Beyond Turing
Every serious journey, if it is honest, brings the traveller back home. This book set out from a map and a method; it has crossed the body, the home, the enterprise, the city and the sky; and now it returns to where every one of its chapters was, in truth, headed from the start: the desk of the person who, on Monday morning, has to sign. The designer: the engineer who draws the system, the manager who coordinates the project, the executive who answers for the result. It is the craft of those who do not have the luxury of questions left open, because at a certain hour of the morning questions become decisions, decisions become signatures, and signatures have a name above them. The two languages of the Preface - the language of technique and the language of meaning - here stop being a theme and become working hours.
There is a precise reason why the book closes on this figure and not on another. The preceding pages have demonstrated, territory after territory, what systems cannot do: they cannot judge, they cannot understand, they cannot answer. But every "cannot" leaves the real question open: then who? The answer is always the same, and it is a person with a role: someone who designs, coordinates and answers. This chapter is dedicated to that someone - to what is beautiful in this craft, to what is hard in it, and to the charter this book hands over to those who practise it.
1. Whoever coordinates answers: the responsibility that exceeds the fault
The international standards of the profession describe the project manager with a definition that sounds administrative and is, on closer inspection, almost dramatic: the person who leads the team, integrates the work of the parts and answers for the achievement of the objectives (ISO 21502). Answers: not "contributes", not "supervises" - answers. The organisation of complex projects has always rested on this deliberate architecture: so that responsibilities do not dissipate along the chain of delegations, there must exist somewhere a single point where they converge - a human being who, when something goes wrong, can be called by name. The reader recognises the design: it is the same logic that the entire law of the algorithmic era has chosen for intelligent systems - the human oversight of Article 14, the human review of the platform work directive - applied, for a century, to the enterprise of man upon man.
But this architecture has a price, and those who practise this craft know it by heart: the coordinator's responsibility systematically exceeds his fault. The supplier who is late, the requirement the client changes midway, the mistake of a tired collaborator, the market that shifts: none of this is his doing, and all of it ends up on his desk, because the role exists precisely to absorb what no contract had foreseen. The chapter on guilt and the algorithm taught the distinction that here reaches its fulfilment: fault is judged, responsibility is assumed. The designer is the man who permanently inhabits the gap between the two - he assumes more than can be imputed to him, he answers for more than he has committed. And in this gap hides one of the most elegant proofs of the book's thesis: one can train a model to optimise any function, but no loss function exists that represents the act of taking another's error upon oneself. The machine can calculate the consequences; shouldering them is a gesture that presupposes a subject - free, exposed, with something to lose. AI functions, and for this reason it does not answer. Man answers, and for this reason there exists the craft this chapter speaks of.
From here, and not from bureaucracy, rigour is born. Whoever answers has the scientific duty to be able to show how he decided: the risks surveyed and weighed, the estimates with their margins, the alternatives considered, the decisions traced. The reader recognises the second thesis of the method - validation is procedural - turned into daily discipline: project documentation is not paperwork, it is the reasoning of the judgement the designer pronounces every day by signing. Scientific rigour, for the one who coordinates, is not an academic ornament: it is the first form of his honesty.
2. The five lenses on the designer's desk
The chapter on the map and the method presented the five lenses as an epistemology; those who direct projects recognise them as the almost embarrassingly precise description of a working day. The morning opens with the lens of engineering: the technical review, the architecture, yesterday's failure and tomorrow's acceptance test. Then comes the lens of the exact sciences: the numbers of the plan - estimates, variances, probabilities - with the professional obligation never to present a figure without its uncertainty. At midday arrives the legal lens: the contract, compliance, the question "who answers?" asked before the court answers it. And within every meeting, if the designer is a true one, the philosophical lens is at work, which is in the end the question no minutes ever record: what are we really building, and for whom?
Then the day ends, and the fifth lens remains. The people remain: the struggling collaborator whom the numbers do not explain, the conflict no process resolves, the choice that tomorrow will touch a life and not just a milestone. The chapter on the method closed with a sentence that here finds its natural addressee: four lenses interrogate, the last one keeps. Here is the full name of the craft. Whoever coordinates is not merely the technical integrator of the project: he is the keeper of what the project contains of the human - the people who make it and the people for whom it is destined. The first four lenses are learned in courses and manuals; the fifth is learned only by carrying it.
3. The wound and the courage: raising people, being wounded, moving on
Why then choose this craft, if the bill of responsibility is always higher than what is owed? Those who have done it for a lifetime know the answer, and it is the part of the job no specification ever lists among the deliverables: making people grow. The true designer is a workshop master: he passes on the part of the craft that is not in the procedures - the tacit knowledge of the chapter on work - teaching at someone's side, correcting without humiliating, entrusting before the other feels ready. To watch a collaborator become, over the years, better than you found him - and perhaps better than you - is the remuneration that never appears on a payslip, and it is the closest thing, in the whole world of work, to the experience of being a father or a mother. It is no accident that it resembles the scene from which this book was born, and which the chapter on the child and the screen reopened: the designer too, when he is a master, accompanies someone towards what they do not yet know they can do - and hands them a key.
But precisely here the craft keeps its wound, and an honest book does not pass over it in silence: whoever makes someone grow hands them, in the same gesture, the power to wound him. It is a law as old as keeping itself, and the canon of our civilisation fixed it once and for all in the most retold story in the world: the Master chose twelve men, formed them for years, carried them higher than where he had found them - and on the night when everything was being decided, one handed him over to his enemies and the most solid of them all, the one to whom he had promised the keys, swore three times that he did not know him. The Gospel records a single detail, the most human of all: the Lord turned and looked at Peter (Lk 22:61). Anyone who has coordinated people for a lifetime knows a small version of that look: the signature that suddenly goes missing, the loyalty that changes sides when convenience changes, the wound that comes - almost always - not from adversaries, but from the very people one had raised the most. It is not an anomaly of the craft: it is its price. Keeping exposes, by definition.
What distinguishes the keeper, then, is not the wound, which comes to everyone: it is what he does afterwards. And here too the canonical story points the way, because it does not end on the night of the denial: it ends on a shore, in the morning, where the Risen One asks three times - once for each denial - "do you love me?" (Jn 21:15-17), and to that man who had betrayed him he hands back the entire flock. Forgiveness, in that scene, is not a feeling: it is an act of governance - the decision to reinstate, to entrust again, to start over. Translated into the language of this chapter: blame no one, because the chapter on guilt has demonstrated that all of us - those who lead and those who are led - are exposed to conditionings we do not choose, and because permanent suspicion is the end of all keeping; forgive as an organisational choice, even before a spiritual one, because no working group survives the bookkeeping of wrongs; and move on, with courage, without fear of tomorrow. The man who opened his service to the world crying "Be not afraid!" (22 October 1978) is the same man who would write, three years later, that work is for man and not man for work: the two sentences belong together, and together they are the keeper's inheritance. Science gives him the tools to decide; faith - for those who believe, and the writer is among them - gives the reason to get up again when the decision has cost dearly; and for those who do not believe, its secular name is trust: in oneself, in others despite everything, in tomorrow. The designer who has stopped being afraid is not the one to whom nothing has ever happened: he is the one to whom it happened, and who went on building.
4. The keeper's charter: ten invariants from the first commit
There remains the book's last promise to keep: to gather in a single place the design invariants the chapters have demonstrated one by one, so that those who design can bring them into the meeting room. Here they are, ordered under the five lenses - not a decalogue of good intentions, but a verifiable checklist, from the first commit to the final acceptance test.
Under the lens of engineering. First: traceability is non-negotiable - every automatic decision that touches a person must be retraceable, from the source datum to the threshold that fired, as the book has seen for pipelines and for scores. Second: the threshold is a decision, not a detail - it is declared, instrumented, its drift monitored, and whoever sets it answers for it.
Under the lens of the exact sciences. Third: never a number without its uncertainty - the score is a statement about a class, the sensor has its measurement error, and presenting them bare is already a dishonesty. Fourth: total automatic control is impossible by theorem, not through the immaturity of the technology - therefore every serious architecture is born already designed for human review, because mathematics itself leaves no alternative.
Under the lens of jurisprudence. Fifth: human oversight must be substantive - a person with the time, the competence and the authority to say no, not a ceremonial signature at the end of the process. Sixth: every decision that affects a person must have a reachable human interlocutor and a possible review. Seventh: personal data must be able to expire - oblivion is not a concession, it is a function to be designed, because an archive that does not forget is a sentence no judge has pronounced.
Under the lens of philosophy. Eighth: the person exceeds the profile - the model of the user is not the user, the map is not the territory, and every system that forgets this difference is already mistreating someone.
Under the lens that keeps. Ninth: whoever holds data or powers on behalf of another holds them in keeping - and answers for them as for something not his own, according to the category that the chapter on the brain restored to design. Tenth: let the system respect rest and return - the unmeasured time in which man is not a user, and the possibility, always, that a person may change and never again coincide with their record.
Ten lines, and not one of them is free: each one costs storage, latency, processes, difficult meetings. That is why the charter needs the keeper, and not a pious exhortation: because someone must defend it precisely at the moment when it costs - when the release is pressing, the budget is tightening and the tenth invariant looks like the first one to cut. The keeper is the one who, in that meeting, does not cut it.
5. The keeper's remainder
At the end of all the pages, one last object remains on the designer's desk that none of the five lenses exhausts, and it is the designer himself. He too - he above all - is what this book has defended in every chapter: a human being who does not coincide with his function, who is worth something before his results, who can err and begin again. The craft asks him for rigour, the charter asks him for firmness, the wound asks him for courage; but none of these demands is his definition. A project, in the end, is delivered and closed. A person is not: a person is kept - beginning with oneself.
And here the chapter hands the book over to the Epilogue, which will unfold in full the difference from which everything began. To the reader who designs, coordinates and answers - to the engineer, the executive, the project manager, the workshop master - these pages leave a single commission, one that holds together the charter and the wound, science and faith, rigour and forgiveness: build systems that know when to stop, and be a man who knows why to start again. Write in the margins of every one of your projects, like the maps of this book, hic est homo. And of tomorrow, simply, be not afraid: because tomorrow is exactly the territory - alive, free, incalculable - for which designing is worthwhile.
References cited in the text: ISO 21502:2020, Project, programme and portfolio management - Guidance on project management; Regulation (EU) 2024/1689, Article 14; Directive (EU) 2024/2831; Gospel of Luke 22:61; Gospel of John 21:15-17; John Paul II, homily for the inauguration of his pontificate, 22 October 1978; John Paul II, Laborem exercens (1981).