What We Left to Luck

I. The house I grew up in

My mother was diagnosed with chronic depression and bipolar disorder when I was eight. She raised three children alone. I have had thirty years to work out how she managed it and I am no closer than I was.

I want to be careful, because she is a person and not an argument. There were long stretches when the house ran perfectly well and stretches when it did not run at all, and a child does not experience those as phases of an illness. A child experiences them as weather.

What I had, and had no idea I had, was a school that taught social and emotional skills properly at a time when almost no school did. Not a poster about kindness in a corridor. Actual practice, on a timetable. Noticing what somebody else was feeling and saying it out loud. Repairing a thing after you had broken it. Sitting beside a person who was upset without trying to solve them.

I took all of it home and used it on my mother.

I was nine, then ten, then eleven, and I was running a small unlicensed clinic out of my own kitchen. I got good at it, in the way children get good at the thing their house requires. I could tell from the sound of her on the stairs what sort of evening it was going to be. Sometimes I could change it.

It took me until my twenties to understand that her happiness had never been mine to control.

That is the part I would cut if I were trying to look good, and it is the part that matters most, because a child who believes he can fix his mother has to explain to himself every single time it does not work. He was not paying attention. He said the wrong thing. He did not try hard enough. I was very good at that explanation. I had it ready before the evening started.

I do not know who decided that my school would teach those skills. Some head teacher, at some point in the eighties, in a small city in northern Denmark, sat in a meeting and argued for something that was not on any curriculum and was not going to show up in any results. I have never known their name. It has occurred to me more than once that the single most useful thing that ever happened to me was an argument I was not present for, won by a person I cannot identify, eleven years before I needed it.

The children a mile away got nothing. Not less than me. Nothing. The only variable separating us was which door our parents happened to walk through when we were five, and I have never been able to make that seem acceptable, no matter how long I look at it.

Twenty years ago I became a mentalist, which is a strange place to end up and an unreasonably good laboratory. I perform about three hundred nights a year. The job is to walk in front of several hundred strangers and work out in roughly four seconds who is guarded, who came willingly, who is holding something they would rather not be holding. Then I have to get people to open up. Quickly, in front of an audience, on demand, and again tomorrow.

Nobody sets out to run that experiment several thousand times. I did it because it was my living. But it has left me with a firm opinion about how two strangers actually become less strange to each other, and that opinion is why I am writing this down.

II. What the numbers say

Before any of that, the scale. It is worse than most people think, and it is worse in two directions at once, which is the part that tends to get missed.

Start with the academic picture. On the most recent National Assessment of Educational Progress, American twelfth-graders posted the lowest average mathematics score since that assessment began in 2005, and the lowest reading score in more than thirty years.¹ Forty-five percent of them fell below the NAEP Basic level in mathematics, the highest share ever recorded.² Among fourth-graders, roughly forty percent are reading below Basic, the largest proportion since 2002.³

This is not simply a pandemic story, though it is convenient to file it as one. The acting director of the Institute of Education Sciences described scores for the lowest-performing students as continued declines that began more than a decade ago.⁴ The distance between the strongest and weakest students keeps widening.

Now the other direction. In 2023 the Surgeon General of the United States issued a formal advisory called Our Epidemic of Loneliness and Isolation.⁵ Not a report, not a white paper. An advisory, which is the instrument reserved for urgent threats to public health. Its central finding is that a lack of social connection raises the risk of early death by roughly as much as smoking fifteen cigarettes a day.⁶

The advisory runs to eighty-two pages and the sentence I keep returning to is not the headline. It sits among the recommendations for schools, and it says: implement peer-led programmes.⁷ The Surgeon General looked at the evidence and concluded that part of the answer is children being taught to reach each other.

For the children specifically, the Centers for Disease Control's national survey found forty-four percent of high school students reporting persistent sadness or hopelessness across the preceding year.⁸ The same advisory notes that young people on social media for more than two hours a day were about twice as likely to report feeling isolated as those using it for under thirty minutes.⁹

Put those beside each other, because the combination is the whole point. We have a generation doing worse academically than the one before it, lonelier than the one before it, and more heavily equipped with technology than any cohort in history.

I should be precise about what I am claiming, because the temptation here is to overreach and the overreach is what lets people dismiss you. I am not saying the loneliness caused the test scores. Nobody can show that from the data we have, and I am not going to pretend otherwise. My claim is narrower. We made an enormous bet on a particular kind of technology in schools. We justified it almost entirely on academic grounds. Both the academic outcome and the human outcome have since moved in the wrong direction, together, for over a decade.

At what point does that stop being a coincidence?

III. What we actually built

Here is the version we told ourselves.

Every child would have a device. Software would meet each student exactly where they stood. The quick ones would stop being held back, the struggling ones would stop being left behind, and the teacher, freed from delivering the same lesson thirty times over, would become a guide. We called it personalisation. It was a genuinely beautiful idea and I do not think anyone was lying.

Here is what got built.

A room in which thirty children sit within arm's reach of one another and look at thirty separate rectangles. We optimised the encounter between one child and one machine, and in the process we quietly removed most of the reasons those children had to deal with each other at all.

Group work became a shared document. A question became a search. The awkward, inefficient, unavoidable business of turning to the person beside you and saying I don't get this, do you became technically unnecessary. A thing that is technically unnecessary in a room of thirty is a thing that stops happening.

I am not making the argument you are expecting. I am not going to tell you to take the screens away, and I will come back to why, because the ban-the-phones position is wrong in a way that matters.

The failure was never the screen. It was the number. One device per child is an individualising technology by definition, and every design decision downstream of it points the same way: inward, toward the single user, away from the person sitting eighteen inches to the left. We spent fifteen years and an extraordinary amount of public money making school more individual, during the exact period when children were becoming more isolated everywhere else in their lives.

Then we expressed surprise.

IV. Room 203

In 1994 a first-year teacher called Erin Gruwell walked into Room 203 at Woodrow Wilson High School in Long Beach, California.¹⁰

The context is not decoration. This was Long Beach two years after the Rodney King riots, in a school where the students had already been written off by people who had never met them. The class had segregated itself along racial lines. It was organised around loyalties that had nothing whatsoever to do with the curriculum and everything to do with not getting hurt on the way home.

What Gruwell did was made into a film. That had one unfortunate effect, which was to convert a method into a personality, and the story now gets told as though a remarkable young woman cared hard enough and the caring was the intervention. That is not what happened. It also does the work a disservice, because caring is not transferable and what she actually did is.

She made the room safe before she tried to make it educational. Every student got a journal. Nobody would read it, she told them, unless they wanted it read, and that promise was the entire foundation of everything that followed. She put books in their hands about people whose lives resembled theirs far more closely than the syllabus was willing to admit. Then she built the whole year around a single repeated practice: tell the truth about yourself in writing, and have it received rather than judged.

They wrote. Then they read each other. Then, slowly, the room reorganised itself around something other than which group you belonged to.

All hundred and fifty of them graduated.¹¹ Most went on to college. They published the journals as The Freedom Writers Diary, having named themselves after the Freedom Riders of 1961, and the book became a bestseller.

The thing I take from Room 203 is not the outcome. It is the order of operations. Gruwell did not motivate those students into caring about each other and then get better work out of them. She gave them a structured, repeated practice of disclosure and reception, and the relationships changed as a result of doing it. The motivation arrived afterwards. In my experience it always arrives afterwards.

V. The ceiling

The Freedom Writers Foundation exists because those students understood what they had been handed and wanted it to reach other people. It has since trained more than eight hundred teachers across all fifty states and twenty countries, and reports reaching over a million children.¹²

Eight hundred teachers. In thirty years.

I mean that as praise. Eight hundred properly trained teachers is a real achievement and a million children is not a rounding error, and the Foundation did the honourable thing, which was to take something that worked and teach it carefully instead of packaging it cheaply and selling it to everybody. That is rarer than it should be.

It is also a ceiling, and it is worth being exact about what kind.

It is not a pedagogy problem. The pedagogy works, demonstrably, and it has been reproduced by hundreds of people who are not Erin Gruwell.

It is a distribution problem. The method travels by training one human being at a time, and training costs time and money and willingness, none of which a school has spare. There are around three and a half million teachers in the United States alone. Run the arithmetic at the rate a training programme can realistically move and it does not arrive anywhere useful inside any of our lifetimes. Not one of them.

Underneath that sits a second constraint, which people are more reluctant to say out loud. Teachers who seek out that training are, almost by definition, teachers who were already going to be good at this. The self-selection is the problem. A child who badly needs a Gruwell is not statistically likely to be sitting in the classroom of somebody who spent their own summer holiday learning to become one.

Which produces the question that everything else in this document is downstream of. What would it take for an ordinary teacher, with no special training, no particular gift for this, and any set of beliefs at all, to run what happened in Room 203? Not a diluted version of it. The actual mechanism.

VI. The mechanism

I said that twenty years in front of strangers left me with a firm opinion. This is it.

It is not credentials. Telling a room who you are and what you have done buys attention and buys distance in the same movement, because the moment you have established that you are impressive you have established that you are separate.

It is not charm either. Charm is a performance of warmth, and people detect the gap between warmth and its performance almost instantly, usually without being able to say how. The detection is not generous.

What works is going first. Saying one small, true, faintly uncomfortable thing about yourself before anybody has asked you to.

Not a confession. Not a wound. Something ordinary that you would normally leave out. The technical term is personal disclosure, and in my experience it is the fastest reliable route into another person that exists. Faster than credentials, faster than charm, and it is not close.

The reason is asymmetry. Whoever discloses first takes a small risk in front of somebody who has not yet earned it, and that risk is the entire gift. It says, without a word of it being spoken, that this is a place where you do not have to be careful. The ordinary human response to being handed that is to hand something back.

This is not folk wisdom. Reciprocal, escalating disclosure between strangers has been studied experimentally for decades and the finding holds: structured mutual disclosure produces measurable closeness inside a single short session, between people who walked in as strangers.¹³ Sustained mutual gaze does related work.¹⁴ The mechanisms are not mysterious. They are just not taught.

I should admit the other half of my evidence, which is that I get it wrong regularly. Three hundred nights a year means several hundred nights a year of reading a room correctly and some number of nights of reading it completely wrong. A room that looks closed is sometimes just cold. A man in the third row with his arms folded is occasionally the warmest person in the building and has simply been sitting on a coach since six. You learn more from those nights than from the ones that go well, and what you learn is that you cannot tell from the outside who is guarded and who is merely tired. You have to give them something first and watch what comes back. There is no version where you work it out by observation alone, which is inconvenient, because observation is the part that looks like a skill.

Now hold that against what we do in schools, which is almost nothing.

We are excellent at naming the outcome. Be kind. Include people. Look after each other. Every school I have set foot in has some version of those words on a wall, usually in a font chosen by an adult. What we almost never do is teach the move. We describe the destination in considerable detail and leave the mechanism entirely to chance, which means the children who arrive are overwhelmingly the ones who were going to arrive anyway. The ones who had it modelled at home. The ones who won the same lottery I won.

Everyone else is told to be kind without ever being shown how, and then judged for failing to be.

VII. What it looks like when the system carries the method

I built something. I am going to describe it plainly, including the unglamorous parts, because a document like this is worthless if the mechanism stays vague — and vagueness at exactly this point is why teachers do not believe documents about social and emotional learning. They are right not to.

It runs on one screen for the whole class. A projector, or the smart board already on the wall. Not thirty devices. One.

Planning a session takes about two minutes. The teacher picks either a morning ritual, which is three fixed questions — a look back, a look forward, and one that holds both — identical every day and never written by a machine, or a custom session, where she sets a focus and the system proposes questions balanced across the five CASEL competencies.¹⁵

Then a screen appears that no student ever sees. Every proposed question, listed, for her to approve, edit or throw out. Nothing reaches a child that an adult has not read first. This is not a setting that can be switched off; it is the shape of the thing.

Then the class starts, and here is where my other life shows up.

Two students stand. They can stay where they are or come to the front, depending on what the room is used to. Four colours appear on the screen. One student picks one in secret while the other looks away. Then they turn back and hold eye contact for three seconds while trying to guess which colour was chosen. The screen says, in words I spent an unreasonable amount of time on: three seconds of eye contact, it's okay to laugh.

Then the guess is revealed, and this is the design decision I am most pleased with.

There is no winner. The screen says same choice or different choices, and then it says who goes first. Guess correctly and you go second. Guess wrong and you go first. Both students answer the same question either way, so the guess determines nothing except the order, and the cost of losing a coin-flip you could not influence is that you have to be brave slightly sooner than the other person.

No child has to volunteer and no child gets to hide. The hardest social problem in any classroom, which is who goes first, has been solved in three seconds by a colour game, and no adult had to single anybody out to do it.

The questions are real. One from the system, verbatim: Do you think everyone puts on a slightly different mask depending on who's watching, even when they don't mean to? They escalate across sessions rather than within them. The first session is favourite colours and easy wins. What a fourteen-year-old is asked in March is not what they were asked in September.

The rest of the class is not an audience. They are practising the other half of the skill, which is receiving. When both students have finished, the room gives two quiet taps on the table. Not applause. Nobody is performed at.

At the end there is a QR code and five questions on the students' own phones, one per competency, worded differently for younger and older children. I listened when my partner was talking. I understood how my partner was feeling. I thought about my answer before I said it. And then one more, which is the important one: how was it working with that person today? Underneath, in small type: only your teacher sees this.

That last question, asked after every pairing, month after month, is a sociometric measure. It tells you whether the relationships in a room are changing, as opposed to whether a lesson was delivered. It is the reason there are any numbers in Section VIII at all.

Two more things and then I will stop describing software.

There are no cameras. No microphones, no emotion recognition, no attention tracking, no sensors of any kind. That is not a policy I could quietly reverse later; it is written into the patent as a claim.¹⁶ Everything the system knows, it knows because a person tapped something. A large part of this industry is moving in the opposite direction and I think they will regret it.

And when a teacher opens the class-dynamics view, the software stops and asks her to confirm the projector is off before it will show her anything. A fourteen-year-old should never accidentally see how the room rated working with them. That screen exists because I sat and thought about what it would do to that particular child, and I would rather build the interruption than trust that it never happens.

VIII. What we know, and what we do not

Between October and December 2025, Svenstrup School in Aalborg, Denmark ran a controlled six-week study across 800 students and 40 teachers, with intervention and control groups.¹⁷ I was the principal investigator. The school's principal, its wellbeing coordinator and its educational psychology counsellor are named on it alongside me.

Class cohesion and relational safety rose 29.0% in the intervention group, from 3.1 to 4.0 on a five-point scale, against 3.1% in control. Engagement rose 30.0%, from 3.0 to 3.9, against 3.3%. Unexcused absences fell 18%, against a 2% rise in control. Teachers reported saving up to 4.3 hours a week with daily use. The longest-running cohort finished at 4.2 on cohesion, 35.5% above where it started.

I am now going to argue against my own study, because if I do not, somebody else will and they will do it less carefully.

It was one school, so generalisability is limited. It ran six weeks, which tells you nothing about whether the effect lasts. Most outcomes were self-reported, with everything that implies. Classes were not randomly assigned, which admits selection bias. Participants knew they were being studied, so some portion of what we measured is the Hawthorne effect in a lab coat. And the principal investigator is employed by the company whose product was under evaluation. That is me. It is a genuine conflict of interest and it is disclosed in the paper.

All six of those limitations are printed on the page before the conclusion, because a pilot that conceals its own weaknesses is not a pilot. It is an advertisement with a methods section.

So what I will actually claim is small. One school, six weeks, eight hundred children: the relationships in the rooms that ran this changed measurably more than the relationships in the rooms that did not, and fewer children stayed away. That is a signal, not a proof. Larger studies with random assignment and longer follow-up are needed, they have not been done, and anybody telling you otherwise about any product in this category, mine included, is selling you something.

I would ask you to hold the absence figure next to Section II, though. The loudest crisis in American schooling right now is that children are not turning up.¹⁸ Something that moved unexcused absence eighteen percent in the right direction, over six weeks, by doing nothing more than getting children to talk to one another on purpose for ten minutes a day, seems to me worth a second look even from a sceptic. Especially from a sceptic.

IX. The middle ground

Which brings me back to the screens, and to the thing I said in Section III that I would explain later.

The ban-the-phones movement is right about the harm and wrong about the remedy. You cannot raise children for a world that has stopped existing, and you cannot hand somebody a device on the first morning of their adult life having given them no practice with it. Prohibition is not a pedagogy. It is a delay, and it is usually a delay that suits the adults more than the children.

Artificial intelligence is the most significant tool we have built since the internet and possibly since the printing press. Refusing it in education would be an act of historic self-harm.

The question was never whether. It was how many, and pointed where.

One device per child, pointed at the child, produced what we now have. One device per classroom, pointed at the space between children, is a different technology built from identical components. In the system I have described the AI never addresses a student. It has no interface a child can reach. It drafts a question, an adult approves it, and then it gets out of the way while two human beings look at each other. That is the whole of its job and it is enough.

That is not a compromise position. It is the technology doing the one thing it is genuinely good at, in service of the one thing only people can do.

While I am here, a small idea I have no plans for and would be delighted to see stolen. Schools across the developed world are about to find themselves holding an enormous surplus of tablets and Chromebooks they no longer want in children's hands. Somebody should start a foundation that collects them and converts them into one good screen per classroom, anywhere with a wall and no budget. The hardware already exists. It is sitting in cupboards.

X. What I believe

Everything above this line I can source. What follows I cannot, and I would rather mark the join clearly than let the footnotes lend it authority it has not earned.

Connection is a skill, not a virtue

Two hundred years ago, reading was something a minority did, and the majority were widely assumed to be incapable of it. Not unwilling. Incapable. The argument was not really about schools or money. It was about whether ordinary people possessed the equipment, and a great many serious men concluded that they did not.

We find that argument embarrassing now. We find it embarrassing because somebody tried the other thing, at scale, for long enough to settle it, and it turned out that reading was a skill. Skills are what happens when you take an activity that looks like a gift and break it into parts small enough to practise. Then you practise them, badly, on a timetable, for years, and at the end of it a person who was not born a reader can read.

I think connection is sitting exactly where literacy sat.

We have filed it under character, next to honesty and courage, and character is the drawer we use for things we admire and cannot teach. So we praise it in assembly and we notice which children have it, and we quietly assume that the ones who do not are made differently. Warmer children and colder children. Sociable ones and difficult ones. We say it in staff rooms with real affection and no malice at all, and it is the same sentence as some people simply do not have the mind for letters.

Break it into parts. Look at someone for three seconds. Say a true small thing first. Receive what comes back without fixing it. Ask one more question than feels comfortable. Those are not personality traits. They are moves, and a nine-year-old can learn them the way a nine-year-old learns anything, which is badly at first and then not.

A great teacher should not be the requirement

I am not saying teachers do not matter. I am saying that a system which only functions when you locate an exceptional individual has failed at the thing systems exist to do.

We do not require every pilot to be Chesley Sullenberger. We build a cockpit in which an ordinary competent pilot is safe on a bad day, and the reason we do that is not disrespect for pilots. It is respect for passengers, who do not get to choose.

Children do not get to choose either. Gruwell proved the method; standing on her shoulders is not a betrayal of what she did. It is the only respectful response available, because the alternative is asking each generation to produce its own Gruwell and quietly writing off the children who do not get one.

It has to work regardless of what the teacher believes

Social and emotional learning has become politically contested. In parts of the United States it is now restricted, and in some places banned outright.

I have watched that fight from a distance, from a country where it is not really a fight, and what strikes me is how much of it is about the term rather than the practice. People who would object furiously to a curriculum are entirely happy for two children to look at each other and answer a question about what they did at the weekend. The objection attaches to the label and the apparatus, and the label is not the thing.

Nothing I have described requires a teacher to hold any particular view about anything at all. She approves a question. Two children look at each other. One of them speaks first because a colour game said so.

A method that requires ideological agreement in order to work will only ever reach the classrooms that already agreed. Those are not the classrooms that need it. If I have to choose between a method that is philosophically pure and one that gets into a room in a state where the word itself is banned, I will take the room every time, and I do not think that is a compromise. I think that is the entire job.

Children should be asked to be vulnerable at school, deliberately

I know the objection, and it is a serious one made by serious people: that this belongs to families, and school should teach mathematics.

Here is my answer, and it is the only place in this document where I am not really arguing.

Some families cannot do it. Not because they are bad families. Because a parent can be ill, or working three jobs, or drinking, or absent, or simply underwater and holding on.

My mother was not a bad mother. She was ill, and there were years when she could not carry that part of raising me, and I was nine and had no idea that was what was happening. I could not rely on her to teach me how to be close to people. I say that with no anger whatsoever. It is just what was true, and it was true through no fault of hers.

School is the only institution that gets every child, whichever house they went home to. That is not a criticism of families. It is an observation about coverage. If we leave this to homes, we are not choosing to trust parents. We are choosing which children get it, and we are choosing them by accident of birth.

You can measure belonging without surveillance

Nobody needs a camera pointed at a child's face. One friendly question after a shared experience, asked repeatedly, kept private, read next to the judgement of an adult who was actually in the room, will tell you more than any sensor ever will. The industry is about to spend a great deal of money learning this the expensive way.

What twenty years actually looks like

This is the part I think about most, and it is the part I am least able to prove.

A child who practises this from seven to sixteen will have done it something like fifteen hundred times. They will not know a fact about connection. They will have a reflex. When somebody in front of them goes quiet, the move will already be happening before they have decided to make it, the way you catch a glass without deciding to catch it.

Now consider what that person is, twenty years later. They are a colleague. They are somebody's manager, which means they are in charge of whether forty adults feel able to say when something is wrong. They are a neighbour. Many of them are parents, and the house they build is not the house they were raised in, because they have a thing to teach their own children that nobody had to teach them on purpose.

And some of them are teachers. That is the loop closing. The generation that learned this becomes the generation that transmits it, and at that point it stops being a programme and becomes something closer to a norm, in the way that washing your hands before surgery was once a controversial proposal by an unpopular Hungarian and is now simply what a person does.

Here is the thing that makes the arithmetic strange. Connection is not an individual attribute. It is not like height, where my having it does nothing for you. It takes two people, which means that one person in a room who can do it well changes the experience of everyone who cannot. A single child in a class of thirty who reliably asks the second question makes that class measurably less lonely for children who have never practised anything.

So it does not scale in a line. A hundred children who can do this are not a hundred children better off. They are however many rooms those hundred children walk into for the rest of their lives, and the effect leaks sideways into people who never sat in one of these sessions and never will, because they happened to be sitting next to somebody who did.

Which means there is a threshold somewhere. Not a market share. A threshold, past which the ordinary background texture of a room changes and stays changed. I do not know where it sits. I am fairly sure of one thing about it, which is that it is not reached by a company growing at the rate companies grow.

The ambition is every child

So let me say the actual ambition, and be exact about the unit, because I have watched people mishear it.

Every school. Every child. Not as a figure of speech, and not as a market.

Section V called a million children a ceiling. I meant it, and I want to be clear about which part was the ceiling. It was never the number. It was that the number had to be carried one trained human being at a time, which sets a rate, and the rate is the thing that does not arrive inside a lifetime.

I am aware of how every child reads coming from a man with one pilot study. So notice what kind of number it is. It is not a forecast. It is not a target I have any means of hitting, and there is no version of it in which I am the one who gets there. A company cannot reach every child. No company ever has. The things that have actually reached every child stopped belonging to anybody a long time ago — the alphabet, the times table, washing your hands, looking both ways before you cross.

Which is the real reason the method is published under a licence rather than sold, and I would rather give you the unsentimental version. It is not generosity. It is arithmetic. I cannot name a number that size while also holding on to the thing. Letting go of it is the only move that makes the sentence honest, and it is why I can write every child without embarrassment when I could not write a user forecast without lying.

And consider the counterfactual honestly, because it is not neutral. On the present trajectory, in twenty years we will have several hundred million adults who were never once required to practise this. Not unwilling. Unpractised. Capable of enormous feeling and with no idea how to open the conversation in which it might be said. I do not think that produces a cruel society. I think it produces a lonely one, full of people who would have helped each other and could not work out how to start.

And the belief I cannot support at all

If children grew up practising this, enough of them and in enough places, I think fewer of them would die.

I cannot prove that. A six-week cohesion study cannot carry a mortality claim and I am not going to dress one up as though it can. What I have is a chain of reasoning where each individual link is evidenced and the whole of it is not.

Loneliness carries mortality risk on the scale of heavy smoking.¹⁹ Adolescent isolation and hopelessness sit at levels that alarm public health authorities.²⁰ The capacity to say a true thing to another person and have it received is the mechanism by which isolation ends. That capacity is teachable.

If those four things are true, the conclusion follows. I think they are true.

I want to be careful here, because this is the sentence that could turn this document into something I would be ashamed of. I am not claiming that a classroom exercise prevents suicide. The causes of a young person's death are specific and multiple and mostly beyond the reach of anything I will ever build, and any founder who implies otherwise is doing something obscene with other people's grief.

What I am claiming is narrower and I think defensible. A child who has practised, fifteen hundred times, saying a true thing out loud to another child and having it received without disaster, is a child who has been given a piece of equipment. Whether they ever use it in the worst hour of their life, I cannot say. But they will have it, and right now we are handing it out by accident, to the ones who happen to walk through the right door.

That is the whole of my case. I have not proven it, and I am telling you so rather than hoping you will not notice.

XI. What I am giving away

Alongside this document there is a second one, called The Connection Protocol.

It is the whole method. The round, the timing, the three seconds of eye contact, the rule about who goes first and why losing is the better half of it, the escalation across sessions, the safety constraints that are not optional, the survey and what it does and does not measure. It is written plainly enough that a teacher can run it tomorrow morning with four coloured cards and a printed sheet, and specified precisely enough that if you hand it to whatever artificial intelligence your school already pays for, you will have working software by the end of the week.

It is published under a licence that lets anyone use it, adapt it, translate it, teach it and build on it, on one condition: whatever you make from it has to stay as open as what I gave you. Nobody gets to take it, improve it, and then close it again. That is the only string attached, and it is there to protect the next person rather than to protect me.

You do not need my permission. You do not need to tell me you are doing it. You do not need anything I sell.

If you teach, that is the whole invitation. Take one class. Ten minutes a day. You will know inside three weeks whether the room has changed, and you will not need me to tell you, because you will be the person standing in it.

If you do not teach, the ask is smaller. Send this to somebody who does.

And if you think I have this wrong, I would like to hear why. I have been wrong about parts of it before, more than once, and every correction so far arrived from a teacher rather than from anybody who builds software.

XII.

I would like this to exist whether or not I am the one who builds it.

That is not modesty. I have a company and a patent and I would very much like both to work, and writing the method down and giving it away makes selling it harder rather than easier. I am aware of that. I did it anyway, and I would rather explain why than let it be mistaken for generosity.

I have just spent seven thousand words arguing that any system which depends on an exceptional individual has already failed at the thing systems are for. It would be absurd to write that and then quietly make myself the exception. To design a method that does not require a remarkable teacher, and then hang the entire thing on one Danish man with a patent, a company, and a finite number of years in him.

So I have not sat here hoping somebody would eventually find this. I wrote the instructions down, all of them, and licensed them so that nobody can close them again. If I build it at scale, good. If the company fails, or I lose my nerve, or I am hit by a bus on the way to a show in some city I have not been to yet, the method does not go with me. It is already out. Somebody in a school I will never visit can pick it up on a Tuesday morning and run it without ever learning my name, and that is not the consolation prize. That was always the point.

The children a mile from my school got nothing, and the reason was a door their parents happened to walk through. Some head teacher I cannot name won an argument I was not present for, and I got a life out of it.

It does not have to be luck. That is the whole of what I have been trying to say.

Notes

Further reading underpinning the approach: Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Schellinger, K. B. (2011). The impact of enhancing students' social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82(1), 405–432. · Blum, R. W., & Libbey, H. P. (2004). School connectedness. Journal of School Health, 74(7), 231–232. · Moos, R. H., & Trickett, E. J. (1974). Classroom Environment Scale Manual.

Jesper K. Lassen is a mentalist who performs roughly three hundred nights a year and was the subject of a three-part documentary series on Danish national television. He is the founder of VEN Method (formerly DeepConexion), which builds CASEL-aligned social-emotional learning for K–12 schools. He grew up in Aalborg, Denmark.