Referencing this Twitter thread: https://twitter.com/TudorAcid/status/1191408828769804289
In the VCV world, you’d feed it input – i.e. jamming on a keyboard, and at a certain point flip a switch and it would use Markov chains to randomly generate output note sequences, that randomly selects a note, and generates a note after it based on the probability in the input that a particular note follows after it.
This is what’s behind ‘travesty’ generators do with text, but do it with notes or groups of notes.
There’s a lot of Markov chain code floating around, e.g.