It’s no secret that I started to learn to sing back in 2014 because I had been very bad a Karaoke. Eight and a bit years later I’m “competent” as a singer. What’s mildly annoying about it that, in far less time, researchers have taught AI entities to sing, and even have them accepted as students in a musical academy.
I am aware of one startups focused on teaching AI to sing and that is Popgun, which confusingly trades as Splash! Back in 2019 they were previewing their results under the Popgun branding, and they were interesting.
Jump forward a couple of years, and the quality has improved quite dramatically and they’re using the Splash branding of the Abelton AI music creation plug-in.
There has been a lot worse pop music sold! That’s the evolution over about four years, projecting forward another couple of years and it will be indistinguishable from human singing.
While I sing for my own enjoyment, a virtual singer Luya has enrolled at the prestigious Shanghai Conservancy of music:
The AI singer, named Luya, will start as a first-year student at the Shanghai Conservatory of Music with the music engineering department next month, the university announced on Tuesday.
https://www.scmp.com/yp/discover/news/asia/article/3191345/beautiful-sweet-voice-ai-singer-enrols-prestigious-chinese
Luya was developed by mainland China AI product company iFLYTEK. Luya has allegedly released a song A Glimpse of Light in the Fog, but I cannot find any link to the song.
As for AI dancing, if robots count then Boston Dynamics did it back in 2020.
Ultimately, I wonder what – if anything – will be left as a human only creative domain?