- Diese Veranstaltung hat bereits stattgefunden.
Interne Konferenz »Die technologische Bedingung«
In unser ersten internen Fachtagung wollen wir uns schwerpunktmäßig mit den technologischen Implikationen unseres Themas befassen, das heißt mit technischen Daten und Fakten, die das Thema KI betreffen und die gelegentlich, im Furor philosophisch-künstlerischer Interpretationen zu kurz kommen. Die Verbundpartner:innen stellen hierzu in Impulsvorträgen und Kurzreferaten ihre jeweilige Position aufs Thema vor, damit wir im Anschluss in eine gemeinsame Diskussion einsteigen können.
Programm
10:00 Graduate-Block
Francisco Uberto (HfMT Hamburg)
P.A.R.T.Y.
In this presentation, I will share the development and challenges of a work-in-progress artistic project named P.A.R.T.Y., conceived as a participatory game blending multimedia installation, theatrical performance, and live-cinema. The project investigates mechanisms that generate interactions between audience, performers, and AI agents, aiming for an organic unfolding of the artistic experience—a central question of this presentation: what does it mean for an experience to unfold organically, and why does it matter?
Technical and conceptual challenges include analyzing text emotion, earning and losing tokens, injecting emotion, translating between digital and analogue, and overlaying digital elements onto a polyphonic live-cinema environment. At the core of the project is the inseparable feedback loop between human and AI: each influences and responds to the other in real time, generating emergent dynamics that cannot be reduced to separate contributions.
Hypothesis: it is in this self-organizing flow,—when participants let themselves be carried by the flow, by the system’s own pace, by the feedback loop—that the interaction truly emerges. How much are we shaped by the process? To what extent are we conscious participants in the dynamics we contribute to generate?
Theresa Büchner (HAW Hamburg)
Viel Lärm
Die rasant fortschreitende Entwicklung generativer KI kann nicht nur künstlerische Produktionsprozesse verändern, sondern auch ästhetische Konventionen. Parallel dazu entstehen neue Narrative über Künstliche Intelligenz, die sich zwischen techno-utopischen Heilsversprechen und Weltuntergangsszenarien bewegen. Verbreitet werden diese Erzählungen von PodcasterInnen und JournalistInnen, CEOs und Tech-DissidentInnen, deren öffentliche Selbstinszenierungen an Figuren aus Das Leben des Brian oder Final Fantasy XVI erinnern. Wahre Dramen! Welche dieser Narrative sich durchsetzen, von wem sie getragen werden und welche Interessen sie bedienen, ist ein Gegenstand meiner Untersuchung.
Ausgehend von der Ausstellung The World Through AI im Jeu de Paume reflektiert der Vortrag außerdem künstlerische Strategien, die von didaktischen und warnenden Ansätzen, über Fluxus-Anleihen und dem Spiel mit Fehlern, bis hin zur Generierung neuer historischer Artefakte und Gesten posthumer Würdigung reichen. Im Fokus steht hierbei die Frage, welche Formen von Autorschaft sich im Zeitalter generativer Systeme neu konfigurieren.
Aus der Perspektive einer bildenden Künstlerin behandelt mein Beitrag die Frage, wie sich künstlerische Praxis verändert, wenn anstatt eigener Produktion das Auswählen, Kuratieren und Kontextualisieren von KI-generierten Inhalten ins Zentrum rückt. Wird das Wissen um die Herkunft bestimmter kulturhistorischer Referenzen an Bedeutung verlieren oder gerade zum Distinktionsvorteil avancieren? Entsteht eine neue Gleichgültigkeit gegenüber Autorschaft und damit eine Art ästhetisches Rauschen? Immerhin scheint es längst zur Praxis geworden zu sein, das Einverständnis zur Verwendung von Material für das Training von Modellen gar nicht erst einzuholen. Oder wird das bewusste Zitieren und Benennen künftig zu einer zentralen künstlerischen Geste?
Lara Stöhlmacher (HCU Hamburg)
Encoding Situated Knowledges: Entanglement as Epistemic Condition in Design Practice
In architectural design, artificial intelligence is most often deployed to accelerate model iterations, optimize workflows, and enhance visualization (»RIBA AI Report,« 2024). Yet these applications remain tool-centric, largely detached from epistemic or cultural critique. By contrast, a growing body of architectural and computational researchers calls for more intentional and reflexive forms of data curation, recognizing the creative and epistemological potential that specialized data practices can unlock (Cardoso Llach and Ozkar 2019).
At the same time, AI has begun to migrate from technical application to conceptual framework. As terms such as entanglement, agency, or response-ability—drawn from feminist technoscience—increasingly occupy space within computational and design research, a rhetorical shift unfolds that exposes a tension between the language of intelligence and the realities of design practice: Who, or what, is entangled in these processes, and under what conditions can such entanglement become productive rather than extractive?
Through a design-led and ethnographically informed approach, the project seeks to examine how specific, subjective, and even contradictory accounts of spatial knowledge can be preserved and made generative within architectural workflows. It aims to develop a counter-methodology to the positivist tradition of spatial datafication, drawing on critical data studies, feminist technoscience, and architectural ethnography. Rather than pursuing efficiency or prediction, the research explores entanglement as an epistemic condition—a framework to study how knowledge is distributed across humans, tools, and infrastructures in the design process, and to probe how AI can participate, provoke, and interfere in situated, collective acts of making space.
12:00 Betreuenden-Runde & Gathering der Graduierenden
13:00 Lunch
14:00 Dozent:innen-Block
Prof. Georg Hajdu (HfMT Hamburg)
AI Diversity
My presentation is about using different kinds of AI in various settings. While the term AI has become nearly synonymous with prompting large language models (LLM), we will learn that the term is much wider and encompasses various use cases including machine learning and other forms of AI which I have applied since 1992. Recently, my focus has shifted to the combination of light-weight agents to achieve my artistic goals. We will discuss the strength and weaknesses of DLNs versus stochastic models—particularly a music machine called DJster based on the groundbreaking work by Clarence Barlow (1945-2023). While deep-learning networks excel at interpolating between learned datasets (corpora), DJster works by combining the mathematical representation of fundamental musical rules, thus allowing the extrapolation into unexplored musical territories. The combination of the two approaches will ultimately ensure that DLNs don’t get stuck in »infinite loops of self-reference«.
Dr. Benjamin Sprick (HfMT Hamburg)
Digital Idiocy #1 – Notes on the symptoms of authoritarian algorithmicity
Occasionally, it is also useful to understand the dynamics of discourses surrounding »artificial intelligence« from the perspective of their opposite. From an artistic perspective, this allows us to argue affirmatively against culturally pessimistic views, just as we question the increasingly uncritical obedience that is also evident in the sciences. Based on current symptoms, this article attempts to paint an artistic picture of digital idiocy. This derives a latent creative potential from it (»artistic intelligence«). The theoretical inspiration for this comes not only from canonical authors such as Fyodor Dostoyevsky, Immanuel Kant, and Martin Heidegger. Reference is also made to the study Cruel Optimism by feminist cultural theorist Lauren Berlant, which allows a connection to be drawn to the current political present.
Prof. Diedrich Wolter (Universität of Lübeck)
AI on Creativity: Case-Based Reasoning, Analogies, and Conceptual Blending
This talk reviews AI techniques with respect to their potential contribution to achieving artistic intelligence, focusing on computational creativity as it is discussed in the AI community. Several general-purpose techniques have been developed in the AI community for solving unseen problems in novel and elegant ways, thereby linking problem solving to creativity. In particular, we discuss three paradigms. Case-based reasoning aims to compose solutions by adapting existing ones that succeeded in related tasks. The field of analogical reasoning focuses on relatedness by discovering structural similarities and using those to construct new solutions. Like analogical reasoning, conceptual blending techniques draw inspiration from human cognition, combining elements of existing solutions to create new ones. As I will argue in the talk, all of these techniques – as well as simpler ones – have the potential to achieve what a human might regard as creative solutions, but judging on the basis of the outcome is misleading.
Prof. Nicola Hein (MHL Lübeck)
(Auto-)Cybernetization, Datafication, and Technodiversity
The talk will take the concept of cybernetization as the technological condition, as proposed by Erich Hörl and Luciana Parisi, as a starting point for exploring perspectives on Auto-Cybernetization in Music and Media Arts. I will explore and discuss my concept of Auto-Cybernetization, which I use to describe the interpretation of human musicians’ musical material through musical agents, using machine learning. This will lead into a discussion of Martin Heidegger’s concept of technology and the contingency and situatedness of technology via Yuk Hui’s Technodiversity and Cosmotechnology. Relating to Chris Salter’s Sensing Machines, Zeynep Bulut’s Voice as Skin, Donna Haraway’s Cyborg and others, I want to raise the question, how these musical practices relate to the increasingly agentic media and the process of Datafication and Cybernetization in society.
16:00 Diskussion
»Artistic Intelligence in Latent Creative Spaces – Welchen Herausforderungen durch KI ist die Künstlerische Forschung aktuell ausgesetzt?«
Moderation: Prof. Sabine Hansmann (HCU Hamburg)
