Artificial Creativity in Theory and Practice
26–27 May, 2026
School of Arts and Humanities
University of Lisbon
Room B112.C
What is creativity? How is it valuable? And how might it be harmful? This conference examines how recent developments in artificial intelligence affect our answers to these questions. It will take a broad look at the relationship between AI and creativity – including creativity in the arts, sciences, business, and beyond. In addition to this, the conference considers how AI is changing our creative practice. How are artists and people working in the creative industries using AI? And what are the social, ethical and political implications of this?
If you wish to attend, please register via this link.
Conference Program
This program is provisional and may be subject to minor changes. Please check the program the week before the conference for the final version.
Day 1 (May 26)
09:00–09:30 Registration
09:30–09:45 Welcome Address
09:45–11:00 Keynote
Caterina Moruzzi (University of Edinburgh)
“From Poiesis to Praxis in AI and Creativity”
Chair: James Pearson
11:00–11:30 Break
11:30–13:00
Filippo Santoni de Sio (Eindhoven University of Technology)
“AI, Creativity and Meaningful Human Control”
Paolo Stellino (NOVA University Lisbon, IFILNOVA)
“Tilly Norwood: The Future of Cinema or a Threat to It?”
Chair: Antonio Oraldi
13:00–14:30 Lunch
14:30–16:00
Moirika Reker (University of Lisbon, CFUL)
“Creativity? Prometheus Lurking, or Looking at AI with Mumford and Assunto”
José Quaresma (University of Lisbon, CFUL)
“The States of Mind that AI can Induce in the ‘Self-Awareness’ of the Artistic Gesture”
Chair: Matthew Dennis
16:00–16:15 Coffee break (informal)
16:15–17:45
Rafael Brundo Uriarte (Kunsthistorisches Institut in Florenz – Max-Planck-Institute)
“Abundant Novelty, Limited Creativity: AI, Memory, and Cultural Selection”
Johannes Stelzer (Champalimaud Foundation)
“From Lunar Ring to the Clinic: Immersive AI as a Therapeutic Environment”
Chair: Robert William Clowes
17:45–19:00 Drinks (informal)
19:30 Conference dinner
Day 2 (May 27)
9:15–9:45 Morning coffee (informal). Café Jardim das Letras (here)
10:00–11:15 Keynote
Robert William Clowes (NOVA University Lisbon; Ruhr University Bochum)
“Generative AI Creativity and the Human Imagination”
Chair: James Pearson
11:15–11:30 Break
11:30–13:00
Matthew J. Dennis (Eindhoven University of Technology)
“Authentic Creativity and the Spectrum of Hybridity”
Maile Costa Colbert (NOVA University Lisbon, IFILNOVA)
“Artificial Atmospheres: AI, Art, and the Activation of Archives”
Chair: Nikos Erinakis
13:00–14:30 Lunch
14:30–16:00
Nikos Erinakis (University of Crete)
“Beyond Humanness and Artificiality: Authenticity, Creativity and the Reordering of Freedom”
Tamara Caraus (University of Lisbon, CFUL)
“The Culture Industry Reloaded: Adorno, AI and the End of Negativity”
Chair: Moirika Reker
16:00–16:20 Concluding remarks
17:00–18:00 Automated art exhibition visit and artist talk
Exhibition: Leonel Moura Lascaux 2.0 (exhibition description available here).
Speaker: Leonel Moura (for more information, see here).
Location: Forum Picoas (a 16-min journey by metro; directions are available here).
Keynote Talks
May 26
From Poiesis to Praxis in AI and Creativity
Dr Caterina Moruzzi (Institute for Design Informatics, The University of Edinburgh)
Generative AI is often presented as a promise of frictionless creation: more content, faster, with fewer constraints. Yet, this frictionless future risks threatening our sense of creative agency, the grounds on which we claim authorship, and the conditions under which audiences perceive something as an authentic creation.
In this talk, I look back over the last decade of AI and creativity and trace how central questions in the field are converging around these three pillars — agency, authorship, and authenticity — and how they have shaped my own research. I suggest that progress on each depends on a productive cycle between theory and practice and, increasingly, intersectoral engagement with creative communities, policy bodies, and technology companies.
In doing so, I argue for a more praxis-driven scholarship on AI and creativity, one that combines the Aristotelian virtues of situated judgment with the Marxian imperative to engage the conditions that shape creative work, rather than merely reacting to each new system. I close with a proposal for a future agenda: in a culture that increasingly equates creativity with productivity and abundance, what is most at risk is responsiveness — to materials, constraints, sources, consequences, and audiences. The challenge ahead is to design, study, and govern creative applications of AI in ways that keep this responsiveness alive.
Bio
Caterina Moruzzi is a Chancellor’s fellow in Design Informatics, University of Edinburgh. Her research lies at the intersection between the philosophy of art, human and artificial creativity, and the philosophy of Artificial Intelligence. She holds a Ph.D. in philosophy from the University of Nottingham and an artist diploma in piano performance from the Conservatorio G.B. Martini, Bologna.
As BRAID Research Fellow, Caterina leads a collaboration with Adobe to promote the responsible integration of AI tools into creative practices. At the forefront of the research on modes of shared agency and creativity between humans, data, and technology, she founded and leads the research cluster “Creativity, AI, and the Human” at the Edinburgh Futures Institute.
✳
May 27
Generative AI Creativity and the Human Imagination
Dr Robert Clowes (Universidade NOVA de Lisboa; Ruhr University Bochum)
Does generative AI threaten or extend the human imagination? In this talk I argue that, properly analysed, contemporary image and text generation systems are best understood as engines of constrained confabulation, operating across the compressed conceptual spaces implicit in their training data and producing novelty through interaction rather than autonomous invention. Drawing on Margaret Boden’s framework of creativity and a 4E cognitive lens, I show how genuinely creative acts can emerge from this interactive setting, even where the systems themselves lack the perceptuo-motor coupling and historical reach that characterise some forms of human creativity. The argument is grounded in a close reading of Elvis Deane’s early use of Midjourney to produce the webcomic Goats: The Dream (2022), a striking case of distributed authorship in which the artist’s intent was continually reshaped through cycles of prompt, generation, evaluation and re-prompt. I close by considering darker possibilities. Where users defer wholesale to these systems, we encounter what I call dead roads: trajectories of deskilling, narrowed exploration and privative unflourishing. The possibility of cognitive flourishing alongside generative AI is genuine but conditional, dependent on maintaining the reflective transparency, ownership and skilled practice that allow cognitive depth to develop with, rather than be displaced by, the new tools.
Bio
Robert W Clowes is a researcher with the Lisbon Mind Cognition and Knowledge Group and the Institute for Philosophy II, Ruhr University Bochum. He works at the intersection of philosophy of mind, philosophy of technology and cognitive science. He received his PhD at the University of Sussex where he also taught for many years. He is principal investigator on two projects: one on Companion Generative AI (funded by the IBM Tech Ethics Lab, Notre Dame University, US) and another on Cognitive Flourishing in the Context of GenAI (funded by FCSH, Nova Lisbon). He recently edited a special issue on the Mind-Technology Problem for the journal Social Epistemology and is currently working on another on Companion AI for Philosophical Studies. Currently he is completing a book on the Cognitive Ecology of Generative AI.
Abstracts
*May 26*
From Lunar Ring to the Clinic: Immersive AI as a Therapeutic Environment
Dr Johannes Stelzer (Champalimaud Foundation)
This talk traces a trajectory from early artistic encounters with generative artificial intelligence to its emerging use as a therapeutic medium. The starting point is Tübingen in 2015–16, where neural style transfer offered a first glimpse of machines not only analysing images, but generating visual forms that felt organic, unstable and unexpectedly alive. For us, the fascination was less the technical novelty than the experiential one: something appeared in front of us that had not existed before, and it suggested that generative systems could create new kinds of encounters.
This intuition became central to Lunar Ring, an artistic collective exploring artificial intelligence through immersive installations. What began with images gradually expanded into environments and worlds, adding movement, interaction, sound, space and embodiment. In public settings, these works often became social situations: people gathered around them, talked, played, interpreted and shared meaning through the presence of a responsive machine-generated environment.
The talk then follows this line of work into the clinic. At the Champalimaud Centre for Restorative Neurotechnology, we are now developing immersive AI systems as tools for clinicians, particularly in movement rehabilitation and exposure-based therapy for obsessive-compulsive disorder. In this context, generative environments are not only aesthetic or speculative. They can become controlled, adaptive and measurable spaces for therapy.
I will argue that the path from AI art to digital therapeutics is not incidental. Artistic experimentation helps us understand what these systems feel like from the inside; clinical research must determine when, how and for whom they can genuinely help.
✳
Tilly Norwood: The Future of Cinema or a Threat to It?
Dr Paolo Stellino (Universidade NOVA de Lisboa)
In 2025, Xicoia, an AI talent studio, introduced Tilly Norwood, widely described as the “world’s first AI actress.” Although the film industry had already been using AI for several years, Norwood’s hyperreal appearance—combined with the claim by Particle6, Xicoia’s parent production company, that her deployment could reduce production costs by 90%—sparked intense debate. Many Hollywood actors called for a boycott of any agency that signed Norwood, while several film critics voiced skepticism about the true potential of AI-generated performers.
Beyond these legitimate concerns, Norwood’s creation—likely the first of many fully AI-generated actors—raises broader philosophical questions about the future of art and its relationship with artificial intelligence. This talk will focus on two central issues: the ethical implications of AI in the film industry and its consequences for artistic creativity. To illuminate these concerns, it will draw a parallel with the rapidly expanding use of AI in academia, where similar debates about ethics, labor, and creativity are unfolding.
[Language editing for this abstract was carried out with the assistance of ChatGPT (Open AI)]
✳
The states of mind that AI can induce in the ‘self-awareness’ of the artistic gesture
Dr José Quaresma (University of Lisbon, CFUL)
This lecture aims to explore different states of self-consciousness during artistic practice using AI (Artificial Intelligence) procedures and environments. The “gaze” of the spectator, as discussed by Sartre, and the “second self,” as proposed by Sherry Turkle, will be the main sources of our discussion regarding the consciousness of the creative self and the demands of the “other” who looks at me during this creative and “artificial” gesture. We also intend to demonstrate how Hegel’s notion of self-consciousness in the “Phenomenology of Spirit” can be updated and considered crucial for understanding the role of AI in new creative processes
✳
Creativity? Prometheus Lurking, or Looking at AI with Mumford and Assunto
Dr Moirika Reker (University of Lisbon, CFUL)
The thought of pressing a button or having a machine that can unload us of our burdens, ease our days by liberating us from those tasks we loathe either because they are too heavy, painful, repetitive, risky or simply dull – leaving us with time and freedom to use at our discretion – sounds dreamlike, and it tempts us time and again. But would it really make us free? Do I really want to press a button and have my paper written for me?
A vacuum cleaner that does its thing while we are away, a technological arm that replaces an amputated one, those are different uses of technology that offer no dispute (or barely any). However, here we are concerned with a specific form of machine or technology use that somehow turns the picture upside down. At a first glance it seems to bind the two legacies that according to Lewis Mumford accompany human development from the beginning: Orpheus (who represents imagination and creative expression) and Prometheus (the constructive impulse, linked to the manufacture and use of tools, and to production). “Artificial creativity” seems to build a bridge between them, bringing creativity and technology into the same plane. Nevertheless, Rosario Assunto warns us of the risks of subjugating Orpheus/Amphion under the domain of Prometheus when discussing the two opposing ways of thinking about the city, and, ultimately, about ourselves, our place in and our relation to the world. The question is, do we want AI to be a tool in our creativity, or to be creative for us, i.e., instead of us?
*May 27*
Artificial Atmospheres: AI, Art, and the Activation of Archives
Dr Maile Costa Colbert (IFILNOVA-CineLab; FBAUL)
Recorded images and sounds carry traces of their past, both materially and in their content. Increasingly, these traces are being reorganized, reinterpreted, and recombined through artificial intelligence. As machine learning systems mediate how media archives are accessed and activated, archives are shifting from static repositories of preservation into dynamic environments shaped by algorithmic processes.
Approaching these transformations as a practicing intermedia artist and researcher, my work
explores relationships between time-based media, particularly sound and moving image, media archives, and machine learning technologies. Rather than treating archives solely as sites of preservation, this work investigates how artistic practice can reactivate archival materials through processes of recombination, interpretation, and creative experimentation. Rather than framing advances in AI as solely dystopian or utopian, the research also engages environmental and ethical questions surrounding AI infrastructures, including the ecological impact of large-scale computation and the responsibility of artists and researchers to develop more reflective and sustainable technological practices.
Through projects developed with the Art and Technology working group at CineLab (IFILNOVA) and collaborations with the Binaural Nodar digital media archive, this presentation examines how artists engage generative AI critically and creatively. In these contexts, archival sound and moving images become material for new artworks, while resulting works return to the archive as new layers within its evolving media ecology. The archive becomes not only a source but an active site of circulation, where media traces are reactivated and returned to cultural memory through unexpected media and artistic practice.
By foregrounding artistic experimentation as both method and inquiry, the talk positions practice-based artistic research as a vital space for exploring how archives, artworks, and AI systems continually reshape one another.
✳
Authentic Creativity and the Spectrum of Hybridity
Dr Matthew Dennis (TU Eindhoven)
Definitions of authenticity have often been contested in philosophy. Nevertheless, there is an emerging consensus that for creative objects to be authentic in the fullest sense of the term, then they must have a special kind of relation to the first-personal experience of their creator (author, artist, designer, etc.). Authenticity in this sense has greater normative weight than simply attributing authorship. We call this kind of authenticity, Au1, such as when an art historian claims a certain painting to be an ‘authentic Rembrandt’, rather than a forgery or an apprentice’s copy. Creative objects, such as paintings, poems, personal letters, authentically express the first-personal experience of their creator by revealing the creator’s true feelings, when they are emotionally honest, or when they disclose the creator’s real existential concerns. We call this type of authenticity, Au2. Recently the importance of authenticity (Au2) has gained urgency with the emergence of artificially intelligent machines that can effortlessly produce creative objects, including prize-winning images, sculpture, and literature. These creative objects are increasingly judged to be similar to those made by human beings, and a growing empirical literature suggests that they are rapidly becoming indistinguishable.
In this presentation, we analyse a range of creative objects that can be produced by machines, showing how, conceptually speaking, there is a strict separation between a group of creative objects that can be legitimately produced using AI and those that cannot. We believe the key to understanding this distinction is to focus on the role of first-personal experience in the creation of the creative object. For many creative objects, the role of first-personal experience is optional or ancillary, but for some it is fundamental. To understand this distinction, we situate a range of creative objects along what we term a ‘spectrum of hybridity’ in which AI is used to an increasingly greater extent. Outside this spectrum, we identify a set of creative objects (e.g., autobiographies, love letters, eulogies) that can only be created using AI at the cost of undermining their very status as authentic creative objects. Understanding the conceptual differences between the creative objects that AI can and cannot produce complements the empirical literature on this topic, as well as showing how philosophical analysis has a unique role to play in understanding how artificially intelligent machines can share our creative lives.
✳
Beyond Humanness and Artificiality:
Authenticity, Creativity and the Reordering of Freedom
Dr Nikos Erinakis (Department of Philosophy, University of Crete)
In the contemporary context between physical reality and digital hyperreality, the prevailing debate on artificial creativity seems to ask the wrong question. By opposing genuine creation to derivative simulation, it remains fixated on the status of the machine while overlooking a more important transformation: that of the human subject. This paper argues that artificial creativity is philosophically significant not because it may produce non-human creators, but because it reorders the conditions under which human beings create, recognize themselves in their acts, and understand their freedom. Generative systems do not simply expand the means of expression; they alter the normative structure of agency by distributing invention across technical operations that exceed the subject’s grasp while presenting this dispersal as empowerment. The result is a paradoxical form of freedom: an unprecedented enlargement of expressive possibility coupled with a thinning of authorship, responsibility, and self-determination. What is threatened under these conditions is not the romantic ideal of solitary genius, but the more fragile possibility of authenticity—of standing in a relation of reflective ownership to what one makes. Artificial creativity thus introduces a historically novel form of heteronomy, one that seduces through convenience, abundance, and immediacy rather than imposing itself through prohibition. The central question is therefore not whether AI genuinely creates, but how artificial creativity reorganizes human authorship and, with it, our concepts of authenticity, freedom, and autonomy. Authenticity is not understood as an individualistic trait of self-expression or self-promotion, but as an existential consequence of responsibility that is realized in relation to others, while freedom presupposes morality and otherness. I argue that while generative systems widen the field of creative possibility, they also redistribute agency across opaque technical infrastructures, producing a tension between increased freedom of operation and diminished freedom of self-determination.
This tension matters because authenticity in creative life depends not merely on producing novel works, but on being able to recognize and justify one’s relation to what one has made. From this perspective, the challenge raised by artificial creativity is not simply metaphysical or aesthetic, but ethical and political: it concerns the conditions under which subjects remain authors of their acts in technologically mediated environments. The paper develops this argument by distinguishing assistance from substitution, augmentation from heteronomy, and collaboration from delegated agency. On this account, AI-assisted creation is not necessarily inauthentic, but it becomes problematic when persons treat generated outputs merely as convenient solutions rather than as material they can creatively and critically appropriate as their own. I conclude by proposing a conception of authentic co-creation (authenticity-as-cocreativity) and criteria for distinguishing autonomy-supporting (freedom-as-equautonomy) from autonomy-undermining forms of artificial creativity. I conceive this within a vision of hyper-dialectical eutopia, a transition from the anthropocene to a symbiocene and noocene perspective, where humans further develop their humanness co-creatively, through a humanizing and democratizing, mildly technophilic, use of artificial creativity, with the aim of a society of relationships between persons rather than spectacles of personas.
✳
The Culture Industry Reloaded: Adorno, AI and the End of Negativity
Dr Tamara Caraus (University of Lisbon; CFUL)
This presentation examines AI-generated art and creativity through two central dimensions of Theodor W. Adorno’s critical theory: his critique of the culture industry and his theory of art developed in Aesthetic Theory. It argues that generative AI represents not simply a technological innovation in artistic production, but a historical intensification of the tendencies Adorno identified within the culture of what he called ‘late capitalism’. Thus, from the perspective of the culture industry critique, AI systems appear as the culmination of industrialized cultural production, characterised through standardization, pseudo-individualization, commodification, and the administration of experience. In the AI-generated artistic outputs creation becomes automated, style becomes extractable, artistic techniques become datasets. Algorithms increasingly optimize cultural production for engagement, speed, content circulation and consumption. In this sense, AI culture radicalizes the logic of the culture industry by extending commodification into the sphere of imagination itself. The presentation further examines AI art through Adorno’s aesthetic theory, particularly his conception of art as a site of negativity. According to Adorno, modern capitalist society produces conformity, instrumental rationality, and domination, life becomes governed by exchange value and efficiency, while artworks resist social domination through contradiction, tension, dissonance, and non-identity. AI-generated art, whose means are the result of instrumental rationality and exchange, threatens to erode art’s critical and oppositional function. The presentation concludes that from an Adornian perspective AI signals the possible ‘end of negativity’: the neutralization of aesthetic resistance within an increasingly automated and administrated cultural order.
✳
Organization: Dr James S. Pearson (Praxis-CFUL)
This event is funded by Portuguese national funds through FCT – Fundação para a Ciência e a Tecnologia, I.P., in the scope of the project UIDB/00310/2025, Centro de Filosofia da Universidade de Lisboa (https://doi.org/10.54499/UID/00310/2025)