If you scroll on TikTok, chances are you’ve seen these terms: “Taylor Swift AI”, “AI cover”, or a song being attributed to an artist who has never sung it before. With the rise of AI-generated content, music is no exception to its theatrics — with a tweak of the lead vocals and some editing with online software, anyone can make Taylor Swift sing “Starboy” by The Weeknd… which they already have, judging by the millions of plays across social media platforms such as Spotify and Youtube. Have you listened to these AI tracks? Do you think AI can replace pop stars?
Right now, the answer hovers between two extremes: skepticism at the extent that AI can be used to replace our favourite singers, and fascination with how far we can afford to push this barrier. While AI-generated music has undeniably captured the attention and curiosity of millions, there remains those who question its ability to replace human pop stars fully. Let’s take a look at some of the world’s opinions about AI-generated music:
The Emotional Pull of Music
Pop stars and musicians possess a unique blend of creativity that draws their fans towards them, be it through their lyrics, soundtrack, or even their stage persona. It’s not just about singing: their music conveys a storytelling atmosphere, authenticity, and a human experience their fans can relate to. While AI can mimic vocal styles and produce catchy tunes, it lacks the raw emotions that shape a true pop artist’s music.
ChatGPT, which is probably a familiar tool to most of us by now, can create lyrics in the style of many famous music artists. However, when a fan sent Nick Cave ChatGPT’s attempt at writing a song in his famous lyrical style, his succinct response was: “This song sucks.” In his opinion, trying to replicate an artist’s lyrics or music is a “travesty”. Songwriting was impossible to truly mimic or duplicate; rather, it required the artist’s “confrontation with one’s vulnerability, one’s perilousness, one’s smallness”. In short, to Cave, writing a song required the true human experience, not just words strung together by an algorithm.
Yet, at the same time, is it that easy for us to differentiate between resonating with a real memory, and one simulated by a machine? Many listeners of AI music have stated that the reason they enjoy such music is because they’re listening to their favourite idol singing their favourite song: two concepts that may not have coincided with each other without the use of AI. Take football legend Cristiano Ronaldo singing “Memories” by Conan Gray as an example, or F1 driver Charles Leclerc singing “Someone Like You” by Adele. While they do fall under the category of “memes”, the number of plays they’ve racked up on TikTok indicate that many users do not look for the rawness of real music: sometimes, all that matters is the voice.
Ownership and Attribution
The very presence of AI has sparked alarm throughout the global industries about intellectual property concerns. In the realm of pop music, many major labels are also taking the necessary precautions to ensure their artists retain the rights to their songs.
Universal Music Group, the largest of the major labels and home to artists such as Drake and the Weeknd, had already flagged such content to its streaming partners this month, citing intellectual property concerns. The label also asked the pertinent question of “which side of history all stakeholders in the music ecosystem want to be on: the side of artists, fans and human creative expression, or on the side of deep fakes, fraud and denying artists their due compensation.”
As the audience, perhaps it is less important for us which songs belong to which singer, to which label; but for the artists themselves, this ties in to a whole host of other problems, such as copyright infringements between the generator and the artist, their publishing rights, and ultimately becoming obsolete in the face of the looming future of AI. The ethics of AI in the arts is an ever-growing question, with legislators only beginning to sort out questions of ownership when it comes to non-humans. In a Black Mirror-esque future, could it be possible that even algorithms are allowed to possess something humans cannot?
Revolutionising the Pop Industry
Of course, AI stealing away the jobs of pop stars is only one side of the equation. There are many examples of pop stars collaborating with AI to create music: in 2019, self-described computer musician Holly Herndon created an AI-powered vocal clone called Holly+ that sings in Herndon’s voice. “I’m trying to present another side: This is an opportunity.” Rather than being afraid of and shunning the use of AI in the music industry, Herndon shows us how this can simply be a new juncture in pop culture, a decade for trying out experimental technologies. Holly+ can sing in languages Henderson can’t speak fluently. If other artists could replicate their own vocals, it would open up whole new markets and possibilities for them; imagine Harry Styles in Spanish, or DPR Ian in Chinese. While it may seem hilarious to us now, there’s no saying how well it could make songs resonate with speakers of other languages when they hear their favourite artists sing in the same tongue.
In addition, pop music is mostly formulaic — you may have heard of the famous “one key, 4/4 time” rule that most pop songs generally stick to. They also measure between three to five minutes, have a chorus that repeats two to four times, and repeat the title multiple times throughout the song. An AI software fed with these pop music algorithms would most probably be able to come up with a catchy song within hours or even minutes, and when paired with a relatively famous artist, could produce the next song of the year. Compared to waiting months or even years for artists to produce one album, which producer wouldn’t be enticed to recruit this new-age employee?
In Conclusion…
AI can change the way we view pop music forever. Yet, it is important to note that the power of technology, as of now, will always remain in the hands of the person using it. Without the pop star’s original vocals, users wouldn’t be able to recreate new songs using it as a template — a paradox within itself. In essence, what we can take away from this is that being afraid of AI isn’t going to solve any existing problems we have right now; rather, we should learn to harness it for our advantage, or at least to keep ourselves relevant and in the now. AI isn’t coming. It’s already here.
Written by: Tan Xin Yee Sarah (23S77)
Edited by: Alicia Ng (23A15), Koh Yu Fei (23A12)
Comments