AIと競合を排除したいという人間の欲求

“Man is the creature who does not know what to desire, and he turns to others in order to make up his mind.” (René Girard, The One by Whom Scandal Comes)David with the Head of Goliath (Caravaggio, 1610)The public conversation about AI has settled into two camps. The optimists say it is the greatest productivity tool ever invented, a technology that will make us faster, smarter, and richer. The pessimists say it is a threat, that it will eliminate jobs, concentrate power, and perhaps endanger the species. Both sides assume the fundamental question is what AI will do to us.Neither side explains why we are so obsessed with it.Not interested. Not cautiously adopting. Obsessed. The speed of adoption, the emotional intensity of the discourse, the hundreds of billions in capital pouring in on terms that no sober analysis can justify, the eagerness with which people hand over functions that were until very recently considered irreducibly human. This is not the behavior of rational actors evaluating a productivity tool. Economists call it a bubble and leave it there. But a bubble is a description, not an explanation.I want to propose a complementary explanation, one that does not replace the economic or technological accounts but sits underneath them. It comes from René Girard, a French-born American literary critic and philosopher of social science who spent fifty years studying the structure of human conflict and desire. His framework, applied to AI, suggests that the extraordinary appeal of this technology has less to do with what it can do for us than with what it does to our relationships with each other.Specifically: it removes them.I should be clear from the outset that this is not a doomsday argument, and it is not an anti-technology polemic. It is an attempt to understand something that the standard explanations leave unexplained. If the argument is uncomfortable, that is because it is about us, not about the machine.For readers unfamiliar with René Girard, here is the core of his thought.Girard’s central claim is that human desire is not original. We do not decide what to want through some private, autonomous process of evaluation. We learn what to want by watching other people want it. Desire is imitative, or as Girard put it, mimetic. I want the job because my colleague wants it. I want the house because my neighbor has one like it. I want the life because someone I admire appears to be living it. Left entirely to ourselves, Girard argued, we would not know what to desire at all.Girard distinguished between two forms of this imitation. The first he called external mediation: the model I imitate is far enough from me, in status, in era, in world, that we can never come into conflict. A medieval Christian imitates Christ. A teenager imitates a musician she will never meet. The distance makes rivalry impossible. I borrow my desires from the model without the model becoming my enemy.The second form is what Girard called internal mediation, and it is where the trouble begins. In a democratic, egalitarian, hyperconnected society, the people I imitate are not distant figures. They are my peers. The colleague who got promoted. The neighbor who renovated. When I want what they want, and they want what I want, we become rivals. The person who taught me what to desire is now standing between me and the thing I desire.Here is Girard’s crucial insight: the rivalry is not an unfortunate side effect of imitation. It is inseparable from it. The closer the model, the more the admiration curdles into resentment. We do not hate our rivals despite imitating them. We hate them because we imitate them. And neither party can acknowledge what is happening, because admitting that your desires are borrowed from the person you resent is psychologically intolerable. So the rivalry intensifies beneath the surface, each side convinced their desire is original and the other person is merely in the way.Girard’s second major contribution is the scapegoat mechanism. When mimetic rivalry escalates to the point of crisis, communities have historically resolved the tension by directing their collective violence toward a single victim. The scapegoat is sacrificed, the community achieves temporary peace, and the cycle starts over. The mechanism works only if the community does not understand what it is doing. “To have a scapegoat,” Girard wrote, “is not to know you have one.”His third insight concerns modernity. The Judeo-Christian tradition progressively exposed the innocence of the victim and made the scapegoat mechanism visible. Once visible, it loses its power. We can no longer sacrifice our way to peace. But the mimetic pressure did not decrease. In a world of increasing equality, visibility, and interconnection, it intensified. Girard’s late work, especially Battling to the End, reads modernity as a civilization saturated with mimetic rivalry that it can no longer ritually contain.“The world is becoming more and more competitive, as individuals become more and more alike.” (René Girard, Battling to the End)Before Facebook, Instagram, X, and LinkedIn, internal mediation was at least bounded by physical proximity. Your rivals were the people you actually encountered: colleagues, neighbors, the parents at your child’s school. The pain was real, but it was contained.Social media destroyed the containment. It turned every human being on earth into a visible model and a potential rival. Your peer group became infinite. The signals of other people’s desire, their vacations, their promotions, their relationships, their confidence, became visible at all hours. Mimetic desire, which Girard had diagnosed in the pages of Stendhal and Dostoevsky, became the operating principle of the global information infrastructure.What followed was not connection. It was escalation. Everyone performing for everyone, everyone measuring themselves against everyone, everyone imitating everyone in a loop that fed on its own intensity. The platforms did not invent mimetic desire. They industrialized it.By the early 2020s, the pressure had become extraordinary. Polarization, status anxiety, professional burnout, the erosion of trust between colleagues, neighbors, citizens. Girard would have recognized the pattern: a sacrificial crisis at civilizational scale, with no functioning mechanism to discharge it.And then, right into this saturated field of rivalry, AI arrived.Here is the argument, stated as precisely as I can manage.AI is not simply a tool that happens to replace certain human functions. It is the instrument through which mimetic rivalry achieves what it has always aimed at: the elimination of the rival.The scapegoat in this story is not the machine. The scapegoat is the other human being. The colleague whose expertise you needed but whose expertise also diminished you, because every interaction with a knowledgeable peer carries the silent weight of comparison. The collaborator whose help you sought but whose competence reminded you of your own limitations. The mentor whose advice was valuable precisely because they understood things you did not, and whose understanding you quietly resented for exactly that reason.This is Girard’s double bind in its most intimate form: the model who helps you is the model who diminishes you. Every act of assistance from a peer is also an implicit assertion of superiority. You cannot receive help from an equal without registering, somewhere beneath the surface, that they occupy a position you wish you occupied.AI dissolves the double bind. It knows things without knowing more than you, because it is not a person and cannot occupy a position. It helps without the implicit claim of superiority that human help always carries. You can ask it anything without the social cost of admitting ignorance to a peer. The transaction is clean precisely because there is nobody on the other side of it.Nobody adopting AI is consciously thinking any of this. But Girard’s entire framework rests on the proposition that mimetic dynamics operate beneath consciousness, and that the people most deeply caught in them are the least likely to recognize what is driving their behavior. What they register is relief. And the relief is not about the time saved. It is about the absence of the person.A growing body of Girardian scholarship has begun to notice pieces of this picture. Paul Dumouchel, in the volume Desiring Machines (Bloomsbury, 2026), argues that the machine constitutes “a perfect model of the inaccessible flesh-and-blood model that fascinates us,” an indifferent, invulnerable figure whose mediating power comes from its perfect absence. This is close to what I am arguing, though I push the implication further: the fascination is not innocent. It is driven by the desire to be rid of the flesh-and-blood model entirely. Liam Magee, in a 2025 article in the journal Subjectivity, reads AI as a kind of fetish onto which collective desires and fears are projected. Magee sees AI as the target of projection. I am arguing that AI is the instrument through which the real target, the human rival, is made unnecessary.One important objection from within Girardian scholarship deserves mention. Jashiel Resto Quiñones, also in the Desiring Machines volume, argues from a Heideggerian perspective that AI cannot engage in genuine mimesis because it is not “truth-apt,” meaning it cannot meaningfully orient itself toward what it imitates. This is a serious point, but it does not touch the argument I am making, because I am not claiming that AI desires or imitates. I am claiming that human mimetic desire produced AI as its instrument.The standard economic account, that capital replaces workers to extract surplus value, is correct as far as it goes. But it cannot account for the emotional texture of what is happening, the strange eagerness, the missionary zeal, the willingness to declare entire human professions unnecessary with something that sounds less like regret and more like satisfaction. Mimetic theory does not contradict economics. It explains the affect that economics cannot reach.“The escalation to extremes is becoming the unique law of history.” (René Girard, Battling to the End)There is a darker layer to this argument, and it requires taking Girard’s logic one step further.Girard showed, across fifty years of work, that mimetic rivalry is inherently escalatory. The rivals become so consumed by their mutual antagonism that they lose sight of everything except the contest itself. In Battling to the End, he read Clausewitz’s theory of war to show that this escalation tends toward a terminal point: each side would rather destroy the field of competition entirely than allow the other side to prevail. He called this “escalation to extremes,” and he believed it was the dominant logic of modernity, a logic that political institutions were increasingly failing to contain.AI follows this logic to its conclusion, only in technological rather than military form.The executive who automates his competitor’s workforce knows, in some abstract way, that executives will eventually be automated too. But the mimetic satisfaction of making the rival unnecessary right now overwhelms the rational calculation about the future. The knowledge worker who uses AI to outperform colleagues knows that the same tool will eventually make knowledge work itself unnecessary. But the immediate advantage over the rival is more compelling than the long-term threat to oneself.This is not irrationality. This is mimetic logic operating exactly as Girard described it: each participant in the rivalry would rather escalate than let the other side win, even when escalation leads to mutual destruction. The destruction is just quieter than Girard imagined. Not warfare. Obsolescence. The rival is eliminated not through violence but through redundancy. Nobody has to feel guilty. The displaced worker, the unnecessary expert, the redundant advisor, they are not sacrificed in any visible sense. They are simply no longer needed.AI is not an exit from the mimetic trap. It is the trap reaching its conclusion.Girard was careful to insist that mimetic desire is not inherently destructive. It is, in fact, essential. “Fortunately,” he said, “people can no more give it up than they can give up food or sleep.” What he meant is that imitation is the mechanism through which human beings learn to desire at all: children learn from parents, students from teachers, apprentices from masters. The problem is not mimetic desire itself but what happens when the distance between model and imitator collapses and the imitation becomes rivalry. Desire is like appetite: necessary for life, but capable of becoming pathological.This matters because of what Girard called the relationship between desire and being. “All desire is a desire for being,” he wrote. We become ourselves, in part, through our models. Even the models who become our rivals contribute to the formation of our identity, because the friction of rivalry is also a form of recognition, the acknowledgment that another human being takes your existence seriously enough to compete with it.If AI progressively replaces the human rival across more and more domains of life, it does not simply remove the friction. It removes the mirror.A student who learns from an AI tutor rather than a human teacher is not merely receiving information from a different source. They are losing the relationship through which intellectual identity develops. The teacher was a model, sometimes frustrating, sometimes intimidating, but a model whose humanity made the encounter formative. The AI tutor reflects the student’s questions back as answers. There is nobody to become through the interaction.A professional who prefers AI assistance to a colleague’s help is not merely optimizing workflow. They are withdrawing from the mimetic relationship that, uncomfortable as it was, gave their professional life its structure and meaning. The colleague was a rival. But the rivalry was also a form of engagement, a sign that the work mattered to someone else.Pope Leo XIV’s recent encyclical Magnifica Humanitas, published just weeks ago, arrives at a remarkably similar concern from an entirely different tradition. Leo frames AI not as a mere technical challenge but as “one of the defining moral tests of the modern age” and calls for the technology to be “disarmed,” arguing that what is at stake is something essential about how human beings relate to one another. He presented the encyclical alongside a co-founder of Anthropic, in a deliberate gesture of dialogue between the Church and the industry. Leo does not use Girard’s vocabulary, but his diagnosis converges with it. When a pope and a mimetic theorist arrive at the same worry from opposite directions, the worry may be worth taking seriously.A civilization that systematically replaces its human mediators with artificial ones does not resolve the mimetic crisis. It dissolves the conditions under which human beings develop into human beings.Girard feared, across his late work, that mimetic escalation would end in apocalypse. He imagined this in the register of war: Clausewitz’s escalation to extremes reaching its terminal point in actual destruction.Perhaps the form is quieter than he imagined.Not a catastrophe but a slow, consensual, perfectly efficient replacement of every human function by a system that performs it without being anyone. Not destruction. Obsolescence. Not the end of the world, but the gradual emptying of everything that made the world a place where human beings could become themselves: the friction, the rivalry, the resentment, and the recognition that came with all of it.We are building a civilization in which you never have to ask another person for anything, in which every human mediator can be replaced by something that mediates without being someone, in which the rival is eliminated so cleanly that nobody notices the elimination is happening.The question to ask about AI is not whether it creates value. It does. The question is why we are so eager to hand over the parts of life that used to require another human being, and what the eagerness tells us about ourselves.Girard would have had a word for it. He would have called it what it is: mimetic desire, doing what mimetic desire has always done, consuming the very thing it cannot live without.MinervaShareNo posts

※ 著作権に配慮し、引用は冒頭3段落までです。続きは元記事をご覧ください。