On 3 April I was invited to speak at the vernissage of Simulacru – Generative Thoughts, an exhibition by the emerging artist Ovidiu Belceanu, opened in the Foyer of the Babeș-Bolyai University Centre in Reșița — a context that generated a series of essential reflections on the profound transformations of contemporary art.
We are, without doubt, at a turning point: the image is no longer, in the traditional sense, created, but generated. We no longer speak exclusively of the relationship between artist and work, but of an extended ecosystem in which the artist, the algorithms, and the data intervene together. This mutation is not only technological, but conceptual — it reconfigures the very nature of the creative act.
Generative art, in the context of artificial intelligence, fundamentally changes the way we understand the artistic process. If, within the classical paradigm, the artist controlled every stage of creation, today the artist becomes rather an architect of systems: establishing rules, parameters, and initial conditions, while the final image is the result of an autonomous process. The artist no longer produces the form directly, but configures the frame from which it emerges.
This direction continues, in a technological form, the idea formulated in conceptual art by Sol LeWitt, who held that the idea is more important than the execution. However, in the case of artificial intelligence, the delegation of execution acquires a radical dimension: it is no longer a matter of another human agent, but of systems trained on massive volumes of visual data.
And here an essential difference intervenes. Algorithms do not create from nothing — they are trained on images, on practically the entire visual culture accumulated globally up to the present moment. Each generated image thus becomes a statistical synthesis of the past, a recombination of what has already existed. In this context, notions such as originality, authorship, or authenticity become profoundly unstable.
This problematic is deepened by the theoretical perspective of Joanna Zylinska, who emphasises that algorithmic generation systems are never neutral. They do not merely produce images — they construct and reproduce cultural structures. In other words, algorithms function as active media for the reconfiguration of visual culture, filtering and redistributing existing meanings.
A telling example in this regard is the viral TikTok phenomenon known as „AI Barbie". Within this trend, users generated versions of the Barbie doll associated with different cultural contexts. The results clearly highlighted that these systems reproduce existing patterns: the American Barbie was represented as extremely colourful and exuberant; the German Barbie included visual elements with sensitive historical resonances; and the Sudanese Barbie was placed in arid landscapes and, at times, associated with military contexts.
These representations are not accidental. They reflect how training data contains and transmits stereotypes, hierarchies, and cultural biases. Artificial intelligence does not „understand" culture, but reconstructs it statistically — and in this process it can consolidate or amplify existing clichés.
At this point, the discussion can no longer remain at the technological level. As Joanna Zylinska suggests, a shift of emphasis towards responsibility is necessary. The contemporary artist is no longer merely a creator of images, but becomes a curator of algorithmic processes.
This responsibility extends to the entire chain of production: from the choice of instruments and data, to the way images are generated and presented to the public. Because these images can carry, sometimes subtly, stereotypes and problematic narratives, the artist becomes a critical mediator between technology and society.
Thus, the fundamental question is no longer „what do we create with AI?", but „how and where do we use these systems?". Because artificial intelligence does not merely generate images — it contributes to generating ways of looking at the world.
And these modes of representation, however sophisticated the technologies that produce them, remain ultimately within human responsibility. The algorithm does not decide meaning — the context in which it is used does. In this equation, the artist does not disappear, but is transformed: from direct creator into curator of the possible, and at the same time into an ethical actor of contemporary culture.
