This project is a group project presented at ircam in 2022 at centre Pompidou.
We are exploring the interplay between restriction and freedom that comes with the normalisation of AI in the music industry. Our idea is to create avatars for our own virtual idol group, pairing with voice/music synthesizing software - replacing real instruments and voices with computer generated sounds. This is a prototype, a start, to see how far we could go.
For generating the lyrics, we found lyric generator websites online. The way these websites work is taking a keyword or keywords and generating lyrics. Our keywords were ‘cyber’, ‘computer’ and ‘dystopia’. Our lyrics combine two languages, English and Chinese. Once we adjusted the lyrics, we used the software 'Synthesizer V’ to create and develop the voices for the track. We built our characters in vroid with customized hairstyles and clothing. The characters, pre-made dance moves and stage that was brought together using Mikumikudance.
In the future we would like to also have AI music made from creating an algorithm in a neuron network. One of our references for this is Sweaty Machines, a team that fed hundreds of Eurovision songs – melodies and lyrics – into a neuron network. Then, algorithms produced thousands of new tunes and lines of verse.