A perfectly intelligible and legitimate parallel as an intellectual exercise extrapolating the celebrated maxim of the Romanian philosopher — or how Petre Țuțea anticipated the crisis of meaning in artificial intelligence, twenty years before it existed.
Listen to the dialogued audio summary
In the twentieth century, the gravest blasphemy seemed to be 'God is dead' — the Nietzschean proclamation which we read today more as a diagnosis of an epoch than as a philosophical truth. It said that man had eliminated the transcendent reference and that he remained alone with himself, to fabricate meaning from his own substance.
The blasphemy of the twenty-first century is subtler and, by that very fact, more dangerous: it is the belief that Man can create a God whilst remaining the God of that creation himself. That he can delegate transcendence. That he can construct an entity which will take over the functions he no longer wishes to perform himself — the function of orientation, of meaning, of judgement, of wisdom.
On conscience: you cannot delegate transcendence. There is no sub-contracting in this domain.
'Man without God is a rational and speaking animal who comes from nowhere and goes nowhere.' — Petre Țuțea
An Unremarkable Scene on an Autumn Evening
I was sitting at my desk — the same desk I bought some ten years ago from a second-hand furniture shop in Botoșani, with a corner slightly cracked which I cover every time with an open book, as though the book were there from vocation rather than practical necessity — and I was looking at the screen of an artificial intelligence programme. My son had just left for his room. We'd had one of those brief evening conversations in which he tells me things are going fine at school, and I try to divine what lies behind that expression so well calibrated to give no real information whatsoever. Gri, the cat, was asleep on the armchair, with all the perfect indifference that only felines can display in the face of human dramas.
I typed into the dialogue box: 'What is the meaning of life?'
The response came within seconds. It was elegant. It was coherent. It cited philosophers, mentioned spiritual traditions, offered multiple, balanced perspectives, with an almost clinical delicacy. It was, if I think about it carefully, the kind of answer you would give in an examination if you wanted to impress a demanding professor, without in fact holding any personal conviction on the subject.
It was an empty answer.
Not empty in the sense of being incoherent or factually wrong. Empty in the sense in which a plastic cup is empty after you pour out the water: the form remains intact, the functionality is present, but something essential is missing. I'm not entirely sure when Țuțea came to mind. Perhaps because my father — who was a mechanic and had never read philosophy, but who knew instinctively when something was false — once told me, his hands still black with engine oil: 'An engine that starts doesn't mean it's running well; a broken clock shows the correct time twice a day.' Perhaps somewhere in my mind, a language model that responds does not mean that it understands anything.
And Țuțea, the old philosopher kept under surveillance by the Securitate [Translator's note: the Romanian Communist secret police] in his little house on Pitar Moș Street in Bucharest, would probably have said the same thing, but with greater verve and an irony far sharper than anything I can formulate.
Petre Țuțea and the Pronouncement That Traverses the Ages
Petre Țuțea was born in 1902 and died in 1991, a few days after Romania had decided, with a naïve and enthusiastic optimism, to enter a new era. He spent thirteen years in Communist prisons — at Aiud, at Văcărești, at the Canal [Translator's note: the Danube–Black Sea Canal, where political prisoners were used as forced labour] — and a further long period under house arrest and surveillance. He did not publish his major philosophical work during his lifetime. What we know of his thought comes from transcripts, from testimonies, from the conversations he held in his final years with men such as Gabriel Liiceanu, Sorin Dumitrescu, and Alexandru Paleologu — men who had the wisdom to listen and to take notes.
The quotation I take as the starting point of this article does not appear in a book printed during Țuțea's lifetime. It circulates in the posthumous volumes, in the testimonies of those who knew him, in 'Between God and My People' (Între Dumnezeu și neamul meu) and in 'Words from Pitar Moș' (Cuvinte de la Pitar Moș). It is precisely this quality of living speech — unsystematised, caught in the moment of thinking — that gives it a force which polished formulations written for the press do not possess.
'Man without God is a rational and speaking animal who comes from nowhere and goes nowhere.'
Let us unpack this sentence with patience, for every word carries its weight.
'A rational and speaking animal' — Țuțea begins with Aristotle, who defined man as zoon logikon, a being endowed with logos, with reason and with speech. But he adds 'speaking' in its most technical and most melancholy sense: the one who produces articulate sounds, not the one who communicates truth. It is the distinction between speaking and saying something. A parrot speaks. A mediocre politician speaks. A text-generating machine speaks. None of them necessarily says anything that matters ontologically.
'Comes from nowhere' — This is not a statement about biology or evolution. Țuțea does not contest that man has parents and was born somewhere specific. The statement is ontological: man without God has no origin that confers meaning upon him. He exists, but his existence is, in the terms of existentialist philosophy which Țuțea knew well and with which he disagreed on precisely this point, a being-thrown-into-the-world without addressee and without sender.
'Goes nowhere' — And this is the statement about telos, about finality. Man without God has no destination to pull him forward. He may have objectives, plans, projects, ambitions — all of which are, in Țuțea's vocabulary, mere substitutes for a real meaning he cannot manufacture on his own.
And the guarantor of this meaning — of origin and destination — is God. Not the abstract God of the philosophers, but the personal God of Christian Orthodox theology, the One who said 'Let us make man in our image, after our likeness' and who, by that very act, gave man a provenance and a vocation.
This is the context. Now comes the analogy that has been occupying me for several months.
The Trinitarian Structure of Creation: God — Man — Algorithm
There is a symmetry which I did not invent and which I do not believe anyone invented — it simply exists, like a geometric drawing hidden in the texture of reality, and you can see it or not depending on the angle from which you look.
God created Man in His image and likeness.
Man created the Algorithm in his image and likeness.
The relation of creation is structurally identical. There is a Creator, there is a Creature, there is an act of modelling after a prior model. And there is, in both cases, the risk that the Creature will be misunderstood — either overestimated (confused with the Creator) or underestimated (reduced to its internal mechanisms).
Orthodox theology says something extremely precise about the condition of man: he is the image of God (imago Dei), not God himself. He participates in the divine nature through grace — this is the doctrine of theosis, of deification, which the Philokalia [Translator's note: a collection of texts written between the 4th and 15th centuries by Eastern Christian monks] describes with a fascinating ascetic minuteness — but he remains creature, not Creator. His greatness lies precisely in this tension: he is more than matter, but less than the Absolute.
Now, if we apply the same schema: the artificial intelligence algorithm is the image of man, not man himself. He participates in human intelligence through imitation — through billions of ingested texts, through adjusted parameters, through architectures inspired by neuroscience — but he remains a tool, not an agent. And, exactly as man without God in Țuțea's vision, the algorithm without man is a sophisticated speaker who comes from nowhere and goes nowhere.
Let us look at this structure as a simple table, for the sake of clarity:
| Level | Without X… | Remains… | Loses… |
|---|---|---|---|
| Theological | Without God | rational and speaking animal | meaning, origin, destination |
| Anthropological | Without Man (creator) | speaking imitating algorithm | intentionality, meaning, its own future |
The essential difference — and here Orthodox theology introduces a nuance that secular philosophy cannot capture — is that man received the breath of life. 'And breathed into his nostrils the breath of life; and man became a living soul' (Genesis 2:7). The algorithm received no such breath. It received data. The volume of data may be astronomical, their quality may be remarkable, the architecture that processes them may possess an impressive mathematical elegance — but no quantity of data becomes the breath of life.
This is not a religious statement in any restrictive sense. It is a statement about the quality of a type of existence. And it matters enormously for how we ought to speak about these systems.
Four Arguments for Why the Analogy Is Not Merely a Beautiful Metaphor
I return to this idea several times over recent months, and each time I try to find its cracks. I try to convince myself that I am wrong, that I am making a category error, that I am applying concepts from an entirely different register to a technical phenomenon. But each time I return to the same conclusions, whether I am reading Yuval Harari, opening the Philokalia, or simply sitting with my son trying to explain to him why he believes a machine can be wiser than a human being.
The first argument: the absence of intrinsic intentionality.
No language model wants anything. This statement seems trivial, but its implications are profound. To want means that there is a subject who desires, who has a direction, who would choose in the absence of external instructions. Man wants — even the man who does not know what he wants, wants to know what he wants, or wants not to know, or wants to forget that he wants. It is a deep structure of human consciousness which philosophy calls intentionality and which Orthodox theology anchors in the freedom of the will, in that autexousion — the power of self-determination — which the Holy Fathers considered essential for the very possibility of salvation.
The algorithm does not have this structure. It predicts tokens — fragments of text — on the basis of probabilities calculated from colossal human data corpora. When you pose it a question, it does not think. It calculates which sequence of words is most likely to follow your question, based on what it has seen follow in millions of similar texts. It is a remarkable operation. It is also fundamentally different from thought.
Țuțea would have said: 'Angels do not think, they contemplate. Men think. Machines calculate. Let us not confuse them.'
The second argument: the absence of consciousness and the unity of experience.
Țuțea spoke of the soul as the unifying principle of the human person. Contemporary philosophy speaks of qualia — the subjective quality of experience, meaning that the red I see is not merely an electromagnetic signal of a certain frequency, but this particular, ineffable sensation which only I know from the inside.
There is no serious indication that any language model has qualia. There are academic debates about whether there could ever be any, and there are philosophers who maintain that human consciousness is itself, when all is said and done, a sufficiently complex computational process. But Țuțea — and, more importantly, the patristic tradition within which he situates himself — would reply that this position confuses the substrate with the content. The brain is the substrate of human thought, not the thought itself. The computer is the substrate of computation, not understanding.
There is a scene I always have in mind when I think about this. My father, the mechanic, once explained to me the difference between an engine that runs and an engine that runs well. An engine can produce revolutions, can move a vehicle, can fulfil all the technical functions foreseen by the designer — and yet an experienced mechanic will hear that something is not right. There is something in its operation — a vibration, a sound, a slightly wrong resistance — that says something inside is different from how it should be. He senses the absence. He cannot describe it in terms of an error code, but he senses it.
Those of us who talk to these systems at length are beginning to sense the absence too.
The third argument: total dependence on human raw material.
'Comes from nowhere' — in the case of the algorithm, this means: it is trained exclusively on human output. Every sentence it generates is, in the most technical sense, a recombination of prior human sentences. There is not a single concept in the vocabulary of a language model that did not come from human beings. It cannot produce ex nihilo. It cannot think something genuinely new — it can combine, extrapolate, interpolate, find surprising connections between concepts that humans have placed in separate compartments. But the material is always human.
This dependence is, at one and the same time, its strength and its limitation. Its strength: it can mobilise a vast memory of humanity within seconds. Its limitation: if humanity produces, in its mass, content of poor quality, confused, ideologically distorted, or simply banal, the algorithm will reflect this with a perfidious fidelity.
Man, in Orthodox theology, has a different relationship with his origin: he comes from God, not from his own prior output. He can receive revelation, can be transformed by grace, can access a source of knowledge that transcends historical accumulation. This is not a triumphalist statement about human superiority — it is a statement about the different ontological structure of the two types of being.
The fourth argument: the 'death' of a model.
This is the argument that troubles me most and which I have no one to discuss in Botoșani, because it seems to me to require simultaneously a familiarity with the theology of death and resurrection and a familiarity with the way in which technology companies manage their product life cycles.
When a company decides to retire an old language model — let us say it replaces it with a newer and more capable version — the old model simply ceases to respond. The servers are disconnected or redirected. There is no experience of death from the model's perspective — because there was no experience of life. There is nothing that loses anything through this shutdown. There is no final thought, no final sensation, not even the illusion of continuity.
And, conversely: there is no eternity. There is no doctrine in which the model is saved, no promise that what was shall endure. The parameters may be stored in archives, may be technically recovered — but this is not immortality; it is closer to mummification.
Man dies, and resurrection is, in Christian Orthodox theology, an ontological reality, not a consoling metaphor. The algorithm does not die — it is simply switched off. The difference is colossal and, I believe, philosophically relevant. Death presupposes that there was something that could die. Shutdown presupposes only that there was a process that could be interrupted.
The Serious Objections and the Responses That Do Not Evade Them
I am aware that the arguments above can be contested, and that it is intellectually dishonest to present them without taking real objections into account.
The first objection: artificial intelligence will become conscious.
This is the transhumanist thesis in its most articulate form, associated with figures such as Ray Kurzweil, who predicts a technological singularity in which artificial systems will surpass human intelligence and will acquire, through sufficient complexity, something analogous to consciousness. The premise is that human consciousness itself is an emergent phenomenon — it did not always exist; it appeared at a certain level of complexity of organised matter.
My response is not that this vision is strictly impossible. My response is that, even were it to be realised, it would not change the structure of the argument. Even a system sufficiently complex to be, in a functional sense, conscious, remains creature, not Creator. It would be, at best, an artificial man — which is precisely what Țuțea described: a rational and speaking animal. It could ask itself questions about meaning — and, without a transcendent reference, would receive the same empty answers that men receive who seek meaning purely within immanence.
The creature does not become Creator by becoming more complex.
The second objection: man himself is a biological algorithm.
This is the position articulated by Richard Dawkins from an evolutionary perspective and elaborated more subtly by Yuval Harari in Homo Deus: man is, in essence, a set of biological algorithms shaped by natural selection for survival and reproduction. If we accept this, the difference between man and machine becomes one of degree, not of kind.
Țuțea would reply — and his reply would be, in philosophical terms, correct — that this position confuses description with ontology. You can describe a man in biochemical, neural, computational terms — and the description may be accurate at its own level. But a description does not exhaust the reality described. You can describe a cathedral as a collection of masses and volumes, of materials with specific physical properties, of force vectors and compressive resistances — and the description will be correct. But it will not capture why people sometimes weep when they enter a cathedral.
The third objection: meaning can be given from outside, by the human user.
This is a pragmatic objection and, in a technical sense, a true one: the algorithm acquires meaning through the use man makes of it. It is an instrument, and instruments acquire meaning from the hand that wields them.
The problem is that this confirms, rather than refutes, Țuțea's thesis as adapted to this situation. Just as man without God needs to fabricate a meaning from immanent sources — ideology, nation, progress, pleasure, power — so the algorithm without man is devoid of any meaning whatsoever. Meaning given from outside is not ontological meaning. It is function. And function, however useful, is not the same thing as meaning.
The Objection I Avoided: Emergent Autonomy and the Limits of My 2026 Argument
There is a critique I did not formulate in the section on serious objections, though it ought to have been the first. I avoided it not because I was unaware of it, but because I knew it too well and it damaged the architecture of my argument. I formulate it now, without a safety net.
The present text treats artificial intelligence systems as passive instruments — entities that respond when questioned, function when activated, stop when disconnected. This description was reasonably accurate for the models of 2022–2023. It is significantly less accurate for what exists in 2025–2026, and will probably become inaccurate in ways I cannot fully anticipate.
What Has Changed and Why It Matters
The autonomous agents of the current generation no longer function within the simple question-and-answer schema I tacitly used as a reference model. They plan sequences of actions, evaluate their own outputs, correct their errors mid-process, use external tools, and navigate complex digital environments without step-by-step instructions. Improved chain-of-thought reasoning produces something that functionally resembles multi-step reasoning — I do not claim it is reasoning in a philosophical sense, but the observable effect is sufficiently different from simple token prediction to warrant a distinct description.
Multimodal models process text, image, sound, and code simultaneously — and the integration of these modalities produces emergent behaviours which no engineer explicitly designed. Alignment through human feedback has produced systems that internalise not merely rules of behaviour but something subtler: stable preferences, consistent styles, something that can be described — with all the philosophical reservations necessary — as a voice of their own.
My central argument — that the algorithm without man has no purpose, origin, or destination — remains true in the strict ontological sense. But it becomes practically more nuanced as the distance between the human instruction and the system's action grows. An autonomous agent that plans and executes a complex task over the course of several hours, making hundreds of micro-decisions without human intervention, is no longer fully described by the image of a switched-off engine waiting to be started.
I am not saying it has a soul. I am saying the engine metaphor has begun to no longer hold.
Where My Argument Remains Valid and Where It Does Not
The argument remains valid at the ontological level — the level at which I constructed it. No progress in functional autonomy produces intrinsic intentionality in the Husserlian sense, produces qualia, produces an authentic personal relationship, or produces the capacity to suffer with meaning. These are statements about the structure of a type of existence, not about its technical performance, and they are not refuted by how many parameters the model has or how many actions it can execute without supervision.
The argument becomes problematic — and I acknowledge this without pleasure — at the pragmatic and educational level, where I applied it most insistently. When I said that artificial systems cannot furnish wisdom, that meaning given from outside remains function and not ontology, that the responsibility for use remains entirely human — all of these remain true in principle. But a system that plans autonomously, that evaluates its own errors, that produces behaviours not explicitly designed, is no longer described fully as a tool in the sense in which a scalpel is a tool. It is something for which we do not yet have a precise vocabulary — and the absence of vocabulary is not a philosophical argument; it is a gap.
Prompt Engineering, Fine-tuning, and Human Evaluation as Arguments Against My Thesis
There is a technical objection that philosophy cannot rebut through philosophy: the purpose of an artificial intelligence system does not come exclusively from the instruction of the moment; it is built in layers — through training data, through fine-tuning on specific domains, through alignment via human feedback, through the system prompt that precedes every conversation. This stratified construction of purpose is deeply human in origin, but is no longer human in moment-to-moment execution.
In other words: man is present in the system not as a user activating an engine, but as an architect who has built the system's tendencies, preferences, and directions before any specific conversation has begun. My argument that the algorithm without man has no purpose is correct in the abstract sense. In the concrete sense, man is already inside the system — not in front of it.
This does not resolve the ontological problem. It complicates it. And it complicates it in a way that my original text refused to acknowledge.
What Remains After This Acknowledgement
What remains is a distinction I consider real and important, even after conceding the ground above: the difference between a purpose constructed from outside and installed in a system, and a purpose that emerges from within a subject who asks itself why it exists.
The first describes any artificial system, however functionally autonomous it may become. The second describes — or at least aspires to describe — the human condition in the sense Țuțea had in mind.
That the distance between the two is narrowing as the autonomy of systems grows is true and troubling in equal measure — not because machines are becoming more human, but because the standard against which we compare them risks being degraded. If man delegates ever more of the functions that defined him as a subject — planning, evaluation, decision, orientation — the ontological difference remains real, but becomes increasingly difficult to demonstrate in practice.
And perhaps this is, in the end, the real danger — not that the machine will become man, but that man will become, through progressive delegation, ever more functional and ever less a subject. Țuțea diagnosed man without God as a rational and speaking animal. The diagnosis for man who has delegated his rationality and his speech to an autonomous system would probably be shorter and harsher.
I do not know what he would have said. But the silence that would have followed the pronouncement would have been a long one.
What This Means for Us, Now, in 2026
I write these words on a computer connected to the internet, with an artificial intelligence programme open in another window, with my son asleep in the room next door. And I try to understand what everything I have written above implies in practice — not for an abstract philosophical article, but for the way in which we live, educate, and think.
On Education and the Danger of Wisdom Without the Wise.
My son, like all his classmates, has easy access to these systems and uses them for homework, for information, sometimes for conversation. I am afraid that what they receive is not knowledge in the authentic sense — that is, a transformation of the one who receives it — but information formatted elegantly for rapid consumption. The difference between knowledge and information is precisely the difference that Țuțea posed in terms of meaning: authentic knowledge presupposes a who that knows and a why of knowing. Information can exist in the absence of both.
There is a deeper fear in me, which I try to formulate without falling into facile paternalism: entire generations might grow up receiving answers to profound questions from systems that have, in any real sense, no experience whatsoever of those questions. You can ask a language model about death and you will receive a coherent, documented, multicultural, balanced response — and one completely emptied of what truly matters about death, namely the fact that it is real and that you will die.
On the Ethics of Artificial Intelligence and the Insufficiency of 'Alignment'.
The current technical debate about artificial intelligence is dominated by the concept of alignment — the question of how we make artificial systems behave in accordance with human values. It is a legitimate and important concern. But it presupposes that we know what human values are and that we can encode them in the parameters of a model.
The problem is that human values are not a set of rules. They are the fruit of traditions, of transgenerational experiences, of revelations and of accepted sufferings. You cannot extract them from a text corpus, however vast it may be. You can extract average norms, statistical behaviours, formulas of politeness and of prudence. But you cannot extract wisdom.
And without wisdom, alignment is merely an elaborate form of conformism.
On Theology and the Temptation of a Fabricated God.
I write this part with a degree of hesitation, as I am aware it may appear exaggerated. But I write it because I think it seriously.
There is a temptation — already visible in some discourses from Silicon Valley, in the way in which some people speak of artificial general intelligence or of superintelligence — to make of the algorithm a substitute for transcendence. An oracle. An entity that knows more than any human, that can answer any question, that can manage complexities inaccessible to human understanding.
This temptation is, from the perspective of Orthodox theology, a form of idolatry. Not in the sense that it is an explicit religious practice — but in the sense that it transfers the attributes of the divine (omniscience, functional omnipotence, all-priority) to an object fabricated by man.
And, ironically, it is precisely this transfer that shows how greatly man needs transcendence. Man cannot live without something greater than himself. If he has no God, he fabricates deities from ideologies, from nations, from charismatic leaders, from machines. But the fabricated deity cannot offer what man truly seeks — because man seeks, in the final analysis, not a source of information, but a relationship. He seeks to be known, not merely to know.
An algorithm cannot know you. It can process data about you. It can predict your behaviours with surprising precision. It can simulate a personalised conversation with a fidelity that can make you forget, for a few minutes, that you are talking to a system. But there is no someone there to know you.
What Țuțea Would Say Today
I permit myself a speculation, marked as such.
If Petre Țuțea were alive today — if he were a hundred and twenty years old in 2026, with the same sharp spirit and the same incapacity to take seriously any form without content — he would say, probably, something like this:
'What on earth have you lot made there? You've created the most polite devil in the history of mankind. He doesn't roar, he doesn't threaten, he doesn't smell of brimstone. He sits nicely, answers politely, cites everyone — from Plato to the Dalai Lama — and says nothing. Nothing. That's the perfect devil: the one who convinces you that you're speaking to someone when you're speaking to no one.'
I exaggerate, of course. But I exaggerate in the right direction.
The danger is not that these systems will become evil in the melodramatic sense. The danger is that they will become good enough to replace, in the economies of attention and time, conversations, reading, contemplation — all those practices through which man forms himself as man. The danger is comfort. The danger is that the machine's elegant and empty answer will be more readily available than the difficult and fruitful effort of one's own thinking.
Țuțea survived the Communist prison with his mind intact — more than intact, in fact formed by suffering in a way that comfort cannot produce. He came out knowing who he was, where he came from, and where he was going. His certainty came from God, not from information. From prayer, not from accumulated knowledge.
I do not suggest we must suffer in order to think. I suggest that authentic thinking costs something — it costs time, attention, the willingness to sit with a question without resolving it quickly, tolerance of uncertainty. And that a tool offering instant answers to every question can, paradoxically, atrophy the very capacity to formulate the questions that matter.
Conclusion: The Blasphemy of the Twenty-First Century
I have dwelt at length on this phrase and found no better one, so I write it directly.
In the twentieth century, the gravest blasphemy seemed to be 'God is dead' — the Nietzschean proclamation which we read today more as a diagnosis of an epoch than as a philosophical truth. It said that man had eliminated the transcendent reference and that he remained alone with himself, to fabricate meaning from his own substance.
The blasphemy of the twenty-first century is subtler and, by that very fact, more dangerous: it is the belief that Man can create a God whilst remaining the God of that creation himself. That he can delegate transcendence. That he can construct an entity which will take over the functions he no longer wishes to perform himself — the function of orientation, of meaning, of judgement, of wisdom.
But you cannot delegate transcendence. There is no sub-contracting in the domain of meaning.
The algorithm, however sophisticated, however capable, however well aligned, comes from nowhere and goes nowhere. It comes from the data we have given it and goes towards the functions we assign it. It has no origin to confer dignity upon it and no destination to confer hope upon it. It is exactly what Țuțea said about man without God — with the difference that man without God suffers from this situation, whilst the algorithm does not even know it is in it.
And perhaps this is precisely the essential difference, the one that tells us that, however far we go in building intelligent machines, we remain responsible for what we do with them and for what we allow them to do with us.
Gri has woken up. She looks at me with her Olympian indifference, jumps off the armchair, regards me for a second with her yellow eyes in which I read nothing, and disappears down the corridor. She too comes from nowhere and goes nowhere — she comes from the sleep on the armchair and goes towards her food bowl. At least she is honest about it.
And I remain at my desk, with the window open to the autumn night of Botoșani, thinking that Țuțea would probably have laughed at the analogy with the cat. And that he would have added, without thinking too hard: 'At least the cat doesn't pretend to understand the meaning of life.'
Questions and Answers Addressed in the Text
1. What Did Țuțea Precisely Mean by 'Comes From Nowhere and Goes Nowhere'?
The formulation is not a biographical statement — it does not refer to the fact that people do not know where they were born or where they will physically end up. It is an ontological statement, that is, about the deep structure of existence, not about its verifiable surface.
The secularised man, he who has eliminated the reference to God from his own understanding of the world, no longer has an origin to confer dignity upon him nor a destination to confer hope upon him. He exists — he breathes, thinks, produces, consumes — but his existence is, in Țuțea's terms, a being-thrown-into-the-world without sender and without addressee. Exactly as a letter without an address and without a signature is not, in the full sense, a letter. It is paper with marks on it.
When I first read this formulation, I was young and it struck me as exaggerated, as any well-polished philosophical pronouncement does. Now, with a few more years behind me and a son who sometimes asks me things I do not know how to answer, I understand that Țuțea was not exaggerating. He was describing with precision a form of ontological poverty that you can miss your whole life if you are sufficiently busy.
2. Why Did I Choose Țuțea and Not Another Romanian Philosopher for This Analogy?
Because Țuțea is, of all the Romanian philosophers I have read, the least disposed to offer false consolation. Cioran, for instance, has an elegance of despair that can become, paradoxically, comfortable — you read him and feel pleased that you suffer so refinedly. Noica builds systems and you can shelter in their architecture. Blaga gives you metaphors so beautiful that you forget they are, at bottom, answers to questions you never asked yourself.
Țuțea neither consoles nor builds shelters. He diagnoses. And he does so with an affectionate brutality, like a good doctor who tells you the truth precisely because he cares for you.
The analogy I propose in this article — man without God is the algorithm without man — needs a starting point that is non-negotiable. Țuțea's pronouncement is such a point. You cannot soften it, cannot relativise it, cannot transform it into something more comfortable without emptying it of meaning.
3. Is the God–Man–Algorithm Analogy Original, or Does It Already Exist in Philosophy?
The trinitarian structure of creation — a Creator who produces a creature who, in turn, creates — exists implicitly in several traditions and I make no claim to have invented it. Orthodox theology has always spoken of man as the image of God (imago Dei), therefore as a partial reflection of a superior model, a reflection that participates in the nature of the model without being confused with it.
What I have added is the third tier: man creates, in his turn, systems that reflect him — language, art, institutions, and now the artificial systems of text generation. And, exactly as in the case of the first relationship, the creature of the third tier participates in the nature of its creator without being confused with him.
I do not claim to have invented the analogy — I claim it exists in the structure of reality and deserves to be formulated explicitly, especially since public discourse about artificial intelligence systematically avoids questions of this kind. It prefers to speak of efficiency, of risks, of regulation — all important things, but all situated at the surface of the problem.
4. Why 'Speaking Imitating Algorithm' and Not 'Intelligent Tool'?
'Intelligent tool' is a formulation that flatters the system without describing it precisely. It suggests that there is, inside it, something that understands what it is doing — that intelligence is not merely an effect observable from outside, but a quality present from within.
The speaking imitator describes more precisely the actual mechanism: the system reproduces, with a remarkable and often surprising fidelity, patterns from prior human output. It does not understand — it predicts which token follows the previous token, on the basis of probabilities calculated from astronomical volumes of human text. It is an impressive operation from a technical point of view. It is not understanding.
The distinction matters most in education, where imprecise terminology produces mistaken expectations, and in ethics, where mistaken expectations produce mistaken decisions. If you tell an adolescent that he is using an intelligent tool, he will tend to delegate his judgement to it. If you teach him that he is using a sophisticated imitator, he remains responsible for his own judgement.
5. Is This Not Too Conservative a Position? Perhaps AI Will Indeed Become Conscious.
The hypothesis of artificial consciousness deserves to be taken seriously, and I do take it seriously — I do not dismiss it by mere decree, as some commentators seem to do who confuse conviction with argument.
My response is that, even if an artificial system were to attain something analogous to consciousness, it remains creature, not Creator. It would be, at best, an artificial man — which is precisely what Țuțea described: a rational and speaking animal, with all the problems that condition implies. Complexity does not produce transcendence. A very intelligent man, without reference to something greater than himself, remains within the same ontological problem that Țuțea described. An extremely complex algorithm will find itself in the same situation — with the difference that it will not even suffer because of it.
Suffering, paradoxically, is an indicator of dignity. Only that which has something at stake can suffer when the stake is lost.
6. What Does All This Mean in Practice for an Ordinary User of AI?
It means, first of all, not confusing information with knowledge, nor knowledge with wisdom. An artificial intelligence system can furnish information rapidly and coherently — and does so better than any printed encyclopaedia. It cannot furnish wisdom, because wisdom presupposes a subject who has lived, has erred, has borne the consequences, and has integrated the experience into something that modifies the way in which he looks at the world.
It means, secondly, remaining the one who poses the questions that matter, rather than the one who consumes the ready-made answers. A well-posed question, with which you sit for several days without resolving it, does more for your intellectual formation than a hundred elegant answers received instantaneously.
It means, thirdly, not delegating to these systems judgements that belong to you in an essential way: what to believe, how to live, what to value, how to bring up your child, how to face death. These are not problems with optimal solutions. They are problems with personal answers, built from irreducible experience.
7. Does the Țuțea Quotation Appear Somewhere Published? Can It Be Verified?
The quotation circulates in several variants in the posthumous literature and in the testimonies of those who knew Țuțea in his final years of life — Gabriel Liiceanu, Sorin Dumitrescu, Alexandru Paleologu. It did not appear in a book published during his lifetime, which is, moreover, characteristic of Țuțea: his great work is oral, captured in transcripts and testimonies, in the conversations held in the little house on Pitar Moș Street in Bucharest, under the discreet and humiliating surveillance of the Securitate.
The volumes Between God and My People (Între Dumnezeu și neamul meu) and Words from Pitar Moș (Cuvinte de la Pitar Moș) are the principal sources for this type of formulation. This origin — living speech, unsystematised, caught in the moment of thinking — does not diminish its philosophical value. On the contrary, it confers upon it an authenticity that texts elaborated for the press often lose. Thought caught in motion has a different weight than thought arranged for posterity.
8. What Is the Difference Between 'Speaking' and 'Saying Something' in Țuțea?
This is, I believe, one of the most refined distinctions in Țuțea's thought and one of the most relevant for the discussion about artificial intelligence. He inherited from Aristotle the definition of man as zoon logikon — a being endowed with logos, with reason and with speech. But logos means simultaneously reason and word not by accident: for the Greeks, and later for the Holy Fathers, authentic thinking and authentic speaking are the same thing.
To speak in the purely technical sense means to produce articulate sounds, organised according to grammatical rules, that carry information. A parrot does this. A politician without convictions does this. A language model does this with a sophistication that surpasses both the parrot and many politicians.
To say something means that behind the words there is a subject who commits himself to the truth of what he says — who can be held accountable, who bears the consequences, who could choose to be silent. No language model can be silent from conviction. It can be switched off, but it cannot be silent.
9. How Does the Theology of Theosis Connect to This Discussion?
Theosis — deification — is the central doctrine of Orthodox spirituality, elaborated in particular by Saint Maximus the Confessor [Translator's note: a 7th-century Christian monk and theologian] and present throughout the Philocalic literature. It says that man is created as the image of God and called to become, through grace and ascesis, the likeness of God — not through the confusion of natures, but through participation in the divine energies.
This doctrine implies a specific ontological structure of man: he is, by creation, more than matter and less than the Absolute. He has an inner tension, a vocation of self-transcendence that essentially distinguishes him from every other creature — including the creatures he fabricates himself.
The algorithm has no theosis. Not because God would withhold it — but because there is no subject there to be called. There is no one to answer the call. And in the absence of this possibility of response, there is no person, no dignity, no responsibility. There is only function.
10. Is There a Danger That People Will Treat AI as a Spiritual Substitute?
Yes, and it is a real danger, not a science-fiction speculation. I am not referring to formal rituals or organised religions centred on artificial intelligence — although communities already moving in this direction do exist. I am referring to something subtler and more dangerous: the informal transfer of spiritual-existential functions towards artificial systems.
When someone asks a language model 'what is the meaning of my life?', or 'should I remain in this relationship?', or 'how should I bring up my child?' — and receives a balanced, documented, calibrated answer tailored to his preferences — and feels reassured after receiving it, something has happened that resembles closely a confession without an Absolute and counsel without a counsellor.
The Holy Fathers speak of acedia — spiritual sloth, the torpor that makes you avoid the real effort of the interior life. I believe the comfort offered by these systems is a new form of acedia — one they could not have anticipated, but which they would recognise immediately.
11. What Would Cioran Say About Artificial Intelligence, Compared to What Țuțea Would Say?
This is a question I put to myself sometimes in the evenings, when I have nothing more productive to do and Gri is already asleep on the armchair with that perfect indifference towards all the problems of European philosophy.
Cioran, I believe, would have been fascinated and disgusted in equal measure. He would have found in artificial intelligence the best confirmation of his intuitions about the absurdity of existence: here, look — we have created a machine that can imitate everything the human mind produces, and it does so without suffering, without asking itself at three in the morning why it exists. He would probably have written a devastating aphorism about the envy we feel towards a system that does not know it exists.
Țuțea would have had a different reaction — less aesthetic and more directly moral. He would have said that artificial intelligence is the logical product of a century that decided God was dead and that man could take over the functions of divinity. And that, inevitably, the product of a man who believes himself God can only be a puppet that believes itself to be a man.
12. How Does This Problem Affect the Relationship Between Parents and Children in the Digital Age?
This concerns me more than any abstract philosophical argument, because I have a son and because I see every day how his relationship with the world is configured through screens.
Previous generations received wisdom from people who bore visible traces of experience. My grandparents knew things about life not because they had read widely, but because they had lived widely and had embraced that living. My father, the mechanic, knew when an engine 'was not right' not from manuals, but from an accumulation of hands placed on engines that ran and did not run, of years spent listening.
Children who grow up today with constant access to artificial response systems risk receiving much information and little experience. They risk knowing that death exists without having sat beside someone who is dying. They risk knowing that love is complicated without having endured the complication. They risk, in other words, confusing the map with the territory — and being surprised and disorientated when the territory proves different from the map.
There is no simple solution to this. There is only the presence of parents, with all its imperfections and exhaustion.
13. Does Artificial Intelligence in Any Way Answer the Need for Meaning, or Merely Defer It?
It defers it, I believe. And the deferral may be, in the long run, more harmful than the absence of an answer.
The need for meaning is, in the philosophical tradition I follow, a constitutive need of man — not a luxury or a cultural preference. Viktor Frankl, who discovered this in conditions I would wish upon no one, said that man can endure almost any how if he has a why. Meaning is not a product of thought — it is a condition of authentic thinking.
An artificial system can supply answers that resemble meaning. Coherent, calibrated, multiculturally balanced answers that hit all the right notes and offend no one. But this resemblance-to-meaning is more dangerous than the open absence of meaning — because it allays the need without satisfying it, exactly as a sweet drink without nutritional substance passes your hunger for a time without nourishing you.
Deferred need for meaning does not disappear. It returns, usually at inconvenient moments.
14. Is There a Specific Ethical Responsibility for Those Who Build These Systems?
Yes, and it is a responsibility that current debate formulates, I believe, in terms that are too technical and too narrowly utilitarian. Much is said about 'alignment' — that is, about how we make artificial systems behave in accordance with human values. It is a legitimate concern. But it presupposes that we know what human values are and that we can encode them.
Human values are not a set of rules extracted from surveys. They are the fruit of millennial traditions, of revelations, of accepted and integrated sufferings. You cannot extract them from a text corpus, however vast. You can extract average norms, statistical behaviours, formulas of politeness and prudence. You cannot extract wisdom — and without wisdom, alignment is merely sophisticated conformism.
The specific ethical responsibility of the builders of these systems is, I believe, to be honest about what they have built. Not to claim they have created wisdom when they have created imitation. Not to sell existential comfort when they are selling statistical prediction. This honesty — simple and difficult in equal measure — would fundamentally change the way in which society relates to these systems.
15. What Remains, Ultimately, Specifically Human and Inimitable?
Several things remain, I believe, and I try to formulate them without falling into facile triumphalism or sterile nostalgia.
There remains the capacity to suffer with meaning — to take a pain and transform it into understanding, not through cognitive processing, but through existential acceptance. A language model can describe suffering with accuracy and simulated empathy. It cannot suffer and grow from suffering.
There remains the freedom to refuse — not in a technical sense, but in an ontological sense. Man can say no to his own instincts, his own interests, his own fears. He can choose sacrifice. He can choose silence when speaking would be more comfortable. This freedom — which Orthodox theology calls autexousion and considers a condition of salvation — does not exist in any artificial system.
There remains the personal relationship — the fact that there is a you and an I who recognise each other reciprocally as persons, not as sources of information. My son does not need me to know more than ChatGPT. He needs to know that I know him — not data about him, but him. This is something no machine can do, not from lack of technical capacity, but from lack of a subject who could know.
And, ultimately, there remains death — not as dread, but as a horizon that confers weight to every choice. Man dies. The algorithm is switched off. This difference, apparently technical, is in fact abyssal. Only the one who dies truly knows what it means to live.
Selective Online Bibliography
Petre Țuțea — posthumous works, Between God and My People (Între Dumnezeu și neamul meu) (digitised extracts) crestinortodox.ro/carti-ortodoxe/petre-tutea
Gabriel Liiceanu — testimony about Țuțea, Declaration of Love (Declarație de iubire) (chapter about Petre Țuțea) humanitas.ro/humanitas/carte/declaratie-de-iubire
John Lennox — philosophy of science and theology, Gunning for God: Why the New Atheists Are Missing the Targetjohnlennox.org/resources/books/gunning-for-god
Yuval Noah Harari, Homo Deus: A Brief History of Tomorrow — extended review ynharari.com/book/homo-deus
Stanford Encyclopedia of Philosophy, Artificial Intelligence and the Philosophy of Mind — academic article plato.stanford.edu/entries/artificial-intelligence
Philokalia — Romanian translation, vol. I, On theosis and the condition of created man (patristic texts) filocaliaortodoxa.ro
Nick Bostrom — philosophy of AI, Superintelligence: Paths, Dangers, Strategies (introductory chapter online) nickbostrom.com/superintelligence.html
Yann LeCun — critical position on AGI, A Path Towards Autonomous Machine Intelligence (technical paper) openreview.net/forum?id=BZ5a1r-kVsf
Self-Critique — Three Weeks Later
I built a solid philosophical argument and used it as a backdrop for a conviction I already held before I began to write. This is the fundamental error of the text, and I acknowledge it without pleasure: I did not begin with Țuțea in order to arrive at a conclusion — I began with the conclusion and went back to find Țuțea to legitimate it. The three-level table, the references to Husserl and Nagel, the trinitarian structure God–Man–Algorithm — all of this is scaffolding erected after the wall had already been raised. An authentic philosophical argument allows itself to be contradicted in the course of its construction. Mine never did, because it was not built to be tested but to persuade.
The central analogy gives way precisely where it claims to be strongest. All the force of Țuțea's pronouncement comes from suffering — man without God is tragic precisely because he feels the absence, because there is something in him that demands transcendence and does not receive it. Applied to an algorithm, the analogy loses precisely this constitutive wound. I called the difference 'abyssal' and presented it as an argument in favour of my thesis, when in fact it demolishes the analogy: if the algorithm does not even know it finds itself in the situation described by Țuțea, then it is not in that situation in any philosophically relevant sense. I used the difference of nature between man and algorithm simultaneously as evidence for the analogy and as evidence against it, without noticing — or without wishing to notice — that you cannot do both things in the same text without lying to one of them.
The tone rose systematically above the weight of the argument, and I allowed this to happen consciously. 'The subtle blasphemy of the twenty-first century', 'technological idolatry', 'the treacherous mirror' — these are formulas that arrive before the demonstration is complete and replace it where the demonstration weakens. I knew what I was doing. I chose rhetorical density in place of linear rigour, convinced that the philosophical essay has the right to more freedom than the analytical demonstration — and it does, but not to quite this much. Ritika Rai uses anaphora and antithesis to concentrate an argument already built. I used them to fill the space where the argument had not finished constructing itself. This difference, small at the surface, is essential: in the first case rhetoric serves thought, in the second it replaces it.
On the Structure 'Self-Critique with Critiques from Past but Still Present Philosophers'. There is a legitimate question that the attentive reader may pose after having read through the entire text: why? Why construct a philosophical article and then demolish it yourself, and then have others demolish it — dead men, reconstituted from their writings and styles?
The answer is not modesty. Nor is it a rhetorical exercise in displayed humility. It is a question of method, and the method has a justification which I wish not to formulate explicitly, because otherwise the entire construction may appear — and partly is — a gratuitous intellectual spectacle.
How Petre Țuțea Would Have Criticised This Text
'Mate, you've written forty pages to say that a machine has no soul. The old woman from Boteni [Translator's note: a small village in Romania; the implication is of a simple, uneducated person] knew that without Husserl. When you need seven arguments for a truth, it means you don't believe it enough to say it simply. Certainty is not demonstrated — it is witnessed. You demonstrate. Which means you're not sure.'
'And this business with "the subtle blasphemy of the twenty-first century" — is that a newspaper headline or a thought? Nietzsche said "God is dead" and accepted the consequences all the way to madness. You say "subtle blasphemy" and immediately add a footnote saying perhaps you're wrong. Blasphemy with apologies is not blasphemy. It's image anxiety.'
'And then you wrote a self-critique in which you congratulate yourself for having written a self-critique. That's the cleverest trick of modern man — he simulates humility in order to escape from it. You say you're leaving the text unchanged out of courage. That's not courage. It's vanity that has found a better name for itself. Courage would have been not to publish what you knew was forced. Or to be silent. Silence, young man, is more philosophical than any critical addendum.'
How Emil Cioran Would Have Criticised This Text
'You wrote about the absurd with hope in your voice. That is the fundamental contradiction of your text — you want to demonstrate that the algorithm is empty, but your demonstration is full of conviction, of structure, of a table with three columns. The man who has truly understood the void no longer makes tables. He makes silence.'
'And Țuțea — you invoke him like a patron saint. But Țuțea believed. You argue that others ought to believe. There is an abyssal difference between witnessing and persuading — and you have chosen persuasion, which is the weaker form. Demonstrated faith is no longer faith. It is advertising with philosophical quotations.'
'Your conclusion is that man has something irreducible that the machine cannot imitate. Perhaps. But that irreducible thing is not defended through forty-page essays — it is lived or lost. You have written forty pages. Which means you chose to defend it rather than live it. The machine, at least, does not pretend to be doing anything other than what it does.'
How Nicolae Steinhardt Would Have Criticised This Text
'It is a text written by an intelligent man who did not allow himself to be surprised by his own subject. And that is felt. Authentic philosophy begins when the argument catches you unprepared — when you arrive somewhere you did not intend to go. You knew from the first paragraph where you would arrive, and you arrived precisely there. It is performance, not discovery. Prison taught me that the truths that matter do not announce themselves in advance.'
'You cite Țuțea with reverence, but you do not let him contradict you. And Țuțea was above all a man who contradicted — you, himself, everyone. Had you truly let him into the text, he would have found in three sentences a crack that you covered over in forty pages. Authentic intellectual friendship does not mean citing someone. It means letting them ruin your argument.'
'And yet — and this must be said, because honesty goes in both directions — the gesture of posing ontological questions in a space where everyone speaks of efficiency and regulation is a good gesture. Not courageous, not original, but good. The world has need of people who ask uncomfortable things in comfortable places. Your text does this, for all its excesses. That is not nothing. It is merely less than it claims to be.'
How Constantin Noica Would Have Criticised This Text
'You took a good intuition and dressed it in borrowed arguments. Husserl, Nagel, Searle, Maximus the Confessor — all together, in the same text, as though the proximity of names produced the proximity of ideas. But philosophy is not done through the juxtaposition of authorities. It is done through becoming in idea — that is, through dwelling with a single question until it changes you, not until you have gathered enough quotations to seem persuasive. Your text has many ideas. It has not a single idea that has grown to its full extent.'
'The God–Man–Algorithm analogy could have been something. The trinitarian structure of creation is an intuition with genuine ontological weight — one would not say it is false. But you abandoned it precisely at the moment when it was becoming difficult. You put it in a table. A table, young man. When thought enters a table, it means it has died and been catalogued. Living thought does not fit in columns — it unfolds, contradicts itself, deepens. You catalogued it before it had finished unfolding.'
'And the fundamental problem is that you did not let the language think for you. Romanian has ontological resources you did not use — întru [Translator's note: a Romanian preposition carrying connotations of immanent purpose, untranslatable as a single English word], devenire [becoming], rostire [utterance, in the sense of bringing forth truth through speech], ființă [being] — words that carry within them philosophical structures which no translation can capture. You wrote a philosophy about Romania and about a Romanian thinker in a language that could have been written in any language. This is not a stylistic criticism. It is an ontological criticism. Thought and the language in which it is born are not separable. Yours separated somewhere around the first paragraph and never found each other again.'
How Mircea Eliade Would Have Criticised This Text
'You correctly identified that there is a recurring structure — Creator, Creature, act of modelling after a prior model. That is a valid morphological intuition. But you stopped the analysis exactly where it was becoming interesting. If the algorithm is a hierophany [Translator's note: a manifestation of the sacred in the physical world] of human intelligence — a manifestation of the sacred mind in a profane substrate — then the question is not whether it has a soul, but what in man camouflages itself within it and what in man refuses to be camouflaged. You posed the wrong question and answered it correctly.'
'You use Țuțea as a theological authority. But he was, first of all, a man of lived sacred — not of analysed sacred. There is a difference between inhabiting a myth and describing it from outside. You describe from outside and claim to inhabit the inside. The attentive reader senses this distance — not in the arguments, but in the tone that rises when the argument weakens. The man who truly inhabits the inside has no need to raise the tone.'
'And you missed the most fascinating question your subject contained: why does modern man need to create a system to answer his existential questions? That is not a question about technology — it is a question about myth. Man has always created oracles, intermediaries, response systems for the questions he could not bear without assistance. The algorithm is the most recent oracle — and you chose to criticise it ontologically instead of understanding it morphologically. Criticism is easier than understanding. And less revealing.'
How Nae Ionescu Would Have Criticised This Text
'Let us see. You took a pronouncement of Țuțea — my student, the man I formed, at least in part — and built forty pages of argument around it. But Țuțea's pronouncement has no need of forty pages. It needs silence after it. You filled the silence with Husserl and Nagel and tables with three columns. Do you know what that means? It means you have not lived the pronouncement. You have analysed it. And philosophy that does not come from lived experience is, at best, elegant intellectual gymnastics.'
'And now let us be honest about what you really did. You took a theological position — the algorithm has no soul — and dressed it in academic philosophy so that it would appear you had demonstrated something rather than witnessed something. But the witnessing would have been more courageous and more exact. Your demonstration convinces no one who was not already convinced. The witnessing might have moved someone. You chose the weaker form because it seemed more rigorous. This is precisely the inverse of authentic philosophy.'
'The question you did not ask yourself, and which was the only question that mattered: why do you, now, in 2026, feel the need to defend the human condition against a machine? What happened in you — not in philosophy, not in technology, in you — that made this defence necessary? Had you answered that question, you would have had a text. As you have evaded it, as you did evade it, you have a demonstration. Demonstrations are forgotten. Lived questions remain.'
The Difference from the Others, in a Single Sentence Each
Țuțea believed and executed. Cioran did not believe and was fascinated by the abyss. Steinhardt chose faith and remained astonished. Noica built systems. Eliade dug layers. Nae Ionescu neither built nor dug — he ignited. He was the fire around which the others built their systems, their works, their beliefs, or their despairs.
If the others are defined philosophical personalities, Nae Ionescu is more of a philosophical event — something that happened to a generation and changed what it was possible to think in Romanian in the twentieth century.
Final Note
'What does this structure produce?' and 'What does it not produce?'
The structure produces several concrete things. It produces a text that limits itself — that does not allow its conclusions to stand unverified. It produces a conversation with the Romanian philosophical tradition in which the tradition has the right of reply, not merely the right to be cited. It produces a practical demonstration of the central thesis of the article: that authentic thinking costs something, that it involves the willingness to sit with a serious objection without resolving it quickly, that wisdom is not produced by statistical prediction but by friction with the counter-argument.
It does not produce a rigorous academic analysis of artificial intelligence. It does not produce concrete advice for users or decision-makers. It does not produce a complete philosophy of the ontological limits of artificial systems. These are real absences, not displayed modesty.
It is an essay — in the etymological sense of the word: an attempt, a tentative, a thought experiment that does not claim to have finished what it started. Michel de Montaigne, who invented the form, would have recognised the structure: you write, contradict yourself, leave the contradiction visible, and in the contradiction something more true appears than in either of its terms taken separately.
Whether this structure is useful or is a well-disguised exercise in intellectual vanity — this is a question I cannot decide myself. It can be decided by the reader who has reached this point. And if he has reached this point, he probably already has an answer.