How AI-music production tools can support the professional user without taking over the process completely

Sammanfattning

Introduction AIG-music production tools such as Suno can now generate full songs based on a short text-prompt. This lowers the entry level to the music industry where almost anyone can create music in this manner without any previous musical skills. However, this raises several questions from a professional artist’s perspective. With the low creative control in the AIG-music production tools artists feel alienated from their work. A possible solution is to integrate tools such as Suno with the artists current music production tools. This could result in greater efficiency as well as usability if done in the right way. Research Question How usable are AIG-music production tools in assisting artists with a more complex task like producing songs similar to their signature styles? Method We adopt a qualitative method with a thematic analysis to explore the effects of AIG-music production tools in a professional context. The tools that will be used in this study are Suno, Udio, Stable Audio and Lyria 3.0. To understand the effects of the tools we will use task-based interaction sessions to capture the interactions in real-time of five professional artists. This will be combined with semi-structured interviews with the purpose of providing deeper understanding for the tools’ actual usability in a professional context. Results We find that the AIG-music production tools have both positive and negative effects from a professional standpoint. The study suggests that the ease of usability does not necessarily equate to high-value outputs. It is clear that professional artists take a strong stand against AI being involved in music production due to the minimized musical skills required in the creative process. Discussion Our results are consistent with the previous research that AI-tools, in this part of the technological era, are only good enough to perform simpler tasks. Tasks that require fine-tuning of parameters, such as music production, are not at this time benefited with AI. This suggests that AIG-music production tools have to be further developed to better match the expectations and necessities of professional artists.