TLDR: A study examined four generative AI music systems (AIVA, Stable Audio, Suno, Udio) to uncover the ideologies driving their their development and adoption, particularly the concept of ‘democratizing music making.’ It found a shared ‘total ideology’ between producers and consumers that is individualist, globalist, techno-liberal, and ethically evasive, often presenting AI as an inevitable, benevolent tool that unlocks dormant creativity, while potentially obscuring commercial interests and the true nature of musical musical practice.
Generative Artificial Intelligence (AI) systems for music creation have become increasingly accessible, allowing individuals without formal musical training to produce music. This rise has been widely promoted as a means of ‘democratizing’ music making, suggesting that anyone can now express their musical ideas. However, a recent research paper delves deeper into these claims, investigating the underlying ideologies that drive the development and adoption of these systems.
The paper, titled Opening Musical Creativity? Embedded Ideologies in Generative–AI Music Systems, was authored by Liam Pram and Fabio Morreale from The University of Auckland. They critically examine how four prominent generative AI music systems – AIVA, Stable Audio, Suno, and Udio – are presented by their developers and received by their users.
What is Generative AI Music?
Generative AI music refers to computational techniques that can create new and meaningful musical content from training data. Historically, methods like musical dice games or chance operations were used to go beyond human limitations in composition. Modern AI systems extend this, producing music that can be indistinguishable from human-made works. These systems can act as passive assistants, co-creative partners, or even autonomous creators. While AI offers diverse uses in music, public engagement often focuses on text-to-music generators that produce complete audio outcomes from simple text prompts.
Despite the excitement, concerns exist about authenticity and intentional agency in AI-generated content, with some studies showing that listeners may unfavorably critique music perceived to be AI-created. The paper suggests that new critical tools are needed to understand how musicians internalize the knowledge frameworks embedded in these technologies, which can profoundly influence musical development.
The Promise of Democratization
The term ‘democratization’ in technology often means making something accessible to everyone. In music, this translates to affordable and sophisticated tools enabling non-experts to create. However, the paper argues that this focus on access might overlook users’ specific desires and that large-scale AI systems often carry a Western bias. Critics suggest that this ‘democratization’ can be a marketing strategy rather than a genuine principle, potentially leading to a ‘deskilling’ of musical tasks and operating in opaque, extractive ways.
Unpacking the Ideologies
The research frames its analysis through the concept of ideology, defined as a comprehensive belief system that links beliefs to practices. Many ideologies driving AI development are rooted in neoliberalism and techno-liberalism. Neoliberalism emphasizes unregulated private competition and individualization, seeing AI as a tool for individual requests and global consumption. Techno-liberalism, a sub-paradigm, places strong faith in technology to solve problems and preserve liberty, often deploying a ‘techno-solutionist’ logic where issues are framed to justify technological development.
How the Study Was Conducted
The researchers used a mixed-method approach. First, an autoethnography involved their own hands-on experience with the four selected AI music systems. Second, a digital ethnography collected data from the companies themselves, including public statements and interviews with key personnel. Third, another digital ethnography focused on the online user communities, primarily on Discord, to understand user experiences and discussions. The data was then analyzed thematically to identify prevalent themes and ideological positions.
What Companies Say
Company representatives consistently promote their systems as democratizing forces, allowing anyone to express themselves musically without needing traditional musicianship. They suggest that creativity is a dormant human trait that AI can ‘amplify’ or help ‘rediscover.’ They also argue that AI will blur the lines between creation and consumption, leading to personalized music experiences. Furthermore, a techno-liberal perspective is evident, with calls for deregulation and framing AI development as an inevitable, unthreatening historical progression, often drawing parallels to past technological advancements in music.
What Users Experience and Believe
Users in the online communities largely embrace AI as a democratizing force, helping them overcome limitations like lack of resources or musical knowledge. Many feel it unlocks a ‘dormant musical ability.’ They often view AI systems as tools or assistants that augment human involvement, integrating them into existing workflows to expedite practices. Users also express belief that AI will innovate the musical landscape, leading to new genres and creative possibilities. Similar to companies, users often see AI as an inevitable part of music’s future, a natural progression from older technologies, and a new kind of instrument.
The Ideological Implications
The study reveals a shared ideology that positions accessible creative independence as an improvement to traditional music-making. This aligns with Michel Foucault’s concept of ‘governmentality,’ where the pursuit of liberty can inadvertently serve processes of control, especially when access is prioritized over depth of engagement. The ideology is distinctly individualist, focusing on personal expression, yet operates within a globalist framework, appealing to a universal ‘humanity’ to obscure specific cultural contexts or profit motives. The framing of AI as an ‘assistant’ or ‘tool’ also encourages users to identify with the technology, complicating critical reflection.
This shared belief system functions as a ‘total ideology,’ presenting AI as an inevitable, unified global process. By framing AI as the natural outcome of technological development, companies legitimize their actions and deflect criticism. This teleological view suggests a movement towards human emancipation, where AI ‘amplifies’ human creativity and requires a ‘new man’ to adapt to this imminent socioeconomic epoch. The anthropomorphization of AI, referring to it with human-like qualities, further naturalizes the technology and limits perceived threats.
Techno-liberalism and Dormant Creativity
The prevailing techno-liberal ideology places faith in technology as the central means of liberal development, prioritizing individual needs and market competition. It suggests that by automating undesirable labor, AI ‘liberates human potential,’ often making unappealing consequences invisible. The idea that musical creativity is a dormant ability, waiting to be ‘made easy’ by technology, is a key part of this discourse. The reliance on language (text prompts) in these systems also shapes how users think about composition, often leading to a focus on genre hybridity as a form of innovation.
Also Read:
- What Users Really Think: Ethical AI’s Surprising Link to Satisfaction
- Advancing AI’s Understanding of Sound Beyond Simple Recognition
Conclusion
The research concludes that the rhetoric used by generative AI music companies significantly influences user experience and fosters a shared ideology across the field. This ideology is individualist, globalist, techno-liberal, and ethically evasive. It leverages the concept of ‘democratization’ as an indirect market force, creating a ‘false consciousness’ that keeps the technology’s value vague but appealing, thereby evading scrutiny. While the study highlights the potential for users to reclaim agency within these systems, it also raises important questions about the true location of creativity and agency in AI-assisted music making.
