I love Brian Wilson and identify strongly with his fear of stages and his love of the studio. This is not intended to disrespect him in any way. It’s such a beautiful song.
And KARL, the best-sounding voice from the Chinese AI voice app I used to make these, did his best.
AI Karaoke is easy to make and it reveals the flawed nature of generative content. I put the a capella into an app that 1) interprets the words and transcribes the song, and 2) outputs MIDI that purports to follow the vocalist’s melody. But both things are done very badly.
I create backing tracks from scratch, and then put the output of the first step back through the app, to reconstitute a new vocal part. Which is inevitably very, very strange.
Then by adding the lime to the coconut, we can envision robot KARL doing his very best, like the character in Disco Elysium, to sing his heart out and emote. With words that are mostly incorrect, and strange melodic artifacts.
I find these hilarious and hope you chuckle too.
/DJIC
Share this post