They work, they study, they learn. Either they cast, participate in talent shows, struggle to emerge: artists work hard at all stages of their career to carve out a space in the music world. And then there's artificial intelligence, which effortlessly churns out a piece of music and gets 100 million listens.
Second a report di Music Business Worldwide, the Chinese streaming giant Tencent Music has recorded and released over 1.000 new songs with AI-generated human voices. One in particular, titled "Today," became the first song with an AI vocal to break nine-figure numbers. Just think, one piece of music alone grossed $350.000 in revenue.
Singers who "come back to life", or entirely new voices
TME's "Patented Text-to-Speech Technology", Lingyin Engine, is the core of these synthetic songs. It is able to replicate the voices of real singers to produce original songs in any genre and language., and Tencent has also used it to develop synthetic voices that recreate the unmistakable timbre of the voices of several departed music stars, such as the late singer Chinese Anita mui.
Here is the song that "resurrects" it. The quality achieved is impressive.
Just like for AIs that create graphics and images based on simple written descriptions, or as for those who recreate missing actors (or rejuvenate those still alive like Bruce Willis o Harrison Ford), this "musical" AI also raises accusations.
Whose AI-created piece of music is it in the end?
La Recording Industry Association of America (RIAA) he has recently sounded the alarm about AI piracy in the music industry. The AI would "fish" from the infinite well of songs already made by human artists to cut, sew, paste, create a piece of music that seems new, but is actually "pilfered here and there".
It's true? Technically yes. Is it possible to charge duties or rights? Maybe it will: maybe this technology will be regulated. Or not, and flesh-and-blood artists will have another cumbersome obstacle on the road to success.