Don’t tell Nick Cave. Who (with a good dose of reason, we’ll talk about it again) had expressed himself in a not too flattering way on a piece produced by the AI that would have really mimicked his style.
Don’t tell him, but even in music, generative artificial intelligence (i.e. capable of producing new content) is inevitably gaining ground. Just as it is happening for texts and images.
And what could happen in the near future would mark a turning pointwhich would open up who knows what new horizons, and which would certainly make purists, lovers of acoustic guitars, unplugged concerts and crackling vinyls turn up their noses.
Two giants respectively of the tech sector and the music industry, Google and Universal Music Group, are in fact in negotiations for music produced with artificial intelligence. Even if the detractors prefer the term deepfake music. Let’s find out better what is happening.
The Google-Universal negotiations for music produced by artificial intelligence begin
It was the Financial Times, on Wednesday 9 August, that first reported the news.
According to which Google and Universal are in talks for the licensing of melodies and also of the voices of the artists represented, to generate music with artificial intelligence.
Said like this, one immediately gets the impression of being faced with a possible epochal turning point, which would concern more than one aspect. Musical, of course, but also contractual, and even cultural.
According to the British newspaper, another record major, Warner Music, is also discussing the same issue with Google. And just the CEO of Warner Music, Robert Kynclrecently stated that “with the right structure” AI could “allow fans to enjoy the music of their favorite artists through a new level of user-driven content”.
How could this work?
Neither Google nor Universal have made any statements on the matter.
But the negotiations would also concern the development of one specific tool which allows the production of music created by artificial intelligence starting from the voice of an artist, from the sounds of a musician, and – it goes without saying – also from the technical specifics of a writer of texts for music.
In short, any listener could create a unique and unrepeatable playlist that echoes the style of their musical idols.
For each song created by the AI royalties will be paid to their respective holders.
But the musicians and singers, in all this?
Singers and musicians free to choose (and divided)
It seems, at the very least, that each individual artist is free to choose whether or not to allow their voice, music or lyrics to be used by artificial intelligence.
Pop stars are divided on this front. Self Bruce Springsteen he declared himself against Madonna he even included in his will a clause that his voice cannot be used by the AI after his death.
On the contrary, Paul McCartney candidly admitted the use of artificial intelligence for creating what he called “the last Beatles record”. AND Grimes she declared herself in favor of anyone using her voice, as long as she is entitled to 50% of the royalties.
And say that In July, Universal Music’s general counsel Jeffrey Harleston spoke out against AI cloning of an artist’s voicewhich “is the most valuable part of his livelihood and his public persona.”
The doubts
On the one hand, if an artist choose freely to feed one’s voice or music to artificial intelligence, no one is harmed, and the discussion should end here.
But we cannot fail to notice at least two negative repercussions of a possible turn in this direction. With works produced by AI in the manner of one artist or another, we would have increasingly standardized songs, less and less vivid and original. And on this, there is little to be done, Nick Cave is right.
Not to mention the fact that we will improvise producers and promoters. And maybe, by virtue of the often inscrutable dynamics of the algorithms that dominate the network, a song that has slavishly cloned the voice of a certain artist will become a hit. Bringing success to the unsuspecting listener who tried to use AI for fun.
In short, true art – in music and beyond – is based on the great personalities capable of creating inimitable masterpieces. On the contrary, deepfake music is based precisely on repetitiveness, on seriality.
Probably, in a few years true art will be joined by modest and indistinguishable operas. But if you think about it, this is already happening right now, without the need for AI.