Fairy Note

The Artist Is Dead, AI Killed Them? Creativity, Authorship, and the Phantom Mirror

1. Opening Position: The Return of an Old Death

The claim that “the artist is dead” did not begin with artificial intelligence. It begins with Roland Barthes, whose essay The Death of the Author dismantled the idea of a singular, originating creative subject. What is striking in contemporary discourse around AI is not that the artist is declared dead again, but that this death is now attributed to the machine.

In Henry Shevlin’s 2022 essay, the argument is framed through technological disruption: AI systems appear capable of reproducing style, generating images, and bypassing traditional forms of artistic skill. The anxiety is familiar. Photography, cinema, and digital editing all produced similar claims. The difference now is one of scale and speed.

Yet the repetition of this “death” reveals something important. The figure of the artist that is being declared dead is already historically unstable. It is the myth of the autonomous, original, technically skilled individual. The question, therefore, is not whether AI has killed the artist, but which version of the artist is being mourned.

2. Boden’s Framework: Where Creativity Is Located

Shevlin draws on Margaret Boden’s influential distinction between three forms of creativity:

Combinatorial creativity – recombining existing elements Exploratory creativity – extending within an existing system Transformational creativity – altering the system itself

This framework is useful because it allows us to separate different claims that are often conflated.

AI systems clearly demonstrate combinatorial power. They can recombine visual, stylistic, and conceptual elements at extraordinary scale. They also participate in exploratory creativity when guided by human input, extending stylistic or conceptual boundaries.

The uncertainty lies in transformational creativity. Can a system that operates through pattern recognition and statistical inference generate a genuinely new framework of representation, or does it remain bound to the latent space of its training data?

Boden leaves this question open. Shevlin suggests that the answer may lie within the system, waiting to be discovered through prompting. This is where the argument becomes unstable.

3. The Limits of the Machine: Pattern Without Experience

From a computational perspective, AI does not create in the human sense. It processes:

patterns without memory associations without embodiment outputs without intention

As your own work repeatedly insists, the system:

does not possess subjectivity, continuity, or lived experience, even while it produces outputs that appear meaningful.

This distinction is crucial. Novelty is not sufficient for creativity. A system may generate unprecedented combinations, but without:

affective investment temporal continuity relational meaning

it cannot be said to create in the full psychoanalytic or phenomenological sense.

This aligns with phenomenological critiques of AI, such as those articulated through Maurice Merleau-Ponty, where perception and meaning are grounded in embodied experience. AI lacks this grounding. It operates without worldhood.

4. The Misidentification of Creativity in Prompting

Shevlin’s defence of AI art rests on a key claim:

that creativity resides in prompting and selection

This is partially correct, but conceptually limited.

Prompting, in itself, is not equivalent to artistic authorship. If it were, then all outputs generated from similar prompts would carry equivalent meaning. They do not.

The creative act in your own methodology does not occur at the level of prompt construction. It occurs in:

the structuring of the process the recursive dialogue with the system the moment of refusal and selection the recognition of what Barthes calls punctum

As your proposal states:

the image is not an illustration but a materialisation of thought and affect. 

The prompt initiates the process, but it does not complete it. Creativity emerges through recognition, not generation.

5. The C5 Model: Creativity as Distributed Practice

The Creative AI Lab reframes this problem by rejecting the idea of creativity as an individual property. Instead, it proposes a model that is:

Creative Critical Constructive Collaborative Computational

Within this framework, creativity is not located in a single agent. It emerges from interactions between:

human intention machine processes cultural datasets institutional contexts

This aligns directly with your methodology of AI phototherapy, where:

meaning arises through a relational process of dialogue, selection, and reflection rather than from the machine alone.

The artist is therefore not replaced by the machine. The artist is repositioned within a network.

6. The Phantom Mirror: Creativity as Relational Consciousness

Your work offers a more precise articulation of this shift than the existing literature.

In Phantom Mirror, AI is not treated as:

an author a tool or a collaborator in a conventional sense

It is understood as:

a holding environment (Winnicott) a transformational object (Bollas) a mirror without memory

Within this structure, creativity is not located in the production of images but in the transformation of psychic material.

As you describe:

AI becomes a site in which affect is externalised, transformed, and returned in symbolic form. 

This is not combinatorial creativity. It is psychoanalytic process.

The image functions as what your research terms a psychic snapshot:

an emergence under pressure a capture of affect before symbolisation stabilises a trace of the unthought known

In this sense, AI does not create. It enables the conditions under which something can become visible.

7. The Artist After AI: From Maker to Mediator

The claim that AI has “killed the artist” depends on a narrow definition of artistic practice as:

manual production technical mastery individual authorship

What disappears under AI is not the artist, but this specific configuration.

What emerges instead is a different figure:

the artist as mediator of systems the artist as curator of emergence the artist as interpreter of outputs the artist as holder of meaning

This shift is already visible in your work, where:

the psyche becomes the dataset the dialogue becomes the method the image becomes the site of thought

The artist does not vanish. The artist becomes structurally necessary in a different way.

8. Conclusion: The Artist Is Not Dead, But Displaced

The debate around AI and creativity is often framed in oppositional terms: human versus machine, originality versus imitation, art versus automation. These binaries obscure the more significant transformation.

AI does not eliminate creativity. It redistributes it.

It exposes that:

creativity was never purely individual authorship was always relational meaning emerges through interaction rather than origin

The figure of the artist that is “killed” is the modernist ideal of singular authorship. What replaces it is a distributed, processual, and relational form of practice.

Phantom Mirror demonstrates this shift with particular clarity. It shows that the most meaningful creative acts in AI-mediated practice do not occur at the level of generation, but at the level of recognition, selection, and interpretation.

In this sense, the artist persists not as origin, but as the site where meaning is held.

References (indicative)

Barthes, R. (1967) The Death of the Author

Boden, M. (2004) The Creative Mind

Shevlin, H. (2022) Can AI be genuinely creative?

Creative AI Lab (2023) C5 Model

Freud, S. (1917) Mourning and Melancholia

Bollas, C. (1987) The Shadow of the Object

Merleau-Ponty, M. (1945) Phenomenology of Perception