Enabling a Paradigm Shift: Intersections of Creativity, Epistemology, and Licensing Authority (Conference Panel Title)

convened by Elaine Chan

Paper Abstract (1)

Distant Reading of Musical Corpora: Generative AI’s Role in Redefining Creative Structures

The emergence of generative AI tools such as Suno has fundamentally altered the way music is written and created. Illustrated by the AI generated band The Velvet Sundown topped the charts of Spotify in 2025. This situation is a direct challenge on the traditional model of humanistic music-making, making us reassess what constitutes artistic worth or novelty.

In light of this disruption, this study investigates the creative impact of AI by conducting a comparative analysis of tracks generated by Suno AI and a human-composed piece. Utilizing "distant reading"—a computational analysis of large-scale musical corpora—AI identifies and remixes prevalent patterns like chord progressions and song structures. However, this method raises a critical question posed by scholars such as Nick Collins: does composition using only pre-defined templates stifle or encourage creativity?

Our findings indicated that while AI excels at generating structurally coherent music by recombining learned patterns, it often lacks the nuanced originality of human composition. We therefore propose a hybrid model for creativity one that intersects AI’s data-infused scalability with human capabilities for “close listening” and “intuitive innovation”. We suggest a hybrid model. This paper argues that this synergy is not a replacement of the human artistry, but a conversion optimizer of the creative process. It gives evidence of such integration of data-driven insight and human imagination, and examines the new musical language created in the era of AI.

Author: Alex Tin-yung Au

Affiliation: Hong Kong Design Institute

Email: alexau@vtc.edu.hk

Bio: Au Tin-yung, Alex was born in Hong Kong. He is an active composer, educator, and promoter of contemporary music. His works have been widely performed in different places such as the United States, China, Ireland, Taiwan, Italy, France, Sweden and the U.K. Au’s works often explore the potential of the audible sonic spectrum by integrating computer technology with contemporary compositions. His work “Dao” has been selected as one of the three finalists in Duosolo Emerging Composer Competition in 2011 (Cortona, Italy).

Au received his Ph.D. in music composition at the University of Hong Kong with support from the University Postgraduate Studentship. His Ph.D. dissertation focuses on audio signal analysis and processing in music composition. This research involves the exploration of digital audio processing, sound synthesis, and spectral analysis of sounds in music composition. Au also received his Master’s Degree (Distinction) in Composition at the University of York. In 2017, he studied computer music at IRCAM under the guidance of Mikhail Malt. He was a council member of Hong Kong Composers’ Guid (HKCG). Au is currently a lecturer at Hong Kong Design Institute (Department of Digital Music and Media).

Paper Abstract (2)

Rethinking the Epistemology of AI in Filmmaking through the ‘Kino-eye’

Since AI-powered media nowadays create images and moving pictures from human language prompts, the technical creative environment for text-to-image generation is based on the publicly available images on the internet. AI's ways to use pre-existing visual data and synthesize new content from learned statistical patterns problematize our understanding of film ontology. AI-generated videos often face challenges in controlling consistently the movements and actions of depicted personalities, especially over longer durations. The output videos frequently deviate from the intended contextual references, leading to inconsistencies in behavior, appearance, and scene continuity. Therefore, it appears that Orson Welles’ film ontology, which conveys a sense of real-time progression through unobtrusive long takes, would no longer sustain as a creative practice in the digital age. Similarly, it seems that the neo-realist filmmaking approach, grounded in subtle portrayals of authentic human experience, can no longer be materialized through language prompts. Epistemologically, the knowledge framework of AI that compiles images from the massive and diverse dataset may probabilistically manipulate images. Many scholars like Charlotte Bird, Eddie L. Ungless, Atoosa Kasirzadeh, Leonardo Nicoletti & Dina Bass are critical of the ethics, authenticity and truthfulness of text-to-image generation. This paper offers an empirical analysis and critical review of the rapidly evolving ontology of filmmaking in the artificial intelligence era, arguing that while the AI-driven video revolution marks a regression, it simultaneously nurtures the potential for enriched machine-human co-creativity. This paper compares and contrasts the visual contents of a contemporary AI generated award-winning short film and Dziga Vertov's ‘kino-eye’ production to rethink the epistemology of AI in filmmaking.

Author: Kim-mui E. Elaine Chan

Affiliation: Hong Kong Chu Hai College

E-mail: kmchan@chuhai.edu.hk

Bio: Kim-mui E. Elaine Chan, assistant professor of Journalism and Communication Department at Hong Kong Chu Hai College, currently teaches communication and media studies. She received her Ph.D degree in film studies from University of Kent, United Kingdom. She taught film theory, film aesthetics, film history, and digital media culture in the MFA and MA degree programmes at Hong Kong Baptist University. Her monograph entitled Hong Kong Dark Cinema: Film Noir, Reconceptions, and Reflexivity was published by Palgrave Macmillan in 2019. The Chinese-language edition of this monograph received in 2019 the 15th Hong Kong Biennial Awards for Chinese Literature, Recommended Prize in the Category of Literary Criticism. Her academic journals and other journal articles appear in Journal of Chinese Cinemas, International Journal of Cinema, Hong Kong Economic Journal Monthly, Ming Pao, Hong Kong Economic Times, Crossover, and Film Bi-weekly. Her columns are published in Tai Kung Pao and Master Insight.

Paper Abstract (3)

Toward a Posthuman Aesthetics: Deconstructing Authorship and Ideology in AI Art

Jason M Allen's AI-generated artwork won the Colorado State Fair fine arts competition yet be deprived of copyright. It seems symptomatic of a profound crisis. It shows us how our aesthetic and ethical systems, being rooted in a romantic and human centered concept of authorship since the Renaissance, are seriously challenged by new technologies.

This paper posits that the transition of AI from a procedural tool (ANI) to a generative cultural agent (AGI) constitutes a categorical shift, moving beyond assisting human expression to actively constituting a new mode of cultural production. It represents an entirely new form of cultural practice, not merely a technological advancement. As philosopher Alice Helliwell from Northeastern University London argues, “if we can consider radical and divergent pieces like Duchamp's urinal as art proper, how can something created by a generative algorithm be dismissed?” The critical question is no longer “Is it art?” but how to understand this new paradigm.

As poststructuralism's "death of the author" (Barthes, 1967) reveals our systems' outdated romanticism, while Critical Algorithm Studies (Noble, 2018) shows AI art is a socio-technical artifact that replicates its training data's biases. The paper makes two central claims.

First, it proposes replacing the myth of the lone artistic genius with a distributed model of agency that includes the programmer, the algorithm, the dataset, and the user. Second, and more crucially, it contends that AI art is a socio-technical construct, one that can disguise and perpetuate the biases and ideologies embedded in its training data under a false veil of neutrality. Uncovering this is an essential ethical task.

By reconceiving AI as an agent of culture, this paper brings forward a critical and original framework for understanding how we will produce, control and disseminate art in an increasingly machine-co-authored age.

Author: Roger Lin

Affiliation: Hong Kong Chu Hai College

Email: rogerlin@chuhai.edu.hk

Bio: Roger Lin is an educator, researcher, and creative practitioner with over three decades of experience spanning arts management, digital media, and sound production. He holds a Doctorate in Professional Studies (Arts Management and Technology) from Middlesex University, UK, and has taught at institutions including Hong Kong Chu Hai College, Hong Kong Design Institute, and Saint Francis University, where he has led courses on cultural production, digital transformation, and creative entrepreneurship.

As a practicing artist, Dr. Lin has contributed to more than two hundred film, television, and theatre productions as a producer, director, screenwriter, composer, and sound designer. His work has been featured at international festivals such as the Berlin Film Festival, Busan International Film Festival, and the Hong Kong Arts Festival. He was also the producer, director, screenwriter, and composer for the VIU TV drama If Love Was Not Timeless.

In the visual arts, Dr. Lin has curated over 60 exhibitions and published art criticism in various media. His research focuses on modernizing traditional performing arts through contemporary sound design and exploring cross-cultural narratives in popular media. Recent publications include Ultraman v Superman: The Metaphors in American and Japanese Pop Culture (2024) and his doctoral thesis, Modernising Cantonese Opera through Contemporary Sound Production Design (2020).