The implications of Ghostwriter’s AI song

JACKIE MARQUEZ / OPINION EDITOR

Earlier this year, an anonymous artist going by the alias “Ghostwriter” released a song that used the “voices” of popular musicians Drake and the Weeknd. The track, titled “Heart on My Sleeve,” used Artificial Intelligence (AI) to mimic the voices of these well known artists, despite never getting authorization to use their vocal likeness. However, the lyrics and background tracks were written and produced by real people. The song quickly caught the audience’s attention on social media, garnering millions of views. Soon after, the team behind the Ghostwriter persona submitted it for Grammy consideration. 

While its submission was questioned by many social media users, the Recording Academy initially embraced it. 

“As far as the creative side, it’s absolutely eligible, because it was written by a human,” CEO of the Recording Academy Harvey Mason Jr. said. 

The song’s popularity sparked a conversation about AI in music, and many media figures are   interested. The New York Times reported that the individuals behind Ghostwriter have “met with record labels, tech leaders, music platforms and artists about how to best harness the powers of AI, including at a virtual round-table discussion this summer organized by the Recording Academy, the organization behind the Grammy Awards.” 

Although the concept of AI-generated music is gaining traction, it cannot replace music created by humans. I mean this both literally and figuratively. In a more literal sense, AI is unable to replace human-created music because it lacks some of the human capabilities required to make music. In a more figurative sense, AI should not replace human music. Such a thing would have reprehensible consequences. 

To start, AI cannot provide the human connection that human-made music can. We are drawn to music that is relatable to us, and AI simply doesn’t have the relatability of a person. If you look at popular musicians and their fan bases, they often have shared experiences. Take for instance the reaction of Olivia Rodrigo’s listeners to her newest album. 

‘GUTS’ by Olivia Rodrigo is    the   most   devastatingly     beautiful album written specifically for     teenage      girls   in  their   twenties,”  @gwenbrina                                             wrote on X.

The album, titled “GUTS,” captures the angst of growing up as a teenage girl. It reflects on experiences like being homeschooled, going through heartbreak and discovering yourself. Rodrigo’s new album has been positively received by women in their late teens and early twenties because of how close it hits to home. 

AI-generated music, however, fails to reflect these shared experiences because it does not experience things for itself. Our chatbots and music generators have never been homeschooled, broken up with or discovered personalities of their own; instead, they’ve learned to imitate it from the people that have. According to Forbes, AI learns by combing through mass amounts of data to discover patterns which can be used to mimic human interactions and emotions. Because it simply attempts to replicate experiences and emotions, AI cannot create songs about original experiences like people can. 

Due to its limited capabilities, AI has only been used as a tool in creating music thus far. However, it still has the potential to wreak havoc on the music industry. Currently, there are two primary uses of AI in music creation. The first type of AI usage is the kind seen in Ghostwriter’s “Heart on My Sleeve.” For this, humans write the music and lyrics, and then they use AI to recreate vocals so that the song sounds like it’s being sung by music superstars. 

This type of usage has become a trend on TikTok and Youtube; social media users have begun using AI to make covers of existing songs that sound like they’re being sung by well-known musicians. One example of this trend is an AI cover posted by TikTok user @1989..repulore of The Weeknd’s hit song “Starboy.” The cover sounds as if Taylor Swift is singing the song, and it’s garnered 38.5k likes. In a way, these covers steal from musicians. AI ultimately replicates the voice and style of artists who have spent their entire careers developing a unique sound. The fact that AI covers use these vocal talents without compensation is vocal plagiarism. 

The second type, AI music generators, are more concerning. Popular music generators like SOUNDRAW, Loudly and Jukebox only require users to pick details like genre, mood and track length, and then click “submit” to generate music. 

Generative  AI  also poses a threat to human musicians who need work. These bots allow content creators and companies to produce custom music, without hiring composers or performers. While this is convenient for people who want custom music at cheaper prices, it displaces musicians who need work and minimizes the value they add to society. Americans for the Arts reports that after the pandemic the unemployment rate for musical artists rose from 3.7% to 10.3%. While that figure is slowly returning to what it was pre-pandemic, it demonstrates the vulnerability of real, working musicians. Musicians rely on projects and gigs to make a living, and choosing to use AI over them pushes them out of the jobs that they need to survive and create more innovative, inspiring music. 

AI replacing human musicians will have a negative impact on the quality of music and hurt the live music scene. Human musicians are often experimental, pushing the norms around what can and cannot be done in music. They are also imperfect. As someone who’s been performing with musical ensembles since she was eight, I can confidently say that musicians never perform a song the same way twice. We miss notes, pause to catch our breath or deliberately mess with tempo and dynamics. This is why seeing music live is so enthralling. Imagine going to see your favorite artists in concert, and instead of performing their music they start playing a recording over the speakers; it’d be pretty disappointing. When you go to a concert or show, you go so that you can experience the music in a way that you haven’t before. If AI were to replace human musicians, there wouldn’t be new music to experience. Because AI simply replicates patterns, it fails to diverge from musical norms and try something new. This is why we need our imperfect, experimental human musicians.

While Ghostwriter’s song “Heart on My Sleeve” pushed AI-generated music into the spotlight, it has most recently reflected AI’s  inability to take the place of music created by humans. After backlash toward Harvey Mason Jr.’s   original statement about the song’s Grammy eligibility and questions about the song’s legality, the Recording Academy reversed its stance. It was announced that Ghostwriter’s   song    was actually   unable    to  be  submitted for a Grammy and that the Recording Academy would stand by human musicians. 

“I’m sure things are going to continue to have to evolve and change,” Mason said in a statement posted on his   Instagram  account, @harveymasonjr. “But please, please, do not be confused. The Academy is here to support and advocate and protect and represent human artists, and human creators period.” 

In taking the side of human musicians, the Recording Academy has taken a step in the right direction. Afterall, AI can never capture the humanness of music. Its algorithmic nature defeats music’s central focus: human connection. It lacks the imperfect, experimental creativity that people put into their songs. It threatens the livelihoods of already at-risk artists. Given these shortcomings, we must recognize that AI-generated music is not a stand-in for human crafted music, just as the Recording Academy finally did.