On Magnifica Humanitas

How should we think about artificial intelligence? As a tool, as a weapon, as an ambivalent technology that could serve emancipatory ends but that, under current conditions of power, will very likely lead to new forms of domination and destruction? These descriptions undoubtedly contain something true. Yet I would like to think of AI as a new historical figure of totality. In many respects, it condenses the contemporary aspiration to translate everything into a functional network of data, correlations, and predictions. AI is therefore not simply another instrument placed before us, but a mediation that reorganizes the very field in which something can appear as meaningful, useful, visible, communicable, or governable.

At this point, among the variety of possible references, I want to recover two names: Heidegger and Wittgenstein. From the former, especially his reflection on the age of technology and his intuition that cybernetics was not merely one discipline among others, but a sign of the modern fulfillment of metaphysics as availability, calculation, and control. Cybernetics is not simply the origin of neuroscience, contemporary cognitive science, and AI, but one of their decisive matrices: the point at which life itself begins to be thought under the model of information, feedback, and the governance of systems. From Wittgenstein, I am interested above all in the Tractatus, especially its culmination: where language reaches the limit of the sayable, a zone opens that cannot be captured by the proposition. Whereof one cannot speak, Wittgenstein says, thereof one must be silent. But that silence, read from Dussel, does not refer only to the mystical or to the unsayable in the logical sense; it can be reinterpreted as the sign of what the totality of language cannot absorb: an exteriority that cannot be translated without remainder into the order of representation, function, or calculation.

At this point, the discussion of artificial intelligence cannot be reduced to the vocabulary of applied ethics, nor confined to risk regulation or to a pedagogy of responsible use. All of that is undoubtedly necessary, but it is not enough if what we want to understand is the core of the problem that AI poses to us. The decisive question lies elsewhere: what happens to that which cannot—and must not—be integrated into the logic of algorithmic, technical, or bureaucratic mediation? What happens to what does not present itself as data, performance, productivity, availability, or “participation”? What happens to that which remains outside totality, not because it lacks meaning, but because its meaning cannot be produced by totality? More radically: what if what remains outside is not a marginal residue, but the very “source” of all meaning?

An aside should be introduced here. AI is not an anomaly, nor an external accident that irrupts from outside into a previously human world. It is, rather, the culmination of a long historical preparation: what Heidegger thought as the consummation of the age of technology, when the real appears above all under the form of availability, calculation, organization, and performance. What AI conceals is not different from what every civil-service, administrative, or bureaucratic regime has always concealed: human life in its nakedness, concrete existence before it is translated into a file, a diagnosis, a protocol, a performance, a competency, a record, an evaluation, or a case. Here it is relevant to recall Agamben’s figure of “bare life,” although displaced toward the terrain of contemporary technical and administrative mediations.

Our ordinary experiences in hospitals, universities, schools, courts, public offices, or police apparatuses already show a rudimentary form of the logic that AI will come to perfect: an operational, functionalist, and constructivist matrix that does not encounter reality, but administratively produces it under categories of management. The patient becomes a medical record; the student, a file; the citizen, a procedure; the migrant, a flow; the poor person, an indicator; suffering, an incident; the word, a form. Exchanges with academics, doctors, judges, police officers, or civil servants often reveal this structure in an almost caricatural way: not because these people are individually perverse, but because they have been trained as agents of a normative and organizational totality that distributes recognition, authority, and privilege in exchange for functional submission.

In this sense, AI does not invent closure. It inherits it, accelerates it, and automates it. It cultivates soil that has long been prepared. Where the institution had reduced life to procedure and intervention, AI introduces calculation; where bureaucracy had replaced testimony with classification, AI introduces an engine of acceleration that eliminates every nuance; where administration had subordinated the face of the victim to the file, AI introduces preventive prediction. Its danger, therefore, does not consist in adding “machines” or devices to our institutions, but in technically consummating a form of world that those institutions were already managing “industrially”: a world in which the exteriority of life appears only once it has already been neutralized, whether as data, case, profile, or function.

At this point, Foucault is indispensable for our analysis. What he showed is that modern power does not operate only by repressing from the outside, but by producing fields of visibility, expert knowledges, classifications, files, diagnoses, and forms of subjectivity. The hospital, the school, the prison, or the public administration are not neutral settings, but apparatuses in which life is captured, ordered, and made into an object of government. AI does not inaugurate this logic: it generalizes it, accelerates it, and automates it. Where Foucault analyzed the disciplinary production of docile bodies and the biopolitical administration of populations, AI introduces an algorithmic governmentality capable not only of classifying what exists, but also of anticipating, modulating, and preforming the field of what is possible insofar as it is sayable. Its power does not consist solely in administering what already appears, but in intervening in the very conditions under which something can appear as legitimate speech, intelligible demand, recognizable need, or life worthy of attention.

Now, the category of “exteriority,” as Enrique Dussel elaborates it from his reading of Levinas, allows us to pose this question with greater precision. Totality is not simply the set of existing things. It is the historical, economic, political, cultural, and symbolic order that organizes the world from a center, distributes places, assigns functions, produces visibility, and defines what counts as rational, useful, true, or human. Exteriority, by contrast, does not designate a simple spatial “outside.” It names that which totality cannot absorb without destroying, neutralizing, or functionalizing it; that which, when fully incorporated into the system, loses precisely what constituted it as an interpellation.

In other words, totality tends to desacralize what it touches, not necessarily in a narrowly religious sense, but in the broader sense of reducing every transcendence to available immanence. Here one might recall Charles Taylor’s critique of the modern “immanent frame”: an order in which life remains enclosed within its own functional conditions of intelligibility, management, and performance. But in Levinas and Dussel the problem acquires a more radical edge: exteriority is not only what conceptually exceeds the field of representation, but the face of the other, the victim, the poor, the excluded, the suffering body, the word that interpellates from a place that has not been produced by the system.

For this reason, exteriority is not simply one more internal difference within the field of totality. It is not a variation, an anomaly, or a singularity that the system has not yet managed to process. It is the ethical interruption of that field. It is what reveals that totality is not innocent, because its order of meaning is sustained by exclusions, silencing, and sacrifices. Exteriority does not simply ask to be included; it demands that totality be judged from what it itself cannot produce.

From this perspective, AI must be understood as a new technology of totalization. Not because it effectively controls everything, but because it installs the promise that everything can be integrated into a single regime of processability. Its horizon is not merely to automate tasks, but to turn reality into a continuous surface of inscription: texts, images, gestures, movements, desires, emotions, bonds, illnesses, decisions, opinions, learning processes.

For this reason, artificial intelligence carries out, in a certain sense, an extreme technical version of what contemporary enactivism has philosophically formulated as the “participatory creation—or construction—of meaning.” From Varela, Thompson, and Rosch, and later in Di Paolo, Cuffari, and De Jaegher, meaning is no longer conceived as the internal representation of a previously given world. Meaning is enacted: it arises in the activity of embodied organisms that are historically situated, affectively involved, and socially coordinated. In the theory of participatory sense-making, meaning emerges in interaction itself. There are not first isolated subjects who then communicate; there are relational dynamics that configure possibilities of experience, understanding, and action.

This intuition has undeniable force. It allows us to overcome the Cartesian subject, representationalist mentalism, and the idea of cognition enclosed in the head. Yet its limit appears when participation becomes the almost absolute name of meaning. If meaning emerges from participation, coordination, and relational dynamics, what happens to that which does not participate? What happens to those who are expelled from the very conditions of participation? What happens to the victim who cannot enter the common space of meaning-production because that space is already organized against her? What happens to the silence of the wounded, to shame, trauma, poverty, colonial exclusion, untouchability, the racialized body, the exploited woman, the migrant administered as threat or surplus?

Here the decisive difference appears between an ontology of participation and an ethics of exteriority. Enactivism can describe how meaning is produced in interaction, but it cannot always judge from where that interaction is already politically organized. It can show how norms, coordinations, and forms of life are stabilized, but it has difficulty thinking what does not appear as a legitimate participant. The victim is not simply an agent who participates differently; the victim is the one who reveals that the order of participation is constituted by exclusions. Her exteriority is not a lack of coordination, but a denunciation of the totality that decides who counts as an interlocutor.

AI radicalizes this problem because it presents itself as a great machine of participation. Everything can enter into it: our words, our images, our questions, our habits, our readings, our desires, our styles of writing. AI seems to carry out a gigantic participatory creation of meaning on a planetary scale. But this participation is deeply ambiguous. It is not a community of speaking bodies, but the massive capture of traces. It is not dialogue, but extraction. It is not recognition, but modeling. It is not communion, but correlation. AI produces the appearance of interlocution from an immense accumulation of voices stripped of their situation, their suffering, their responsibility, and their face.

For this reason, the question is not only whether AI understands or does not understand. The decisive question is what had to happen for the human word to be transformed into available material for a machine that responds without having lived, without having suffered, without having hoped, without having loved. AI speaks, but it does not respond in the ethical sense. It can simulate attention, but it cannot be summoned by the face of the other. It can produce linguistic consolation, but not compassion. It can organize information about pain, but not take responsibility for pain. It can recombine the words of love, but not love. It can describe mercy, but not give itself as gift.

Here it is essential to distinguish the private from the intimate. The private still belongs to the order of property. It can be legally protected, sold, encrypted, administered, monetized, or violated. Privacy is a necessary category, but an insufficient one, because it remains within the possessive logic of modernity: there is something that is mine, that I can reserve, defend, exchange, or withdraw. The intimate, by contrast, is not simply the private. It is not an inner property. It is not sensitive data. It is not hidden content. The intimate is the reserve from which something can be given without being reduced to function. It is what is not given because it has been requested by the system, but because it emerges as excess, as gratuity, as love, as compassion.

Intimacy is not deprivation, but the possibility of gift. It is not the narcissistic retreat of the proprietary individual, but the non-appropriable place from which the person can give herself without being captured. For this reason, intimacy cannot simply be “protected” as one private good among others. It must be recognized as an irreducible dimension of human existence. What is at stake in it is something AI cannot produce: the gratuity of a word that does not seek performance, the patience of a presence that optimizes nothing, compassion that does not calculate, love that is not translated into measurable interaction.

The algorithmic totality threatens precisely this dimension. Not only because it collects personal data, but because it accustoms us to thinking every human manifestation as an available inscription. The problem is not only that our preferences are captured, but that the very structure of life begins to be organized as if everything meaningful had to be externalized, recorded, communicated, shared, evaluated, archived, or processed. Intimate life is pushed into becoming public or private, visible or protected, exploitable or encrypted. But the intimate does not belong to that alternative. The intimate is not what I hide for myself, but what can appear only as gift.

From Dussel, this question acquires a greater ethical-political depth. Exteriority is not a bourgeois interiority threatened by technology. It is not a matter of defending the small subjective refuge of the modern individual against the machine. It is a matter of affirming that the person, especially the violated person, is not exhausted by the functional relations into which totality inscribes her. The poor person is not merely surplus labor power. The migrant is not merely a population flow. The child is not merely human capital in formation. The sick person is not merely a burden on the health care system. The elderly person is not merely dependency. The exploited woman is not merely a productive or reproductive body. And the human word is not merely training material.

Here we also encounter the limit of a certain contemporary bourgeois subjectivity, even when it presents itself in spiritual, therapeutic, or contemplative forms. This is not a matter of denying the value of meditation, silence, or inner attention. These practices can open real spaces of lucidity, de-automatization, and freedom in the face of the functional noise of the world. But when they are separated from the exteriority of the suffering other, they are easily integrated into the same logic they claim to resist: they become techniques of self-regulation, subjective hygiene, emotional optimization, or adaptation to the system. Interiority, then, ceases to be a place of conversion and becomes the administered refuge of the proprietary subject.

True openness does not occur when the subject simply turns back upon itself, but when it is interrupted by what does not proceed from itself: the face of the victim, the cry of the poor, the wounded body, the discarded life, the word that cannot be translated into the language of efficiency. Exteriority is not a psychological depth, but an ethical irruption. It is not accessed through pure introspection, but through exposure, responsibility, and response. Where bourgeois subjectivity seeks to preserve its inner equilibrium, the victim reveals that the very equilibrium of totality is founded on injustice. For this reason, against spiritualities of adaptation and against technologies of systemic participation, Dussel forces us to displace the center: there is no liberation in pure interiority, but in openness to what totality cannot receive without being judged.

Exteriority names that excess. It names the impossibility of reducing the other to his or her place within the system. For this reason, when AI becomes the general infrastructure of mediation, it does not threaten only certain jobs or certain cognitive competencies. It threatens the very structure of encounter, because it introduces a mediation that tends to translate the appearing of the other into processable information.

Against this, the encyclical Magnifica Humanitas offers a powerful image: the alternative between Babel and Jerusalem. Babel represents the false unity of a humanity that seeks a single language, a single technique, and a single direction. It is the city of self-sufficiency, homogenization, and domination. Jerusalem, by contrast, represents the patient reconstruction of a coexistence in which plurality is not absorbed by uniformity, but oriented toward communion, justice, and dignity. AI can be placed at the service of one city or the other. But to discern this it is not enough to ask about its efficiency. We must ask what it does with the weak, with the poor, with the voices and bodies that do not fit, with the words that do not produce quantitative performance.

The problem, therefore, is not whether we should participate or not participate. Participation is necessary. There is no human life without language, relation, community, history, and mediation. But not all participation is liberating. There are obligatory, captive, administered, and commodified forms of participation. There are forms of inclusion that are modes of subjection. There are systems that invite everyone to participate only in order to turn every difference into material for the reproduction of totality. For this reason, the criterion cannot simply be to participate more, communicate more, interact more, connect more. The criterion must come from exteriority: from those who cannot participate, from those who are forced to participate under unjust conditions, or from those who demand to preserve the silence that makes possible a nonfunctional gift.

The critique of AI from Dussel, then, does not consist in defending an abstract humanist nostalgia against technology. It consists in asking whether technology remains subordinated to life or whether life becomes subordinated to technology; whether participation opens a place for exteriority or closes it off; whether the word serves encounter or capture; whether intelligence is oriented toward justice or toward the expansion of a faceless power. The core of the problem is not artificial intelligence, but the totality that finds in it its new metaphor, its new infrastructure, and its new promise.

The human is not saved simply by regulating machines. It is saved by safeguarding what no machine can produce: the capacity to receive the other as other, to listen to a word we have not generated, to respond to the suffering that interrupts us, to give without calculation, to love without utility, to feel compassion without turning compassion into performance. There, in that nonfunctionalizable remainder, exteriority begins. And from there the new algorithmic Babel can be judged.

References

Agamben, G. (1998). Homo sacer I: El poder soberano y la nuda vida. Pre-Textos.

Cincunegui, J. M. (2026a). La vida en la historia: Más allá de la biología, la fenomenología y las ciencias cognitivas [Doctoral dissertation, Universitat Pompeu Fabra].

Cincunegui, J. M. (2026b). La vida en la historia: Más allá de la biología, la fenomenología y las ciencias cognitivas. Dado Ediciones.

Di Paolo, E. A., Cuffari, E. C., & De Jaegher, H. (2018). Linguistic bodies: The continuity between life and language. The MIT Press.

Dupuy, J.-P. (2009). On the origins of cognitive science: The mechanization of the mind (M. B. DeBevoise, Trans.). The MIT Press.

Dussel, E. (1998). Ética de la liberación en la edad de la globalización y de la exclusión. Trotta.

Dussel, E. (2016). 14 tesis de ética: Hacia la esencia del pensamiento crítico. Trotta.

Foucault, M. (n.d.). Seguridad, territorio, población: Curso en el Collège de France (1977–1978). Fondo de Cultura Económica.

Foucault, M. (n.d.). Vigilar y castigar: Nacimiento de la prisión. Siglo XXI.

Heidegger, M. (1994). La pregunta por la técnica. In Conferencias y artículos. Serbal.

Larrabeiti, G. (2026, 30 de mayo). «Magnifica Humanitas», o de los nuevos límites. InfoLibre. https://www.infolibre.es/opinion/plaza-publica/magnifica-humanitas-nuevos-limites_129_2200334.html

Leo XIV. (2026, May 15). Magnifica Humanitas: Encyclical letter on the safeguarding of the human person in the age of artificial intelligence. Libreria Editrice Vaticana.

Levinas, E. (1977). Totalidad e infinito: Ensayo sobre la exterioridad. Sígueme.

Taylor, C. (2007). A secular age. Harvard University Press.

Varela, F. J., Thompson, E., & Rosch, E. (2017). The embodied mind: Cognitive science and human experience (Rev. ed.). The MIT Press.

Wittgenstein, L. (2002). Tractatus logico-philosophicus. Alianza.