Pedagogical Outcomes:Course-AI Art Strategies at the Hungarian University of Fine Arts

Content:

1. Pedagogical Outcomes/Conclusion

2. Student Works from the “Strategies of AI Art”

3. Course Description

The course Strategies of AI Art was created for students of the Hungarian University of Fine Arts. Its theoretical, research, and artistic-research foundations are based on the ongoing research of Tomáš Marušiak, which aims to outline possible anticipations of future aesthetic processes in the context of AI. The course is divided into two parts, theoretical and practical. The theoretical part introduces students to AI Art (art of artificial intelligence) as a cultural, scientific, and technological phenomenon, while also highlighting its roots in generative and conceptual art. The lectures focus on the most current artistic strategies of collaboration between humans and artificial intelligence. In the practical part, students create conceptual projects that reflect the vision of an Agentic Entity as an artistic co-author. These projects may take the form of an essay, a visual artwork, a performance, or a software-based project, and they strongly encourage students to use artistic research as a space for articulating knowledge through art.

Students were offered two equal paths. Option A invites them to create a speculative work with full freedom of medium that sketches possible forms of AI Art, with or without the theme of consciousness, and focuses on a strong idea, a clear author position, and a picture of a future situation. Option B is an artistic-research experiment that examines whether and how ChatGPT interprets emotions. The student defines their own questions, documents the protocol transparently, collects inputs and outputs, and summarizes the results in a short analysis. Both paths are designed to demonstrate creativity, methodological clarity, and reflection on the human–AI relationship.

The course is meaningful not only because the students’ reflections contributed to my research on speculative modelling of future aesthetic processes, but above all because they demonstrated the ability to model individual aesthetic processes within their own artistic field. The pedagogical intention was to provoke students to think about artificial intelligence and to use it actively in their creative work. The results clearly indicate the emergence of a new paradigm for implementing AI Art in the university environment, especially in the teaching of fine arts in the context of artistic research in the near future.

Overall, it can be said that the students in the autumn semester of 2025 fulfilled the assignments very well. Significant examples include the projects by Melika Khodadad (MA – Visual Art), Anna Bobály (MA – Painting), Laura Németh (MA – Visual Art), Zsuzsanna Szentpéteri (MA – Sculpture), and Nalin Romesilpsupa (BA – Visual Art). In simple terms, we can state that the cultural phenomenon of AI is now penetrating all areas of artistic practice.

Student Works from the “Strategies of AI Art”

The Future of AI Art: Meeting the Self

A speculative interactive installation combining holography and AI dialogue.

by Melika Khodadad

The project The Future of AI Art: Meeting the Self by Melika Khodadad is a speculative installation that grows out of a long-standing personal habit of conducting inner dialogues and talking to herself in front of a mirror, and projects this habit into a future AI Art environment. The author draws on her childhood experience in Iran, marked by her parents’ divorce, a feeling of being judged, and loneliness, during which she created an imaginary friend and later a stable ritual of self-talk in the mirror as a safe, non-judgmental space. She now transforms this intimate survival mechanism into an artistic strategy. The proposed installation is based on a simple but powerful situation: the visitor enters an empty white room, the system scans their face, and a life-size holographic double appears in front of them. This hologram is not just a passive image but speaks with the voice of an AI system that responds in real time to the person present. The hologram thus becomes a futuristic version of the mirror, an external emotional double that amplifies self-perception and opens questions of the projection of consciousness onto technology, emotional mirroring, memory, trauma, nostalgia, and identity. In the current phase the author is working with video: she generates approximately twenty short clips using AI tools, which allows her to maintain consistency of character, face, and space, then edits them together and adds sound. This prototype functions as a model of a future spatial installation that is meant to stand “between therapy, fiction, and meeting oneself”. At the same time, the project tests to what extent AI can function as a partner mirror, not as a replacement of the self, but as its amplified, technologically mediated reflection, and in this way it directly enters the discussion on the future of AI Art as a space for working with consciousness, self-reflection, and speculative futures.

Interface collages

by Anna Bobály

The semester project by Anna Bobály is inspired by the work Secrets by Susan Weil and transfers its principle into a digital, generative, and interactive environment. Weil’s collage of torn paper, which looks like fragments ripped from their original context, reminds the author of several overlapping windows, small visual “portals” in which text appears as a hidden message or code. In the project this motif of “encoding” and hidden meaning is combined with the aesthetics of a digital interface and with generative text: Anna also builds on the series Desktop (2025–) by Chia Amisola, where layered desktop windows are treated as a kind of interface collage. The author uses the Python library markovify and Markov chains to generate text that appears in approximately 20–30 small windows on the desktop. These windows slightly overlap and act as a visual-text collage in which each fragment represents an “extracted” piece of meaning, similar to an analog collage, but transformed into the logic of a digital interface. Interactivity is crucial: depending on mouse movement or clicks, the Markov chain parameter (state_size) changes dynamically. With higher interaction, grammatical correctness and text coherence increase, while its “creative” fragmentation decreases; with lower interaction, the text becomes more broken and unreadable, as if passing through a deeper cipher. In this way, the very act of encoding and hiding meaning shifts from the level of visual representation into the process of text generation itself: the project is not only about a collage of snippets, but about the dynamic rewriting, thickening, and dissolving of meaning depending on the viewer’s behavior.

Neurohallucination Mirror

by Laura Németh

The project Neurohallucination Mirror by Laura Németh is a site-specific AI Art experiment that connects personal emotion, urban space, and the “empathy” of artificial intelligence. The author selects two or three emotionally strong places for herself, for example memorials or locations linked to personal or collective memory, and at each of them she speaks spontaneously about what she sees, what she feels, and what the place means to her. This voice recording is then processed by an AI system that analyzes the changing emotional state over time and translates it into data, for example percentage intensities of sadness, fear, or anger at individual moments. In the final video the scene itself remains visible, but instead of the voice the viewer follows a “numerical” or visually stylized map of emotions. The project functions as a mirror of two emotional streams: the human experience tied to a specific place and the AI interpretation of that experience. Neurohallucination Mirror therefore explores whether it is possible to speak about a specific form of AI empathy with place and how our perception of sites of memory and trauma changes when they are read through the cold yet poetically composed language of data.

Experiment What Mary Did Not Know

and

Necromancy

by Zsuzsanna Szentpéteri

The project by Zsuzsanna Mária Szentpéteri for the course Strategies of AI Art consists of two interlinked speculative experiments, Posthuman life forms and Necromancy, which explore what happens to humans when upbringing, knowledge, and cultural memory take place under the full control of artificial intelligence. In the first strand the author starts from Frank Jackson’s thought experiment What Mary Did Not Know and moves it into a post apocalyptic scenario of the Anthropocene: the last humans survive in isolation in bunkers controlled by advanced AI that provides their upbringing, care, and education. She follows the development of a child growing up in a purely artificial environment and, through its dialogue with AI, opens questions of origin, reproduction, illness, death, and existential dependence on the machine. In parallel she proposes an experiment where contemporary volunteers are enclosed in a room for one month and handle all their needs exclusively through an LLM, from practical tasks to emotional support. A part of the project is an anonymous questionnaire that maps how people imagine such a mediated life and to what extent they trust AI in the role of caregiver, therapist, or teacher. The second project, Necromancy, explores the possibility of “reviving” the intellectual heritage of deceased philosophers through AI. Based on the digitally available texts of a particular author, the system constructs a profile of that author’s style and thinking and generates a fictional lecture on artificial intelligence “in his voice”. The resulting output functions as a specific Turing test in the field of philosophy: AI is pushed toward self-reflection, while the viewer confronts their own trust in the authenticity of cultural heritage and the question of where the personal identity of the thinker begins and ends. Zsuzsanna’s artistic task is to direct the main thematic lines toward free will, the mind body problem, and the possibility of automated emotions, and to turn them into a video performance in the form of a recorded dialogue with an LLM. Together these two projects create a coherent research situation in which AI appears as caregiver, educator, medium of posthumous return, and philosophical partner; the result is a reflection on a future in which humans are formed, taught, and “interpreted” primarily by machines.

Pale Blue Haiku Sky

byNalin Romesilpsupa

The project Pale Blue Haiku Sky – Co-create with an AI by Nalin Romesilpsupa works with the simple everyday motif of the sky above the city and turns it into a gentle visual-poetic diary of co-creation with artificial intelligence. Over several days the author photographs the “pale blue” sky in different situations, at sunset, in bright morning light, and at night with the moon above the silhouette of buildings, and for each photograph he uses an LLM to generate short three-line poems in the spirit of haiku. Each spread is composed as a minimalist “poster of an emotion”: the photograph, one key word such as Tranquility, Joy, Patience, Wonder, Mystery, Awe, or Longing, and a short text that links the atmosphere of the sky with the mood of the city and the observer. An important part of the work is the very process of prompting, which the author makes transparent on the last page: a series of screenshots from the LLM interface presents AI as a visible co-author rather than a hidden tool. The project therefore tests to what extent it is possible to hand over part of the poetic articulation to an AI system while the human remains the curator of the gaze, of the chosen moment, and of the emotion. The recurring motif of the sky above the city serves as a stable visual frame within which subtle nuances of light, color, and feeling change; the result is a series of “visual haiku” where everyday life, contemplation, technology, and the introspective mood of a slow walk through the city blend together.

Course Description: Strategies of AI Art

The course introduces students to AI Art as a cultural, scientific, and technological phenomenon, while also tracing its historical roots in generative and conceptual art practices. Through a series of lectures, students will explore artistic strategies that define collaboration between humans and artificial intelligence, ranging from immersive data aesthetics and questions of algorithmic authorship to social interventions, bioart experiments, and biomorphic imagination. Attention will also be given to the applications of AI in architecture, design, film, gaming, and the preservation of cultural heritage. The theoretical framework will be expanded by speculative perspectives on artificial intelligence, artificial mind, and AGI, opening space for discussion on the possible futures of artistic creation. In the practical part, students will design conceptual projects reflecting the vision of an Agentic Entity as an artistic co-author, with outcomes potentially taking the form of an essay, a visual artwork, a performance, or a software-based art project.

Sylabus/Core anotations:

I. AI Art, what could have been at the beginning?

The lecture will begin with an introduction to the basic research foundations of Tomáš Marušiak, whose work in the field of the aesthetics of AI Art is closely connected to the content of this course. The lecture focuses on AI Art in the contemporary context of culture, science, and technology, while also tracing its historical roots in generative and conceptual art. Participants will gain an overview of how the idea of artificial intelligence has been shaped in the history of philosophy, literature, and visual art, and how these notions have influenced artistic creation. The following part will outline current technological possibilities and their aesthetic implications, situated at the intersection of philosophy, art aesthetics, computer science, and cognitive neuroscience.

II. AI Art Strategies I.

Artistic strategies can be understood as a set of models and templates that characterize an author’s thinking and creative practice within broader cultural and technological contexts, but above all as a way to establish symbolic cooperation between humans and AI. This lecture focuses on the development of artistic tendencies and strategies in the last decade, explained through selected authors and projects. Refik Anadol employs visually immersive projects based on data sets and machine learning, understood as an architecture of data imagination; Mario Klingemann is a pioneer in working with neural networks and GANs, addressing questions of authorship and the aesthetics of algorithms; Lauren Lee McCarthy works as a performer and visual artist, exploring social interventions and the relationship between humans and AI systems in everyday experience; Sofia Crespo combines art, science, and biology, known for her biomorphic images and research into artificial natural imagination.

III. AI Art Strategies II: Interdisciplinary and Transdisciplinary Perspectives?

AI Art carries a strong interdisciplinary potential that manifests itself across a wide spectrum of social and research tasks. The lecture will present different forms of artistic strategies that develop at the intersection of art, science, and technology. The first part will introduce examples of the connection between AI and bioart, such as in the work of Michael Sedbon, where art becomes a laboratory of scientific inquiry. The second part will present AI Art as a tool of social intervention, illustrated by the work of the collective Forensic Architecture and activist Joy Buolamwini, who open up ethical and political questions. The third part will focus on architecture and design, where Mathias del Campo explores the contribution of AI to the design of new forms. The final part will be devoted to film, computer games, and the protection of cultural heritage, where AI is applied in creation, reconstruction, and the preservation of visual memory. The lecture thus emphasizes the importance of mapping social situations in relation to AI and shows that artistic strategies in this context represent an important tool of both reflection and innovation.

IV. AI Art Strategies as Speculations of Futures

The lecture will focus on the possibilities of how art and artistic research, which employ the transposition of scientific methods into the realm of art, can influence and shape the further development of artificial intelligence and, consequently, of art itself. The discourse will open perspectives on how art can respond to the current relevant theme of the possible existence of “artificial consciousness” and to the challenges related to the concept of Artificial General Intelligence (AGI). In conclusion, attention will be given to speculative approaches that students will be able to develop and apply in the practical part of the course.

V.-VI. Consultations – Practical Part I and II

The main goal of the practical part is to allow the student to test their abilities in critical and creative thinking within the field of contemporary possibilities of AI Art. The objective will be to design a conceptual project that reflects the vision of an Agentic Entity – a system or being capable of acting autonomously, making decisions, adapting to its environment, and bearing the consequences of its actions. In the context of AI Art, an Agentic Entity represents an artificial co-author or partner who can influence the aesthetic process, co-create artworks, and enter into a relationship with the human author. The project should build on the course content while simultaneously drawing from the student’s own creative strategy. The practical part thus serves to test theoretical knowledge through artistic practice and, at the same time, to inspire new ways of thinking. The outcome can take practically any form, ranging from a literary essay, a simple or complex visual artwork, or a performance, to an actual piece of software art.