Music, humans, and machines: initial reflections for the development of research with collaboration between composers and artificial intelligence in the creative process of Brazilian music

Resumen

In the 1950s, Americans Hiller and Isaacson pioneered computer-generated music: “Iliac Suit”. Despite advancements in artificial intelligence (AI) systems, current music generation through machines still employs the paradigm established by Hiller and Isaacson (STEELS, 2021). Concurrently, emergent research on human-machine co-creation is reshaping the creative industries, enabling computers to contribute to music, art, and cultural production in ways that were previously unimaginable. Computers now create (make) music, art, and culture with potential for consumption (COMITÊ GESTOR DA INTERNET NO BRASIL, 2022). These transformations may alter music in epistemological and even ontological terms, restructuring the role of the composer. Computer science researchers and those from other technology fields have sought foundations in the humanities, particularly in art-based research, to underpin their studies (CARAMIAUX; DONNARUMMA, 2021). In this regard, there is an urgent need for research stemming from the academic field of music to establish a balanced dialogue with the field of computer science and other technologies. It is worth noting that in a context where music generation and AI projects are funded by professional software-producing companies, with economic investment driven by “usability” (RUTZ, 2021), it is arduous for the music field to conduct practical research on collaborative music generation between humans and machines since most current systems are not available for free experimentation. Therefore, this work aims to discuss the possibilities and challenges faced by researchers in the music field to conduct practical research on human-machine collaboration for music generation. This discussion has proven to be crucial from the challenges found in investigating whether collaborations between composers and AI music generation systems can preserve Brazilian cultural elements in musical outputs. Moreover it is vital as it addresses both the methodological barriers and the broader implications of integrating AI with cultural and creative expressions. The research aims not only to assess the feasibility of such collaborations but also to explore their potential to expand creative and cultural boundaries.