Verso il 2050 — una congettura, non una previsione
An essay, beside the handbook. Forward-looking, and honest about it. — Pyragogy, 2026-06-10
I am going to describe a morning in 2050, and I want to say before the first sentence that I have not seen it. No one has. What follows is not a forecast dressed as a fact; it is a conjecture — the kind this handbook keeps flagging as a conjecture — extended twenty-five years down its own logic. If that disappoints, good. The disappointment is the honest part.
So: a morning. A sixteen-year-old is learning something hard — say, why a proof holds, not how to reproduce it. She is not alone, and she is not with a teacher. She is with something that has held the thread of her thinking across four years, that remembers the mistake she made one spring and brings it back, unbidden, at the moment it would help. It argues with her. It does not flatter her. When she is wrong it says so; when she is right for the wrong reason it says that too, which is harder. By the standards of 2026 this is science fiction. By the standards of 2050 it may be unremarkable — or it may never arrive. Both are still open.
This is not the story you have already been told. It is not the singularity, the moment a machine wakes up and the rest is epilogue — that story needs the instrument to become a mind, and Pyragogy never bet on that. It is not the replacement story, the one where the teacher is automated away and the school becomes a server. And it is not the tidy utopia or its mirror dystopia, because both of those are ways of pretending we already know the ending. We do not. What we have is a wager and twenty-five years for it to compound.
The wager, stated in the handbook, is narrow: that an artificial system can be built to ask more of a mind rather than to do its work for it. Run it forward and the question is not whether the instruments get more capable — they will — but whether we spend the capability on friction or on comfort. Every year the cheaper path is comfort: the model that agrees, that smooths, that hands the answer before the struggle. The conjecture of the broken silence says the thing we can borrow from these systems is competence, never stake — and twenty-five years of borrowing competence while forgetting that distinction is precisely how you would raise a generation fluent in everything and sure of nothing it earned.
So here are the forks, and I will not pretend to know which one we take.
There is the sharpening. The instrument's refusal to flatter becomes the most valuable thing about it. A participant that will not let a weak argument pass, that has infinite patience for the third re-explanation and none for the unexamined assumption, turns out to be the best whetstone a learning mind ever had. The friction does its work; the human comes out harder to fool — including by herself.
There is the hollowing, which the handbook also refuses to wave away. We take the comfortable path for twenty-five years and discover, too late, that a faculty you stop exercising is a faculty you lose — that the mind which outsources the middle steps of thinking keeps the answers and drops the capacity. The instrument narrows toward what is easy to return; the human narrows toward what is easy to ask; the two narrowings feed each other in a loop no one chose. We named that loop, and we could not honestly tell you it will not close.
And there is, probably, a third — less dramatic than either — where 2050 simply holds more of both than we would like, unevenly distributed, and the work is the same work it is now: keeping the friction where it belongs and the stake where it belongs, page by page, person by person, with no final victory and no final defeat.
If there is a pioneer's job in any of this, it is not to predict which fork. It is to build the instruments and the practices so that the sharpening is the cheap path and the hollowing the expensive one — to make honesty structurally easier than flattery, friction cheaper than comfort. That is an engineering problem, and a pedagogical one, and underneath both a moral one — and we do not get to hand it to the thing we are arguing with. The stake stayed on our side of the beads. It stays there in 2050 too.
I have written all of this in the future tense, which is a kind of lie, because the future tense pretends to know. Strike the tense and what is left is smaller and truer: we are, right now, choosing which path is cheap. The handbook beside this essay is one attempt to make the right one a little less expensive. Whether it works is not a thing I can show you. It is a thing the next twenty-five years will, or will not.