- AI has shaken the music industry with weird and wonderful results.
- Some are supportive and optimistic, but work must be done to protect artists.
- Copyright debates remain largely unresolved, exposing artists and producers to legal risks.
2023 will go down as the year where generative AI seized the headlines.
Language models like ChatGPT became household names, leading to a cascade of tools designed to handle all forms of media – including audio.
Stability AI’s “Stability Audio,” Meta’s “AudioCraft,” and Google DeepMind’s “Lyria” are three examples of advanced AI music generation tools.
Stability Audio produces high-quality audio from text prompts, aiding musicians in creating unique samples. AudioCraft creates varied audio and music from text, and Lyria excels in generating full music pieces, including vocals, as part of initiatives like YouTube’s Dream Track.
These technologies and others have supported AI-automated and AI-assisted music production. While AI has been a part of audio production for years, the crucial distinction with generative AI is that it challenges the nature of musical creativity. Just like you can ask ChatGPT to write a novel, you’ll soon be able to ask AI tools to produce near-complete tracks with minimal input.
Earlier in the year, the AI-generated song ‘Heart On My Sleeve’ ignited debate surrounding AI’s role in the music industry and its impact on artists. This track, created by an anonymous artist, Ghostwriter, using AI, mimicked the voices of Drake and The Weeknd.
The song’s instrumental parts, including the piano and drums, were human-created, with the AI’s role primarily in modifying the voices. It rapidly went viral, picking up millions of likes and views across Spotify, YouTube, and TikTok. It was later removed from streaming services due to copyright concerns.
Overall, the song was a huge hit, and people loved it. It showed AI’s ability to completely overhaul the creative process, not to mention the politics and negotiations required to create such a collaboration. Of course, Drake and The Weeknd had no say in the track and would have come across it by chance, the same as anyone else.
Commenting on an unofficial upload of the song, one commenter says, “Now Drake and The Weeknd know how it feels when a Drake and The Weeknd song drops,” and another, “New album came out and this is still the best Drake song to come out this year.”
In a similar case, Bad Bunny firmly dismissed speculations of a collaboration with Justin Bieber. In a TIME interview, he stated, “That’s fake,” emphasizing the unpredictability of his musical endeavors. However, a twist occurred when a TikTok song, seemingly featuring Bad Bunny and Bieber’s vocals, gained massive popularity.
The catch? The song was an AI creation by FlowGPT, incorporating the voices of Bad Bunny, Bieber, and Daddy Yankee. Bad Bunny’s response was clear and blunt – he denounced the song as a “shit of a song” and discouraged his fans from engaging with it.
Despite its removal from TikTok, the song had already garnered millions of views.
AI artists secure record deals
2023 saw AI-generated pop stars, including the animated artist Noonoouri, rise to international stardom and secure record deals.
Noonoouri, a creation by German designer Joerg Zuber, described as a “blue-haired, doll-faced virtual influencer,” clinched a record deal with Warner Music. The AI tools used to craft her vocals have led to her being labeled an “AI popstar,” though, in response to criticisms, Zuber said she was better described as “CGI.”
Reflecting on her creation, Zuber said, “We started with the hair, the drawings of the fabric, the movement and everything, and I slipped into a motion capture suit to walk as her… to define her movements, her gestures and everything.” In fairness to Zuber and Noonoouri, a huge amount of work goes into the project.
Noonoouri represents yet another progression in AI music, but the ramifications differ from ‘Heart on My Sleeve.’
Here, rather than a producer using someone else’s identity to platform their own tune, the project is original by ‘inhuman.’ This isn’t unprecedented, though – Gorillaz is a prime example of how you can artfully blend animations and authentic music.
However, in the future, AI might automate the entire process of creating music, leaving little to the human imagination.
There is an economic advantage to AI artists, too, as they’re exceptionally cheap and easy to manage.
Marec Lerche, head of business development at Warner Music Central Europe, described exactly that, stating, “You can appear in different places at the same time, you can change her style in a minute – we can make Noonoouri fly if we want, because it feels natural to her.”
There’s no overblown technical rider to meet, no trashed hotel rooms, and few overhead costs associated with producing and publishing the music.
A polarized debate
Some see AI as democratizing music creation, breaking down barriers for aspiring artists.
Others, however, view the co-opting of personal elements like a voice or style for commercial gain as deeply problematic.
Within that, many are questioning the boundary at which music stops being authentically creative.
Lex Dromgoole, a musician and AI technologist reflects on this, asking, “If there’s a huge explosion in music created at infinite scale and infinite speed, will that return us to thinking about what we are actually bringing to the table as humans? Where does imagination exist in this? How do we bring character to our own creations?”
On the other hand, Musicologist Dr. Shara Rambarran highlighted how AI poses an economic risk to real-life musicians “who want to put their music out there and not have to have that [additional] competition.”
The ramifications of AI in music are also attracting legal scrutiny. Universal Music Group and other major labels have initiated lawsuits against AI models replicating copyrighted lyrics.
In terms of copyright law, the primary issue revolves around the definition and rights of “authors.” According to current guidelines by the U.S. Copyright Office, only works created by humans can be copyrighted.
Therefore, material entirely generated by AI without human intervention does not qualify for copyright protection. However, if a human is involved creatively, such as selecting or arranging AI-generated material, that work could be protected by copyright.
Songs like ‘Heart on My Sleeve’ raise two critical questions in the realm of copyright law: First, the legality of using copyrighted music to train AI models (input question), and second, whether AI-generated works based on these models are considered derivative works and thus infringe on the original copyright (output question).
The current legal framework does not answer these questions clearly, indicating that this is an evolving area of intellectual property and copyright law.
Artists react to AI in the music industry
AI in the music industry has elicited a wide range of reactions from the public, artists, and influential figures. Here’s a few key opinions:
- The Chainsmokers: The DJ duo, consisting of Drew Taggart and Alex Pall, have expressed an interest in using AI to clone their voices. Their venture capital company, Mantis VC, has substantial investments in AI, indicating their belief in the potential of this technology in music creation. Taggart sees AI as a tool to enhance artistic expression and is interested in exploring new AI-generated voices based on their own.
- Selena Gomez: She has expressed concerns about the illegal use of her vocals through AI technology. Her worries highlight the potential for AI to infringe on artists’ rights and control over their creative output.
- Ed Sheeran: The Grammy Award-winning artist has voiced concerns about AI endangering human livelihoods, calling it “weird.” He views AI as a threat to human jobs, reflecting a broader unease about the impact of automation and AI on employment in creative fields.
- Barack Obama: The former U.S. President has commented on AI in music, suggesting that it lacks the spontaneity and nuance essential for disruptive creative expression.
- Grimes: The Canadian pop artist has been a vocal supporter of AI in music. She launched an AI tool that allows users to transform their voices into hers and has even offered to share royalties for any successful AI-generated song using her voice.
- Pharrell Williams: The mega-producer and artist has adopted a stance of acceptance towards AI in music, suggesting that it’s a natural progression of technological development.
- Hit-Boy: The hip-hop artist and producer known for his work with Jay-Z, Kanye West, and Drake has shown interest in AI’s potential. He acknowledges its increasing realism and suggests integrating AI into workflows.
- Damon Albarn: Damon Albarn from Blur and Gorillaz says, “We’re gonna need more drugs to get through absurd AI.”
As we can see, the road ahead for the music industry is one of exploration and adaptation as it navigates this new era of AI-augmented music creation. There are vast benefits and some big risks, reflecting AI’s impact on society as a whole.
In 1996, electronic music visionary Brian Eno said of digital technology, “Whatever you now find weird, ugly, uncomfortable, and nasty about a new medium will surely become its signature. CD distortion, the jitteriness of digital video, the crap sound of 8-bit – all of these will be cherished and emulated as soon as they can be avoided.”
AI will test the industry’s ingenuity just as technologies before it.