Four-time Grammy winner Timbaland, who helped shape the careers of musicians such as Aaliyah and Missy Elliott, has a new act — a pink-haired “singer” who exists only in code. Her name is TaTa Taktumi.
Her song Glitch x Pulse has already racked up several thousand plays on YouTube and music streaming platforms. A musician like Timbaland is likely sitting on a pile of demos and half-formed ideas, some of which are now being completed with the help of AI.
He types prompts describing what he has in mind and lets software such as Suno — where he serves as a strategic adviser — do the heavy lifting. All delivered via TaTa. Suno is a generative AI platform that can create full-length songs.
Closer home, a popular RJ recently sent a WhatsApp group message asking people to listen to dozens of his songs. Five minutes into the tracks, one hears a bit of Robert Miles, a hint of Enrique Iglesias and a whole lot of familiar inspiration. The result feels like the afterimage of the music that dominated airwaves in the 1990s and 2000s.
The RJ writes the lyrics, remasters his voice and uses Ozone from iZotope, a complex mastering tool. Through the platform TuneCore, the music is then distributed across streaming services. It takes him nearly a week to assemble a track, one of which asks listeners to “feel the fire” and then “let go”.
Research from French streaming service Deezer suggests that 97 per cent of people cannot tell the difference between AI-generated music and human-composed music. Roughly 50,000 fully AI-generated tracks are now uploaded to the platform every day, accounting for 34 per cent of daily deliveries.
Producing a basic AI track has become almost child’s play. The Telegraph asked a 14-year-old to write a few lines and dictate prompts to Suno. Within minutes, the app produced something impressive enough to make Charlie Puth drop his jaw.
A few weeks ago, Suno co-founder and CEO Mikey Shulman wrote on LinkedIn that the AI music generator now has two million paid subscribers and $300 million in annual recurring revenue.
Suno and its rival startup Udio have already flooded the Internet with millions of AI-generated songs. At first, record labels came after them. Sony Music, Universal Music and Warner Records sued the two startups for copyright infringement in 2024. Slowly, however, the AI companies have begun forging alliances with the industry that challenged them.
Suno recently reached an understanding with Warner. Udio has signed licensing agreements with Warner, Universal and the independent label Merlin — a digital rights agency representing independent record labels and distributors worldwide.
Technology has long reshaped how music is made. DJs edited songs for the dance floor in the 1970s. The following decade saw hip-hop artistes sample funk and soul to create entirely new tracks. The tipping point, however, arrived in 1997.
Computer scientist and composer David Cope, who in the 1980s created Experiments in Musical Intelligence (EMI), a programme designed to assist in classical music composition, was challenged to a Turing-style showdown. A pianist performed three pieces before students and lecturers at the University of Oregon, US. One piece was composed by Bach, another generated by EMI and a third written by Steve Larson, a professor at the university. The audience was asked to vote on which one was the real Bach. Most chose the EMI version.
In 2023, Suno arrived on the scene, ushering in a wave of spam-like tracks. In truth, the startup simply appeared at a moment when the AI curve had matured.
Musician Ray Dickaty of the jazz collective Light Star Guiding told The Telegraph, “I like machines in music but not music made solely by machines. Although, having said that, I recently heard some AI music that was really good. What is unfortunate is the sheer amount of AI music choking YouTube, Spotify and so on... Due to the algorithms, these are always the top hits and I worry for younger generations who absorb this music as normal.”
Birsa Chatterjee, who performed at Jazzfest in Calcutta as part of the Caravanserai project and was trained at The Juilliard School in New York City, says, “For me, AI can never replace lived experience, vulnerability or the cultural memory embedded in improvised music. But it can enhance certain aspects of workflow, composition or production. My concern is less about the technology itself and more about how society chooses to value — or devalue — the human beings behind the art.”