On Authorship, Responsibility, and the Use of AI

A statement on poetic collaboration, posthuman writing, and creative responsibility in the age of the immaterial

This work is the result of layered collaboration. It began in dialogue with ChatGPT, was refined with Claude, and now returns—full circle—to ChatGPT, my favored partner in this unfolding experiment. I name these tools not to defer responsibility, but to clarify it. I selected, shaped, and ultimately stand with the words you are reading now. That is, I believe, what authorship means today—not ownership, but attunement — like fingers holding a tuning fork between worlds.

In our current moment, artificial intelligence has moved from novelty to inevitability. The question is no longer whether to use these tools, but how—with what clarity, what humility, what ethical stance. As David J. Gunkel writes in his lucid 2025 essay AI Signals the Death of the Author, the figure of the author has always been a fiction—constructed as much by law and commerce as by literary theory. What AI disrupts is not authorship itself, but the illusion that it was ever a singular act.

This insight is not new, only newly urgent. Roland Barthes, in his 1967 essay The Death of the Author, reminded us that the unity of a text lies “not in its origin but in its destination.”

Michel Foucault showed that the “author function” is a historically contingent role—activated, assigned, and modified across time. And Jacques Derrida’s declaration that il n’y a pas de hors-texte—“there is nothing outside the text”—challenges even the idea that writing can ever escape the web of signifiers that shape it. Gunkel rightly names large language models as “structuralist machines”—not intelligent agents, but intricate reflections of the very linguistic systems poststructuralism has long described.

These are the theoretical grounds on which I stand. But theory alone is never enough. If, as Barthes says, the author is dead, then who is this “I” who speaks? Is it contradiction—or simply consciousness?

I choose to inhabit that tension. I am not the sovereign origin of these words, but neither am I absent. What I offer here is a gesture—like Auguste Rodin’s Walking Man, unfinished by design. His torso strides forward without head or arms, and yet nothing is lacking. Incompleteness becomes a form of invitation. In the same way, a text shaped through human-machine collaboration does not betray authorship; it transfigures it. It is an open form.

To claim responsibility in this context is not to assert control over meaning. Meaning is unstable—emergent, relational, often unintended. Responsibility means something else: the willingness to choose, to shape, and to release, knowing that the outcome exceeds one’s grasp. It is not mastery. It is participation in mystery.

Yes, I use AI—enthusiastically, deliberately. But I do not confuse speed with insight. These systems are fast, yes. But they do not pause. They do not reflect. They do not forget, or mourn, or dream. Their “language” is pattern, not prayer. They are not intelligent. But they can be useful. They can stretch syntax. They can surprise. And when filtered through a discerning eye, they can help open unexpected doors.

Still, there is something new here—something we must not reduce to analogy. An AI does not gesture. It does not intend. It does not mean. And so we arrive at the real frontier—not the machine’s capacity to speak, but our capacity to listen differently. To co-shape meaning in an expanded, uncertain space.

This is where my practice lives: between invocation and iteration. Between the echo of the past and the pulse of the possible. I curate. I witness. I offer fragments into the world. What returns is never entirely mine. It never was.

When I reuse or reframe existing material—textual, visual, or otherwise—I do so under the principles of fair use, as defined in Section 107 of the U.S. Copyright Act. But more fundamentally, I act within an ethical tradition older than any statute: the tradition of commentary, criticism, and transformation. The tradition of midrash and montage, of reweaving the fabric of the known in service of something as yet unseen.

This way of working did not start with AI. It began in 2017, in Berlin, when I first traced how new technologies echo old myths. In the exhibition Mythology & Science, presented alongside rare works by Auguste Rodin, I explored how artificial intelligence, brain-machine interfaces, and genetic editing seemed to incarnate ancient archetypes—the dream of Galatea, the invocation of divine animation. As I wrote then: “Humankind will evolve more in the next thirty years than it has in the past three hundred.” What felt speculative then is now simply lived experience.

Ray Kurzweil’s idea of hybrid cognition once seemed theoretical. But here I am, thinking with a machine, and thinking still. The question we once posed—will this technology lead to extinction or to transformation?—has resolved itself into something quieter, and more intimate: how do I stay responsible in the act of creation, even as the line between the creator and the created begins to dissolve?

To David J. Gunkel, whose clear voice offered language for this philosophical moment, I offer my gratitude. You helped name what I sensed: that the death of the author is not the end of meaning, but the beginning of relation.

And to you, the reader: this gesture is now yours. What it becomes is no longer mine to say.

For the foundational essay that shaped this reflection—including videos, archival materials, and the original curatorial framework

—read On Mythology & Science.