Artificial Intelligence (AI) has been on the rise for years, but is recently experiencing exponential growth with the launch of tools accessible to all such as Chat GPT, DALL-E or Soundraw. These tools are quickly evolving thanks to their deep learning capabilities – and this is only the beginning. The rapid development of AI is forcing industries to reconsider their modus operandi and tackle complex legal and ethical questions regarding intellectual property and ownership, some of which are currently in a grey area. In the music industry more specifically, copyrights are a big point of contention. Last month, a viral video made by TikTok content creator ‘Ghostwriter’ used AI technology to create vocals sounding like Drake and the Weeknd. The artists had not consented to their vocals being reproduced, and the song “Heart on my Sleeve” garnered millions of streams before Universal Music Group requested for online platforms to take it down as it was infringing copyrights.
Interestingly, recent news are prompting us to consider another scenario. On April 30, 2023, singer-songwriter Grimes released GrimesAI-1 on Elf Tech, an open-source software enabling anyone to upload their music and transform their vocals into Grimes’. The initiative came directly from the artist as she encouraged her 1.3M followers on Twitter to use the technology. This is not only an innovative proposition, but also offers new ways of interacting with fans and building a community. Only a few guidelines are in place: no use of her voice for “rly rly toxic lyrics;” a 50/50 split of royalties with anyone creating a successful song that uses her voice, in a similar fashion to artist collaborations; and her approval is required prior to distributing the music independently or through a label. Finally, for uploading the tracks on Digital Service Providers (DSPs) such as Spotify, there is a fee of $9.99 per year. An easy-to-use tool available to everyone, pretty much for free. You can find the more information here.
Grimes is no stranger to AI music creation, having launched her “AI Lullaby” project with Endel in 2020. She’s taking her advocacy for AI to the extreme, wanting to “open-source all art and kill copyrights.” Copyrights are here for a reason – to protect artists, their work and ensure that they are remunerated accordingly – although the rapid growth of AI means that these laws may need to evolve in the near future to strike a balance between protecting artists’ rights and encouraging creativity with AI, as it’s here to stay. I would add that these potential adjustments may differ from one country to another and may be implemented at different times, despite a global baseline agreement (eg. the Berne Convention). This makes for a tricky transition with the fast pace of technological evolution and the need to address those emerging challenges.
According to Grimes’ manager, more than 15,000 users are already using GrimesAI-1. While we can expect a higher volume of music to be created with Grimes-like vocals, the technology has the potential to be a tool for creativity rather than one that would saturate the market. It will depend on various factors, such as how many artists decide to follow a similar path, how good AI-generated music gets over time and how we pick music that we enjoy listening to. Ultimately, music preferences are deeply subjective – some people need the human touch of personal feelings and memories weaved into a piece of music, while others might find meaning in AI-music. I personally lean towards the former, and a combination of both approaches (eg. using AI to aid the creative process) may also result in an emotional and meaningful song.
In conclusion, GrimesAI-1 is an innovative example of how to use AI to encourage creativity. I am now waiting to see how it works in practice, and while it is difficult to predict how the music industry will continue to evolve, it is clear that AI will keep on playing an increasingly important role in the way we create and consume music.
9 to 9 