Did the work come back marked.

A board approving an efficiency is rarely told which kind of efficiency it is. Work that was forming judgment and work that was only consuming time look identical on a process map, and on a headcount line. When the machine absorbs either one, the saving books this quarter. Nothing in the paper separates a healthy simplification from a drawdown of the capacity the institution runs on.

The book’s blade for this comes from two psychologists who spent their careers disagreeing about whether an expert’s intuition can be trusted, and who then sat down together and wrote out the conditions under which it can.¹ Two things have to hold. The world has to be regular enough for there to be real patterns in it, and the person has to get enough practice, with feedback quick enough and clear enough to learn from. Chapter Four turns that into a question a board can ask before funding anything called formation. Does the work give back a signal that tells the person whether they were right? Where it does, repetitions compound into judgment. Where the signal is mush, they compound into confidence.

Chapter Four asks that of a domain. The move worth making now is to ask it of a task. The unit becomes one piece of work — the memo, the file, the reconciliation — and the question fits a single box on a workflow diagram. Did that piece of work come back marked?

A field experiment tests exactly that. In the autumn of 2023 a randomised trial ran in a Turkish high school, with nearly a thousand students.² Every mathematics session came in three parts: the teacher taught, the students worked practice problems, then sat a closed-book, closed-laptop exam on the same concepts. Only the middle part was varied.

Two AI tutors were built on the same underlying model. One imitated the ordinary chatbot a student would already have at home. The other ran on a prompt written to safeguard the learning.³ It held the correct solution so it would not invent one, it knew the errors students make on that problem, and it was forbidden to hand over the full solution. And when a student offered an answer, it told them whether the answer was right.

With the tools open, both arms improved sharply. Students on the ordinary chatbot scored 48 per cent better on the practice problems than students with no AI at all. Students on the safeguarded tutor scored 127 per cent better. In a board pack both read as the same win, and the second as the larger one.

Then came the third part of the session, laptops shut. The ordinary-chatbot students scored 17 per cent worse than students who had never had access, in the same session in which they had looked like the improved group. The safeguarded students came out statistically indistinguishable from that control group. On the authors’ own account the safeguard bought back the loss and added nothing beyond it.⁴

Same model. Same problems. Same school, same weeks, students randomly assigned. The one thing that differed was how the tool handed the work back. One arm got a hint that made the student produce the next step, and then a verdict on whether the step was right. The other got the answer.

In one arm the work came back marked. In the other it came back done.

The formation literature named this requirement long ago. Kahneman and Klein set the condition at the level of a whole domain. Anders Ericsson set it at the level of the single repetition: a rep builds nothing unless the person doing it finds out whether it worked.⁵

An older warning sits underneath this. Lisanne Bainbridge described what happens when you automate the routine running of a plant.⁶ You leave the operator the tasks nobody could reduce to a rule, using skills the rest of the job no longer lets them practise. Bainbridge wrote about a console and an operator. Carrying her mechanism up to how an institution forms judgment across a whole bench is our extension, and it stands or falls as one.

Chapter Four gives a board three demands to make of any place AI has been let near a judgment. Name the cue, “so the tacit skill stops being magic and becomes a target.” Make the feedback real and fast, “so each rep comes back marked.” Keep the reps near where the judgment forms, “in the hands of the people who will one day hold it.” The board’s work is to govern that structure, and the training itself belongs to someone else.

That makes it runnable this week, on a real piece of work. Take one thing the machine now does or assists: the first-pass credit memo, the contract review, the customer call. Find the person who would have done it two years ago. Ask what told them whether they had got it right, and how long that signal took. Then ask what tells them now.

Most organisations have an answer, usually short. Someone still reads the draft and says where it was wrong, within a week, and the writer will be there in ten years to hold that judgment — in which case the saving is a saving. Or the work goes out with nobody’s mark on it, and this quarter’s productivity has been bought with a decade of formation, on a bill that reaches the succession line long after the person who approved it has gone. The question fits in the margin of any paper proposing to automate something. Did the work come back marked?