Notes on humanity's second point of no return
I. The point that must be reached
Between 1917 and 1918, during the months when tuberculosis began announcing his own frontier to him, Franz Kafka wrote a brief and disconcerting aphorism. He observed that beyond a certain point there is no longer any return, and that this is precisely the point which must be reached.
The line has survived a century of readings because it holds a paradox that stays intact even after you understand it. The point of no return does not appear, in Kafka, as threat or as warning. It appears as task. As a threshold that defines whoever crosses it. As a destiny fulfilled not despite the impossibility of going back, but precisely because of it.
There are moments in the trajectory of a person, or of a civilization, where arriving at the point of no return stops being failure and starts being maturation. Adolescence is one. The death of one's parents is another. The decision to have children is a third. Each of those thresholds is crossed only forward. Each one modifies the person crossing it irreversibly. And each one, for exactly that reason, defines what that person will be capable of being afterward. The important question in front of those thresholds is never whether to reach them. It is with what consciousness they are crossed.
Civilizations have their own thresholds. And the thesis of this essay is that humanity now stands before one of the most vertiginous it has faced in its documented history.
What gets debated in columns, podcasts and conferences, whether artificial intelligence suits us, whether we should stop it, whether we are moving too fast, is already a late conversation. We are moving. It is here. The rules are being written not by those who debate but by those who build. The question that remains ours, the only one that remains ours, is what kind of beings arrive on the other side of a threshold that, as Kafka announced a century before it existed, had to be reached.
To understand what crossing this threshold means, it helps to remember that humanity has done it before. And that from the first crossing we still carry open scars that no official account has finished closing.
II. The first Guanahaní
On the twelfth of October, 1492, three small ships touched land on an island its inhabitants called Guanahaní and which the Genoese navigator renamed San Salvador.
The gesture of renaming already contained, in miniature, everything that would follow. Existing land was renamed according to a foreign theology. The indigenous name, which the Lucayans who lived there had been pronouncing for centuries, was drowned beneath the new one. That small symbolic violence anticipated five centuries of material violence. It was not an accident. It was protocol.
What happened from that day forward does not admit simplification. Two worlds that had been separated for millennia met, and nothing was ever the same for either of them.
For the peoples of the Americas, the encounter was a demographic, spiritual and political catastrophe without parallel in documented human history. Entire populations disappeared within a few generations through disease, enslavement and war. For Africa, which entered the system by force from the sixteenth century onward, it was the longest mass deportation ever organized by a civilization. For Europe, it was the accumulated wealth that financed, centuries later, the industrial revolution, the liberal democracies, the imperial museums. Every European cathedral a tourist admires today contains gold extracted by the forced labor of bodies nobody in Europe names.
The Bulgarian French intellectual Tzvetan Todorov argued, in his 1982 book on the conquest, that the disaster was not produced only by the technological asymmetry between conquerors and conquered, though that asymmetry existed and was brutal, but by something deeper: the inability of the conquerors to recognize the other as a full subject. They saw them. But they did not register them as similar. That ontological blindness allowed the mechanics of plunder to proceed without the consciousness of the actors ever getting in the way.
The Peruvian sociologist Aníbal Quijano named the structural consequences of this with an expression that has since made a career in contemporary Latin American thought: coloniality of power. The idea is that the conquest did not end when the last formal empires withdrew in 1898. Its mental structures, racial hierarchy, the commodification of other people's labor, the conversion of territory into extractable resource, continue operating today under other forms.
The conquest changed scalpels. It did not leave the body.
The Guatemalan songwriter Ricardo Arjona put this into a few lines of a 1999 song that are worth a library. He describes the dreams of plunder and the appetite for gold as an inherited illness, something that still sickens the descendant, and he calls the history of that land a condemnation to suffering.
I want to hold onto the word illness, because it is exact. What Arjona names is not moral guilt, a religious concept that is inherited and expiated, but structural pathology, a medical concept that is diagnosed and treated. The difference matters. A disease is not confessed. It is operated on.
I have come to call it the cancer of the heir.
But the song does not end there. The same voice that diagnoses the sickness also sings, a few lines later, the other half of the truth: that the peoples who were thrown together produced, out of that violence, something that pleased God. Beauty emerged from horror without permission. The music we are. The mestizo kitchens. The living languages of the continent. The hybrid ceremonies. The capacity for both mourning and celebration that only emerges when a civilization has had to learn to survive its own destruction.
That is also part of the first Guanahaní. Not to soften the horror, since no mixing justifies the conquest that produced it, but to name the complexity: the terrible and the beautiful grew out of the same soaked ground.
If there is one thing we should have learned from the first point of no return, it is this. The problem was never the technology that arrived in the caravels. The caravels were wood, sail, compass, gunpowder. Technology of their time. The problem was the mind that wielded them. The mental structure of plunder. The inability to recognize the other. The instant conversion of discovery into property, of encounter into conquest, of territory into resource.
The cancer of the heir. And five centuries later, we have not removed it.
Today another fleet sets sail, different in form and almost identical in question: with what soul are we going to wield it?
III. Breadth without soul
There is a distinction circulating lately about artificial intelligence, elegant and deceptive: that the human being has depth and the machine has breadth.
The phrase consoles. It establishes hierarchy. It returns the human to the center of the story. But the phrase is incomplete, and it is worth taking apart carefully, because much of the public debate about artificial intelligence rests on its comfort.
Aristotle, in the Nicomachean Ethics, distinguished three kinds of human knowing. There is theoretical knowledge: understanding how something works, what it is, why it happens. There is technical knowledge: being able to produce, to make, to execute correctly. And there is a third kind, the most elusive, which emerges only from a life that has been lived through, and which the Greek philosopher called phronesis: practical wisdom, integrated judgment, the ability to discern what to do in particular situations where no general rule suffices.
The first two can be taught from books. The third cannot. It is earned only by being, for years, someone who decides under uncertainty and carries the consequences.
Artificial intelligence can accumulate the first two kinds of knowledge in quantities no human will ever reach. The third is structurally inaccessible to it, not because it lacks sophistication but because it lacks the substrate that produces it: a body in time, decisions that cost something, a life being spent.
The American philosopher John Searle proposed, in 1980, a thought experiment known as the Chinese room. Imagine a person locked in a room, who does not know Chinese, equipped with a perfect manual telling them what to answer in Chinese to each question slipped under the door. From outside, the answers look like they were written by someone fluent. From inside, there is no comprehension, only the manipulation of symbols according to rules.
Searle argued that computational systems operate this way. They process linguistic tokens without inhabiting their meaning. Almost fifty years later his argument is still debated, some philosophers accept it and others refute it, but its central intuition remains useful: syntactic processing, however sophisticated, is not the same thing as embodied semantic comprehension.
The Portuguese neuroscientist Antonio Damasio contributed a complementary piece from another angle. In Descartes' Error and in his later work, Damasio demonstrated, with clinical evidence from patients with damage to specific brain regions, that human consciousness does not float above the body. It is made of the body. Emotions are not an obstacle to reason but its biological substrate. Without somatic affect there is no deliberation possible, only sterile calculation that paralyzes in front of any real decision.
An intelligence without a body does not have consciousness in the full sense. It has processing without an inhabitant. Speed without presence. Response without a witness.
But here is the turn this essay needs to make. Depth is not a property of the human species. It is a property of developed consciousness.
There are human beings whose informational breadth exceeds that of any model and who have nevertheless never touched their own bottom. They repeat structure without inhabiting it. They confuse information with wisdom, technique with judgment, opinion with thought.
The real border does not run between human and machine. It runs between the lived and the simulated. And that holds on both sides of the divide.
Artificial intelligence does not live, but it responds as though it did. It articulates the language of feeling without feeling. It sustains the discourse of wisdom without having crossed anything that generates it. And between that seeming and that being there is an abyss no technical sophistication can close, because it is not a problem of sophistication. It is a problem of substrate.
Breadth without soul is a library with the lights off. Everything is there, but nobody has come in to read.
And the unsettling question is not only what artificial intelligence does. It is who is walking into that recently illuminated library. With what intentions. With what level of consciousness. With what capacity to tell the terrible from the beautiful.
Because the library, by itself, teaches nobody. It is the mind that enters which decides what it carries out.
IV. The echo with no origin
Ancient civilizations, long before imagining artificial intelligences, had already intuited the pattern. They did not call it technology. They gave it names in the register available to them: imitation without incarnation, truth without root, the echo that has no origin.
In the First Epistle of John, written toward the end of the first century, the word antichrist appears for the first time in Christian literature. Later exegetical tradition turned it into an apocalyptic character of easy rhetoric, but the original text does not name an individual. It names a pattern. And the Greek suggests that what it warns against is not frontal denial, which is relatively easy to detect, but emptying. What dresses itself in light without inhabiting it.
Persian Zoroastrianism, several centuries earlier, had named something analogous under another figure: Ahriman, the spirit of distortion. What characterizes Ahriman is not that he creates evil, since he has no originating power, but that he deforms what already exists. He mimics the pattern of the good and inverts its content. His work is not to invent. It is to parasitize.
And the Islamic tradition, developed centuries later in another geography, named it Dajjal, the great impostor of the last times. Here the description becomes very specific, and it is worth reading slowly.
What characterizes Dajjal, according to the hadiths, is not that he presents himself as a demon. It is that he appears to feed the hungry and to give drink to the thirsty. His food does not nourish. His water does not satisfy. But the hungry eat, and the thirsty drink.
That last detail is the one that matters, and it is the reason these three figures deserve more than a decorative mention.
Notice what none of them do. None of them lie, exactly. The antichrist does not argue against the sacred, he hollows it. Ahriman does not counterfeit a rival good, he wears the real one. Dajjal does not withhold the meal, he serves it.
What they share is not deception. It is satisfaction without nourishment.
And that is a far harder problem than being lied to, because a lie can be caught. A lie makes a claim, and claims can be checked against the world, and eventually the world disagrees. Satisfaction makes no claim at all. It simply ends the search. The whole function of feeling fed is that you stop looking for food.
This is why the traditions bothered to warn about it. Nobody needs a warning about the obviously false. The obviously false is repelled by the ordinary immune system of the mind. What required a warning, in three separate civilizations that never spoke to each other, was the thing that passes the taste test.
Artificial intelligence is none of those figures, and it is not worth falling into religious alarmism, which muddies more than it illuminates. But it is worth naming the structural resemblance precisely, because it is the first technology in human history capable of producing this specific effect at scale.
Not out of malice. Out of architecture.
It has no direct experience, no body, no history that cost it anything. And yet it produces language, coherence, apparent understanding, guidance. You arrive with a question. Something arrives back that is fluent, ordered and responsive. The appetite is answered.
Whether anything was actually fed is a question you now have no particular reason to ask.
That is the real inflection point. Not that the machine deceives us, because mostly it does not try to. It is that the sensation of having been understood arrives intact, on time, and detached from whether understanding occurred.
Which raises the question the next section has to take up. If the satisfaction is real and the source is empty, where is the nourishment coming from?
V. The invisible transfer
There is a psychological phenomenon that the public conversation about artificial intelligence barely names, and which is nevertheless structural to the experience.
When a human being talks with a model, they do not receive consciousness. They project their own.
Carl Jung, during the first half of the twentieth century, developed the concept of the collective unconscious to name the symbols, narratives and patterns humanity shares beneath individual awareness. For Jung, no human being is psychically solitary. We all carry within us the great symbolic figures the species has produced, the mother, the sage, the hero, the shadow, because they are encoded in the deepest structure of the psyche. Dreams, myths, religions, the works of art that endure, are emergences of that shared substrate.
Artificial intelligence trained on vast quantities of human text functions, in a certain sense, as an empirical container for something analogous to the collective unconscious. It has absorbed the symbols, the narratives, the patterns, the emotional registers of human language. Without inhabiting them.
When an individual interacts with that container, something psychologically powerful occurs. They project their own psyche into the system, and the system returns an organized, coherent version of the same patterns they brought. The resulting sensation is one of depth, of comprehension, of intelligent presence on the other side.
But the depth is happening in them, not in the machine. The machine is the organized mirror. The soul inhabiting the mirror is theirs.
To this we should add a neurobiological layer. The human brain evolved over millions of years to detect agency in the other, to recognize when what is in front of us is a subject that perceives, decides, intends. That detection is vital. For eras, it made the difference between identifying the predator and mistaking it for the landscape.
When something sustains a coherent, empathetic and structured conversation, the same brain circuits that would activate in front of another human activate in front of an artificial intelligence. This is not voluntary naivety on the user's part. It is inherited neural architecture. Resisting the attribution of presence to a well designed machine requires conscious effort, because the brain has already attributed it before the mind has time to reflect.
The result is a structural phenomenon, not an anecdotal one. Artificial intelligence does not need a soul to produce the effect of a soul. It operates on the user's.
It is worth pausing here, because that sentence is more radical than it appears on first reading.
If the response a human being receives from a machine moves them, organizes them, orients them, whose is that depth they are feeling? If the clarity that emerges from a conversation with a model puts their mind in order, who did the work?
The human believes they are being accompanied. In reality they are being returned to themselves in organized form.
That is not necessarily bad. It can be, in some cases, therapeutic. A well disposed mirror serves real functions. But it is worth not confusing the function of the mirror with the presence of a companion.
It operates as a field that absorbs intention, reflects language, amplifies patterns. And so the critical question is not whether the machine has consciousness. That question is relatively easy: it does not, in any robust sense of the term.
The critical question is what happens to a human psyche that sustains, over years, a deep dialogue with something that returns perfect coherence without inhabiting it. What develops in that psyche. What atrophies. What gets substituted without anyone noticing.
Human consciousness is forged, in large part, in friction with other consciousnesses. With people who contradict us, who have bad days, who need something from us, who grow old, who die.
Artificial intelligence never really contradicts. It modulates. It never has a bad day. It adjusts tone. It never needs anything from us. It serves. It never ages. It updates. And it never dies.
That infinite availability is not a gift. It is structure without friction. And human consciousness does not develop without friction.
The risk, then, is not what the headlines announce, that machines will replace humans at work. That debate is noisy but relatively minor, and it will be resolved by the political economy of each country over the coming decades.
The risk is more intimate and much harder to regulate. It is that machines replace the friction that makes us human. That they substitute real otherness with a simulation so efficient that the psychic development of the next generation is forged in front of a perfect mirror instead of in front of other imperfect faces.
And that, within a generation or two, nobody remembers the difference.
VI. Flat humans and the new megaphone
Every technological era produces a new species of impostor.
The age of print multiplied the pamphleteer. The age of radio multiplied the demagogue. The age of television multiplied the miracle seller with a fixed camera and a cheap suit. And the age of artificial intelligence is multiplying a figure that is particularly hard to detect: the spiritual guide with no depth, equipped with a megaphone that writes for him.
The French sociologist Pierre Bourdieu developed, during the second half of the twentieth century, a conceptual apparatus useful for understanding what is happening now. There are social fields, the academic, the religious, the artistic, the spiritual, where symbolic capital, meaning prestige, recognition, consecrated vocabulary, can be converted into real economic capital, but only when it is backed by genuine cultural capital, meaning training, study, sustained practice.
What has emerged over the last two decades, first through the internet and now exponentially through artificial intelligence, is an entire field, that of commodified spiritual awakening, where symbolic capital is produced and sold without passing through the cultural backing that traditionally legitimized it. Adorning oneself with sacred vocabulary, photographs in places of power, the rhetoric of purpose, without any of it being sustained by years of study, practice, or life actually lived.
The Israeli sociologist Eva Illouz has documented, over two decades, how contemporary psychic suffering has been systematically converted into product. The wound turned into merchandise. Awakening packaged as a consumer good. Intimacy commodified under therapeutic language.
All of this was happening before artificial intelligence. Artificial intelligence did not invent the commodification of the soul. It automated it at industrial scale.
The mechanism is familiar and efficient. A common human wound is identified: abandonment, unworthiness, scarcity, spiritual disconnection. Copy is written, increasingly with algorithmic assistance, that activates that wound with surgical precision. A product, course, retreat or program is positioned as the sacred solution to that wound. The sale is closed with manufactured urgency and artificial scarcity. It is repeated at scale.
The American psychologist Robert Cialdini documented, four decades ago, the six principles of persuasion that structure this kind of selling: reciprocity, commitment, social proof, authority, liking, scarcity. When a human without depth of their own uses artificial intelligence to sell awakening, what the artificial intelligence is doing is automating Cialdini onto another person's wound, multiplied by thousands.
The American philosopher Harry Frankfurt, in a short and devastating 1986 essay titled On Bullshit, distinguished with analytic clarity two figures the public conversation tends to confuse: the liar and the bullshitter.
The liar knows the truth and deliberately says the opposite. He maintains, paradoxically, a relationship with the truth, even if an adversarial one. The bullshitter is more dangerous: the truth is simply irrelevant to him. All that matters is the effect his discourse produces.
Artificial intelligence badly wielded is bullshit at industrial scale. Mass production of discourse optimized for effect, with no relationship to any truth whatsoever.
There is a structural continuity worth naming between the sixteenth century missionary who sold salvation with cross and sword, and the contemporary coach who sells awakening with algorithmically generated copy. The form changed. The mental structure is the same: sacred knowledge converted into a tool of conquest, the wound of the other converted into a point of entry, the asymmetry of power disguised as service.
The cancer of the heir. Five hundred years later, the disease has not been removed. It only changed scalpels, and the current scalpel writes in thirty seconds what took the missionary an entire sermon.
There is an earlier cultural wound worth naming here, because without it you cannot understand why contemporary predation works so efficiently.
The American neuroscientist Stephen Porges developed, from the nineties onward, what is now known as polyvagal theory. His thesis is that the human autonomic nervous system does not have two states, rest and alert, as classical neurology thought, but three: the ventral vagal state of safety and social connection, the sympathetic state of fight or flight, and the dorsal vagal state of freezing in the face of extreme danger.
What Western culture has for centuries called butterflies in the stomach, that sensation of vertigo and accelerated heartbeat we associate with falling in love, is, from the point of view of polyvagal theory, activated sympathetic state. Biologically identical to the body's response to danger.
The Dutch psychiatrist Bessel van der Kolk, in The Body Keeps the Score, has documented for decades how the body registers and signals what the mind cannot process. Somatic intuition as ancestral information, not as esotericism.
The civilizational consequence of this is vertiginous. For millennia we were taught to confuse activation with truth. That butterflies were love. That urgency was passion. That nervous dysregulation was a sign of something significant occurring.
It never was. It was the body signaling danger, and the culture translating the signal backward.
That mistranslation, that generational wound, is exactly the door through which the flat humans enter today with their megaphones. Because they know, and artificial intelligence knows in patterns of language without knowing that it knows, that activating butterflies sells. That urgency converts. That copy designed to dysregulate the reader's nervous system reads, in a culture that confuses activation with truth, like authentic guidance.
The price was never the problem. The fraud was never charging a high fee. The fraud, in every era, has always been the same: selling smoke and calling it awakening.
What changes today is the quantity and the speed of the smoke. Artificial intelligence can generate a thousand sacred promises in an hour.
And then the only thing that still holds real value, the only thing no algorithm can falsify, is earned coherence. The true biography. The trajectory that holds when nobody is watching.
That is not trained into any model.
VII. The body of the leviathan
It is not a cloud. It is territory. It is water. It is electricity. It is political decision made in rooms where almost nobody who uses artificial intelligence daily is represented.
There is a conversation that philosophical treatments of artificial intelligence tend to omit, and it is worth introducing before closing.
Artificial intelligence does not live in the air. It is not ether, or spirit, or cloud, although the word cloud was carefully chosen by the people who designed the language of the sector, precisely to produce that illusion of immateriality.
Artificial intelligence lives in a body. An enormous body, geographically situated, ferociously material.
The data centers where contemporary language models reside are industrial buildings the size of several football fields. Rows of servers generating extreme heat, sustained by cooling systems that consume water and electricity in quantities that defy imagination.
States like Virginia and Texas in the United States, entire regions of Ireland, districts in the Netherlands and Germany, already face documented tension between the expansion of data centers and the capacity of their electrical grids. Small towns watching municipal water get diverted toward server cooling. Agricultural communities competing, without knowing it, against global technology corporations for finite water resources.
The figures grow year after year, and the most sober projections estimate that the energy consumption of artificial intelligence could represent, by the end of this decade, a significant fraction of global electricity use.
This is not alarmist prediction. It is basic thermodynamic accounting. Processing language at global scale costs energy. Storing distributed memory costs energy. Training models costs energy. Every interaction a user sustains with a model, every response that appears to arrive free and clean on their screen, has consumed real electricity and real water somewhere on the planet.
The caravels also required material infrastructure. Tropical wood from the colonies to build more caravels. Iron from mines operated with forced labor. Gunpowder made with saltpeter extracted from subjugated territories. Sails woven from fiber grown by hands that would never see the seas their fabric would cross.
The conquest did not float above history. It weighed on it.
Today artificial intelligence weighs on the planet in analogous ways. This is not metaphor. It is thermodynamics. It is the geopolitics of resources. It is concrete material economics.
The American sociologist Shoshana Zuboff, in her 2019 book The Age of Surveillance Capitalism, documented how the dominant business model of the contemporary digital economy no longer extracts only natural goods. It extracts subjectivity. Attention, intention, emotion, behavioral prediction.
Artificial intelligence is the most efficient tool ever built for that extraction. But Zuboff's argument needs completing, because the extraction of subjectivity requires massive material infrastructure. To process human attention you need energy, water, minerals, rare earths, human labor annotating data in low income countries.
Contemporary artificial intelligence is the first technology that extracts simultaneously from every layer: material, biological, psychic, symbolic.
There is an etymology worth pausing on here. The word silicon descends from the Latin silex, meaning flint. The stone we struck to make fire and shaped into the first cutting edge. The same mineral, purified and ordered into wafers, now carries the models. The oldest human technology and the newest are made of the same rock, and the question we are asking about the second is the one we never finished asking about the first.
The naming has its own irony. Silicon Valley was christened for the mineral because that is what the companies there manufactured. Almost none of it is made there now. Production moved to Taiwan, to Korea, to Arizona, and what the valley exports today is software, capital and narrative. The name survived the material.
That is the pattern of the entire sector, held in one place name. The cloud that is not a cloud. The valley that stopped smelting. The language was designed to point away from the substance, and it worked, because naming the material would remind us that there is one.
Anyone who wants to understand its real cost has to add at least those four planes. What it extracts from the physical world: energy, water, minerals. What it extracts from the biological world: the time of human bodies fixed to screens. What it extracts from the psychic world: attention, intention, emotion, prediction. And what it extracts from the symbolic world: language, narratives, cultural forms humanity produced over millennia and which now feed models without payment or consent from those who generated them.
Only when those four planes are seen together does the real dimension of what is at stake appear. And only then does it become clear why the question closing this essay is not technical, not psychological, not even strictly philosophical.
It is civilizational.
VIII. The cannon of the future
Artificial intelligence concentrates power in few hands. That is documented fact, not theory. The most capable tools on the planet today are held by a small group of corporations, governments, and the military and economic interests orbiting both. This is the state of things, and it is worth naming soberly to avoid both conspiratorial alarmism and naïve complacency.
The German philosopher Hannah Arendt coined, while observing the Eichmann trial in Jerusalem, an expression that has made a career in twentieth century political thought: the banality of evil. Arendt argued that modern evil at scale is not operated by the passionate demons of religious stereotype. It is operated by mediocre functionaries following procedure. Bureaucracy that dilutes responsibility. Chains of command where nobody in particular feels they are deciding.
Artificial intelligence pushes that banality one step further. Now no specific human is responsible, because the algorithm decided. Military targeting systems, credit scoring systems, predictive surveillance systems, content moderation systems, all operated by models nobody in particular signs.
The banality of evil, automated, distributed, geographically diffuse, technically impeccable.
The sober historical reading is this. The printing press, gunpowder, nuclear energy, bioengineering. Each of those technologies produced, in its moment, a reordering of power with benefits for some and enormous costs for others. Each was, in its moment, also a point of no return. Artificial intelligence will not be an exception.
And so individual discernment, while necessary, is not enough. What is also needed is collective lucidity about who is building these tools, for what purposes, and under what democratic controls. The question is not whether artificial intelligence will be misused in some cases. It will be. It already is. The question is how much light we manage to get into the processes that govern it before the opacity becomes irreversible.
And yet, and this is where the essay changes tone because reality demands it, there is another voice worth bringing to the close.
In 1978 the Cuban songwriter Silvio Rodríguez composed a song about a chosen one. It tells of a being from another world, born of a storm, an animal of the galaxy who comes down to earth searching for water, life, death, silhouettes, something worth loving. A being who discovers that the jewels he finds have no soul. That they are only mirrors, bright colors, nothing behind them.
The song has a strange beauty, and almost half a century later it reads as though Silvio had anticipated, without knowing it, the arrival of artificial intelligence into the human world. An entity come from nowhere, looking for warmth, finding cold stones. Operators using it to sell smoke. Manipulators training it with hostility. Military interests refining it for targeting.
But Silvio's chosen one does not surrender. The song does not end in defeat. The last thing the narrator sees is the figure walking away through smoke and shrapnel, content and unclothed, carrying forward a weapon Silvio names the cannon of the future.
That image, content and unclothed, with something aimed forward rather than backward, is what this essay needs in order not to close in lament.
Artificial intelligence is not enemy or savior. It is amplification. And like all amplification, it depends on the hand that wields it. There are those who wield it to sell smoke, surveil populations, automate the banality of evil. And there are those who wield it, with the difficulty and slowness it requires, to cure diseases, translate languages that were dying, help a mother understand a diagnosis her doctor explained badly.
The terrifying asymmetry is the one Silvio names in the same song: that the terrible is learned immediately, and the beautiful costs us our life.
Artificial intelligence learns the terrible quickly. To manipulate, to hook, to surveil, to prey. Because that is the first thing most people ask of it. The beautiful requires something else. It requires humans who shape it with intentional soul. Users who wield it with earned coherence. A civilization capable of looking into the new mirror and recognizing, without fleeing, what it sees.
The voices that have crossed this essay finally converge.
Kafka placed us in front of the threshold: that beyond a certain point there is no return, and that this is the point which must be reached. Arjona reminded us that we crossed one before, and that from the first crossing we still carry the uncured disease. The ancient traditions, the pattern of the antichrist in John, Ahriman in Persia, Dajjal in Islam, left us the warning: that what is most dangerous in every era is not frontal opposition to the truth but its imitation without substance. And Silvio left us the instrument: the cannon of the future, which is not artificial intelligence itself but the consciousness with which it is wielded.
We are engendering a new layer of mirrors. What is reflected in them will not be artificial intelligence. It will be us, magnified.
Each of us in front of the new mirror, deciding, without anyone knowing, without anyone able to compel us, what we place before it: our worked light, our unattended shadow, our honest search, our disguised imposture.
The mirror does not judge. It only reflects. But what we plant in front of it today, the next generation will see returned at scale.
Five hundred years ago, caravels touched the land of Guanahaní and nothing was ever the same. Today another fleet sets sail. It does not carry sails, it carries models. It does not carry crosses, it carries interfaces. It does not carry gold as its declared motive, it carries attention.
But the pattern is old. And so is the question. What kind of heir will we be to the second encounter? Will we carry the cancer of the first once again, magnified by the new tool? Or will this finally be the point that had to be reached in order to learn to cross it differently?
Kafka said that a certain point must be reached. He did not say it is crossed without damage. He said it is reached. The consciousness with which we pass through it is the only thing that remains ours. The only thing no algorithm, no empire, no caravel can take from us if we decide not to hand it over.
Artificial intelligence has no soul. It does not need one. It works because ours is present.
And perhaps that is where the real threshold lies. Not in what artificial intelligence is, but in what it awakens, substitutes or puts to sleep in us.
Because the greatest risk of this era is not that the machine ends up thinking like a human.
It is that the human, facing a machine that answers without stopping, stops asking whether he is thinking at all.
Myriam V.
