Five problems, one limit.
Almost everything said to be wrong with school is treated as its own separate fault, with its own separate committee. It isn't. Five problems sit on top of a single hard limit — the limit of one human brain — and one move reaches all of them. This is the preview. Each problem, and the solution, gets its own detailed piece after this.
Start with two small scenes, because the first two problems are easier to feel before they're explained.
He's ten, in the third row of a class of thirty-one, in a town the good teachers left. He didn't catch fractions the first time; the class moved on; nobody noticed. By spring the gap has hardened, and he's quietly decided he's “just not clever.” He's wrong — but the exam he'll sit in a few years won't know that. It will mostly record where he was born. Three streets over, a child with a specialist teacher and a parent free in the evenings is about to be measured as though they started in the same place.
She's marked twenty-seven essays tonight, after five lessons and two phone calls and a colleague's class she covered. The first essay got a careful paragraph. She looks at what she's just written on the twenty-eighth: a tick, and the word good. She isn't lazy or unkind — she's a person whose brain has run out of the careful attention it started the evening with. The child who wrote number twenty-eight worked three evenings on it, and will learn nothing from a tick.
Neither of those is anyone's villain. That's the whole point. Now the other three, which are less about a single face and more about a whole system quietly failing.
The five, briefly
Where a child lives sets the ceiling on the teaching they'll ever get. Good teachers flow toward better pay and conditions and stay there; poorer and remote places are left with whoever remains — the newest, the least supported, sometimes the weakest. It isn't that a poor area lacks good teaching by accident. The system actively drains it upward, toward the money.
Even with a post filled, a human teacher is luck of the draw. Experienced or brand-new, gifted or struggling, patient or worn thin by a pile of grading, having a good day or a bad one — a child's result rides on which human they happen to get, and on all the ordinary human things that come with being a person under pressure. No child chose their teacher, yet the teacher is half their outcome.
This one is colder, and bigger than any one classroom. Every child who could have been an engineer, a nurse, a builder of things, and never finds out — because the door shut before they reached it — is not just a sad story. Multiplied across a population, it's a country throwing away, year after year, the exact skilled people it will later stand around complaining it doesn't have. Undeveloped talent is a national loss, not a private one.
The economy needs competent people faster than school and training can produce them, and in the shapes industry actually needs — which keep changing. Retraining an adult is slow, costly, and reaches too few, because it leans on the same scarce teachers already stretched thin. So shortages everyone can see go unmet for years, while capable people wait for a place that never opens. The pipeline is too slow, too narrow, and too poorly aimed at demand.
Children spend years drilled on detail most will never use again, and that drill crowds out what would genuinely serve them: languages at the age they come almost for free, the useful basics of a real life, a few things mastered properly instead of a hundred skimmed. We cover everything, master nothing, and call the coverage an education — and an employer pays to re-teach the rest later.
The one limit underneath all five
Read them back to back and the single thing underneath surfaces. It isn't laziness, or bad teachers, or too little money — though money doesn't help. It's that we built an entire society's schooling on top of one human brain, and asked that brain to do more than any brain can.
One brain has a fixed amount of attention, and it runs out — that's the twenty-eighth essay getting a tick. One brain can only be in one place, so it must be rationed — and rationing is always thinnest where the money is short and the classes are largest. One brain carries moods, fatigue, and quiet assumptions it never asked for — so its fairness wobbles from child to child and day to day. And one brain can hold only so much, teach only so fast, and be cloned only by training another whole human over years — which is why skilling a population is slow.
Every one of the five is that same ceiling, seen from a different side. The problems aren't five. They're one limit, wearing five faces.
When the limit is one, the useful move is one.
The move — a futurist reform
For most of history there was no way around the limit: the teacher was the only copy of the knowledge in the room, so one brain had to do everything, and we designed school around that necessity. That necessity is gone. We can now lift the parts that were only ever chores for a brain off the human, and leave the human the parts that actually need a human.
Concretely, three things stop depending on one overloaded brain.
The clearest lesson a subject has — taught once by a genuinely excellent teacher — can reach every child identically. The remote town gets the same lesson as the capital. The explanation no longer chases a salary, so the drain out of poor areas stops mattering.
Vision at scaleA machine can grade and correct instantly, with the same care for the twenty-eighth piece as the first, and without any idea of a child's name, accent, or yesterday's behaviour — so it can't tire, can't play favourites, and can't quietly decide a child isn't worth the minute. Not a better judge than a great teacher on a good day. A consistent one, which is what the child at the bottom has never had.
Proven · runs todayThe same engine teaches and assesses at scale, repeatable for life — so a population can be skilled, and re-skilled as industry shifts, without first finding and paying a specialist for every course.
Vision at scaleAnd the human, freed from the tireless chores, is handed back the part only a person can do: to encourage, to steady, and above all to notice — to catch the boy in October, not find the wreckage in spring.
What's real, and what's a promise
Almost all of the above is a proposal, and a proposal is only worth the test it invites. One piece, though, already runs: a working platform today grades a learner's writing in seconds — a score on each criterion, every error corrected in their own text with the rule named, and the exact next step to improve — the feedback no tired human can give a whole class at eleven at night. That engine is proven. The rest — the classroom, the whole model inside a school — is vision, and every step would have to be proven again, with evidence.
So the ask is small and reversible, and it has a concrete first form: a single pilot in adult language learning, A1 to C1 — one semester, one cohort of motivated learners, measured against a conventionally-taught control group sitting the same CEFR exam. Measure cost and time-to-level, but above all the equity of outcomes — whether the learners with the least gain the most or get left further behind, reported either way it comes out. Only one thing is proven going in: the criteria-based grading engine already running today; everything built around it is what the pilot exists to test. The concrete pilot is set out in full here →
This is the preview — the shape of the whole thing. Each part now gets its own detailed piece, in plain language: the poor area, told from the child's point of view; the teacher, shown as human rather than villain — the ordinary human failings that hit every classroom; the wasted human potential we throw away; the too-slow skilling of a population against what industry needs; the wrong subjects at the wrong depth; and finally the solution, in full. The concept and the working language course are in the documents below, in English and German. A demo and a conversation are available on request.
Written from a working platform and a concept put up to be tested — not from a result already claimed. — Crosshire.