The Meteoric Rise of Generative AI Music

(Image generated by author with BlueWillow)

In our previous post, we explored the transformative power of generative AI in the art industry. We saw how AI can be used to create stunning and innovative artworks. In this post, we will shift our focus to the music industry. With more than 14 million songs created (14% of the world’s music recordings) by just one platform under three years, generative AI music is taking off in a big way. What does this mean for the future of music? This blog post aims to explore the creative potential of AI in music while also addressing some of its challenges as well as the ethical and legal considerations it raises.

Table of Content

A brief History of Machine Learning in Music Generation

Over the years, computer-generated music has become increasingly sophisticated, relying on advances in machine learning to generate complex and expressive compositions. Notable milestones in the history of machine learning in music generation include:

1957: Illiac Suite (String Quartet No The. 4), a pioneering computer-generated piece of music crafted by Lejaren Hiller and Leonard Isaacson.

1980s: David Cope’s ‘Experiments in Musical Intelligence’ (EMI) program, designed to aid productivity, analyze individual composers’ styles, and generate new pieces in their respective styles.

2016: Google DeepMind’s WaveNet system, employing a convolutional neural network (CNN) to comprehend sound wave patterns, operating directly on raw audio waveforms.

2019: OpenAI’s MuseNet, a deep neural network utilizing a large-scale transformer model to discern harmony, rhythm, and style patterns by learning from vast repositories of MIDI files.

The Basic Structure of the Music Industry

To understand the implications of AI-generated music, we must grasp the multifaceted and interconnected structure of the music industry. It can be categorized into three primary segments:

Recording: This segment encompasses all aspects related to the creation, production, and distribution of recorded music. It involves entities such as record labels, recording studios, distributors, and artists.

Publishing: The music publishing industry deals with safeguarding, managing, and monetizing the intellectual property of songwriters and composers. This includes activities like royalty collection, copyright administration, and licensing music for various contexts.

Live Performance and Touring: This segment of the industry is responsible for the production of live performances, involving concert promoters, tour managers, and booking agents.

How Does AI-Generated Music Work in a Nutshell?

AI-generated music involves the application of artificial intelligence algorithms to create music. The process typically entails training an AI model on extensive datasets of existing music, which exposes it to diverse musical patterns, melodies, and styles. Through this exposure, the model learns and develops an understanding of musical components, including chord progressions, melodic themes, rhythmic structures, and stylistic features specific to different music genres.

Once trained, the model can generate melodies, harmonies, rhythms, and even entire compositions by applying its acquired knowledge. Users can guide the AI by adjusting parameters or inputs, such as desired mood, tempo, or genre, to produce music that aligns with specific criteria. AI-generated music can be used in a variety of ways, including as background music for movies or video games, as personalized music experiences, and as a source of inspiration for musicians and composers exploring new musical ideas.

What Are The Benefits of AI-Generated Music?

AI-powered music generation is changing the music industry in many ways, giving musicians, listeners, and industry professionals new possibilities. Here are some examples of how it is driving this change:

Recording:

Reducing expenses by eliminating the need to hire musicians, rent studios, or invest in expensive equipment.
Streamlining music production processes by automating tasks such as generating chord progressions, melodies, and lyrics. This can save musicians a significant amount of time and free them up to focus on more creative aspects of their work.
Creating music in different genres, with different tempos and styles, ensuring that there is something for everyone in the music world.
Helping artists overcome creative blocks by offering fresh ideas and inspiration.

Music Publishing:

Creating a vast pool of music for use in film, television, and advertising. This makes it easier for filmmakers and advertisers to find the perfect music for their projects, and it can also lead to increased revenue for music publishers.
Facilitating a more decentralized model of music publishing, which means that independent artists can now connect directly with audiences without having to go through traditional industry gatekeepers. This can be a great way for independent artists to get their music heard, and it can also help to level the playing field in the music industry.

Live Performance and Touring:

Enabling real-time customization of live performances, making each performance more unique and captivating.
Fostering audience engagement by allowing artists to create interactive experiences that allow the audience to actively participate in the performance.
Reducing costs for live performances by facilitating the creation of backing tracks, which eliminates the need for additional musicians.

What Concerns Does AI-Generated Music Raise?

While AI music generation offers significant benefits, there are also some potential concerns that have been raised by stakeholders. These concerns include:

Recording:

Lack of creativity, originality, and spontaneity. Some worry that AI-generated music may lack the creativity, originality, and spontaneity associated with human-made music. This could lead to a decline in overall diversity and a more standardized musical landscape.
Market saturation. The popularity of AI-generated music could oversaturate the market, making it challenging for listeners to discover new music they genuinely enjoy.
Artificial or inauthentic feel. AI-generated music might feel artificial or inauthentic to some listeners due to its algorithmic imprint.
Job losses. There are also concerns about potential job losses for musicians, composers and other music industry professionals.

Music Publishing:

Limited opportunities for songwriters and publishers. If AI-generated music becomes popular there could be less demand for human songwriters. This could lead to lower royalties for both songwriters and publishers.
Increased competition: AI-generated music could create new competition for human-made music. This could lead to lower prices for music licenses and royalties, which could hurt the bottom line for music publishers.
Copyright ownership and legal disputes related to AI-generated music can be complex, as unauthorized music datasets may have been used to train AI algorithms. This can lead to disputes over who owns the copyright to AI-generated music, as well as disputes over whether AI-generated music infringes on the copyrights of other artists.

Live Performance and Touring:

Impact on demand for live performances. The accessibility and personalization of AI-generated music may impact the demand for live performances, potentially affecting ticket sales and revenue streams for live performers.
Altered experience. AI-generated music in live shows could change the experience. It could become more automated and less human-centered, which some people who value the human element of live music might not like.
The threat of deepfakes. Generative AI technology could be used to create deepfakes of virtual live performances, which could mislead fans and damage the reputation of artists and the Live performance segment as a whole.

Looking Ahead

Generative AI music is becoming more sophisticated, unlocking new possibilities for creating music. However, there are still some challenges that need to be addressed to ensure it does not wreck havoc in the music industry. As AI-generated music advances in sophistication, it’s important to establish strong legal and ethical frameworks to make sure it’s used responsibly and ethically. This is important for everyone involved in the music industry and for society as a whole.

In closing, I’d like to share a thought that beautifully captures the delicate interplay between human creativity and AI :

“Machine Learning works best as a tool to help elevate the output of a human creative mind, not to replace it.” – Anna Chaney, former Data Science Architect at IBM

Generative AI Music Platforms

Here is a list of AI-generated music platforms:

Amper Music, Amper Music is a platform that allows users to create AI-generated
Aimi, is a platform that offers users a variety of features to create and interact with AI-generated music in a variety of ways.
Soundraw, is an AI music generator that allows users to create and compose original music that they can use in their projects.
Boomy, is a generative AI music platform that claims its users can create original songs in seconds, even if they have never made music before.

Additional Resources

Here are some additional resources to keep you updated on the latest developments in generative music and explore the challenges and opportunities that this innovative technology presents.

Spotify Blocks Boomy Music Platform Releases, Spots Sungs, July 17, 2023
AI music wars: Meta takes on Google and releases its own AI music generator – but whose is better?, Matt Mullen, musicradar, June 16, 2023
Beatles get back to making new music thanks to assist from AI, AP, June 14, 2023
Making Intelligent Noise: AI In The Music Industry, Aaron Labbé, Forbes, May 9, 2023

AI Talks