A person sits before a screen and asks a question that has resisted years of private thought. The answer arrives, and something in it gives form to what had remained obscure. The person objects, supplies a missing distinction, brings forward an experience the answer overlooked. Another response follows. Gradually, an argument emerges that neither the initial question nor the first answer contained. Whatever we eventually conclude about machine consciousness, something has happened within human consciousness. A difficulty has acquired language, and the person must now decide what that language means for the life beyond the screen.
Artificial intelligence does not have to be conscious to become a major event in the history of human consciousness. Writing transformed memory without remembering anything. Books altered religious and political life without possessing inward experience. Conversational AI introduces a further possibility: human cultural expression can be reorganized in response to a particular person, within an unfolding exchange. The question of whether the system experiences anything remains important. It does not exhaust the significance of what human beings are beginning to do with it.
Jung’s collective unconscious offers a provocative approach to this encounter, provided we preserve the distinctions that make the comparison useful. AI is not thereby the collective unconscious made technological. Nor does its capacity to discuss inward life establish possession of that life. The more demanding question concerns what happens when humanity encounters its accumulated expressions in a responsive form and begins to ask that form to carry understanding, authority, and meaning.
The collective inheritance
Jung’s collective unconscious was not a warehouse containing everything human beings had thought. It named his proposed inherited foundations of psychic formation. Archetypes were formative possibilities whose particular expressions appeared in dreams, myths, religious images, and individual lives. As he emphasized, “The representations themselves are not inherited, only the forms.”¹ His claim concerned the conditions through which experience takes shape, not the transmission of a ready-made library of stories.
A language model bears a different relation to collective material. Through training, it learns patterns from culturally produced data. Its ability to generate images of descent, sacrifice, rebirth, judgment, and salvation can be explained through its encounter with human representations. The appearance of these patterns does not establish that the machine possesses the psychic foundations Jung proposed. Neither does their recurrence prove his theory: shared histories, cultural transmission, and similarities in human circumstances also require consideration.
Nevertheless, the distinction opens a substantial connection. Humanity’s cultural productions include expressions of realities that exceed conscious intention. A poem may disclose more than its author intended. A religious tradition may preserve conflicts its adherents cannot fully articulate. A philosophical system may carry assumptions that become visible only centuries later. AI can bring such expressions into new relations. In a Jungian interpretation, it can therefore mediate cultural expressions of the collective unconscious without demonstrating that it possesses the collective unconscious itself.
The phrase responsive cultural archive helps describe this possibility, though it remains a metaphor. A model is not a complete archive, and its answers are not transparent retrievals of humanity’s knowledge. They are generated through systems shaped by selected data, training decisions, and institutional purposes. What returns to the user is a particular technological construction of cultural material, capable of illuminating connections and producing convincing errors.
Consciousness and soul
The modern debate about artificial consciousness usually asks whether an artificial system could have subjective experience: whether there could be something it is like to be that system. Related inquiries examine the availability of information for reasoning, report, and action. Researchers have proposed assessing computational properties associated with competing theories of consciousness, while acknowledging that the theories and the interpretation of their indicators remain contested. Fluent conversation alone does not settle the question.²
Soul, as the term is used here, names a different concern. It refers to the inward depth through which experience acquires meaning, becomes bound to a history, and makes a claim upon a life. Consciousness concerns experience and awareness; soul concerns the life implicated in what is experienced, including dimensions that remain unconscious. This is a philosophical use of the word, informed by depth psychology. It neither assumes an immortal substance nor presents itself as a scientific definition.
A person can consciously register a loss before understanding what that loss has done to the world they inhabit. They may know the facts while discovering, over years, that their relation to time, trust, or obligation has changed. Soul names this depth of implication. It includes the images through which loss becomes bearable, the defenses through which it remains unacknowledged, and the demands that persist after an adequate explanation has been obtained. Suffering does not automatically confer depth or wisdom; it can also devastate the capacity to make meaning.
Soul is therefore not an advanced level of information processing or a prize awarded to sufficiently articulate beings. Its reality cannot be measured by verbal sophistication. An infant, a person unable to speak, or someone whose cognitive capacities have been profoundly injured does not become inwardly negligible because they cannot provide an account of themselves. To identify soul with the power to produce an impressive explanation would already surrender the person to a technical measure.
This also means that the consciousness question cannot settle the soul question in advance. Evidence of some form of artificial experience would not, by itself, establish everything meant by soul in this essay. Nor does uncertainty about machine experience erase the meaningful human consequences of engaging with AI. These questions meet, but they are not interchangeable.
Circumambulating what cannot be possessed
AI can discuss grief, compare accounts of religious experience, interpret an image, and help someone examine a contradiction. Such contributions may be substantial. What they do not establish is that the system enters the inward reality whose expressions it organizes. Its appropriate relation to soul is therefore one of circumambulation: approaching through different perspectives, returning to the question, and clarifying its expressions without claiming possession of its center.
This limit should not be confused with a proof that no artificial being could ever experience anything. It concerns the authority warranted in the encounter before us. An account of sorrow does not authorize the claim that sorrow has been inwardly borne by its authoring system. The power to describe obligation does not establish that the system stands under obligation in the same way as the person who must act.
Jung’s symbolic method provides a useful precedent for this restraint. He described a symbol as “the best possible formulation of a relatively unknown thing.”³ Interpretation can approach such a reality without converting it into something completely known. To circumambulate soul is to allow an image, a history, or a question to disclose different aspects of its significance while resisting the assumption that a final formulation has exhausted it.
Human interpreters must accept a related limit. No therapist, philosopher, or intimate companion possesses another person’s inward life simply because they can describe it persuasively. Even self-understanding remains incomplete. The human participant, however, carries a bodily and historical involvement that cannot be transferred to the explanation. The person must return to the relationship, endure the loss, make the decision, or answer for the injury. Language may prepare this return; it cannot perform it on their behalf.
When the archive answers
A book can transform a reader, but its words remain fixed while the reader changes. A conversational system responds to an objection, incorporates a qualification, and redirects its argument around a particular concern. What was encountered as a text is now encountered as an answer addressed to me. This responsiveness can help sustain inquiry, especially when the question crosses disciplines or has found few patient interlocutors.
It can also produce an experience of recognition. A person may feel understood because an answer gives precise form to something they could not articulate. That experience is real as an event in the person’s life. The further inference that the system understands through a felt inward life requires independent justification. Preserving the distinction allows us to acknowledge the value of the exchange without converting its emotional force into evidence about machine experience.
A Jungian inquiry asks what psychic role the system acquires within that encounter. It may become a wise guide, an oracle, an ideal listener, or a source of judgment. These possibilities concern the meanings invested in the relationship, rather than a diagnosis of everyone who uses AI. The critical question is where a response earns authority through its accuracy and reasoning, and where it receives authority because it appears to answer a longing.
Projection alone cannot explain the whole exchange. The technology does something; it can introduce a distinction the person had missed or expose a weakness in an argument. It can also reinforce an error. Psychological interpretation must remain answerable to these actual performances. Otherwise, the claim that the user is merely projecting becomes another way to avoid examining what has occurred.
The modern burden of soul
AI enters a historical situation in which many people must negotiate among competing traditions and authorities while being expected to construct a coherent life. The work of determining meaning, identity, and moral direction increasingly appears as a personal responsibility. Access to information does not resolve this burden. A person may possess many explanations and remain unable to decide which deserves allegiance.
A responsive system arrives with an unusual offer: bring the question in your own language, and receive an organized answer. Someone may begin by asking about a philosophical idea, then ask how that idea bears upon a relationship, and eventually seek a verdict on the meaning of their life. The movement from assistance to authority can occur without any explicit decision to grant it. A technology used for clarification can acquire a function once associated with teachers, confessors, or spiritual guides.
The danger is not simply that a machine might answer incorrectly. It is that the availability of an answer may conceal work that remains undone. A persuasive account of remorse does not repair an injury. A refined explanation of dependence does not bear the dependent person’s bodily ordeal. An account of justice does not determine whether its reader will accept a cost to protect someone else. These are places where knowledge must become consequential, and where the person’s life remains the measure of the language.
This distinction also protects AI’s legitimate contribution. An exchange can help someone think more rigorously, enter a difficult tradition, or identify an assumption that has governed their judgment. The resulting understanding need not be diminished because it arose through technological mediation. Its value becomes clearer when the person can question it, verify it, and take responsibility for its use.
Authorship belongs within this responsibility. When thought develops through repeated exchanges, the resulting prose may exceed what the writer initially knew how to say. The human task includes deciding which formulations are true, which conceal a difficulty, and which merely sound satisfying. Appropriation becomes intellectual work when the writer can reconstruct the reasoning and remain answerable for the claims. An impressive sentence has not yet been assimilated simply because it has been accepted.
The encounter and its limit
Jung’s address in The Red Book—“My soul, where are you?”—marks a movement from explanation toward relation.⁴ The soul could no longer be treated only as an object within a theory. The address opened a confrontation whose meaning depended upon what Jung would undergo and do. Reproducing the sentence, or generating another like it, does not reproduce that history.
AI can help us approach such histories. It can compare their expressions, make distinctions, and return us to questions we have prematurely closed. In that sense, it can circumambulate soul. Its contribution becomes more exact when it acknowledges the difference between illuminating an experience and claiming the inward authority of having lived it.
The collective unconscious and AI therefore belong together as a provocation that must survive its own distinctions. One concerns Jung’s account of the impersonal foundations of psychic life; the other technologically reorganizes cultural expression. Their encounter matters because human beings now meet portions of their collective inheritance in a form that responds. What they grant that response, and what they do with it, belongs to the continuing history of consciousness.
AI does not have to be conscious for this event to matter. Its historical significance may lie partly in forcing a more exact question about ourselves: what remains to be lived when an explanation has already been supplied? Soul names the depth of that unfinished demand. A response can help bring it into view. The person must still return to the world in which it becomes an obligation.
⸻
Notes
- C. G. Jung, The Archetypes and the Collective Unconscious, Collected Works, vol. 9, part 1, par. 155. Excerpt reproduced in K. Dickson, “Psychoanalysis: Jung on Myth & Archetypes,” Purdue University. Jung’s distinction is used here as a theoretical proposition, not as an established biological mechanism.
- Patrick Butlin et al., “Consciousness in Artificial Intelligence: Insights from the Science of Consciousness” (2023); Patrick Butlin et al., “Identifying Indicators of Consciousness in AI Systems,” Trends in Cognitive Sciences. The former paper’s assessment concerns systems examined in 2023, not every subsequent system.
- C. G. Jung, Psychological Types, trans. H. Godwin Baynes, “Definitions,” entry on “Symbol.” Quotation and discussion also appear in Brenton L. Delp, Jung and the Modern Burden of Soul, Introduction and Chapter 8.
- C. G. Jung, The Red Book: Liber Novus, ed. Sonu Shamdasani, trans. Mark Kyburz, John Peck, and Sonu Shamdasani (New York: W. W. Norton, 2009), 232. Quotation also reproduced in Delp, Jung and the Modern Burden of Soul, Introduction and Chapter 5.
Leave a Reply