I operate on eyes. Specifically, I remove cataracts and rebuild the delicate architecture of the eyelid and orbit. Cataract surgery, when you have done thousands of them, becomes something close to choreography. The best cases are over in five minutes. The wait between one patient and the next is often longer than the operation itself.

What I do during that wait matters, and it is the reason I want to start this article with a scene from the operating theatre.

While the healthcare assistant brings the next patient through, the nurses reset the instruments, and the anaesthetist prepares her tray, I stand at the edge of the theatre and I do nothing. Or rather, I appear to do nothing. What I am actually doing is running the next case in my head. I can see the patient already on the table. I can see the specific shape of their pupil, the density of the lens, the angle at which I will place the main incision. By the time they are wheeled in, I have already completed the operation once. The real case is a second performance.

I did not invent this. Every good surgeon I have trained with does something like it. It is called visualisation, and it is only possible because the mind, when left unstimulated, will obligingly rehearse the thing it is about to do. Boredom is the raw material of preparation. The empty ten minutes between cases is not empty. It is the most important ten minutes of the morning.

Now watch what the young doctors and medical students in the theatre do during those same ten minutes.

They reach for their phones. They scroll without expression, thumb moving in the same small arc, face unlit by anything in particular. If the phone is in a pocket, they drift to the theatre computer and scroll through a news aggregator like the BBC or a messaging app instead. The scrolling has no subject. They are not reading. They are not preparing. They are topping up.

These are some of the brightest people I have ever met. They got into medical schools that turn away nineteen applicants for every one they accept. They have passed exams that would defeat most of the country. Their capacity for sustained intellectual work is, in principle, extraordinary.

And yet they cannot sit with ten minutes of nothing.

I do not say this to indict them. They are not weak. They are the first generation of doctors who grew up with the smartphone, and the smartphone was engineered (deliberately, expensively, by some of the largest companies in human history), to make sitting with ten minutes of nothing feel almost physically painful. They are casualties of a design decision, and the design decision was made by other people a long time ago.

But here is the thing I cannot stop thinking about. If my resident doctors cannot forward-play a cataract operation during the quiet between cases, they will never become the surgeons I trained to be. That is not sentimental. It is mechanical. The craft cannot be acquired by a mind that is being constantly interrupted. It simply cannot.

And if this is true in surgery, a profession with perhaps the most concentrated selection pressure for sustained focus of any in modern life, what is it doing to everyone else?

In 2004, a researcher at UC Irvine named Gloria Mark began following knowledge workers around their offices with a stopwatch.

She was not in an operating theatre, but what she found describes mine precisely. She shadowed software engineers, financial analysts, project managers. Every time one of them moved from a document to an email, from an email to a phone call, from a phone call back to the document, she clicked the stopwatch. Start. Stop. Start. Stop. For three and a half days, everything each person did was timed to the second.

When the numbers came back, they were strange enough to seem wrong. The average person spent two and a half minutes on any single thing before switching to something else. That was 2004. The iPhone did not exist. Slack did not exist. Gmail had launched that April and most of her subjects were still on Outlook.

Mark kept measuring. By 2012 the average was 75 seconds. By the late 2010s it was 47 seconds. The median (the midpoint, the number that tells you half of all observations fall below it) had collapsed to 40.

Forty seconds.

That is how long the average person can now stay on a single screen before their mind or their machine pulls them somewhere else. It is not long enough to finish a paragraph. It is not long enough to follow an argument. It is not long enough to have an idea.

This is not a productivity problem. It is a civilisational one. And almost nobody is treating it like one.

The framing we have been handed is that this is your fault.

You lack discipline. You have poor habits. You should try time-blocking, or the Pomodoro technique, or install one of the seventeen apps that promise to protect you from the other apps. The industry that built the problem has generously offered to sell you the solution.

This is the lie, and it is useful to say so plainly. The collapse in sustained attention did not happen because three billion people simultaneously lost their willpower between 2004 and 2020. It happened because the most profitable companies in human history poured a trillion dollars into engineering devices and interfaces whose explicit commercial purpose was to interrupt you as often as physically possible.

They succeeded. Forty seconds is the measure of their success.

The engagement economy is not a metaphor. It is a business model with a target variable, and the target variable is the number of times per hour you look at a screen. Everything else - the infinite scroll, the pull-to-refresh, the red notification badge, the algorithmic feed, the push alert that arrives while you are trying to read a chapter of a book - flows from the optimisation of that single number.

The CEOs who built these devices know exactly what they have done. Kaiwei Tang, who runs the Light Phone (one of the few devices designed deliberately to do less) put it like this: “Why do I have to bring a mini computer that tries to drive engagement 24/7? Why do I take it into bed? Wake up with it, eat my dinner with friends with it? We don’t stop.”

We don’t stop. And so we cannot begin.

Here is the number that has stayed with me since I first read it. When you are interrupted at a task (by a notification, a ping, someone walking past your desk, or by yourself reaching for your phone out of habit) it takes an average of 23 minutes and 15 seconds to fully return to what you were doing.

Gloria Mark again. Published in 2008. Replicated many times since.

Now pair that with the other finding: the average knowledge worker is interrupted, or interrupts themselves, roughly every three minutes. The arithmetic is obvious and grim. We are never not recovering. Most of us spend our working lives in a permanent state of partial concentration, halfway back from the last interruption when the next one lands.

This is not the same thing as being busy. It is the opposite. Busyness is what happens when you have too much to do. This is what happens when you cannot do any of it.

Count the things that become impossible at this cadence. You cannot write a novel in 47-second increments. You cannot develop a mathematical proof. You cannot compose a piece of music. You cannot read Middlemarch. You cannot hold a complicated argument in your head long enough to see whether it is true. You cannot visualise the next case. You cannot, in a meaningful sense, think.

And these are not trivial losses. The entire accumulated inheritance of Western thought - the essays, the novels, the theorems, the symphonies, the legal arguments, the scientific papers, the theologies - was produced by minds that could hold a thought for longer than 40 seconds. Every one of them. Descartes did not write the Meditations in between Slack pings. Virginia Woolf did not draft To the Lighthouse while also checking LinkedIn. Darwin did not develop natural selection during brief windows of concentration between push alerts from the Royal Geographical Society.

If the historical record of human intellectual achievement was produced by people with attention spans of two minutes and up, and the current average is forty seconds, we should at least be asking the question: what are we no longer capable of producing?

The strange thing is how quickly this became normal.

A person born in 1990 has now spent their entire adult life inside the engagement economy. They do not remember being bored in a queue. They do not remember the specific quality of a Sunday afternoon before the smartphone: the slow, almost uncomfortable expansion of time, the mind turning inward because there was nowhere else for it to go. For them, the constant low-level fragmentation of attention is not a recent condition. It is just what being alive feels like.

This is how civilisational shifts work. They do not feel like shifts from the inside. They feel like weather. The Romans who lived through the slow collapse of the Western Empire did not think they were living through the slow collapse of the Western Empire. They thought they were living through a bad few decades.

We are, I think, living through something like that now, though the thing collapsing is not a political order but a cognitive one. Specifically: the capacity of ordinary educated people to sustain serious thought for long enough to produce serious work. That capacity was never universal and never easy, but for roughly four centuries it was at least available to anyone who wanted it. It is not available now. Not in the same way. Not at the same scale.

And the trajectory is downward, not upward.

I have started to think of the forty-second figure as a kind of civilisational vital sign. Not a diagnosis (you cannot diagnose a civilisation, regardless of what Dr. Trump thinks) but a reading, the way a doctor takes a blood pressure before they know what is wrong. The reading is bad, and it has been getting worse for two decades, and almost none of the institutions that should be paying attention to it actually are.

Schools hand children the devices that do the damage and then complain that the children cannot read. Universities issue laptops that run the same four apps as everyone else’s laptops and then wonder why the essays are worse. Publishers lament the death of the novel while serialising their authors on platforms engineered to fragment the very attention novels require. Governments commission reports on productivity decline while declining to regulate the single largest extractive industry operating on the human nervous system. Hospitals (my own profession!) hand their brightest trainees tablets and email accounts that are functionally indistinguishable from slot machines, and then wonder why the next generation of Consultants seems harder to form than the last.

The absurdity is so total that pointing at it feels almost rude. But someone has to. And the person who can see it most clearly is not the policymaker and not the academic but the practitioner: the individual human being who sits down each morning to try to hold one coherent thought in their head for long enough to do the work. The surgeon between cases. The writer at their desk. The student at their books. Each of them knows. They know in their body. They know because every morning, for a few minutes or a few hours, they are in a losing fight with the machine in their pocket.

This publication is for that person.

Not because I have a solution. I do not. I’m trying to build one, but a solution at civilisational scale would require coordinated action across industries and governments and individual households, and I am not holding my breath. But at the scale of a single life - your life, my life, the life of the person reading this on their phone on a train - something is still possible. You can protect one hour. You can build one ritual. You can choose one tool that was designed to help you rather than harvest you. You can, against the grain of everything the economy is pushing at you, insist on finishing the thought.

That insistence is the only starting point I can see. And it is the reason I am writing.

In the essays that follow, I want to try to tell this story properly. What was lost. What the machines were designed to do, and who designed them, and when it became clear that the design was working. What the alternatives look like (the dumb-phones, the typewriters, the monasteries, the four-hour focused morning, the deliberate, almost embarrassing insistence on doing less). What writers and thinkers and makers across history have said about attention and solitude and the conditions of real work, and why almost none of those conditions are present in the average 2026 knowledge-worker’s day.

I want to document the collapse, because it is not yet documented. And I want to describe, carefully, what it would take to live on the other side of it.

There is no ten-step list at the end of this essay. There will not be one at the end of any of them. The problem is not that you lack tips. The problem is that something has been taken from you, and the first thing anyone can do is notice it has been taken.

Forty seconds. That is the measure.

Start there.

Sources and further reading. The headline figures in this essay come from Gloria Mark, Chancellor’s Professor of Informatics at the University of California, Irvine, and are drawn from two decades of fieldwork: the shadowing studies published in the mid-2000s, the screen-logging work that followed, and her 2023 book Attention Span: A Groundbreaking Way to Restore Balance, Happiness and Productivity. The 23-minute-15-second refocus figure is from Mark, Gudith & Klocke, “The Cost of Interrupted Work: More Speed and Stress,” CHI 2008. Kaiwei Tang’s remarks were given in an interview with Mobile World Live, March 2026.