← Writing

Essay · August 12, 2026 · 6 min read

Decision Architecture

Decision Architecture

Deciding is not the result of long thinking.

A decision is not what you get at the end of intense thinking and planning everything step by step.

Deciding is not arguing with every doubt that shows up.

A decision is 100% certainty without knowing all the details. It is delegation, and trust in the process that carries it.

It is like dropping a letter in a mailbox and leaving it there. Not checking on it. Not pulling it back out to read it one more time.

You do not need to know how

When you drop a letter in the box, you do not know what bike the postman rides. You do not know his route, his shift, which van the bag goes into, or what the sorting office looks like at three in the morning.

You do not need to. You need the letter to arrive.

That is the part people invert. They believe a decision requires knowing the whole path, so they stand at the mailbox trying to model the entire postal system before they let go of the envelope. Some of them try to build the post office themselves, and then audit it, just to be certain it works.

That is not diligence. That is a letter not being sent.

A decision is a result you have named. Everything between here and that result is a problem, and problems get solved when you arrive at them. You cannot solve them in advance from the pavement in front of the mailbox.

The architecture is the mailbox

You write the letter. You drop it in. You forget about it.

What happens next is not your job. The box is built to hold it. The postman comes. The letter travels. It reaches someone. Something comes back, or nothing does, and either way the loop closes.

That is decision architecture. The construction that carries a decision after you release it, and filters what happens at each stage without you standing over it.

Your only job is writing letters worth sending.

A well written letter needs no translation. It is a clear command with a clear answer available. If the person receiving it has to guess what you meant, you did not send a decision. You sent a question wearing a decision’s clothes.

And if it does not arrive, you write another one. That is the entire disaster you were protecting yourself from.

How is the easy part

How is comfortable. How has no consequences. You can spend a month on how and nothing in your life is on the line, because how does not ask you to decide anything. It only asks you to be busy.

Why is the expensive one. Why comes with responsibility, and responsibility comes with consequences that arrive in your name.

So people escape into how. Into tasks, into tools, into the new project, into a full calendar. It looks like work. Often it is good work. It is still an escape, and the thing being escaped is a decision sitting unmade in the background.

Busy is the most respectable way to avoid deciding. Nobody questions it. You do not question it yourself.

What not deciding actually looks like

You rarely refuse outright. Usually you half decide.

You leave it open. You will get to it. You need more information first. You want to see how the next two weeks go.

That is the letter that gets written and stays on the desk. No stamp. Not today. Later.

It is always easier to start something than to end something. A new project costs you energy. Ending one costs you a conversation, a position, and your name on the outcome. So the new thing gets started and the old thing stays open.

And an open loop feels like a door left open. Optionality. Nothing lost, nothing foreclosed.

It is not optionality. A door you never walk through is not a choice you are preserving. It is a decision you are not making, and it charges you rent for the privilege.

One letter on a desk is nothing.

Twenty letters stop being twenty letters. They fuse. They become one mass of unfinished business and you cannot see the individual decisions inside it anymore. You only feel the weight.

That is when you get stuck without knowing why. Nothing dramatic happened. You just have too many open loops running in the background, all consuming the same nervous system you need for the next move.

That is why people are stuck. Not workload. Unsent letters.

Getting out

Two ways.

Clear the whole thing. Throw the stack out and start with an empty desk. Most of what is on there died months ago and you have been carrying a corpse.

Or filter it. Take the stack one letter at a time and answer one question about each: does this get sent, or does this get thrown away.

Both work. What does not work is keeping the stack.

And notice what is true here. Deciding not to do something is a decision. Throwing the letter away closes the loop as cleanly as sending it. The RAM is released either way.

What surprises people is the last part. You expect clearing the desk to be brutal. Mostly it is not. Most of those letters lost their weight a long time ago and were being carried out of habit.

What is left afterwards is quiet, and an empty desk, and considerably more room than you thought you had. Room for things you actually want on it.

That is what deciding buys. Not a better outcome on any single letter. Peace, and space.

The two tools

Asymmetric Decisions is three products in one. The book, and two standalone files that do the actual work.

Shadow Patterns is the diagnostic. It catalogs the six mechanisms that keep letters sitting on the desk: architecture as avoidance, half open loops, negotiating against yourself, over responsibility, confusing pressure with importance, dragging unclean deals. It comes with a prompt. Paste it into whatever AI you already talk to, give it your context, and it will name which pattern is holding your specific letter. Not blockers in general. Yours.

The Decision Filter is the instrument. Five levels, each one closing a layer:

Recognition. Name the decision in one sentence. If you cannot say it in fifteen words as a yes or no, you are still holding a cloud, and you cannot filter a cloud.

Distillation. Strip out fear, social pressure, sunk cost, borrowed assumptions. This is where I feel like this is not right gets separated from this is structurally unsound. They look identical until you do this.

Structural Check. Four questions, scored. Is the structure clean. Does the other side carry their own weight. What does maintaining this cost me. Does it expand my freedom or create dependency. The number tells you before your feelings do.

Selection. Four possible answers. Yes. No. Yes with changed conditions. Not now until a named trigger. Maybe is not on the list. Maybe is avoidance in formal dress.

Movement. One physical action within 48 hours. Send, sign, decline, delegate, delete.

One tool tells you why the letter is still on the desk. The other filters out the false assumptions and ends with one decisive move.

The book is here: book.lucashubert.me

If you already bought it, both files came with it, separate from the main text. If you read the book and stopped there, you have the philosophy and not the instruments. They are still in your download folder. Open them.

And if you want this construction built around your actual context instead of in general, that is the work I do. Not coaching. Architecture.

Lucas

— Lucas Hubert

Originally published in Beyond Noise on Substack.

Keep reading

Beyond Noise

Join the community.

The essays, Open Office every Friday, and the Decision Room twice a month. Free to join; the room is for members — $18/month or $180/year. What Beyond Noise is →

Join Beyond Noise