UiB AI #11: AI, Ethics, Aesthetics

How can we rethink the integration of artificial intelligence into creative human activity? Meet our researchers and participate in our next UiB AI seminar, this time organized by the Faculty of Fine Art, Music and Design.  

Hearing Bias? The Impact of AI-Driven Technologies on music production and listening by Jill Halstead, Grieg Research School in Interdisciplinary Music Studies 

The shifts in music technology that have driven musical innovation in the last 80 years have largely consolidated historic gender imbalances in the creation and production of music. Today, AI technologies are amplifying imbalances in ways that extend existing marginalisation and shape biases in music listening. This presentation considers these issues in relation to the Gender, Technology, Participation Project, a collaboration between Grieg Academy and Center for Women and Gender Studies at UiB. The project seeks to develop interdisciplinary research collaboration networks with the music sector, which will raise and address questions of gender bias in music technology design, music production and distribution.

Perceptions of AI and evaluations of authenticity online by Brita Ytre-Arne, Department of information science and media studies 

This talk presents a new pilot project that will explore how cultural imaginaries of AI influence people’s conceptions of what is real and fake in online media content. The possibility that images could be AI-generated changes public discourse and social negotiations of authenticity in communication, as seen in recent controversies from celebrity culture to war reporting. The pilot is funded by the UiB Humanities strategy, and the project is a collaboration between Bergen Media Use Research Group and Gabriele de Seta at Center for digital narrative, as part of broader initiatives to research folk theories of AI and emergent technologies.

AI-Artist (inspired by Sakana's AI-Scientist) by David Johnston, Department of Linguistic, Literary and Aesthetic Studies 

On August 13, 2024, Sakana released "the first comprehensive framework for fully automatic scientific discovery, AI Scientist". On August 17, 2024, I began to playfully build an autonomous AI-Artist capable of prompting itself to generate a conceptually relevant artwork: title, image, description, context. This talk will review the process, output, limitations, cliches, and strengths of these rapid experimental AI-Artist explorations, while considering the implications of non-prompting semi-autonomous generative AI that operate within a proposed ethical constraint of 'whole-AI' use.

Dhvāni: Sacred Sounds and Decolonial Machines by Budhaditya Chattopadhyay, Department of Contemporary Art 

This talk provides an entry into a decolonial approach to AI-driven music and sound arts by describing an ongoing artistic research project, Dhvāni. The project is a series of responsive, self-regulating, and autonomous installations driven by a custom-made artificial neural network and a machine learning model trained with ritual and sacred sounds from South Asia. With a decolonial approach, such technological mélange re-emphasizes and advocates for the values of interconnectivity, codependence, participatory networks, and community. By giving AI an autonomous agency, the project aims to reimagine the future of AI with inter-subjective reciprocity in human-machine assemblages, transcending the technologically deterministic approach to AI-driven live art, sound and media arts, and an expanded sense of improv music. By unpacking the project, this talk underscores the necessity of dehegemonizing the AI-driven arts field to a transcultural exchange, transcending the field’s Eurocentric biases.

Politics of Algorithmic Representation by David Rych, Department of Contemporary Art 

This presentation examines the shifting dynamics of visibility in the digital age, focusing on how the widespread use of image interpretation technologies can perpetuate prejudice, discrimination, and false identity attributions. The talk will explore the potential of artistic counter-narratives as strategies to resist algorithmic bias. The case study of the art project "Untitled (Authentic)," will highlight the critical role of human input in correcting distortions within AI systems and underscore the broader implications for power dynamics and political awareness in technological development. 

Moderator: Frans Jacobi, Faculty of Fine Art, Music and Design