FDays after the attempt on his life, Donald Trump’s voice resonates over loudspeakers in Montego Bay, Jamaica: “If they needed an assassin, they should have called Bodyguard… about to commit a quadruple murder at Montego Bay Sumfest.” The audience, who had been prepared to hear a reggae beat, is surprised and laughs.
The Bodyguard crew just took the stage at Sumfest Global Sound Clash, a musical gladiator competition where sound systems compete against each other with creative mixes, highly animated presenters, and exclusive (often incendiary) recordings featuring star guests and inside jokes. But AI vocalists like this fake Trump are shaking up a decades-old musical tradition where authenticity and originality are paramount and sound systems pay artists premium rates to get vocals for clashes.
“AI is going to revolutionise the industry,” says Fabian Anderson, a dub agent who liaises with artists and sound systems to secure those exclusive tracks. He has refused to touch the technology, but knows of studios that are dabbling in it, so much so that he now sends his clients videos of artists’ recording sessions to verify their legitimacy.
Clashes arose in Jamaica’s music scene in the 1950s, when selectors imported records from the United States and played them to crowds through custom-built mobile sound systems. “Jamaica became the loudest island on Earth, with ever-larger crowds gathering for performances,” Island Records founder Chris Blackwell wrote in his 2022 memoir. As more sound systems emerged on the scene, they had to fight for audience attention: “Who had the best tunes? Who had the loudest, loudest sound?”
To prepare for a showdown, sound systems work with artists to record exclusive dubplates, often covers in an aggressive dub style with lyrics rewritten for battle scenarios. A famous example is the Fugees’ cover of Roberta Flack’s Killing Me Softly, originally written by the group. like a battle plate with the lyric “killing a sound boy with his sound.” Commissioning an artist to appear on one of these tracks costs between $150 and $800, Anderson says, and the most exclusive dubplates can fetch into the thousands of dollars: Super Cat and Shabba Ranks are among the most coveted artists.
To the joyous, loud vuvuzelas of the Sumfest crowd, German sound system Warrior Sound releases an exclusive dub recorded by Jamaican upstart Nigy Boy, featuring the lyrics of his viral success The continent shifted to chants of Warrior: “Sumfest / We will kill them / We will win the trophy.” Other artists heard at the clash included Bounty Killer, Damian Marley, Capleton, Beres Hammond and the Heptones.
Most selectors only use seconds of a dub “to get the message across,” Anderson says. Songs can’t be repeated, so a large number of dubplates are needed to ensure momentum. It’s an expensive process, which has led to a common practice of splicing — copying dubplates and erasing the names from the original sound system — to get cheap tracks. With the advent of generative AI, the game is getting even dirtier. “AI is even worse[than splicing],” Anderson says.
Veteran British reggae/dancehall vocalist Paul Scott Levy, aka General Levy, records dubs “almost every week.” He relies on established sound systems to respect the rules of the clash. “It’s not just about using a voice. It’s about how you got the voice, the relationship, how much money you paid for it.”
Levy cites the recent rap battle between Kendrick Lamar and Drake as an example of generative AI failing: When Drake released Taylor Made Freestyle, a song featuring AI covers of 2Pac and Snoop Dogg that dissed Lamar, “Drake got slammed for that. He didn’t give[the AI]any merit in the realm of confrontation.” Tupac’s estate threatened legal action against Drake, and the song has since been removed from all music platforms.
Sound systems play their cards close to home (Notorious hired a “top secret” artist to record a dub version of Bob Marley’s Buffalo Soldier (with the lyrics changed to “Japanese Soldier”)), but stress that they would never use AI to resurrect Marley, nor would they fake any living artist. “It’s to respect the artist. If people want to do it, let them do it, I’ll find out and curse them,” Notorious selector Bad Gyal Marie smiles.
Fellow contender Dynamq, who was the 2023 Sumfest champion, has a more ambivalent view. “The splicing and AI… I don’t think[the public]cares or minds,” he said. “As long as it sounds good, people will accept it. I don’t support it; I’d rather lose with integrity than win.”
At Sumfest, the vuvuzelas are louder for Jamaican-Japanese sound system Notorious International, which takes home the grand prize of one million Jamaican dollars (about $6,300). Bodyguard finishes second after a thrilling “tune-fi-tune” duel with Notorious. Bodyguard founder Courtney Singh, a veteran of sound clashes active in the scene since the early 1990s, says he doesn’t want to use AI beyond parodies like the Trump dub. “It’s a real danger, I’m not sure how it will be controlled,” he says. “There’s an unspoken ethical code, but (now) there’s a different generation that’s out to win at all costs.”
Singh notes that next-generation sound systems may rely on AI to break into the scene, due to the high fees charged by many artists, who often sell to richer sound systems in Japan and Europe. “In Jamaica, we’re paying US dollars for our dubplates! Artists are charging 10, twenty times more than when I started, which makes it almost prohibitive,” he says. “They’re playing a major role in their own demise.”
Perhaps inevitably, Levy disagrees. “AI will be used to infiltrate a genre that was once all dancehall and reggae, taking away the real vibe,” he says, defending his flesh-and-blood voice. “Our voice is our bread and butter.”