Imagine Kishore Kumar yodelling a Badshah banger.
Or Mohammed Rafi’s dulcet vocals replacing Arijit Singh’s brooding baritone.
Open YouTube and you will find countless creations like these. Long gone singers voicing current chart toppers. Artists who never worked together featuring on the same song. All of it made possible by AI tools at the disposal of creators.
Tech may never come close to mimicking the full breadth of Sonu Nigam’s legendary talent though. Among the defining playback artists of the late 90s and the aughts in Bollywood, Nigam’s is a generational voice. Countless of his songs remain the soundtrack to people’s lives. “Tu Fiza Hai.” “Kal Ho Na Ho”. “Suraj Hua Madham”. He’s the winner of the National Award. He’s received the Padma Shri. He’s sold out live performances and world tours. And he sang Tees Maar Khan's title track, incredibly, in 54 different voices.
Human creativity, especially music, is held as the pinnacle of our inner expression. Generative AI may present the greatest challenge to it yet. Early in 2023, a version of Rafi singing “Haule Haule Ho Jayega Pyaar,” a Sukhwinder Singh tune from Rab Ne Bana Di Jodi went viral. Nigam, who heard the song, was “baffled and fascinated”. Another time, some online tributes to deceased singer KK used his voice without permission, a move that upset several prominent artists. Copyright violations. Invasion of privacy. Insensitivity to the artist’s legacy.
Nigam takes a pragmatic stance towards any new tech. “AI can be a beautiful tool but it can also quickly become a gimmick.” Once the “look at this trick” aspect of AI wears off, it might get tiring, says Nigam. A tech wave often arrives as an intruder into music-making at first, but then gradually assimilates. Auto-Tune, that allows for vocal modifications. Cutting-edge sound mixing. “AI will be very useful also. For instance, if a composer has made a song and wishes to present it to a label in my voice, they don’t have to call me now,” says Nigam.
Still, the artistry will have to speak for itself. “Live shows cannot work on AI,” he says. “The human connection to music will remain, at least as far as I can see.” The biggest tussle looming on the horizon is the indiscriminate use of an artist’s likeness by AI – without their consent. Recently, a song created using AI software was removed from Spotify after the two artists featured on it – The Weeknd and Drake – raised objections. Listeners, however, were unable to tell the difference, and the song became the first generative AI streaming hit. In fact, a first ever lawsuit has been filed by music publishers over the scraping of copyrighted music lyrics by Google- and Amazon-backed AI company Anthropic for their large language model, Claude.
AI has crash landed into the world of music and entertainment. Are we prepared for the artistic world that’s coming? At SYNAPSE, Sonu Nigam will address the possibilities and limits of tech in art. Compare artist vs. algorithm. Music from the soul vs. music by numbers. Creativity vs. copyright infringement. Will generative AI erase music’s individuality, or simply be another tiny river pouring into the vast ocean of creativity?
