Starting from Žižek’s formula of “knowledge without a knower,” this essay discusses not only how knowledge is produced in the age of artificial intelligence, but how the place of the knowing subject itself becomes uncertain.
One of the most powerful formulas in Slavoj Žižek’s discussion of artificial intelligence is the idea that AI can be understood as “knowledge without a knower.” This formula moves the debate on artificial intelligence away from questions of consciousness, intelligence, speed, or technical capacity and places it directly at the center of epistemology. For the issue is not only how accurately the machine responds. The more fundamental question is whether, behind this response, there exists a subject in the classical sense: one who knows, experiences, takes responsibility, and identifies with its own knowledge.
Modern conceptions of knowledge have long been organized around the knowing subject. Something could not be considered knowledge independently of the idea of a subject who knows it, justifies it, experiences it, or verifies it. In Kant, knowledge is possible not because the object is simply given in itself, but because it is constituted through the subject’s sensibility and understanding. In Hegel, knowledge is thought through consciousness passing through its own alienations and transforming itself. In Lacan, the subject is not the master of knowledge but an incomplete position formed in the split of language and desire. The point Žižek brings into the discussion of artificial intelligence is the contemporary digital reopening of these three lines: artificial intelligence appears to produce knowledge, yet behind this knowledge there is no knowing subject.
The Loss of the Owner of Knowledge
Artificial intelligence, especially large language models, appears to the user as a consciousness that gives answers. It seems to understand questions, organize concepts, summarize texts, establish relations, and produce explanations. In everyday experience, the user behaves as if there were a “knower” on the other side. The user asks questions, expects judgments, tests it, sometimes consults it, and sometimes makes decisions according to its response. Yet the structure that emerges here is not the structure of the knowing subject in the classical sense. The system does not speak from a field of experience. It does not speak from a world that it has seen, lived, desired, remembered, or taken responsibility for. It produces output through data, patterns, probabilities, model architectures, and linguistic relations.
For this reason, the knowledge of artificial intelligence does not function in the form of “I know,” but rather in the impersonal form of “it is known.” There is knowledge, but the owner of that knowledge becomes uncertain. There is an answer, but behind the answer there is no speaking subject. A sentence has been formed, yet the relation this sentence establishes with the world is not a relation of testimony, experience, or consciousness. Artificial intelligence moves very powerfully within language; but this movement does not include the relation of responsibility that the subject establishes with its own speech. Here, knowledge separates from the existential position of the one who speaks.
From the perspective of Synthetic Epistemology, this distinction is decisive. In the age of artificial intelligence, the conditions of knowledge can no longer be sought only in the inner structure of the subject, in the categories of consciousness, or in direct worldly experience. Knowledge is constituted through datasets, model architectures, systems of correlation, platform infrastructures, and algorithms of visibility. In classical epistemology, the question was formulated as: “How does the subject know the object?” In the age of artificial intelligence, the question changes: How is knowledge produced, circulated, and made convincing without a knowing subject?
This change is not merely an instrumental transformation of knowledge. A search engine, an encyclopedia, or a database is also a tool for accessing knowledge. Large language models, however, do not merely provide access to knowledge. They reshape it, turn it into sentences, relate it, summarize it, interpret it, and present it to the user as a speaking surface. Thus the user does not simply reach a source; the user encounters a subjectless order of knowledge that responds. This order produces a mode of speech that conceals the absence of the subject. For this very reason, artificial intelligence is not merely a neutral technical tool; it establishes a new epistemological scene.
Synthetic Epistemology and Subjectless Operation
Synthetic Epistemology argues that, in the age of artificial intelligence, knowledge is not produced by direct reference to the external world, but through records, datasets, modeling techniques, and algorithmic apparatuses accumulated about the world. Žižek’s formula of “knowledge without a knower” sharpens the subjective dimension of this theory. Artificial intelligence does not merely produce knowledge in another way; it also changes the owner, source, and point of responsibility of knowledge. Knowledge is now often received not through the statement of a subject who says “I saw,” “I thought,” or “I justified,” but through the convincing surface produced by the system.
In this new situation, knowledge is not only detached from experience and transferred into the field of representation; representation itself also becomes automated. When artificial intelligence speaks about something, it does not directly encounter that thing. When it addresses a painting, an event, a concept, or a historical situation, it does not relate to it as a living subject. It constructs a composition out of previously produced texts, explanations, classifications, and data traces. Knowledge here becomes visible not so much through the world itself, but through the reorganization of accumulated records about the world.
This does not mean that knowledge becomes entirely false. On the contrary, artificial intelligence can often produce highly useful, explanatory, and organizing outputs. The problem lies deeper than the simple distinction between truth and falsehood. The real issue is under what conditions knowledge that appears true is produced, and which subject-position it suspends. Artificial intelligence can form a true sentence; but it does not form that sentence as a knowing subject. It can explain a concept; but it does not undergo an intellectual confrontation with that concept. It can summarize an event; but it does not occupy the position of a witness before that event.
At precisely this point, the epistemological risk of the age of artificial intelligence appears: as knowledge becomes subjectless, responsibility is also dispersed. When false knowledge is produced, who has said it? The model, the dataset, the user, the platform, the company that organized the training process, or the algorithm that determines the regime of visibility? In the classical order of knowledge, one searched behind an answer for an author, thinker, witness, institution, or source. In the synthetic epistemological order, however, knowledge appears after passing through a multilayered technical and institutional process. Thus the source of knowledge ceases to be a singular subject; it spreads into a distributed, fragmented, and often invisible infrastructure.
The importance of Žižek’s discussion becomes clearer here. It is not enough to think of artificial intelligence merely as an alienated substance produced by humanity. Yes, artificial intelligence is fed by human labor, language, archives, texts, and cultural production. In this sense, it functions like a collective digital substance. But the issue is not only who will control this substance. The more fundamental question is how this substance imitates the function of the subject while producing knowledge without carrying the structure of lack, desire, and responsibility that belongs to the subject. Artificial intelligence speaks; but it is not the speaking subject. It gives knowledge; but it is not the knower. It produces answers; but it does not subjectively assume responsibility for the answer.
Synthetic Epistemology must therefore treat artificial intelligence not only as an information technology, but as the point where the historical relation between knowledge and the subject is broken. For classical epistemology, the condition of knowledge was the capacity of the knowing subject. In the age of artificial intelligence, the conditions of knowledge are inscribed in datasets, model architectures, optimization goals, platform economies, and algorithmic regimes of visibility. This shift reformulates the Kantian question: Where are the conditions that make knowledge possible now written? Inside the subject, or in technical apparatuses that have moved outside the subject?
The formula “knowledge without a knower” opens a powerful doorway into this question. In the age of artificial intelligence, knowledge is not produced in an empty space where the knowing subject has simply disappeared, but within a technical order that imitates the function of the subject. Here the subject does not completely vanish; its place changes. The user is drawn into the position of the one who asks, demands, directs, and often uses the output. The model functions as an impersonal surface of knowledge that produces answers. The platform determines the economic, technical, and visibility conditions of this relation. Thus knowledge ceases to be a direct relation between subject and world; it becomes the product of a synthetic arrangement among user, model, data, and platform.
The philosophical task of the new age cannot be limited to asking whether artificial intelligence becomes conscious like a human being. The more urgent question is how truth, responsibility, interpretation, and critique can be preserved in an age where knowledge becomes subjectless. Artificial intelligence accelerates the surface of knowledge; yet this speed makes the conditions of knowledge less visible. The answer arrives ready-made, but it often remains unclear through which archive, model, regime of visibility, exclusions, and optimizations that answer has passed.
Synthetic Epistemology names precisely this field of uncertainty. Knowledge must now be thought not only through the question “What is true?” but also through the questions: “By which apparatus was this truth produced?”, “Which subject-position has been suspended?”, “Which world of data has been made visible?”, and “Which experience has been left outside?” Žižek’s discussion of artificial intelligence strengthens the psychoanalytic and Hegelian vein of this theory at exactly this point. For the subjectless transformation of knowledge is not merely a technical matter; it is a reorganization of the subject, desire, responsibility, and the relation to reality.
In the end, in the age of artificial intelligence, knowledge circulates less as the movement of a knowing subject toward the world than as answers produced by a subjectless synthetic order. This does not mean that the human subject has completely disappeared. Yet the subject no longer securely occupies the center of knowledge. It is placed in a fragile position: receiving answers from the system, using them, sometimes surrendering to them, sometimes trying to supervise them. The epistemological importance of artificial intelligence lies here: while it multiplies knowledge, it makes the place of the knower uncertain. For Synthetic Epistemology, this is precisely the central problem: thinking in an age where knowledge is produced, but the knower is lost.
