Is Generative AI a disrupter in music composition field?
Until six months ago, no one knew about ChatGPT, or even imagined it could automate one’s life. Since then, the world has seen a drastic transformation by artificial intelligence (AI) from an imaginary dystopia into a depressingly fast-changing reality.
The song “Heart on My Sleeve,” sung by an AI-generated clone of Drake and the Weeknd vocalist, was posted in April 2011 by the then-unknown TikToker “Ghostwriter977”. The song by “Ghostwriter977” had been viewed millions of times, but was targeted for deletion which claim was issued by Universal Music Group (UMG). This case is just the beginning. Spotify recently deleted massive amounts of AI-generated songs from its streaming platform, and a lot of similar cases are happening.
With such generative AI (Generative AI) used in the music industry, it is clearer than ever that AI software has the potential to dramatically change the way music is conceived and recorded. They may say AI will create automated creative tools but the real question is, does this mean AI is a disrupter to put the jobs of an entire industry at risk?
Sean Everett, a Grammy-winning recording engineer, tried using AI tool while working on an unreleased song by The Killers. Everett recalls that when he loaded a chord progression into AI and asked it to continue the emotional progression in the style of Pink Floyd, it output an unexpected melody. “I had never heard anything generated before. Of course, this kind of talk is frightening to anyone,” Everett says, “but I’m sure it will become a reality.
Rick Beato, a veteran producer of music-based contents, said “the first job that cease to exist will be the mastering engineers. The mixers are the next most in danger,” Beato said. AI tools for mixing and mastering could mimic any engineer’s style. Microphone-stands that adjust microphone height via an app are already in use by engineers, and these apps will be controlled by AI. These types of jobs will surely disappear from the world.
However, there is always an upside. Beato believes that the advent of AI will put an end to a long-standing problem in music production that we have been working to solve: the limits of what one person can do. Taking lyric writing as an example, you could write lyrics with other lyricists but, today’s lyricists can enter lyrics into ChatGPT and have them rephrased in the style of other lyricists, so you don’t have to meet your collaborator in person.
Also, the AI music is thought to upgrade the creativity of human beings to the next level. Sono AI, the music composition app, enables us to combine totally different genres into one music by simply typing meaningly disparate words randomly. Nowadays, so many composers are using this app for “brainstorming” their ideas, for their composition.
The main reason people are reacting emotionally to AI is because of the potential threats to their position and the financial trouble it could cause. The rise of these tools will drive people out of the creative industries, but the most important thing to consider is whether AI will increase TAM (total addressable market) in the long run.
The thing we can say for sure at this moment, is that emerging AI tools can radically democratize songwriting. People who were not previously considered musicians will be able to write and produce songs, as was the case with sampling in early hip-hop.
In the short term, AI paved a way for the digitization of music on a different scale than software such as Pro Tools or Auto Tune (pitch correction software) did in the past. Nevertheless, if one takes a longer view of AI’s development, it is not easy to make a conclusion.
Let’s imagine the future where if you want to hear a Metallica song with elements of Lithuanian music mixed in, all you have to do is press a button of AI, and ask them to make that song. Even if you could use such a tool, and the song made by AI could be close enough to the one human made, would it really have any artistic value?
Music, which is thought to be having power to unite people, is only effective when the real sympathy is felt among the musician/audience. The music made by AI lacks this empathy, since robots can’t experience real up and down that surrounds human life, and the lyrics superficially made by learning a ton of data, does not necessarily bring the atmosphere of bonding.
~Other information source~
https://www.theverge.com/2023/4/18/23688141/ai-drake-song-ghostwriter-copyright-umg-the-weeknd
https://wired.jp/article/generative-ai-music/
https://wired.jp/article/ai-generated-music-streaming-services-copyright/
