WCMM is the result of a creative process making use of Machine learning. to show an artificial artificial intelligence at the piano while painting on canvas.

WCMM is based on Magenta's performance_rnn, which gives some sensitivity to improvisation.

Although, of course, that you consider the performer sensitive and generate emotions is up to you. He only wants to tell stories and therefore offers you four different narrative frames with a unique and unrepeatable result. It's up to you to listen to each story independently or mix them together.

Which one do you start with?