Blob Opera 

David Li and Google Arts and Culture made this little machine-learning experiment to create opera-singing blobs.

The music is created with a machine learning model that learned about opera by listening to Tenor, Christian Joel, bass Frederick Tong, mezzo‑soprano Joanna Gamble and soprano Olivia Doutney (plus Ingunn Gyda Hrafnkelsdottir and John Holland-Avery). It’s not playing their audio–rather, it’s playing what the model thinks opera music sounds like based on analyzing 16 hours of their singing.

I like how the eyes watch the cursor - that’s one of those little procedural animation touches that’s simple but adds a lot of life.

https://artsandculture.google.com/experiment/blob-opera/AAHWrq360NcGbw?cp=e30.