The 21st century can truly be called the era of artificial intelligence. AI is actively utilized not only in various specialized fields but also across all aspects of human daily life, and the field of music is also included in this trend. In particul...
The 21st century can truly be called the era of artificial intelligence. AI is actively utilized not only in various specialized fields but also across all aspects of human daily life, and the field of music is also included in this trend. In particular, it is noteworthy that AI is now actively applied not only in the appreciation and performance of music, but also in the creative process.
Today, with the emergence of various AI composition models such as Suno, Udio, and AIVA, AI music composition is rapidly evolving, and interest in algorithmic development and technological achievements is growing accordingly. However, compared to these technology-centered discussions, academic discourse that analyzes and examines AI-generated music from a musicological perspective still remains limited. This situation further emphasizes the necessity of attempts to regard AI music not merely as a technological product, but as a work of art and to explore its meaning and value.
From this perspective, David Cope’s ‘Emmy’(Experiments in Musical Intelligence), developed in 1981, deserves attention. Emmy is an AI composition model that generates new works based on data analyzing the musical style of a specific composer. While most AI composition models focus on short instrumental music or pattern-based music, Emmy has distinct significance in that it challenged the operatic genre, which requires large-scale orchestration and narrative structure.
Among Emmy’s works, this study focuses on the opera Mahler, which was composed based on the life and music of Gustav Mahler(1860–1911). This work was completed by having Emmy generate music based on Mahler’s songs and symphonies, while Cope wrote the libretto reconstructing Mahler’s life. To analyze this work, the study first organized the composition algorithm and musical features of Emmy, particularly the concepts of musical style and signature, and then analyzed the structure and development of each scene. Furthermore, the study comprehensively examined the quoting techniques and the interaction between music and narrative throughout the work, and from this, derived the aesthetic implications, limitations, and possibilities of AI composition.
As a result, the musical features of the work can be summarized in two aspects. First, Mahlerreproduces Mahler’s musical world through various quotations and similar references centered on his music, showing that AI can engage in creative activity beyond simple imitation. Emmy reinterprets traditional compositional styles actively by recombining existing musical data in new ways and actively quoting Mahler’s musical language. Second, through close interaction between libretto and music, emotion and narrative are organically constructed, demonstrating that collaboration between humans and AI can be realized even in narrative-centered art forms. These musical quotations and interactions between humans and AI were unprecedented in existing AI composition models, and can be seen as a key point demonstrating Emmy’s imaginative capacity.
In this context, one might ask the question: “Can this opera be considered Mahler’s opera?” In Mahler, Mahler does not merely serve as a subject of quotation or a source of data; rather, he functions as a central axis mediating the creative process of Emmy and Cope. That is, a new creative identity that cannot be reduced to a specific creator becomes clearly visible, which Eleanor Selfridge-Field(1940–) has described as a “synthetic identity.” Moreover, Emmy’s compositional approach reveals a dimension of “data unconscious,” as it does not simply follow explicit rules or algorithmic outcomes, but constructs potential associations based on Mahler’s vast musical data. From this perspective, Mahlercan be seen as a work that proposes a new type of creative identity in which not only Emmy and Cope but also Mahler are included as co-creators.
However, criticisms also exist. Mahler’s emotions, experiences, and historical context are unique human traits that AI cannot replicate, and AI creation is criticized for being unable to capture this existential depth. In other words, although Mahlertechnically imitates Mahler’s style with precision, it ultimately remains music that merely resembles Mahler. This connects to the argument that Emmy’s accurate analysis and reproduction of Mahler’s style, due to its excessive similarity, makes it difficult to be recognized as an independent creative work.
In response, Cope emphasizes that imitation has historically been an important method of musical creation, interpreting Emmy’s composition as “Recombinant Creativity.” This is evaluated as an active creative act that arranges and constructs existing musical elements in new ways. Mark Coeckelbergh(1975–) also argues that AI can be a subject of art creation by positively viewing the artistic potential of AI through the concept of “mimesis.” In this way, Mahlerpresents a new creative identity that intertwines AI, humans, and data from past composers, and holds significance as a case that reconsiders the boundaries of traditional music creation.In conclusion, this study highlights that AI composition, through the opera Mahler, can move beyond mere reproduction, transform and reconstruct the styles of existing composers, and create new aesthetic meanings, leading to multifaceted aesthetic discussions. This suggests that instead of applying existing aesthetic standards directly to AI-generated music, it is necessary to seek new aesthetic frameworks suitable to its particularity and creative context. AI music possesses unique modes of creation and context distinct from traditional aesthetics, and its evaluation requires new theoretical approaches.
AI prompts reflection on the subjectivity and boundaries of musical creation, bearing important aesthetic and historical implications. In an era where art and technology intersect, it is hoped that discussions that continuously track the development of AI music and explore its meanings from diverse perspectives will become more active.