Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.

Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium, totam rem aperiam eaque ipsa quae ab illo inventore veritatis.

The keeper of the gate

Nemo enim ipsam voluptatem quia voluptas sit aspernatur aut odit aut fugit, sed quia consequuntur magni dolores eos qui ratione voluptatem sequi nesciunt. Neque porro quisquam est, qui dolorem ipsum quia dolor sit amet consectetur adipisci velit.

She did not tend the garden because it was beautiful. She tended it because forgetting was the only thing she feared more than being alone.

Ut labore et dolore magnam aliquam quaerat voluptatem. Ut enim ad minima veniam, quis nostrum exercitationem ullam corporis suscipit laboriosam, nisi ut aliquid ex ea commodi consequatur quis autem vel eum iure reprehenderit qui in ea voluptate velit esse.

Quis autem vel eum iure reprehenderit qui in ea voluptate velit esse quam nihil molestiae consequatur, vel illum qui dolorem eum fugiat quo voluptas nulla pariatur. At vero eos et accusamus et iusto odio dignissimos.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.

Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium totam rem aperiam.

The paradox of abstraction

Nemo enim ipsam voluptatem quia voluptas sit aspernatur aut odit aut fugit, sed quia consequuntur magni dolores eos qui ratione voluptatem sequi nesciunt.

We do not understand the tools we build, yet we trust them with our lives, our attention, and increasingly — our thinking.

Ut labore et dolore magnam aliquam quaerat voluptatem. Ut enim ad minima veniam, quis nostrum exercitationem ullam corporis suscipit laboriosam nisi ut aliquid ex ea commodi consequatur quis autem vel eum iure reprehenderit qui in ea voluptate velit esse quam.

The fence had been there longer than anyone could remember, and so had she.

Most of the village had stopped asking about the garden after the third drought. There was nothing left on the other side of it, they said — nothing worth the walk, nothing worth the care. The soil was too old, too tired, too full of whatever it had already given. But every morning before the heat came, she was there with her tin watering can and her unreasonable patience, coaxing water from a well that everyone else had written off.

Her name was Mira. She had been born in the house at the garden’s edge and had never left — not because she couldn’t, but because she had seen what happened to the things that were not tended. She had watched the orchard to the east go grey. She had watched the river to the west slow to a colour she didn’t have a word for.

She kept tending.

The keeper of the gate

There was a boy who came sometimes, in the late afternoons when his mother thought he was doing homework. He would sit on the wall and watch her work, not speaking, as if afraid that words might disturb something fragile.

One day he asked her why she bothered.

She did not answer immediately. She deadheaded a rose — the last rose; she knew it was the last rose — and set the dried petals in the pocket of her apron.

“Because forgetting,” she said finally, “is the only thing I fear more than being alone.”

The boy turned this over for a long time. Then he climbed down from the wall and picked up the second watering can.

She did not thank him. She just moved to the next row.

Some things are not said. Some things are only done, again and again, at the edge of the world, in the cool hour before the heat arrives. That is enough. That has always been enough.

She wrote it on the back of a receipt — a grocery receipt, three days old, already soft at the fold lines from being in her coat pocket. She used a pen that was almost out of ink, pressing hard.

She did not address it. There was no point in addressing it.

What she wrote was not long. It was not beautiful. It was the kind of writing that happens when there is no one left to explain yourself to, and the silence has become a weight you need to distribute somewhere outside your own body.

She folded it twice, the way you fold paper when you are not sure what to do with it next. She put it in the same pocket as the pen.

What the letter said

That is the part I will not tell you.

Some things are kept not out of secrecy but out of dignity — hers, and the star’s, and the particular intimacy of speaking to something that cannot answer and will not judge. There is a kind of honesty only possible in a one-sided conversation.

What I will tell you is that she kept the receipt for four years. That she moved twice, and the receipt moved with her, tucked into the same inner pocket of the same coat, which she repaired three times rather than replace.

And that on the fourth year, on a night when the sky was very clear, she took it out, unfolded it, read it once, and then folded it again exactly the same way.

She put it back in her pocket.

She went inside.

She made tea.

The star, for its part, continued being a star — which is, perhaps, all anyone can be asked to do.

The technician said he had three minutes. Maybe less, depending on the cloud cover over the relay station.

He had been waiting fourteen months for three minutes. He had practised what he would say. He had written it down, thrown it away, written it down again, and finally stopped writing it down because the act of writing it had started to make it feel like a script, and what he wanted to say was not a script.

He picked up the receiver.

The static had a particular quality at this distance — not empty, but full, the way silence in a house is full of everything the house has ever held.

“It’s me,” he said. And then, because there were only three minutes: everything.

He said the things that had been sitting in his chest like stones since the last time they had been in the same room. He said the things he had translated into ordinary conversation for fourteen months — into questions about sleep and weather and whether the supplies had arrived. He said them plainly, without architecture, without the careful arrangements he had spent so long constructing.

He said: I was wrong about the important thing.

He said: I know what the important thing was, now.

He said: I have been thinking about the afternoon in the kitchen with the bad light, and I think I understand what you were trying to tell me.

The technician held up two fingers. Two minutes.

He stopped talking. He listened to the static, which was full.

Then he said the last thing, the simplest one, the one that needed no context and no explanation, the one that had been true the entire time and would remain true after the signal dropped and the relay station went dark and the cloud cover moved in over the mountains.

He said it carefully, the way you say something you want to survive the distance.

Then the line went quiet.

He kept them in jars on a shelf above the sink. None of them were the same.

There was the silence after his father’s last sentence, which had a particular density to it, almost amber. There was the silence at the end of a concert he had attended alone at nineteen — lighter, almost effervescent, the kind that pops against the skin. There was the silence of the empty flat on the first morning after she had left, which he had bottled carefully, using both hands.

He was not sure when he had started doing this. The collecting had preceded the awareness of it, as these things often do.

The taxonomy

He had tried once to organise them. There were the expectant silences — the held breath before news arrives, before a door opens, before a name is spoken. These tended to be compact, pressurised. He kept them on the left side of the shelf.

The finished silences occupied the middle. These were the ones that had come after something — after applause, after arguments, after long meals. They were spacious and slightly luminous. He liked them best.

On the right: the silences he could not categorise. The one from the afternoon in the museum when he had stood in front of a painting for forty minutes and not thought about anything at all. The one from the garden of a stranger, which he had bottled without permission, guilty and grateful in equal measure.

He was not sure what he planned to do with any of them.

Most likely nothing. Most collections are not for any purpose beyond the having — the acknowledgement that something happened, that it was worth keeping, that the world makes particular shapes of quiet that deserve to be noticed.

He washed his hands above the shelf, carefully, every morning.

He dried them on a towel that had been folded the same way for eleven years.

He looked at the jars.

He went to work.

The lift between the seventh and sixth floor stopped, and for seven seconds nothing was certain.

In those seven seconds: the woman in the corner calculated whether her phone had signal. The man by the door thought, involuntarily, of his childhood dog. The teenager with the earbuds looked up for the first time, which is how they noticed each other — three strangers who had been sharing four square metres of space without any acknowledgement that the other two existed.

For seven seconds, they were all in the same situation.

The woman’s phone had no signal.

The man’s childhood dog was called Biscuit, and was probably not relevant.

The teenager took out one earbud.

Then the lift shuddered, the numbers resumed their climbing, and the doors opened on the sixth floor to let someone in who had been waiting and did not notice that anything unusual had happened.

The three of them did not speak. They did not exchange a glance that meant something. They arrived at their floors, which were different floors, and they went about their days, which were different days.

But for seven seconds they had all been aware, simultaneously, that the ground was somewhere far below and that very little — structurally speaking — stood between them and it.

That is the whole story.

It happened on a Tuesday.

The river had been running through the village for two hundred years and had never spoken once, which made it, by most reckonings, the most trustworthy thing in the valley.

People brought it everything. They brought it the names of people they had loved and stopped loving, wrapped in paper, weighted with stones. They brought it the plans that had failed, the letters that had been intercepted, the confessions that were too heavy to carry to the priest. The river accepted all of it without comment or condition, carried it downstream, and delivered it eventually to the sea, which had room for everything.

What the old women knew

The old women of the village had an understanding with the river that was not discussed openly. They had learned — from the women before them, who had learned from the women before that — that the river did not merely receive. It also returned.

Not the things themselves. The river was permanent about the things themselves. But a quality — a residue of having known something — did sometimes come back in the water on certain mornings, when the current ran slower than usual and the light was the particular colour of forgiveness.

On those mornings, the old women filled their buckets early and used that water for cooking.

The food made on those days tasted slightly different. Not better, exactly. More complete.

The younger people did not notice, or if they noticed, they attributed it to the season, the soil, some variation in the local grain. They did not think to ask about the river.

The river did not take offence.

It carried what it was given. It returned what it could. It kept running — through the springs that were cold and the summers that were not, through the dry years and the flooding years, through the building of the new bridge and the falling of the old one — patient and continuous and full of everything it had been trusted with.

It is still running now, past the village and beyond it.

It has not spoken yet.

Every tool I use daily is, at some level, beyond my comprehension.

I do not understand the compiler that turns my code into something the processor can execute. I do not understand the processor. I do not understand the physical properties of the silicon, or the manufacturing process, or the logistics chain that brought the materials from the ground to the fabrication plant to the device that is now warm in my hands.

This is not unusual. This is the condition of living in a technological civilisation. But I have been thinking lately about what we lose when we stop asking — when the gap between capability and comprehension becomes so wide that we no longer notice the gap is there.

The paradox of abstraction

Abstraction is the mechanism by which all progress is made. You abstract the hardware into an operating system. You abstract the operating system into a programming language. You abstract the programming language into a framework. You abstract the framework into a product. At each layer, you gain leverage — the ability to do more with less understanding.

But leverage is not free. Every abstraction is also a promise: that the layer below will behave as described. That the contract will hold. That the edge cases have been considered by someone, somewhere, at some point.

We do not understand the tools we build, yet we trust them with our lives, our attention, and increasingly — our thinking.

This is not an argument against abstraction. It is an argument for epistemic humility — for being precise about what we know versus what we assume, for maintaining some thread of curiosity that reaches down through the stack to ask: what is actually happening here?

The people who build the next layer of abstraction are, usually, the ones who understood the previous layer well enough to see its limits. That understanding — the earned kind, the uncomfortable kind — is what makes the difference between an abstraction that holds and one that fails.

What I am not saying

I am not saying we should all become systems programmers or materials scientists. Division of labour is real and valuable.

I am saying: pick one thing. One layer. One mechanism. Follow it down further than you think you need to. The fluency you develop will be specific, but the habit of following-it-down will transfer.

The most dangerous sentence in technology is “I don’t need to know how that works.” Usually, you don’t. Until suddenly you do.

The last time I was genuinely bored, I mean uncomfortably, productively bored with no exit available — I was nine years old, sitting in the back of a car on a long drive with nothing to do, and I ended up inventing a game involving the shapes of road signs.

That game eventually became a small obsession, which became a notebook, which became one of the clearer memories I have of early childhood creativity. It came entirely from boredom and a captive hour.

I cannot remember the last time I had a captive hour.

The filling of every gap

The design of the modern phone is fundamentally a design against boredom. Every moment of potential vacancy — the lift, the queue, the thirty seconds between meetings — has been identified as an opportunity, and a product has been built to fill it.

This is not accidental. Boredom is uncomfortable. Discomfort drives behaviour. Behaviour can be monetised.

What is not accounted for, in this economy, is what boredom was for.

Boredom is the mind’s idle state — and idle states are not wasted states. They are integration states. The brain, freed from directed attention, begins to connect: to consolidate memory, to process unresolved things, to wander in directions that directed attention would never allow. The daydream, the shower thought, the idea that arrives while you are walking with nowhere particular to be — these are boredom’s outputs.

We have eliminated the input and are surprised to miss the output.

What I notice in myself

I have become worse at sitting still. I have become worse at being in a room with nothing happening. When I attempt sustained focus, I feel the phantom reach — the gesture toward the phone that is no longer in my hand — with a frequency that I find, on reflection, alarming.

This is not a productivity essay. I do not have a system to propose.

I am simply noting that something was there — an uncomfortable, generative, deeply human state — and that it has been quietly replaced, and that most of us did not notice it happening, and that I am not sure we have fully counted the cost.

The road signs are still out there. No one is looking at them anymore.

The phrase “attention economy” has become so familiar that it has mostly stopped meaning anything. We say it the way we say “late capitalism” — as a gesture toward something we sense is true but have stopped bothering to examine. Let me try to examine it.

Attention is scarce. There are approximately sixteen waking hours in a day, and a meaningful fraction of them can be directed at one thing with sustained focus. This fraction is smaller than most people believe, and smaller still than it was before the smartphone. What’s scarce is valuable. What’s valuable attracts competition.

This much is understood. What I want to think about is the mechanism.

The rental model

Real estate was the original attention economy. You owned land; you charged rent; people paid to exist in your space. The landlord’s power came from scarcity — there was only so much land, particularly in places people wanted to be.

Attention operates similarly, with one crucial difference: you cannot expand land, but you can manufacture urgency. You can create the conditions under which people feel they must be somewhere — must check, must refresh, must respond. The scarce resource is artificial, maintained by design.

This is not an accident or an externality. It is the product.

What we are selling, exactly

We do not sell our time in this economy. Time is the wrong unit. We sell the moments in which we are not paying attention to anything else — the state of openness, the cognitive surplus, the readiness to receive. Advertisers have always known this; the insight is not new. What is new is the precision of the extraction.

Previous attention economies — broadcast television, newspapers, radio — worked with aggregate attention, measured in households and circulation numbers. The contemporary model works at the individual level, in real time, with a feedback loop that allows constant refinement.

The result is a system optimised, to a degree no previous medium has achieved, for the capture and monetisation of your specific cognitive state at any given moment.

You are not the customer. You are the inventory.

This is worth sitting with, not as a reason for outrage, but as a description of the actual situation — so that any choices you make about your attention are made with clarity about what those choices are.

I have eighteen unfinished notebooks. I counted them last week, which is itself a suspicious activity — the kind of thing you do when you are avoiding starting something new.

The notebooks contain stories that reached the middle and stopped. Essays that ended mid-sentence, not from interruption but from the particular fatigue that sets in when you have understood something enough to say it but not enough to say it well. Lists of intentions. Lists of lists of intentions. The geological record of a dozen projects that seemed essential in January and had been replaced by April.

I used to think of these as failures. I have started to revise that.

The unfinished as process

There is a school of creative thought that holds finishing as the only meaningful outcome — that the draft, the sketch, the note are only valuable insofar as they lead to the completed object. Shipping is the point. Done is the deliverable.

I understand this argument. I have worked in environments where it was correct. There are categories of work where the incomplete has no value — software, legal documents, the construction of bridges.

Creative work is not always this kind of work.

The essay that stops in the middle often stopped because the thinking it was generated by reached its natural boundary. I understood the thing. I did not need to perform understanding for an audience. Moving on was not abandonment; it was efficiency.

The suspicion I am sitting with

I am not sure this is fully true. I suspect I also abandon things out of fear — the particular fear that the finished version will be less than the imagined version, that delivery will be the moment of reckoning, that the gap between what I wanted to make and what I made will become visible.

The unfinished is still potential. The potential is safe.

I am learning to distinguish, in myself, between the essay that stopped because it was done and the essay that stopped because I was afraid to find out. The former I leave alone. The latter I return to.

Most of these notebooks are still on the shelf. But I am asking better questions of them now.

I have been designing interfaces for long enough that I sometimes forget they are arguments.

An interface is not neutral. The decision to show a confirmation dialog before deletion is a claim about human error and irreversibility. The decision to make “unsubscribe” grey and small is a claim about whose convenience matters. The decision to put the settings menu three levels deep is a claim about what kind of user you are designing for and how often you expect them to stray from the expected path.

Every button is a sentence. Every menu, a paragraph. Most software is, in this sense, a very long essay — one that most of its authors have not read as a whole, because it was written by many hands over many years, and the argument has become internally incoherent in the way that arguments do when they are assembled rather than composed.

The designer as author

The hardest part of my job is not interaction design or visual design. It is the maintenance of a coherent position — a set of beliefs about the user that remain consistent across the surface, even as the surface changes.

When that position is clear, the interface has integrity. Not in the moral sense, but in the structural sense — the sense in which a bridge has integrity when all its components are in agreement about where the load is and how to carry it.

When the position is unclear, or compromised, or abandoned at some point in the product’s history and replaced with a different position without the old one being fully removed — you get the kind of experience most people have on most days with most software. The sense that the tool is occasionally arguing with itself.

What I think about when I review designs

I try to ask: what does this surface believe about the person using it?

Not what it says it believes — the marketing claim, the mission statement — but what it actually claims, in the placement of elements, in the defaults, in the things it makes easy and the things it makes hard.

This is an uncomfortable question, because the answer is usually complicated. We believe the user is intelligent, except on this screen, where we do not trust them with advanced options. We believe the user’s time is valuable, except in this flow, where we have added three steps we know most people will not need.

An interface is an argument. Like most arguments, it benefits from knowing what you are trying to say.

The algorithm knows I like a certain kind of film. It knows this from watch history, from pause and rewind patterns, from the time of day I tend to watch and the device I watch on. It has built a model of my preferences that is, by any measurable metric, more accurate than the model most of my friends have.

And yet when a friend recommends a film, I watch it differently.

I have been trying to understand why.

The archaeology of recommendation

When a person recommends something, they are not merely transmitting information about a cultural object. They are also transmitting: why they thought of you when they encountered it, what it made them remember, the particular evening or mood in which they found it, and an implicit claim about the kind of person they believe you to be.

This is surplus content. It cannot be extracted from watch history or quantified in a training set. It arrives as subtext, or sometimes as text — “this reminded me of the conversation we had about your father” — and it changes the experience of the thing being recommended.

The algorithm has no surplus content. It has a match score.

What taste actually is

Taste is often described as preference — the set of things you like. I think this is incomplete. Taste is more like a lens: the particular way everything you have experienced has shaped the way you see what you encounter next.

It is the residue of the bad films you watched with someone you loved, which contaminated your relationship to a genre. It is the book your teacher handed you with a note inside, which meant you could never read it without hearing her voice. It is the album that played on a specific night, which is now inseparable from what happened on that night.

Taste is biographical. It accumulates through the specific accidents of a specific life. It cannot be reconstructed from behavioural data, because the data does not contain the life.

What the algorithm is good at

This is not an argument against algorithms. Personalisation solves a real problem — the discovery problem, the infinite-shelf problem, the impossibility of finding anything good without a guide.

The algorithm is very good at finding more of what you already like. This is useful, regularly.

What it cannot do is introduce you to something that will change what you like. That requires a force from outside the feedback loop — a person, a chance encounter, a recommendation that violates the pattern precisely because the recommender knows something about you that your history does not record.

Curation, in the fullest sense, is an act of interpretation. You cannot automate an act of interpretation. You can only automate the appearance of one.