Author: Alejandro Díaz Garreta

  • When Reality Taught Us How to Think 

    A Journalist Notices Something

    Good journalism often performs an unusual service. It notices historical change before anyone has found the concepts to describe it. This is rarely because journalists are better theorists than philosophers or historians. Quite the opposite. They arrive earlier because they are not searching for theories at all. Instead, they pay attention to ordinary conversations, recurring frustrations and subtle changes in people’s habits. They notice that familiar activities begin to feel different long before anyone can explain why, and only later do those observations acquire broader historical significance.

    That was my reaction when I read Arnaud Leparmentier’s recent reporting from Silicon Valley for Le Monde. On the surface, his article concerns familiar themes: artificial intelligence, the changing labour market, the difficulties facing young graduates and the uncertainty experienced by software engineers wondering how long their expertise will retain its value. Companies reorganising around increasingly capable AI systems have become almost routine subjects of discussion. Yet one detail in the article stayed with me long after I had finished reading.

    Several engineers described subtle changes in the way they approached difficult problems. They found themselves becoming stuck less often than they once had. When they reached an impasse, they were less likely to spend hours wrestling with it alone or to interrupt a colleague for another perspective. Instead, they reformulated a prompt, consulted another model or waited for the next iteration of the software. None of them presented this as a loss. On the contrary, they regarded it as an entirely rational adaptation to more capable tools. If a machine could help solve in minutes what previously demanded an afternoon of work, why insist on preserving the longer route?

    Nothing about those remarks appears especially profound. Had I encountered them in isolation, I might have passed over them without a second thought. Taken together, however, they suggested that something more interesting was taking place. The engineers were not merely describing a faster way of working. They were describing a different relationship with intellectual difficulty itself.

    My first instinct was to interpret this simply as another story about productivity. Every major technology reduces the effort required to perform certain tasks. Calculators transformed arithmetic, search engines changed access to information and GPS altered navigation. Artificial intelligence seemed merely to continue a familiar historical pattern in which the distance between a question and an acceptable answer steadily contracts.

    The more I reflected on Leparmentier’s reporting, however, the less convincing that explanation became. Productivity measures how quickly we arrive at an answer, but it tells us remarkably little about the experience of getting there. What struck me was not that the engineers were finding solutions more quickly, but that they were spending less time inhabiting uncertainty itself. They encountered fewer moments in which their own understanding had genuinely reached its limits, and they relied less often on another person’s perspective when it did. Gradually, I realised that what interested me was no longer the answer at the end of the process but the process that seemed quietly to be disappearing.

    Technology has always changed the way we solve problems. What artificial intelligence may be changing, however, is the way we experience having one. At first, that distinction seemed almost semantic, yet the more examples I considered, the harder it became to dismiss. Throughout most of human history, intellectual work possessed a certain stubbornness. A scientist whose experiment failed could not immediately ask reality for an alternative explanation. A lawyer confronting a novel case had relatively few ways of escaping uncertainty. An architect facing an unexpected structural problem could not instantly generate dozens of plausible alternatives. Progress depended upon remaining with a difficulty long enough for something to change—sometimes the evidence, sometimes the question itself, but often the person trying to understand it.

    No educational institution invented that rhythm, and no profession consciously designed it. It simply reflected the way the world answered our questions: slowly, incompletely and often only after our first explanations had proved inadequate. As I continued thinking about Leparmentier’s article, I found myself wondering whether the most significant consequence of artificial intelligence might not be that it changes our answers, but that it changes our encounters with reality itself.

    I had not intended to think about education. Indeed, I was not thinking about education at all. Not yet.

    The Lessons Reality Used to Give Us

    The more I reflected on this, the more I realised that what fascinated me was not artificial intelligence itself but the kind of world within which previous generations had learned to think. Reality has always resisted us in ways that education never had to invent. A student struggling to understand a mathematical proof, a physician confronted with contradictory symptoms or an engineer unable to explain an unexpected result were not participating in carefully designed learning experiences. They were simply encountering a world that refused to yield its answers on demand.

    That resistance shaped the rhythm of intellectual life. Before arriving at a solution, we often had to exhaust our own understanding, pursue explanations that ultimately proved inadequate and reformulate the question because the original one no longer made sense. Eventually, we turned to another person, not because collaboration was inherently virtuous, but because we had reached the limits of what we could achieve alone. Looking back, it is tempting to describe these episodes as educational, yet they rarely felt that way at the time. They felt frustrating, inefficient and occasionally humiliating. Their educational value became visible only in retrospect, when we recognised that what had mattered was not the failure itself but the sequence of experiences it had made unavoidable.

    Those experiences shared a common characteristic. They forced us to discover that our first explanation was incomplete, to distinguish confidence from understanding and to recognise when another person’s perspective contained something we could not yet see ourselves. None of these lessons had been deliberately taught. They emerged because reality imposed conditions that no teacher had designed and no curriculum had specified.

    It was at this point that I realised I had been asking the wrong question. Until then, I had been wondering what artificial intelligence might change about learning. Gradually, however, another question displaced it. What if many of the experiences through which human beings have learned to think were never educational inventions in the first place?

    The possibility was unsettling because it inverted one of our most familiar assumptions. We tend to think of education as the institution responsible for intellectual formation. Schools design curricula, universities organise courses and teachers create learning activities. Formation therefore appears to flow from intentional educational design. Yet the engineers in Leparmentier’s article suggested a different possibility: perhaps educational institutions have always depended upon something they did not themselves create.

    Consider a student who finally walks into a professor’s office after struggling with an idea for several days. The educational significance of that conversation does not arise because asking for help was written into the syllabus. It arises because the student first encountered a difficulty that could not easily be bypassed. The conversation becomes transformative precisely because reality had already done part of the educational work before the teacher entered the picture.

    The same pattern appears well beyond classrooms. Scientists often remember the experiment that refused to behave as theory predicted, lawyers the case that forced them to abandon an elegant argument, doctors the patient who did not fit any familiar diagnosis and architects the project that exposed a flaw in assumptions they had scarcely realised they were making. Years later, these moments frequently stand out as turning points in professional formation, even though none of them had been designed with education in mind. Their formative power derived from the fact that reality insisted on resisting human expectations.

    This, I began to suspect, is what artificial intelligence may be changing. The technology does not abolish uncertainty; anyone who works seriously with today’s systems knows that they generate new forms of ambiguity of their own. Models hallucinate, disagree with one another and require careful interpretation. Yet they also alter something more fundamental. They reduce the frequency with which reality itself obliges us to remain inside uncertainty before offering a possible way forward. There is almost always another prompt to try, another model to consult or another explanation to consider before we have fully exhausted our own understanding.

    The point is not that these answers are necessarily correct. It is that the experience of moving from not knowing to knowing begins to change. For centuries that interval was largely dictated by the world itself. Increasingly, it is mediated by technologies designed to shorten it.

    When I reached this point in my own thinking, I realised that Leparmentier’s article was no longer simply about software engineers. Without intending to do so, it had led me towards a much older question: whether education had ever fully understood where some of its deepest formative experiences actually came from.

    When Reality Changes

    If many of the experiences through which judgement developed emerged because reality itself imposed them, then artificial intelligence raises a question that extends well beyond education. The issue is no longer simply whether students will rely too heavily on AI or whether professionals will delegate too much cognitive work to increasingly capable systems. The deeper question is what happens when the environment itself ceases to generate, with the same frequency, the experiences through which human beings have traditionally learned to think.

    This is not an argument against technological progress. Every significant technological advance has removed forms of difficulty that previous generations simply accepted as part of ordinary life. We no longer calculate logarithms by hand, spend days searching library catalogues or navigate unfamiliar cities with paper maps, and few would wish to recover those conditions merely because they demanded greater effort. The point is not that difficulty is inherently valuable, but that some forms of difficulty proved unintentionally formative. Until recently, we rarely needed to distinguish between obstacles that merely slowed us down and obstacles that quietly shaped our judgement because both disappeared together. Artificial intelligence forces us to recognise that they are not necessarily the same thing.

    The distinction becomes clearer if we consider the difference between obtaining an answer and arriving at an understanding. An intelligent system may provide an elegant explanation, propose alternative interpretations and even anticipate objections that its user has not yet considered. In many circumstances this is an extraordinary intellectual resource. Yet it also changes the sequence through which understanding develops. Instead of discovering the limits of our own reasoning before encountering a better one, we increasingly encounter sophisticated reasoning before we have fully explored where our own thinking might have failed.

    Education has always depended upon that sequence. Intellectual maturity is rarely produced by exposure to correct answers alone. It develops through the gradual recognition that apparently convincing explanations conceal unnoticed assumptions, that problems first thought to be simple demand different questions, and that genuine understanding often begins with the recognition that one’s own interpretation is inadequate. Such experiences are remarkably difficult to design artificially because they depend upon authentic encounters with resistance rather than the mere transmission of information.

    Seen from this perspective, the challenge posed by artificial intelligence is not primarily one of knowledge but of intellectual formation. Knowledge can often be transferred with remarkable efficiency; formation cannot. It develops through repeated encounters with uncertainty, revision and responsibility, experiences that educational institutions undoubtedly cultivate but that they have never supplied on their own. For centuries, schools and universities operated within a world that continuously generated these formative encounters for them. The classroom did not exist in isolation from reality; it depended upon the epistemic conditions of the wider world.

    This is why so many contemporary debates about AI in education feel oddly incomplete. We devote considerable attention to assessment, academic integrity, personalised tutoring and future skills, all of which are important questions. Yet these discussions often assume that the educational environment itself remains fundamentally unchanged and that the challenge is simply to decide where artificial intelligence should or should not be introduced. If the argument developed here is correct, that assumption can no longer be taken for granted. The environment within which intellectual formation occurs is itself changing, not because schools have been transformed, but because reality is increasingly mediated by systems designed to reduce the very forms of resistance through which judgement has so often developed.

    None of this allows us to predict the future with confidence. Human beings have always learned under changing technological conditions, and every major transformation has produced intellectual virtues that earlier generations could scarcely have imagined. Artificial intelligence may ultimately foster forms of judgement that are different rather than diminished. The mistake would be to assume, however, that because answers remain available, the experiences that once led us towards understanding remain essentially unchanged. What is changing is not merely the destination of thought but the terrain across which thought travels.

    By this stage I realised that I had almost stopped thinking about artificial intelligence itself. Leparmentier’s engineers had not offered a theory, nor were they trying to explain a civilisational transformation. They had simply described small changes in the texture of their working lives. Yet those observations had gradually led me towards a question that reached far beyond software engineering: what happens when reality itself no longer educates us in quite the same way as before?

    The Educational Question

    None of this suggests that education should attempt to recreate every intellectual hardship that previous generations endured. There is little virtue in preserving inefficiency simply because it is familiar, and few would argue that scientists should work more slowly, engineers solve fewer problems or physicians delay diagnoses in the name of intellectual formation. If artificial intelligence enables people to discover treatments more rapidly, design safer infrastructure or devote less time to routine tasks, those are achievements to be welcomed rather than regretted.

    The challenge, however, is subtler than deciding which tasks should remain human and which can safely be delegated to machines. It lies in recognising that many of the experiences through which judgement developed were never deliberately protected because nobody imagined they could disappear. They seemed inseparable from the very act of understanding the world. Reality itself ensured that difficult questions demanded patience, that uncertainty could not always be escaped and that our first explanations often proved inadequate before better ones became possible.

    For centuries, education quietly relied upon that world. Schools and universities undoubtedly shaped knowledge, values and habits of mind, but they did so within an environment that continuously supplied its own forms of intellectual resistance. Teachers did not have to create every formative experience because many of them arrived uninvited, embedded in the ordinary act of trying to understand something that refused to yield. Education certainly cultivated judgement, but it did so in partnership with a reality that was itself profoundly educational.

    Artificial intelligence may be changing that partnership more deeply than we yet appreciate. Not because it eliminates uncertainty altogether, but because it increasingly mediates our encounters with it. The world continues to resist us, yet its resistance is now filtered through systems designed precisely to make that resistance more manageable. In many circumstances this is unquestionably beneficial. The question is whether, in making reality easier to think with, we may also be altering the conditions under which people learn to think in the first place.

    If that is so, then the central educational question of the coming decades is not simply how artificial intelligence should be incorporated into classrooms. It is whether we can recognise which formative experiences belonged to a world that no longer exists in quite the same form, and whether new experiences capable of cultivating judgement will emerge within the world that is replacing it. History offers no reason to believe that human beings cease developing wisdom whenever their tools improve. Every technological transformation has altered the conditions under which people think, and each has produced intellectual virtues that earlier generations could scarcely have anticipated. There is every reason to expect that artificial intelligence will do the same.

    If this account is broadly correct, the distinctive responsibility of educational institutions may not be to preserve access to knowledge, which intelligent systems will make increasingly abundant, but to preserve encounters with realities that remain genuinely independent of the learner. Difficult texts that resist immediate interpretation, scientific problems whose answers cannot simply be prompted into existence, historical events that refuse present-day assumptions, and other people whose perspectives cannot be reduced to our own all confront us with a world that does not reorganise itself around our immediate purposes. Such encounters are often inconvenient. They are also among the most powerful conditions under which judgement has traditionally developed.

    What seems less certain is whether the experiences through which judgement develops will remain unchanged simply because intelligence itself has become more accessible. Answers may arrive more quickly than ever before, while the journey that once led us towards them quietly contracts. If part of our intellectual formation depended upon that journey, then preserving the destination may not be enough.

    I have often found myself returning to Leparmentier’s engineers. They were neither celebrating artificial intelligence nor lamenting its arrival. They were simply describing small, almost unremarkable changes in the texture of their professional lives: they became stuck less often, asked colleagues for help less frequently and spent less time inhabiting uncertainty before moving on. Those observations stayed with me because they seemed to capture something larger than the transformation of software engineering. They hinted that a change in our tools might also represent a change in the kind of reality within which human judgement has long been formed.

    Whether that intuition ultimately proves correct remains to be seen. Yet if good journalism sometimes notices historical change before we possess the concepts to describe it, perhaps Leparmentier’s engineers were describing more than a new way of working. Perhaps they were offering an early glimpse of a world in which reality no longer teaches us in quite the same way as before.

  • Why Students Still Need One Another in the Age of Artificial Intelligence

    A Conversation in Colombia

    Some time ago, I visited a school in Colombia whose educational model differed from almost every school I had known: students were not grouped by age but by autonomy. As they demonstrated greater independence, they were given increasing responsibility for organising their own learning. They decided when to arrive, how to structure their day, and how quickly to progress. Teachers remained present, but gradually shifted from directing learning to accompanying it.

    One student particularly caught my attention. He had reached one of the highest levels of autonomy. Watching him move confidently through the school, it was easy to imagine that he could continue learning almost entirely on his own.

    So I asked him a simple question: “If you are already capable of learning independently, why do you still come to school?”

    He smiled, almost as though the question itself had caught him slightly off guard. Then he answered without hesitation: “Because when I’m here, talking to other people, I learn things about who I am.”

    Only much later did I realise that I had probably heard the most important answer of the day.

    At first, his words puzzled me. Schools help us understand mathematics, literature, history, and science. They help us acquire knowledge and develop skills. But how could conversations with classmates teach someone who they are?

    The more I reflected on his answer, the more I realised that it pointed towards something we rarely notice. Most of the time we do not look at our own way of thinking. We look through it. Like a pair of glasses worn every day, our assumptions quietly shape everything we see while remaining almost invisible themselves. They feel like reality because they are the only perspective from which we are looking. Perhaps this is why genuine conversations can be so unsettling. Someone else notices what we overlook. They find significance where we saw none. They question distinctions that had always seemed obvious to us. They expose assumptions we did not even know we were making.

    The student had not come to school because he needed more information. He had come because other people were making his own way of seeing visible to himself. Once I understood his answer in those terms, I found myself thinking about Aristotle.

    Why Aristotle Still Matters

    The student’s answer stayed with me for a long time. The more I reflected on it, the less it seemed like an isolated remark and the more it reminded me of something I had first encountered many years earlier in Aristotle. At first, the connection may seem unexpected. Yet Aristotle was trying to understand precisely the same question: how do human beings come to know themselves? His answer was striking. Human beings, he argued, do not flourish in isolation. We become ourselves fully only through life shared with others. This is the deeper meaning of his famous description of the human being as a political animal. It is not simply that we live in societies. It is that some of our most important human capacities can only mature within them.

    The idea becomes even clearer in his account of friendship. Aristotle describes a friend as another self. At first, the phrase sounds puzzling. Another person is not another version of us. Quite the opposite. They are someone whose life, experiences and assumptions have developed independently of our own. Precisely because of that independence, they allow us to discover aspects of ourselves that would otherwise remain hidden.

    Listening to that student, I slowly realised that he had understood this instinctively. When he said that conversations with his classmates helped him discover who he was, he was not describing the exchange of information. He was describing the experience of seeing his own way of understanding become visible through contact with someone else’s.

    Left entirely to ourselves, our thinking often remains internally coherent. We refine our ideas, organise them more carefully and become increasingly consistent. But consistency is not the same as self-knowledge. We rarely notice the assumptions that organise our thinking while nothing genuinely independent interrupts them.

    Another person changes that. They ask questions that would never have occurred to us. They care about things that seem irrelevant to us and sometimes find our own certainties surprisingly difficult to understand. In doing so, they make visible the invisible framework through which we have been interpreting the world. For Aristotle, this was not an accidental benefit of friendship. It was one of the reasons friendship mattered in the first place.

    The more I thought about that conversation in Colombia, the more convinced I became that the student had expressed exactly the same insight in ordinary language. He had gone to school not because he lacked autonomy, but because other people were helping him become visible to himself.

    What Artificial Intelligence Changes

    Reflecting on that conversation in Colombia gradually led me to another question. If so much of education depends on encounters with people whose perspectives have developed independently of our own, what happens when a student can increasingly learn without needing those encounters?

    This, I believe, is the educational question raised by artificial intelligence. It is not primarily a question about information, because AI already provides extraordinary access to information. Nor is it simply a question about problem-solving, since many intellectual tasks can now be completed more efficiently with its assistance. The deeper question concerns the structure of learning itself.

    For the first time, students can make substantial intellectual progress while spending far less time encountering minds that were not chosen by them, that do not naturally adapt to them and that have not been mediated through systems designed around their own purposes. That possibility represents a profound educational shift. It changes not only how students learn, but also the conditions under which they encounter other ways of understanding the world.

    Artificial intelligence is built to help. It answers our questions, adjusts its explanations, remembers our preferences and increasingly adapts its responses to the way we think. This is one of its greatest strengths. It makes learning more accessible, more efficient and, in many cases, more enjoyable. Yet every strength reshapes the educational environment in which it operates.

    A classmate does not reorganise their thinking around ours. A teacher may try to understand us, but their perspective ultimately remains their own. A difficult book does not rewrite itself because we misunderstand it. History does not unfold in order to make our learning easier. Reality itself has no obligation to conform to the categories through which we happen to interpret it.

    Other people remain, gloriously, independent of our purposes, and that independence is not an obstacle education has traditionally had to overcome. It is one of the reasons education has always worked as it does.

    When another person’s understanding refuses to fit comfortably inside our own, we are confronted with something more demanding than disagreement. We are forced to consider the possibility that it is not merely our conclusions that require revision, but the assumptions from which those conclusions emerged. We begin to recognise that our way of framing the situation may itself be incomplete.

    Artificial intelligence can certainly challenge us. It can present counterarguments, introduce unfamiliar ideas and even sustain remarkably sophisticated forms of dialogue. None of this should be underestimated. Even so, those challenges arise within an environment whose fundamental orientation is user responsiveness. Its intelligence is exercised in the service of our questions.

    Human encounter is different. When another student disagrees with us, they are not performing disagreement as an educational strategy. They disagree because they genuinely see the matter differently. When reality frustrates our expectations, it is not trying to improve our understanding. Its resistance exists independently of our purposes.

    That distinction may appear subtle, yet it has profound educational consequences. The most formative encounters are often those that no one arranged. They surprise us precisely because they emerge from realities that have their own integrity, their own history and their own logic. They do not exist to complete our thinking. They exist independently of it.

    For that reason, they often reveal the limits of our understanding more effectively than any challenge deliberately designed to improve it.

    Why Students Still Need One Another

    If students can increasingly learn, produce, and solve problems alone, then the purpose of schools deserves reconsideration. Schools have never existed simply because information was difficult to obtain. Libraries have always contained more knowledge than any individual could master, and today, artificial intelligence places much of that knowledge within immediate reach. If education had only ever been about access to information, its central institutions would already have lost much of their purpose. Their deeper purpose has always been different.

    Schools bring people together who would not naturally have chosen one another. They place students in sustained contact with classmates whose experiences, priorities and ways of understanding the world have developed independently of their own. Living alongside those differences is not merely a social benefit that accompanies learning. It is one of the conditions through which learning becomes formative.

    It is in those encounters that students gradually discover not only how to answer questions, but also how to recognise which questions deserve to be asked in the first place. They begin to realise that every perspective, including their own, is necessarily partial, and that understanding often starts with the unsettling discovery that someone else has seen something they themselves could not.

    This is why the arrival of artificial intelligence does not diminish the importance of schools. If anything, it makes their distinctive human purpose more visible. The more capable AI becomes of supporting individual learning, the more valuable schools become as places where students regularly encounter minds that do not simply respond to them.

    Their unique contribution will lie less in delivering knowledge than in creating opportunities for students to live with perspectives that surprise them, contradict them and occasionally force them to rethink not only their answers, but the assumptions from which those answers emerged.

    I often find myself thinking again about that student in Colombia. Only afterwards did I realise that he had answered a question I had not even known how to ask. He did not come to school because he needed more information. He came because, in the company of others, he was slowly discovering the person who was doing the learning.

    Artificial intelligence may become an extraordinary intellectual companion. It cannot become another self.

  • Why Judgement Requires Independent Reality

    AI, Education, and the Limits of Designed Challenge

    Artificial intelligence is often discussed in terms of what it can do. It can generate information, explain concepts, answer questions, summarise arguments, and increasingly assist with forms of reasoning that once appeared to require considerable expertise.

    These developments naturally raise questions about knowledge and learning. Yet the most important educational question may lie elsewhere.

    Education has never been concerned solely with what people know. It has also been concerned with the formation of judgement: the capacity to recognise what matters, to interpret situations responsibly, and to determine when familiar ways of understanding are no longer adequate.

    The significance of AI for education depends not only on what it allows students to know, but on how it affects the conditions through which this capacity develops.

    Beyond Information

    Judgement is often misunderstood as a matter of choosing between available options. In practice, it begins earlier than that.

    Before any decision can be made, someone must determine what kind of situation they are facing, which distinctions are relevant, what should count as evidence, and whether the available options are themselves adequate.

    These questions cannot always be answered through the application of existing rules. Sometimes the difficulty lies not in selecting the right answer but in recognising that the framework through which the situation is being understood has become inadequate.

    Judgement becomes necessary precisely at that point. The capacity to recognise such moments does not arise through information alone. It develops through experience.

    The Importance of Irreducible Encounters

    Not every challenge contributes to the formation of judgement. Many difficulties can be overcome through greater effort within an existing framework. A student may solve more complex mathematical problems, learn additional historical facts, or acquire greater technical expertise while leaving their underlying assumptions unchanged.

    There are, however, encounters that cannot be resolved in this way. A scientific anomaly may resist established explanations. A historical event may challenge familiar narratives. A moral dilemma may expose tensions that existing principles cannot easily reconcile. A conversation with another person may reveal a perspective that cannot simply be translated into one’s own terms. In such moments, additional effort is no longer enough. The framework itself becomes part of the problem.

    This introduces what might be called irreducibility: an encounter in which resistance emerges from beyond the learner’s existing framework, such that the framework itself rather than merely the content within it becomes inadequate.

    At that point, understanding is no longer a matter of extension but of reorganisation. One does not simply know more than before. One begins to see differently. Assumptions must be revised, distinctions that once appeared obvious become uncertain, and new ways of understanding gradually emerge.

    Difficulty operates within a framework. Irreducibility places the framework itself under pressure. Reorganisation is the response to that pressure. Judgement develops as the capacity to recognise when such pressure is occurring.

    How Judgement Develops

    What develops through these encounters is not simply a better understanding of a particular subject.

    Each time a person discovers that a framework that previously seemed adequate no longer fits the reality before them, they must step back from that framework and reconsider it.

    Initially, this appears to be a local problem. A particular interpretation fails. A familiar explanation no longer works. An assumption proves inadequate. The ability to recognise framework failure can only develop through experience of it.

    What matters in such moments is not simply that a belief changes. Something more fundamental occurs. A person discovers that a way of understanding the world which previously appeared adequate was inadequate in a manner they could not fully recognise from within it.

    This is more than a revision of knowledge. It is an experience of becoming aware of a framework as a framework. What had previously functioned as an invisible lens through which reality was interpreted becomes visible as something contingent, limited, and open to revision.

    Attention shifts from the content of a particular judgement to the conditions that made the judgement possible in the first place. The learner is no longer simply thinking within a framework. The framework itself becomes an object of reflection.

    This is why repeated encounters matter.

    Each experience of framework failure develops sensitivity to the possibility that current assumptions may also be incomplete. What transfers across domains is not knowledge of when particular frameworks fail, but a disposition to treat frameworks themselves as provisional and open to evaluation.

    The transfer is not primarily intellectual but reflective. A person who has repeatedly discovered the limitations of one framework becomes more capable of recognising the possibility of limitation in others. What develops is not expertise in particular failures but sensitivity to framework failure itself.

    Judgement emerges from this process. It is the growing capacity to recognise that the way one is currently understanding a situation may itself require reconsideration.

    Why Independence Matters

    This helps explain why certain encounters are especially formative.

    Many educational traditions have recognised that learning often depends on confronting realities that do not adapt themselves to the learner.

    Hannah Arendt emphasised the importance of plurality. Hans-Georg Gadamer described understanding as something transformed through genuine dialogue. John Dewey argued that inquiry begins when habitual assumptions encounter situations that resist them. Despite their differences, all point toward a similar insight.

    What forms judgement is not simply difficulty. It is an encounter with something whose resistance exists independently of our own purposes.

    Other people are often among the most reliable sources of such encounters. Their perspectives emerge from experiences, assumptions, and concerns that were not organised around our own. Historical events do not unfold for our intellectual benefit. Difficult texts were not written to confirm our expectations. Scientific results do not adapt themselves to our preferences.

    Their resistance does not arise because someone designed it for us. It arises because reality is not organised around us.

    When resistance comes from something independent of our purposes, framework failure can no longer be dismissed as a feature of a system intentionally designed for our benefit. The inadequacy of our understanding is revealed by something that is indifferent to whether our framework survives or not.

    Independent resistance does not make self-deception impossible. People can always defend familiar assumptions or explain away anomalies. What it removes is a particular interpretive escape route. The resistance can no longer be understood simply as a feature of an educational design. It appears instead as evidence that reality itself may not conform to the categories through which it is being interpreted.

    That independence is what gives such encounters much of their formative power.

    Artificial Intelligence and the Reduction of Independence

    This is where artificial intelligence introduces an educational question that goes beyond information.

    Most AI systems are explicitly designed to serve user purposes. They respond to intentions, adapt to preferences, and organise information around the user’s questions. This is one of their greatest strengths. It is also what makes them educationally significant.

    The issue is not that AI eliminates the challenge. It clearly does not. Nor is it that AI cannot introduce unfamiliar ideas, opposing arguments, or deliberately Socratic forms of interaction. The deeper question concerns the nature of the resistance being encountered.

    A challenge deliberately generated within a designed system remains part of an environment organised around the user. It exists because someone decided that this particular challenge should appear at this particular moment for a particular purpose.

    Human encounters are different. Other people are not organised around our learning goals. Historical events are not structured for our intellectual benefit. Difficult texts are not written to accommodate our assumptions. Their resistance exists independently of our purposes. For that reason, they confront us with the possibility that our framework itself may be inadequate in ways we did not anticipate.

    Nor is the issue whether an AI system could be deliberately designed to challenge users. It certainly could. The deeper issue is that even deliberately adversarial systems operate within a space of challenges anticipated in advance by their designers. Independent reality is different. Its resistance is often unexpected, unplanned, and unconcerned with educational objectives. Part of its formative power lies precisely in the fact that no one arranged for it to appear.

    The question is therefore not whether AI can create difficulty. The question is whether designed challenge is equivalent to encounters with realities whose independence from us is precisely what makes them formative.

    Judgement in an AI-Mediated World

    The central educational significance of artificial intelligence does not lie primarily in its ability to provide information, explanations, or even sophisticated interpretations. Increasingly, it can do all of these things remarkably well.

    The deeper question concerns the formation of the person who receives them. If judgement develops through repeated encounters with realities that expose the limitations of existing frameworks, then educational institutions cannot focus exclusively on access to knowledge or efficiency of learning.

    They must also preserve opportunities for students to encounter realities that were not organised around them, perspectives that resist immediate assimilation, and situations that require genuine reorganisation of understanding.

    The future of education may depend less on preserving access to information than on preserving encounters with forms of resistance that remain genuinely independent of the learner’s purposes.

    Because judgement is formed through learning that our frameworks are not the same thing as reality.

  • The Formation of Judgement in the Age of AI

    Why judgement depends on formation, not information

    The Missing Question

    Much of the current discussion about artificial intelligence in education is framed in terms of capability: what systems can do, how they perform, and how they might be integrated. A secondary layer addresses pedagogy: how teaching must adapt to a world in which many forms of execution can be delegated.

    What remains largely unexamined is a more fundamental question: how judgement is formed.

    This omission is not accidental. It reflects a deeper assumption, that judgement is a function of knowledge, or at most of reasoning, and therefore that any system capable of producing knowledge or assisting reasoning can contribute, directly or indirectly, to its development.

    This essay begins with a different premise: judgement is not produced by knowledge or by reasoning alone, but by formation.

    Formation as a Concept

    Formation is not the accumulation of information, nor the acquisition of skills in isolation. It is the gradual structuring of how the world appears to a person — what is noticed, what is ignored, what is taken to matter, and what is not.

    Formation can be defined as: the structuring of perception and disposition through repeated engagement with situations that require responsible responses, resulting in the capacity to discern what matters in a situation and to recognise when established frameworks no longer apply.

    Each element of this definition matters.

    Formation structures rather than accumulation. It changes how situations are perceived, not merely what is known about them. It concerns perception and disposition, not knowledge alone. It shapes both how one sees and how one is inclined to respond. Its effects are developmental rather than additive. What is formed cannot simply be retroactively substituted by later descriptions of experiences that were never actually lived through.

    Formation also depends on a responsible response. The individual’s engagement matters because their judgement has consequences for which they must answer. What is formed is therefore not a store of answers, but a capacity: the ability to recognise what matters, to orient oneself in situations where rules do not suffice, and to act accordingly.

    Formation, in this sense, cannot be reduced to increasingly sophisticated participation within an already established framework of practice. It concerns the transformation of the criteria by which situations themselves are interpreted — including the recognition that existing frameworks may no longer be adequate.

    Experience and Transformation

    Hubert Dreyfus’s account of skill acquisition provides a precise description of how this transformation occurs. The movement from novice to expert is not simply a matter of internalising more rules, but of a qualitative shift in perception. The expert does not merely apply rules more efficiently. The expert sees the situation differently.

    This transformation depends on a history of encounters in which the individual’s responses mattered, where misjudgement had consequences, and where reality resisted reduction to pre-given categories. It is this exposure to resistant reality, rather than repetition alone, that produces the shift from rule-based reasoning to situated discernment.

    What matters here is not practice in general, but a specific form of engagement: one in which the individual is answerable to the situation itself, rather than operating within a controlled system designed to guide or optimise performance. For this reason, not all forms of practice are equivalent.

    Formation depends on participation in situations whose structure is not organised in advance around the learner’s development, and in which consequences arise from an independent world rather than from a system calibrated to the user.

    AI-mediated environments, by contrast, are necessarily designed around responsiveness, optimisation, and alignment with user purposes. They reduce friction, guide action, and structure possible pathways through the environment. This is part of their usefulness. But it also creates a structural limit.

    A system calibrated to the user cannot at the same time be genuinely independent of the user, because its responses remain oriented, at every level, toward the purposes for which it was built. Designed resistance remains generated within that horizon, and therefore cannot reproduce the indifference of a world that does not answer to the learner.

    Even when resistance is deliberately introduced, it remains resistance organised for a formative purpose, not resistance arising from a world indifferent to whether formation occurs.

    This is not primarily a limitation of current AI systems. It follows more broadly from the fact that any purposive system necessarily organises its responses in relation to the purposes for which it was designed.

    The distinction at stake is therefore not between intelligent and non-intelligent systems, nor between digital and non-digital environments. It concerns whether the responses encountered are structured in relation to the learner’s formation or arise independently of it.

    The issue is not whether AI systems are useful (they clearly are), but whether they can provide the kind of encounter through which judgement is formed. The answer depends not on computational sophistication, but on the structure of the relationship itself.

    Difficulty and Formative Friction

    A central distinction follows from this account: the distinction between difficulty and formative friction.

    Difficulty operates within a stable framework. A problem is difficult when it requires effort, time, or skill to resolve, while the situation’s underlying structure remains intact. The individual works harder within an already defined horizon.

    Formative friction occurs when the framework itself is no longer sufficient, when the situation resists interpretation according to the categories available to the individual. In such moments, the task is not to solve a problem, but to reconsider what the problem is.

    Not all encounters with friction are formative. Friction becomes formative only when the individual remains engaged with the resistance, assumes responsibility for responding to it, and undertakes the work of reconstructing understanding in light of that resistance.

    Absent this orientation, friction produces not formation but paralysis, avoidance, or mimicry without understanding.

    Formation, therefore, depends not only on exposure to resistance but on a mode of engagement in which that resistance is taken seriously as a demand upon one’s own judgement.

    Habituation and Structured Sensitivity

    Aristotle’s concept of hexis provides the classical account of what is formed through such processes. Hexeis are not automatisms or habits in the narrow sense, but structured sensitivities to what matters in a situation.

    Through habituation, the individual becomes attuned to relevant features of situations, not by mechanically applying rules, but by perceiving significance directly. This is the basis of phronesis, or practical wisdom: the capacity to act appropriately in situations that cannot be fully specified in advance.

    Crucially, Aristotle’s account also establishes that the quality of what is formed depends on the quality of the situations encountered. Formation is not independent of context. It is shaped by the norms, practices, and expectations embedded within the environments in which it occurs.

    Institutions therefore do not merely host formation. They actively constitute the conditions under which formation becomes possible.

    Hexeis are shaped through what institutions reward, ignore, sustain, or leave unchallenged. For this reason, the character of an institution cannot be separated from the character of the judgement it forms: what an institution does not require, recognise, or sustain cannot be developed within it.

    Plurality and Irreducible Encounter

    Judgement is not formed in isolation. It emerges within a shared world in which others appear, respond, and resist.

    Hannah Arendt’s account of plurality clarifies what this requires. Thinking is not simply the processing of information. It involves taking into account perspectives that are not one’s own, perspectives grounded in the fact that others occupy distinct positions within the world.

    What makes these perspectives irreducible is not merely their unpredictability, but their origin. Each perspective reflects a situated existence rooted in a distinct history, position, and experience of the world, which Arendt describes through the idea of natality.

    The world appears differently from each such position, and no perspective can be fully substituted for another.

    Formation, therefore, depends on encounters with perspectives that are not generated on demand and whose responses are not structured around producing a particular effect on the learner.

    A system that presents multiple perspectives can simulate content diversity. But it cannot reproduce the encounter with a perspective grounded in a history of situated existence that is not reducible to representation within the system itself.

    The decisive point is not only that representations lack the existence they represent, but that they are generated within a system whose responses remain oriented toward the learner’s experience.

    What gives such encounters formative force is precisely the fact that the other’s response is not organised for the learner’s development. The resistance emerges from an independent position, not from a pedagogical function.

    The difference concerns the structure of the relationship between the learner and what they encounter, whether that encounter is organised toward the learner’s development or indifferent to it.

    Educational Implications

    If formation depends on encounters with situations that resist existing categories, then educational institutions cannot be organised exclusively around optimisation, guidance, and controlled progression. Students must encounter situations in which interpretation itself becomes necessary, situations where frameworks no longer fully settle what matters, and where judgement cannot be replaced by the efficient execution of procedures.

    This has consequences for how educational environments are structured. Not every form of difficulty is formative. Repetition, challenge, and complexity may strengthen performance within an established framework while leaving the framework itself untouched. Formation requires something more demanding: encounters with situations whose meaning is not already stabilised in advance for the learner.

    For this reason, educational institutions cannot be understood merely as systems for transmitting knowledge or coordinating competencies. They shape the conditions under which students learn to perceive significance, respond to irreducible plurality, and assume responsibility for interpretations whose consequences are real and not fully pre-structured for them.

    The question raised by AI in education is therefore not simply how intelligent systems can improve learning outcomes. The question is whether educational institutions will continue to preserve the kinds of encounters through which judgement becomes possible in the first place.

    Conclusion

    If formation requires engagement with a resistant world, exposure to irreducible plurality, and the development of structured sensitivities through responsible action, then educational transformation must be evaluated in light of its effects on those conditions.

    What is ultimately at stake is the formation of the capacities through which individuals recognise what matters, orient themselves amid uncertainty, and exercise judgement.

    Education, therefore, faces a deeper question: whether it will continue forming human beings capable not only of operating within frameworks, but also of discerning when those frameworks no longer adequately respond to the situations they encounter.

  • Edgar Morin and the Educational Question of the AI Age

    Fragmented Knowledge and Complex Thought

    The death of Edgar Morin last week marks the end of a remarkable intellectual life. For more than a century, he witnessed wars, revolutions, ideological conflicts, technological transformations, and the slow reorganisation of modern knowledge. Few thinkers remained intellectually active for so long while continuing to engage seriously with the central questions of their time.

    Many tributes will rightly focus on his theory of complexity, his critique of reductionism, and his call for a more integrated understanding of knowledge. Yet his significance may be even greater today than when many of his most influential works were written. Not because he anticipated artificial intelligence, but because he identified a problem that artificial intelligence now returns to us in a sharper form.

    Morin spent much of his life arguing that education had become fragmented. Schools and universities divided reality into disciplines, specialisms, and isolated domains of expertise. Students learned pieces of the world without necessarily learning how those pieces related to one another. Knowledge accumulated, but understanding often remained fragile.

    The world, he argued, cannot be understood through separation alone. Human beings must learn to perceive relationships, interdependencies, feedback loops, contradictions, and the dynamic interactions that connect apparently unrelated phenomena. Education should not merely transmit knowledge. It should cultivate the capacity to situate knowledge within a wider whole. For decades, this was a profoundly important educational challenge.

    The New Paradox of Artificial Intelligence

    Today, however, a new paradox appears. Artificial intelligence is extraordinarily good at producing connections. A modern language model can relate history to economics, biology to ethics, philosophy to technology, and literature to politics in a matter of seconds. It can generate syntheses across fields, propose interpretative frameworks, identify analogies, and produce apparent wholes from scattered fragments of information.

    In one sense, AI seems to fulfil part of Morin’s educational ambition. It overcomes fragmentation with astonishing speed. It connects what traditional schooling often kept apart. Yet this conclusion arrives too quickly.

    The ability to generate connections is not the same thing as the ability to judge their significance. A system may propose relationships between ideas. It may produce interpretations. It may generate frameworks through which a situation can be understood. Yet none of these answers a more fundamental question: which of those interpretations matters? Which framework should guide action? Which relationship is genuinely significant and which is merely plausible?

    The more connections become available, the more important this question becomes. Complexity does not eliminate judgement. It increases the need for it.

    The Knower Within the Known

    This is where Morin’s legacy becomes particularly important for the age of AI. His concern was never simply that knowledge had been divided into fragments. His deeper concern was that human beings might lose the capacity to orient themselves within complexity. To think complexly is not merely to connect more things. It is to understand one’s position within a system of relations that one is also helping to shape. That is a crucial point.

    Morin’s thought does not place the observer outside the world being observed. The knower is part of the system that is being known. Human beings do not judge complexity from nowhere. They judge from within histories, institutions, responsibilities, loyalties, limits, and consequences.

    This is precisely where AI-generated synthesis differs from human judgement. A system can produce connections across domains, and often brilliantly. It can even generate meta-connections, compare interpretations, rank possibilities, and suggest which frameworks might be more useful. But it does not inhabit the situation whose significance must be judged.

    To inhabit a situation is not merely to be present within it. It is to be shaped by it and answerable to it. The consequences of one’s interpretation become part of one’s own history. Judgement emerges from this double condition: being situated within a reality and remaining responsible for what one decides about it. AI systems generate interpretations without being constituted by them or accountable for them. Judging significance is therefore not merely another layer of connection. It involves orientation from within complexity.

    Complexity, Responsibility, and Education

    The above ideas matter for education. If machines can increasingly generate relations between ideas, then the educational task cannot be simply to teach students to make connections. That remains important, but it is no longer sufficient. Students must learn to ask which connections deserve attention, which interpretations distort more than they reveal, and which frameworks carry responsibilities that cannot be reduced to explanatory elegance.

    A brilliant synthesis may still be morally irrelevant. A plausible analogy may still be misleading. A coherent framework may still conceal what matters most.

    Morin helps us see why this is not a secondary problem. Complexity is never only epistemological. It is also practical and moral. Understanding a complex situation is not only about describing its parts and relations, but also about deciding how one should respond within it.

    Artificial intelligence changes the conditions under which this response is formed. It increases the availability of possible interpretations while making it easier to bypass the uncertainty of which interpretation begins. The danger is not that students will lack connections. The danger is that they will receive connections before they have learned to assume responsibility for their significance.

    A Different Educational Problem

    This is a different educational problem from the one Morin first confronted. 

    In the twentieth century, education had to resist the fragmentation of knowledge. In the twenty-first century, it must also resist the illusion that automatically generated synthesis is the same as understanding. The problem is no longer only that students may fail to see relationships. It is that relationships may appear too quickly, too fluently, and too persuasively, before judgement has been formed.

    This does not make Morin obsolete. It makes him newly necessary.

    Morin’s Last Question

    His legacy reminds us that education is not simply the organisation of knowledge, nor even the production of interdisciplinary fluency. It is the formation of human beings capable of orienting themselves responsibly within a reality that exceeds any single framework. And that may be the most important educational question Morin leaves us with in the age of artificial intelligence.

  • Pope Leo XIV, Plato, and the Conditions of Human Formation

    The Technocratic Paradigm

    Pope Leo XIV’s recent encyclical on artificial intelligence is being discussed primarily as a theological document. That is understandable. Yet it would be a mistake to read it only within ecclesial or doctrinal categories. What makes the text important is not simply that the Church has entered the AI debate, but the level at which it enters it.

    The encyclical treats artificial intelligence not merely as a technological development requiring regulation or ethical oversight, but as part of a broader civilisational transformation affecting how human beings understand responsibility, agency, judgement, and formation itself. Its central concern is not whether intelligent systems are useful. Clearly, they are. The concern is whether societies organised increasingly around optimisation, automation, and immediacy gradually weaken the conditions under which human beings develop the capacity to govern such systems wisely.

    What the encyclical calls the technocratic paradigm at the level of institutions and social organisation is the institutional expression of what delegated thinking describes at the level of individual cognition. In both cases, interpretative responsibility progressively disappears behind systems optimised for efficiency, prediction, and procedural execution.

    The convergence itself is intellectually significant. Different anthropologies and entirely different philosophical premises arrive at the same educational intuition: human judgement forms through encounters that resist immediate assimilation. This convergence may itself constitute evidence that both traditions are tracking something structurally real about human cognitive formation rather than merely expressing a cultural preference. When traditions beginning from incompatible premises converge not merely on similar conclusions but on the same description of what those conclusions require educationally, the independence of the routes strengthens the case that the conclusion tracks something real.

    Whether one begins from the encyclical’s theological anthropology, grounded in the irreducibility of the person created in the image of God, or from a more philosophical account centred on framework interpretation and human agency, both traditions identify the same structural vulnerability in AI-mediated environments: the displacement of the conditions under which judgement forms.

    Responsibility Cannot Be Delegated

    One of the recurring illusions surrounding AI is the belief that automated systems somehow remove human accountability because they appear objective. The encyclical rejects this directly and correctly.

    Algorithms do not dissolve moral agency. They relocate it.

    Someone still defines the parameters. Someone still determines what is measured, optimised, ignored, or prioritised. Someone still answers for the consequences. The apparent neutrality of automated systems often conceals rather than eliminates the human decisions embedded within them.

    What is specific to algorithmic systems is not simply that responsibility is delegated. All institutions delegate responsibility. The distinctive feature of algorithmic systems is that the interpretative choices embedded in their design become progressively invisible during operation. The system appears neutral while silently reproducing prior assumptions, priorities, and exclusions.

    This differs even from ordinary bureaucratic diffusion of responsibility. In institutional systems, responsibility may become distributed across many actors, but the chain of decisions remains in principle traceable through procedures, records, and identifiable acts of judgement. In algorithmic systems, by contrast, interpretative assumptions become embedded in technical architectures whose operation may be opaque even to those who deploy them. The concealment is structural rather than merely practical.

    As these systems scale, the distance between the interpretative choices embedded in design and the consequences generated in practice increases dramatically. Responsibility becomes technologically mediated at scale. Power concentrates within infrastructures that increasingly appear detached from visible human judgement, even as they continue to embody it at every level.

    The educational significance of this is difficult to overstate. A society that increasingly relies on systems whose embedded assumptions are hidden from ordinary perception requires citizens capable not merely of using systems competently, but of interrogating the frameworks within which those systems operate.

    Plato, Friction, and the Conditions of Learning

    The encyclical’s educational section is its most philosophically interesting part. It warns repeatedly against what it calls a culture of immediacy: the progressive disappearance of the interval between question and answer.

    Remarkably, the text invokes Plato in this context. Genuine understanding, it argues, emerges through the intellectual friction produced by rubbing concepts and experiences together like flint until insight appears. Learning requires resistance, dialogue, uncertainty, and effort. Understanding is not downloaded intact into consciousness. It forms through sustained engagement with situations that do not immediately yield resolution.

    This converges closely with a philosophical tradition extending from Aristotle through Dewey and phenomenological accounts of expertise such as Hubert Dreyfus’s. In each case, formation depends not simply on information acquisition but on encounters that resist immediate assimilation into existing frameworks.

    Immediacy weakens this process because it removes the interval in which thought begins.

    That interval is not merely time. It is the period during which the absence of an available answer generates the pressure that makes interpretation necessary. The question forms because the situation remains unresolved long enough for uncertainty to become intellectually productive. When answers arrive before this pressure fully develops, the question itself never fully forms either.

    What fails to form is not the surface query but the deeper question of whether one’s existing conceptual frameworks are actually adequate to the situation at hand. That question only emerges when reality resists immediate resolution.

    The encyclical’s educational insight therefore goes beyond a generic defence of slower learning. It identifies something structurally important about human formation itself: judgement develops through sustained encounter with situations that cannot immediately be absorbed into available categories.

    This is why one of the encyclical’s most striking claims is that educating for the digital age also means educating people for when and why not to use AI.

    Prediction Is Not Judgement

    The relevant distinction is not between optimisation and judgement in the abstract. Increasingly capable AI systems can clearly assist in sophisticated forms of deliberation across competing considerations.

    The relevant distinction is between operating within predefined frameworks and determining whether those frameworks remain adequate to the situation itself.

    Even highly sophisticated AI systems remain structurally framework internal. They may propose reframing, identify tensions, or detect inconsistencies within the objectives they have been trained to optimise. But the adequacy of those objectives themselves remains outside the horizon the system can independently establish.

    This does not mean human beings stand outside all frameworks in some absolute sense. Human judgement is also historically situated, fallible, and shaped by inherited assumptions. The difference is that human beings can commit themselves to revising even their framework evaluating assumptions through encounters with resistant reality in ways not fully determined by prior optimisation structures.

    Artificial intelligence can predict. It can optimise. Increasingly, it can even assist sophisticated forms of deliberation. But judgement appears precisely where rules, distributions, and procedural criteria no longer fully settle the matter.

    The issue is therefore not whether AI can assist deliberation. Clearly it can. The issue is whether the conditions through which human beings develop the capacity to question frameworks themselves remain preserved.

  • Who Sets the Questions? Artificial Intelligence and the Delegation of Thinking

    I was recently speaking with a colleague about artificial intelligence and education. At one point, he made a remark that is becoming increasingly common: what really matters now, he said, is learning how to write good prompts. If machines generate answers, then the quality of those answers depends on how we ask. Better prompts lead to better outputs, and as models improve, that skill will probably become even more valuable.

    There is an obvious truth in this. Yet I left the conversation thinking about something else entirely. Not about how we ask questions, but about who defines them in the first place, and what happens when that task slowly ceases to belong fully to us.

    Authorship and Orientation

    The distinction that increasingly matters is not between humans and machines, nor even between knowledge and ignorance. It is the distinction between initiating thought and refining what has already been structured.

    An editor works within an existing form. He clarifies, reorganises, strengthens and improves. All of this requires intelligence and judgement. But the editor operates on something that already exists.

    The author begins somewhere earlier. He gives shape to something that did not yet fully exist before the act itself. The difficulty is not only finding a good answer, but determining what deserves to become a question at all.

    Artificial intelligence is extraordinarily powerful at the editorial level. It can refine, summarise, reformulate, extend and optimise with remarkable speed. Increasingly, however, it also provides orientation before inquiry has properly begun. It suggests how a problem might be approached, which distinctions appear relevant, and what kind of structure the question should take.

    Once that orientation is accepted, the nature of the intellectual task has already changed.

    From Producing Answers to Selecting Them

    One of the clearest shifts introduced by artificial intelligence is that intellectual activity increasingly moves from producing answers to selecting between them.

    Where thinking once involved constructing a line of reasoning from an initially unclear problem, we are now often presented with organised alternatives that can be evaluated, refined or combined. This is not necessarily passive. Selection still requires judgement. But it operates within an important limitation: it presupposes that the relevant possibilities have already been presented.

    Choosing between three AI-generated framings of a problem is not the same intellectual act as determining whether those framings are adequate in the first place. The prior question, whether the structure itself is sufficient, whether something important has been omitted, whether the problem should be approached differently altogether, belongs to another level of thinking.

    That distinction matters because the capacity to recognise that a framing is inadequate does not emerge primarily through evaluating already structured possibilities. It develops through earlier encounters with problems whose relevance and shape have not yet been organised in advance.

    Neil Postman observed, in a different technological context, that every medium reshapes not only the answers available to us, but the kinds of questions that become natural to ask. Something similar may now be happening at the level of intellectual orientation itself. Certain paths of inquiry become immediately visible and frictionless, while others quietly recede before they have fully formed.

    Techne and Phronesis

    Aristotle distinguished between techne and phronesis, and the distinction remains useful here.

    Techne concerns production, execution and method. It refers to forms of knowledge that can be systematised, transferred and evaluated according to effectiveness. Artificial intelligence dramatically expands this domain. It allows us to generate, structure and optimise with a speed and scale that would have been difficult to imagine only a few years ago.

    Phronesis, by contrast, concerns judgement. It is the capacity to orient oneself in situations where rules alone do not determine what matters. The question raised by AI is therefore not simply whether machines perform certain tasks better than we do. In many domains, they already do. The deeper question is whether, in expanding technical capability, we gradually weaken the conditions under which human orientation and judgement are exercised and formed.

    Delegated Thinking

    What emerges from this shift can be described with some precision. It is a form of delegated thinking, although not in the simplistic sense often imagined.

    The issue is not that people stop thinking. In many cases, they remain highly active intellectually. They compare, evaluate, refine and select. What changes more subtly is where thinking begins. The initial act of orientation, deciding what the problem is, what deserves attention, what framework should govern inquiry, increasingly arrives pre-structured.

    This transfer rarely occurs consciously. No one decides to relinquish the capacity to initiate thought. The shift happens gradually through repetition, convenience and habit. The system becomes the place where inquiry begins by default.

    What makes this difficult to perceive is that the resulting activity still feels intelligent, and often is intelligent. Yet intelligence exercised within a structure already organised by another process is not identical to the experience of confronting a problem before its meaning has stabilised.

    Beginning and Education

    Hannah Arendt described one of the defining human capacities as the capacity to begin: to introduce something new into the world rather than merely continue what is already given. That insight acquires a different weight in educational contexts shaped by artificial intelligence.

    The central challenge is not merely to teach students how to use AI effectively. It is to preserve the conditions under which they still learn to orient themselves before the structure of inquiry has already been supplied.

    The criteria by which significance is recognised are not simply given in advance. They are formed through the effort of distinguishing relevance from irrelevance, of deciding what deserves attention, of remaining with uncertainty long enough for a problem to emerge from experience itself rather than arriving already interpreted from outside.

    A student may, therefore, produce sophisticated work while remaining increasingly dependent on frameworks they did not generate and cannot fully interrogate. The issue is not an absence of intelligence. It is a gradual displacement of intellectual initiative away from the learner’s own encounter with the problem.

    The question, then, is not whether artificial intelligence helps us think. In many respects, it clearly does. The issue is whether, in helping us think, it also begins to shape what counts as worth thinking about in the first place.

    More subtly still, which questions quietly cease to arise at all.

  • Re-reading Émile in the Age of AI: Where Thinking Begins

    One possible reading of Rousseau’s Émile is to understand it not primarily as a theory of knowledge transmission, but as an inquiry into the conditions under which judgement can be formed. This is not the only legitimate interpretation, but it brings into focus a structural feature of the text: the insistence that the educator must not anticipate the learner’s process. The tutor does not organise learning as a sequence of explanations, but as a sequence of experiences in which the learner is required to confront the resistance of things without premature mediation.

    This claim, that the interval between encounter and response is constitutive of the formation of judgement, has been developed in earlier essays of this project. Here, it is sufficient to observe how Rousseau’s educational design depends on preserving that interval.

    Within this framework, error acquires a different status. It is not simply something to be corrected, but the moment in which the learner encounters the limits of their own understanding. Similarly, the delay of answers is not a deficiency, but a deliberately constructed condition that allows a question to emerge. Learning is not defined by the speed with which a correct solution is reached, but by the structure of the process that leads to it. What is being protected is not a body of content, but the possibility that thinking has its point of departure in the learner rather than being supplied from outside.

    From social interference to systemic anticipation

    The problem Rousseau seeks to avoid is, however, specific. The threat to the formation of judgement comes from social interference: the premature introduction of opinions, conventions, and external criteria that replace the learner’s direct relation to what is to be understood. Education is therefore conceived as a form of temporary protection, allowing a degree of independence in judgement to take shape before such interference becomes unavoidable.

    The contemporary situation presents a different configuration of this structural problem. The difficulty no longer lies primarily in others imposing their judgements, but in systems anticipating the moment at which judgement would arise. In AI-mediated environments, the issue is not only that answers are available, but that the space in which a question might emerge is already structured. Intervention does not follow experience; it precedes the conditions that would make questioning necessary.

    This alters the temporal order that underpins the formation of judgement. Where Rousseau sought to preserve a sequence, encounter, resistance, question, response, contemporary systems reorganise it so that response comes first. In doing so, they remove the condition under which the question would have been genuinely formed.

    Two trajectories: identical content, different structure

    This shift can be clarified by comparing two trajectories that are equivalent in content but divergent in structure. In the first, the learner encounters a problem without immediate mediation. Their initial attempt fails, not because of a lack of intelligence, but because the situation exceeds their current framework. This failure is not bypassed but sustained: the learner remains within it long enough for its inadequacy to become explicit. The question that emerges is not a generic request for information but a necessity arising from the experience itself.

    Any subsequent guidance enters a process already structured by the learner’s engagement. The response does not substitute the question, but answers one that has been formed through confrontation with difficulty.

    In the second trajectory, the learner encounters the same problem within an AI-mediated environment. The system provides an immediate pathway: a sequence of steps, a reformulation, or a prompt that directs attention before the learner has confronted the indeterminacy from which the question would have arisen. The learner remains cognitively active and may arrive at a correct result. What is absent is not thinking, but the condition that would have made that thinking necessary. The activity begins within a framework already organised.

    The difference lies not in the observable outcome but in the internal structure of the process. In one case, thinking originates in necessity; in the other, it unfolds within a necessity that has already been interpreted.

    The displacement of the origin of thought

    The decisive shift can be described as a displacement in the origin of thought. What is delegated in AI-mediated environments is not only the answer, but the conditions under which questioning arises. Thinking still occurs, but it begins within a structure that has already organised the indeterminacy from which the question would otherwise emerge. The issue is therefore not the presence of assistance, but its position: when intervention precedes the formation of the question, it alters the process through which judgement is formed.

    What is at stake is not the quantity of knowledge acquired, but the formation of a capacity. Judgement, understood as the ability to orient oneself in situations that exceed predefined frameworks, is formed through processes in which initial understanding fails and must be reorganised. When those processes are systematically pre-empted, the capacity they would have formed does not emerge in the same way.

    Implications for educational design

    The consequences of this analysis are structural rather than technological. The question is not whether artificial intelligence should be adopted or rejected, but whether its integration preserves the interval between encounter and response on which the formation of judgement depends.

    This requires that the integration of AI into education be evaluated not only in terms of efficiency or performance, but also in terms of whether it preserves the moment when the learner must confront the limits of their own understanding. The issue is not whether systems can support learning, but whether they occupy the space in which thinking would otherwise begin.

    The role of the teacher: sustaining friction

    Within this framework, the role of the teacher becomes more precisely defined. It is not primarily to provide answers or transmit content, but to sustain the conditions under which thinking can begin. This involves a deliberate management of time, difficulty, and intervention: knowing when not to respond, when to allow error to unfold, and when to introduce guidance that does not replace the question the learner is still in the process of forming.

    This entails not only restraint but also the deliberate design of situations in which the learner is required to confront what does not yield immediately and, in doing so, to initiate thought within that encounter.

    Far from becoming obsolete, the teacher assumes responsibility for maintaining a space that other systems tend to close. Their role is not to compete with the availability of answers, but to ensure that such availability does not eliminate the process through which thinking finds its point of departure in the learner.

  • Wonder, Aporia, and the Conditions of Thinking in the Age of AI

    Where Thinking Begins

    Philosophy begins in wonder. Aristotle’s claim is well known and often repeated, but too quickly softened, as if it simply meant that curiosity leads to learning. It means something more demanding than that.

    Wonder, for Aristotle, is not the desire to know a little more about something already understood. It is the moment when what seemed adequate no longer suffices. Something previously taken for granted becomes strange. The familiar stops fitting; the experience stops fitting, and the mind can no longer proceed as it had been proceeding.

    This is different from confusion, which is merely a failure to process, and different from aesthetic fascination, which may suspend attention without directing it anywhere. Wonder holds on to its object as something that calls for understanding, even when understanding is not yet available. It interrupts thought and holds it at the threshold of a question not yet formed.

    That is why wonder belongs to the beginning of thinking, not to one stage within it. Before anyone reasons toward an answer, something has to make the question feel necessary.

    Aporia: The Structure of Being Stuck

    Plato gives this condition a sharper shape. In the dialogues, especially the Theaetetus, Socrates repeatedly leads his interlocutors to a point at which what seemed clear reveals itself as contradictory or insufficient. The result is aporia.

    Aporia is not simple ignorance. Most of us do not know many things, and this usually causes no disturbance. Aporia is narrower and more uncomfortable: it is the experience of being unable to move forward with the concepts one currently has.

    The dialogues are full of this moment. Someone offers an account of justice, knowledge or piety. Socrates examines it patiently until it can no longer stand. By the end, the speaker cannot defend the original position, but has nothing better to replace it with.

    That suspension matters. Whether one remains in it long enough for thought to deepen, or escapes into the first plausible answer available, is often the real fork in the road.

    From Interruption to Inquiry

    Aporia is not yet inquiry. It is the condition from which inquiry may begin.

    What turns difficulty into investigation is not logic alone, but something more basic: a mind genuinely engaged with what it does not understand does not simply remain passive before it. Aristotle notes in the Metaphysics that human beings are by nature oriented toward understanding. The impulse to know is not externally imposed. It belongs to the kind of beings we are.

    But the transition is never automatic. Difficulty can also lead to withdrawal, distraction, dependence, or passive acceptance of whatever account is offered next.

    John Dewey explains this transition with particular clarity. In Logic: The Theory of Inquiry, he argues that inquiry does not begin with solving a ready-made problem. It begins when the person must determine what, in a confused situation, actually requires resolution.

    The problem itself has to be formed before it can be solved.

    This prior activity matters more than many educational systems recognise.

    Initiation and Evaluation

    There is a distinction here worth preserving.

    Evaluation works within an established frame. The problem has been identified. The relevant variables are visible. Reasoning proceeds toward a conclusion. Much schooling is organised around this model: students are given a problem and asked to solve it.

    Those are real intellectual skills. But they are not the whole of thought.

    Initiation is something different. It is the activity through which the frame itself is established. What matters here? What is the difficulty? How should the situation be approached?

    In practice, initiation and evaluation interact constantly. Solving can reveal that the original framing was wrong. New questions can emerge during analysis. But the distinction remains important.

    An education that develops evaluation while neglecting initiation may look successful while omitting its most demanding task.

    What AI Tends to Provide

    Artificial intelligence does not eliminate thinking. Nor does it automatically destroy initiative. Its effect is usually more subtle.

    AI systems often provide already-shaped framings of problems: summaries, suggested pathways, relevant factors, likely interpretations, and structured next steps. A student asks a question and receives not only information, but an organised field in which to proceed.

    For someone already capable of independent thought, this can be useful. The framing itself can become material for criticism, comparison or refinement.

    But that depends on prior formation.

    The student who has not learned to remain with uncertainty, to define the problem, to distinguish the central from the peripheral, may simply accept the frame and begin working inside it.

    The visible output may look entirely competent.

    What has been skipped is harder to see.

    What Is Worth Preserving

    Intellectual independence does not mean refusing tools or thinking in isolation. It means being present at the point where relevance has not yet been established, where the difficulty is still unclear, and where one must help determine what the question really is.

    A student who rarely encounters this does not merely miss an exercise. They miss the repeated practice through which a specific human capacity is formed.

    And this is not a capacity easily acquired later, once convenience has become a habit. It develops through use, or it weakens through neglect.

    What education should preserve is not primarily the correct answer to a predefined problem.

    It should preserve the capacity to stand before a situation not yet organised, to endure the discomfort of not yet knowing what matters, and to begin responsibly.

    Wonder is often where that begins. Aporia is the form it takes when it becomes serious. Inquiry, in the fullest sense, depends on both having had time to do their work.

  • Prediction, Decision-Taking, Deliberation, and Judgement. What Education Must Preserve in the Age of Intelligent Machines

    Much of the current debate about artificial intelligence treats different human capacities as if they belonged to a single scale: more data, faster processing, better outputs. From that perspective, systems that outperform us in some tasks appear to be gradually approaching, or even surpassing, human judgement. But the activities usually grouped together under that heading are not the same kind of thing. To predict, to decide, to deliberate, and to judge are related capacities, but they are not interchangeable.

    Prediction: Estimating What Is Likely

    Prediction concerns likelihood. It identifies patterns in past data in order to estimate what is likely to happen next. Weather forecasts, fraud detection systems, recommendation engines, and language models all operate substantially in this way. They estimate probable continuations, probable behaviours, or probable outcomes.

    Predictions may inform action, sometimes decisively, but they do not settle it. A model may predict that a borrower from a particular postcode is more likely to default. That may be statistically useful. It does not by itself determine whether granting the loan would be fair, prudent, or socially desirable.

    Decision-Taking: Selection Within Given Criteria

    Decision-taking means selecting among alternatives according to criteria that have already been set. A navigation system chooses the fastest route. A hiring filter ranks candidates based on specified criteria. A school admissions algorithm allocates places according to published priorities.

    This can be highly effective because the system is not determining the worth of the criteria themselves. It is operating within them. Choosing the fastest route is not deciding where one ought to go. Ranking candidates by chosen metrics does not decide whether those metrics deserve priority.

    Decision in this sense is powerful, but it remains bounded by a prior human act: someone has already decided what counts.

    Deliberation: Weighing What Cannot Be Fully Measured

    Deliberation begins when no single rule or metric resolves the matter. It involves weighing competing goods, considering consequences, interpreting circumstances, and recognising that some values cannot be reduced to one scale. 

    Should a hospital prioritise urgency, age, prognosis, equality, or waiting time? Should a school prioritise examination performance, inclusion, character formation, or social mobility? Should a company protect jobs, maximise profit, or preserve long-term trust?

    These are not merely technical questions. They involve rival goods that cannot be fully compared. Deliberation does not process considerations toward an optimum. It weighs what cannot be fully measured and reaches a position that remains contestable.

    Judgement: Commitment Under Responsibility

    Judgement is deliberation that has issued in commitment. It includes elements of prediction, decision, and deliberation, but it cannot be reduced to any of them, because what distinguishes it is not the quality of the reasoning alone, but the nature of the commitment it produces.

    Judgement occurs when someone determines what ought to be done here and now, under conditions where certainty is unavailable, and consequences matter. A machine may recommend, rank, or simulate an argument. But it cannot bear responsibility in the human sense.

    A system can be audited, corrected, regulated, or even blamed instrumentally. That is not the same as answering for a determination. Responsibility requires a subject capable of recognising fault, offering justification, and bearing the consequences of being wrong.

    Systems may be part of the chain of action, but they do not stand within it as accountable agents.

    A Simple Example

    Imagine a university awarding a scholarship. 

    A predictive model estimates which applicants are most likely to complete the degree with high marks. That is a prediction. An algorithm then ranks candidates based on grades, income levels, and predefined weightings. That is decision-taking. A human committee notices that one student’s lower grades followed a year of serious family difficulty. Members discuss whether resilience, disadvantage, or future potential should matter more than raw attainment. That is deliberation. The committee then chooses and publicly stands behind the decision. That is judgement.

    The acts are connected, but they are not the same. Each requires something different of the human being involved.

    Why This Matters for Education

    If education is understood mainly as the production of correct outputs, much of its domain becomes vulnerable to automation. Systems can already predict, classify, rank, retrieve, summarise, and generate competent responses at speed. If these are treated as the highest aims of learning, machines will increasingly outperform students in the very tasks schools reward.

    But education has a deeper purpose.

    Students need to learn what predictions can and cannot tell them. They need to understand how decisions are shaped by criteria chosen in advance. They need practice in weighing goods that cannot be reduced to a formula. Above all, they need opportunities to commit themselves to conclusions they can justify and defend.

    That kind of formation does not arise through exposure to information alone. It develops through use, through practice that requires one to take a position and answer for it.

    Conclusion

    Judgement begins where no metric fully determines what ought to matter, where someone must decide, and where that decision must be defended and owned.

    The more we distribute execution to intelligent systems, the more essential it becomes to form individuals capable of governing what those systems do not and cannot settle.

    That capacity, to deliberate under genuine uncertainty and to answer for the determination reached, is what education must still form. No system can form it on our behalf.