The recording studio, long imagined as a sanctuary of pure human creativity and analogue warmth, is now a laboratory where artificial intelligence plays an increasingly active role in the composition, production, and distribution of music. AI-powered tools can analyse the vast catalogue of recorded music to generate chord progressions, suggest melodic variations, or even clone a vocal timbre so precisely that it becomes a new instrument in a producer’s palette. For artists in Aotearoa, from bedroom pop creators in Christchurch to hip-hop producers in South Auckland, these tools lower the barrier to achieving a polished, radio-ready sound without the prohibitive cost of hiring session musicians or renting expensive facilities. The technology is not a futuristic novelty; it is embedded in the digital audio workstations and plugins that are updated with each software iteration, quietly reshaping the workflow of an entire industry in real time.
Advertorial
The most immediate and practical application lies in the realm of audio mastering and stem separation. Tasks that once required a highly specialised engineer with decades of experience, such as cleaning up a noisy vocal track or giving a mix the loudness and clarity demanded by streaming platforms, can now be accomplished by an AI service in minutes. Stem separation tools, which can deconstruct a fully mixed track back into its constituent parts (drums, bass, vocals, and other instruments), have revolutionised remixing, sampling, and restoration work. This democratisation of high-end audio engineering is a double-edged sword: it empowers independent artists to compete sonically with major label releases, but it also disrupts the livelihood of mastering engineers and session players whose irreplaceable human nuance must now be argued for, not assumed. The value proposition shifts from technical execution to taste, curation, and the emotional intelligence that a machine, however well-trained, cannot originate.
Generative AI models that create full compositions from text prompts have sparked the most intense debate about authorship and originality. A producer can type “upbeat synthwave track with a melancholic saxophone solo in the style of the 1980s” and receive a royalty-free piece of music that serves as a temporary placeholder in a video edit or, controversially, as the final soundtrack for a commercial project. This capability raises profound questions about the nature of creativity when the line between inspiration and algorithmic recombination becomes blurred. Copyright systems, built on the premise of human authorship, are straining under the weight of legal challenges regarding whether an AI trained on copyrighted material produces derivative works. For composers of library music and production music, the existential threat is acute, as a cheaper, instant alternative floods the market. The counterargument is that these tools will push human composers towards more daring, less formulaic work that an AI, which is inherently backward-looking, cannot conceive.
