Revival
Revival Artwork: Collaborative Interaction Between Human and AI
The Revival artwork is a live audiovisual performance that exemplifies the dynamic interaction between human artists and AI agents in the creative process. It integrates real-time music improvisation with AI-generated audiovisuals, featuring a collaboration between a percussionist, an electronic music performer, and AI musical agents. The agents, including MASOM and SpireMuse, are trained on compositions from deceased composers and the collective's own works, allowing them to respond dynamically to human input and emulate complex musical styles. The AI-driven visual synthesizer, Autolume, interacts with the music through an AI-guided process, shaped by a human VJ, to create evolving visuals synchronized with the musical performance.
The Revival performance demonstrates the potential of AI to act as a co-creator in real-time, enhancing the human artistic experience. The AI agents don't simply provide accompaniment but engage in a reciprocal creative process, adapting to the performers' inputs, making decisions autonomously, and shaping the music and visuals as part of the artistic flow. This interaction challenges traditional notions of authorship and creativity, inviting the performers and the audience to experience AI as a dynamic, creative collaborator.
Research-Creation Methodology in Music and Artistic Practice
The research-creation methodology in the Revival project blends artistic practice and academic research, where the creative process informs and is informed by the technological developments. In this case, the focus is on the co-creation between human musicians and AI agents, with the aim of exploring new forms of improvisational music-making and audiovisual art. The methodology emphasizes several key principles:
Collaborative Co-Creation: Human musicians and AI systems interact in real-time to produce music and visuals. This dynamic collaboration is central to the methodology, as it allows for the exploration of new musical ideas and creative possibilities that neither the human nor the AI could achieve alone.
Iterative Exploration: The methodology allows for ongoing refinement and experimentation, where feedback loops from both musicians and the AI agents shape the outcomes of the performance. The AI's response to human inputs is an essential aspect, as it learns from and adapts to the evolving performance environment.
Ethical and Sustainable Practices: The project incorporates a "small data mindset," which involves using small, curated datasets for training the AI systems. This approach reduces the computational resources needed for AI training, making it more environmentally sustainable. It also ensures transparency in the use of data, helping avoid ethical issues like copyright infringement or the misuse of artists' work. Small datasets also allow for more personalized and artist-centered AI models.
Human-AI Synergy: A core aspect of the research-creation methodology is the recognition that AI can enhance human creativity. The AI agents do not replace the human performer but act as creative partners, amplifying the potential for novel musical and visual expression. The methodology encourages a reflective understanding of how AI alters the traditional roles of composers and performers, especially in the context of improvisation.
Interdisciplinary Collaboration: The integration of music, visual art, and technology highlights the interdisciplinary nature of the research-creation process. The project brings together musicians, VJs, AI researchers, and technologists to create an interactive, co-creative space, demonstrating the evolving role of AI in artistic practices.
In essence, the research-creation methodology in Revival allows for an in-depth exploration of human-AI collaboration, with the goal of pushing the boundaries of both musical and audiovisual art forms while ensuring ethical practices and sustainability.
Credits:
Artist Collective K-Phi-A
Keon Ju Maverick Lee: Drummer, Music System Design
Philippe Pasquier: Live Electronics, Director
VJ Amagi (Jun Yuri): VJing, Visual System Design
Related Publication:
Keon Ju Maverick Lee, Philippe Pasquier, and Jun Yuri. (December 2024), Revival: Collaborative Artistic Creation through Human-AI Interactions in Musical Creativity, NIPS Workshop on Creativity & Generative AI for Creative Work Track.
Artwork Paper link: http://arxiv.org/abs/2503.15498
