Advertisements

AI-Assisted Compositions to be Featured at China Now Music Festival

by Madonna

The 7th annual China Now Music Festival, titled “Composing The Future,” will take place from October 12 to 19, 2024, at Carnegie Hall.

In line with the future-focused theme of this year’s festival, the event organizers have invited the Central Conservatory of Music’s Department of Music Artificial Intelligence in China to contribute an ‘AI Suite’ to the concert program. This suite will be partially composed by AI, aiming to foster an innovative dialogue between the composer, the orchestra, and the interplay of human creativity with machine assistance. Additionally, the festival seeks works that incorporate AI technology in live performances.

Advertisements

Audiences can look forward to hearing Li Xiaobing’s AI Suite, which features a “Cloud Chorus” comprising 1,000 voices gathered from around the globe, alongside a piece by Sun Yuming, where a traditional guzheng zither is played on stage without direct human interaction.

Advertisements

The festival offers insights into this new approach to composition through a conversation with composer Li Xiaobing, who serves as a Professor and Director of the Department of Music Artificial Intelligence and Music Information Technology at the Central Conservatory of Music in China.

Advertisements

Li Xiaobing explained that three key elements are vital in AI composition: data, algorithms, and computing power. The process begins with data, likening it to a warehouse filled with diverse musical styles—classical, pop, rock, and more. This “big data” is used to train the AI. Initially, the AI listens to various music forms, learning melodies, rhythms, and styles.

Following this, the AI translates the music into a “symbolic language,” transforming the audio data into a “symbolic database” that is tailored for AI models. By drawing from this database, the AI learns the underlying logic and structure of music.

Regarding algorithms, Li stated that the AI employs a sophisticated music model. The learning process involves predicting the next “symbol,” akin to how a writer predicts the next word in a sentence. In this context, the AI predicts the next note or melody.

The automatic composition model allows machines to learn and create music similarly to human composers. While the process may appear complex, it fundamentally revolves around learning and prediction. The AI can efficiently uncover deep patterns in notes, melodies, and harmonies through self-supervised learning, enabling it to produce entirely new artistic works, similar to human creativity. Thus, AI composition can be viewed as a process of learning from data to generate new music.

When asked about the final outcomes of AI compositions, Li noted that different models yield different types of works. For instance, a song composition model produces songs, while a twelve-tone composition system creates modern music, and an orchestral model generates works for orchestras.

On how to guide AI in achieving specific outcomes, Li explained that composers can input lyrics, styles, instruments, or themes, allowing the AI to compute and generate new pieces of music based on these inputs.

Addressing the concern of AI replacing human composers, Li remarked that from a humanistic perspective, this idea could be alarming. With the rapid advancement of AI globally, there is a possibility of job displacement across various fields, not just music. However, he expressed confidence that historical technological revolutions, while disruptive, ultimately lead to progress. He emphasized the importance of being prepared for the future.

Li believes that AI will not replace top composers in the near future; rather, it will assist them in achieving their goals efficiently, particularly for foundational tasks. He encourages composers to embrace AI as a collaborative tool. He outlined three potential directions for the future of music:

Traditional music (including modern music) will continue to evolve.
AI will facilitate innovation within traditional music.
AI will spawn entirely new forms of music that will develop independently.
Li concluded that AI in music is intended to support and empower humanity, not to replace it. He hopes that composers will collaborate to explore the future of music enriched by AI.

In addition to Li’s work, the concert will also feature a piece by Sun Yuming, a Lecturer on Electronic Music Composition at the Central Conservatory of Music, China. Sun’s composition involves a traditional guzheng zither being played on stage without direct human touch. He collaborated with AI engineer Zhang Xinran from the Department of Music Artificial Intelligence and Music Information Technology to achieve this innovative approach.

Sun explained that his concerto for guzheng and orchestra, titled “Starry Night,” aims to transcend the instrument’s limitations by using two guzhengs during the performance and incorporating the movement of the bridge. However, to vary the playing technique of the guzheng, he sought the expertise of an AI engineer.

Zhang Xinran elaborated on their process, stating that they designed a virtual instrument using computer vision technology. This system captures specific movements through optical motion capture, triggering certain musical materials, thereby providing composers and performers with a broader range of performance options.

During the performance, Sun noted that the guzheng player engages with two instruments—one tuned to a pentatonic scale and the other to a heptatonic scale. Traditional sounds emerge from the instruments, while non-traditional sounds, enhanced by electronic elements, are activated through AI motion capture technology.

Zhang further clarified that during the live performance, the player’s hand movements are tracked by a computer to trigger musical materials. Different trigger logics are established for various sections of the music, enabling real-time interaction between the performer and the system.

When asked about the potential for AI to replace composers, Sun emphasized that AI is an inevitable outcome of technological progress. Rather than fearing its integration, he advocates embracing it as a tool to accelerate human civilization’s advancement. In music, AI should initially serve to inspire composers with new ideas and possibilities, functioning as a partner in creativity. As AI technology continues to evolve, new creative methods, works, and aesthetic standards will emerge. For composers, it will be crucial to adapt to these changes by enhancing their skills, sensibilities, and creativity in conjunction with AI.

On October 22, a pre-concert event at Carnegie Hall will feature a panel of composers and music researchers at the second annual US-China Music Forum. They will explore the intersection of technology and music in contemporary composition.

Related Topics

Advertisements

You may also like

blank

Musicalinstrumentworld is a musical instrument portal. The main columns include piano, guitar, ukulele, saxphone, flute, xylophone, oboe, trumpet, trombone, drum, clarinet, violin, etc.

【Contact us: [email protected]

Copyright © 2023 musicalinstrumentworld.com