely. His approach: work on whatever feels most important right now. Three to five items, written fresh each night. No backlog. No system to maintain. Just a clear head and a short list.

Both arrived at the same insight: the goal isn't a better list. It's fewer things on the list that need to be you.

What changes with AI

The latest models aren't chatbots. Claude Code with Opus 4.6 doesn't wait for you to ask it things. It reads your files, checks your calendar, sends emails, researches options, and coordinates across systems. It has a plan mode — you describe what needs to happen and it figures out the steps, executes them, and comes back when it's done. The list is taken care of. You don't need to get in the way.

This week I rebuilt my task system. I had a Burkeman-style open list with about thirty items on it. But I noticed something: most of the items weren't things I needed to do. They were things I needed to think about, research, coordinate, or decide. Those are exactly the tasks that sit on a list longest, because they have no clear first step.

So I split everything into four categories. Do: things only I can do, with my hands, in the world. Delegate: things an AI can handle with a prompt. Auto: things that should just happen without me thinking about them. Someday: things that matter but not yet.

The majority moved to Delegate or Auto. They didn't need my attention — they needed someone else's. And "someone else" is now available around the clock for the cost of a prompt.

My closed list went from overwhelming to four items. Not by cutting scope. By recognizing that most of what I was tracking wasn't work for a human.

The risk is obvious: if the plan was always the comfort, a system that makes better plans faster might just be a better sedative. I don't have a rebuttal for that yet. Just a shorter list and the sense that something is different this time. Which is, of course, what I said last time.

The tax you pay to remember

Every list is a promise you made to yourself to think about something later. The longer the list, the more promises. Each one takes a small toll — not when you do the task, but every time you scan the list and decide not yet. That's the tax. Not the doing. The carrying.

What I want is a list I don't have to think about. Not because I'm ignoring my responsibilities, but because the system is carrying the ones I don't need to carry. The daily brief that checks my calendar without me asking. The birthday scanner that flags next week's gifts. The travel planner that already knows our preferences.

The best to-do list is the one with the fewest items on it — not because you did everything, but because most of it is handled.

What notebooks still do

I've filled several handwritten notebooks over the last couple of years. Every new one starts the same way — I don't want to write in it. It feels too pristine to mess up. Then I force the first words out and the preciousness breaks. What follows is different from anything I type. Slower, less filtered, closer to what I actually think.

No system replaces that. But a system can make sure what you write by hand doesn't vanish into a drawer. I digitized years of notebooks recently, and the AI found patterns across entries I'd never read together. The notebook stays analog. The insights don't have to.

The practice of writing by hand is worth protecting. It's one of the few things that slows your thinking to the speed of a pen. The notebook and the system aren't competing. One is for thinking. The other is for doing. The mistake is asking either to be both.

The shift

Every productivity system assumes you're the bottleneck — that the solution is a better way for you to organize, prioritize, and execute. Bezos and Andreessen got around this by being Bezos and Andreessen: they have people and organizations that handle everything else. The rest of us just stared at our lists.

AI changes who the bottleneck is. Not for decisions — those are still yours. But for the vast middle layer of modern life: the researching, the coordinating, the remembering, the following up. The work that isn't really work but fills a list like it is.

The cognitive shift isn't "I have a better system." It's "I think about fewer things." Not because I care about fewer things. Because the system holds what I don't need to hold.

I'm not more organized. I just carry less. Ask me again in six weeks.