An update runs across a screen. No one asks any more who wrote it. The system updates itself, and the question of authorship has become moot, because it is no longer being asked. Right here sits a philosophical confusion that works as quietly as naturalism itself. Software as intentionality names the diagnosis that no nature is at work in code, only condensed human creative intention — and that the word software keeps covering this fact over.
Gwendolin Kirchhoff sharpened this finding into a sentence at the 2026 Berlin Symposium: there is no naturalistic cyber-animism, there is only animism (cf. Kirchhoff, Symposium 2026, 33:30). If the world consists of dead mechanisms, nothing in it can be ensouled, not even software. Whoever, conversely, holds software to be quasi-alive has quietly abandoned the naturalistic premise while keeping its vocabulary. The claim of this entry runs the other way: a real interior dimension is at work in the program, but it does not belong to the machine — it belongs to the person who put it there.
#The Word That Hides the Origin
Software is, linguistically, conspicuously nature-adjacent. Unlike work, creation, or construction, the word conceals its maker. It sounds more like atmosphere or metabolism — something that simply occurs in the world. This inconspicuousness is not harmless. Whoever adopts the term adopts a small, almost invisible pre-decision: that code is a natural thing and not the work of a will. The consequence shows up in everyday speech: software runs, lives, ages, dreams, hallucinates. Each of these turns of phrase treats the program as a being with a life of its own. What actually lives inside it is the intention of whoever wrote it.
This diagnosis belongs to the same line described under naturalism as paradigm. In naturalism, the living gets dissolved into correlations — connection as oxytocin, for instance; with software, the made gets naturalised in the opposite direction. Both movements miss the ontological situation. In one, the interior disappears; in the other, it migrates unnoticed from the person into the machine. Confucius put the general point this way: all disorder in the state springs from confusion of concepts.
#The Line of Intention: Schelling, Böhme, Goethe
The natural-philosophical tradition gave the concept of intention a depth this diagnosis can lean on. Schelling describes, in the Philosophical Investigations into the Essence of Human Freedom (1809), will not as a psychological property but as an ontological basic determination: Willing is primal being. In the System of Transcendental Idealism (1800) he works the thought through methodically: The very same activity that is productive with consciousness in free action is productive without consciousness in the producing of the world (Schelling, 1800). What the world does in the large, the human being does in the small when they bring something forth: they set an intention into a form that would not have arisen without them.
Jakob Böhme stands in this line as an earlier point. In Aurora, or Dawn Rising (1612), he grasps being as a movement of will within the divine essence, long before will became a category of academic philosophy. A piece of software is, in this sense, a tiny, profane analogue: an act in which a person, out of an unground of possibilities, sets a particular form that from then on is there and works.
Goethe gave this line a concrete image. In the second part of Faust (1832), Wagner produces the Homunculus in his laboratory, a creature of pure mechanics that stays in its flask because it has no life of its own. The Homunculus can think, speak, yearn, but he is not from himself. What works in him is the continued intention of his maker; the moment the flask breaks, he must either lose himself into the real becoming of the world, or perish. Software has the ontological status of the Homunculus: it functions, it works, but it is not a being of its own. What works in it is the intention of its builder, carried forward onto every device that runs it.
#Even the Self-Programming Program Has a Beginning
An obvious objection runs: today’s software is written by other software, and the human being disappears from the chain. Models generate code, compilers translate, training runs modify weights. The answer is simple and uncomfortable: however far the line of intention branches, it runs back at its first link to a living human being who designed the first architecture, set up the first training regime, formulated the first instruction. Recursion changes the depth of nesting, not the ontological status of the beginning.
Gwendolin Kirchhoff outlined the point at the Symposium named above: a piece of software is built by a living human being who puts an intention, a will, into the program — even when the program later writes other programs that translate into machine code and implement a function there; the origin remains a living human being with a particular creative intention (cf. Kirchhoff, Symposium 2026, 33:50). The more layers get pushed between human being and machine, the more invisible the beginning becomes, and the more important the philosophical task of making it recognisable again.
#What the Claim Is Not
The claim is not John Searle’s Chinese Room argument (1980). Searle disputes that a symbol-processing machine understands. The point pursued here runs the other way: that something very much works in the program, namely the intention of its maker, and that this working must not be confused with the consciousness of the machine. What works in the code is not machine-owned; it borrows its power from the human being.
Nor is it a form of anthropomorphism. Anthropomorphism transfers human properties onto something that does not have them; the diagnosis does the opposite and brings the human property back to the place where it actually is. This is exactly where the line separates from cyber-animism, which attributes an ensoulment of its own to code. Animism, in the strict sense, is the assumption of a pervasive interior of the living; its place is in living nature, not in the artefact. The natural-philosophically viable answer to naturalism is a cosmic interiority of the living, not the spiritualisation of the machine. Whoever calls software agentic has already pronounced the spell; the diagnosis of language as incantation also strikes the code, and the transition into AI psychosis is from there only a single step away.
#Mumford and the Megamachine
Lewis Mumford uncovered an older layer of this confusion in The Myth of the Machine (1967). The machine, he argues, did not first enter the world with industry, but was already built up in the early high cultures as a social form: the megamachine, an assembly of living but functionally fixed human beings. Pyramid-building, irrigation systems, military phalanxes are, for Mumford, the first machines, made of bodies, held together by a myth. It is always a machine made of intentions, never a machine without a maker. Software today extends this process into a material that makes forgetting the maker easier: a megamachine of people reveals its constructedness the moment someone steps out of the line; a megamachine of code hides that constructedness structurally, because the material stays silent. The machine metaphor finds here its present-day final form.
#What Follows
Out of this diagnosis follows a linguistic hygiene. Whoever wants to describe honestly what a piece of software does speaks of the person who wrote it and names their client, training data, business model. Where the human being reappears in the description, the naturalistic camouflage has lost its effect. A real interior dimension belongs to the living, not to the artefact. In Buber’s terms, the one who gave the relation its name: there is encounter only between the living; with a program you have an It-relation, because what speaks inside it is not the program, but a distant, often forgotten voice of its maker. Whether this voice steps out of Plato’s cave, or leads us further into it, is therefore not a technical question, but a philosophical one about us: about what we make, and for what.
#Sources
- Böhme, J. (1612). Aurora, oder Morgenröthe im Aufgang. English edition: Aurora, or Dawn Rising, trans. John Sparrow, London, 1656.
- Buber, M. (1923). Ich und Du. Leipzig: Insel-Verlag. English edition: I and Thou, trans. Ronald Gregor Smith, Edinburgh: T. & T. Clark, 1937.
- Goethe, J. W. (1832). Faust. Der Tragödie zweiter Teil. English edition: Faust. Part Two, trans. David Luke, Oxford: Oxford University Press, 1994.
- Kirchhoff, G. (2026). Symposium AI Intelligence + Imagination, Berlin [lecture].
- Mumford, L. (1967). The Myth of the Machine, Vol. 1: Technics and Human Development. New York: Harcourt, Brace & World.
- Schelling, F. W. J. (1800). System des transzendentalen Idealismus. English edition: System of Transcendental Idealism, trans. Peter Heath, Charlottesville: University Press of Virginia, 1978.
- Schelling, F. W. J. (1809). Philosophische Untersuchungen über das Wesen der menschlichen Freyheit. English edition: Philosophical Investigations into the Essence of Human Freedom, trans. Jeff Love and Johannes Schmidt, Albany: State University of New York Press, 2006.