Three Entistatios (Parts and Score)
'Three Entistatios' is scored for 12-part chamber ensemble. This download includes the score and parts.
This three-movement work is a piece inspired by, created in collaboration with, and responding to artificial
intelligence. Most research into AI begins with a model that has learned as much as possible from its dataset – where its training is ‘complete’. For this piece, I was interested in examining the AI at different stages of learning music.
The first movement is an exploration of the beginning of the learning process, where an AI has not yet learned what we consider the basics of coherent composition and musical technique. Its neurons fire randomly, producing many unrelated ideas at an extremely fast rate. It will also often ‘overfit’, meaning that it copies directly from its dataset, like a quote, or it might become stuck in a repeating cycle of one or two notes. Some of the material from this movement is derived directly from an AI in the early stages of training; other ideas are original, conceived to fit this overall narrative.
The second movement is a conversation between myself, the composer, and a state-of-the-art AI network that is fully trained upon my music. I would write music, then the AI would respond in a style it had learned from studying my previous works. I decided which of its ideas were interesting, reworked them into new forms, and then gave this as the algorithm’s next input.
The final movement moves more towards science fiction. It is an imagined soundscape of a future-AI that has reached ‘singularity’ – the ability to rework and improve itself at an exponential rate. The incoherence and randomness of the untrained model is now totally replaced by a single-minded, obsessive world of whirring tones passing by at an ever-increasing rate. Eventually the computer ascends into a plane that is too rapid and fleeting to be heard by human ears.
An ‘Entistatio’ is a word devised and defined by the artificial intelligence network ‘Lexiconjure’, trained on the Merrian Webster English Dictionary and created by computational artist Ross Goodwin. Its definition is given as “a short piece of music or metal”.
With thanks to my collaborator Christine Payne (OpenAI), and her model MuseNet.
This three-movement work is a piece inspired by, created in collaboration with, and responding to artificial
intelligence. Most research into AI begins with a model that has learned as much as possible from its dataset – where its training is ‘complete’. For this piece, I was interested in examining the AI at different stages of learning music.
The first movement is an exploration of the beginning of the learning process, where an AI has not yet learned what we consider the basics of coherent composition and musical technique. Its neurons fire randomly, producing many unrelated ideas at an extremely fast rate. It will also often ‘overfit’, meaning that it copies directly from its dataset, like a quote, or it might become stuck in a repeating cycle of one or two notes. Some of the material from this movement is derived directly from an AI in the early stages of training; other ideas are original, conceived to fit this overall narrative.
The second movement is a conversation between myself, the composer, and a state-of-the-art AI network that is fully trained upon my music. I would write music, then the AI would respond in a style it had learned from studying my previous works. I decided which of its ideas were interesting, reworked them into new forms, and then gave this as the algorithm’s next input.
The final movement moves more towards science fiction. It is an imagined soundscape of a future-AI that has reached ‘singularity’ – the ability to rework and improve itself at an exponential rate. The incoherence and randomness of the untrained model is now totally replaced by a single-minded, obsessive world of whirring tones passing by at an ever-increasing rate. Eventually the computer ascends into a plane that is too rapid and fleeting to be heard by human ears.
An ‘Entistatio’ is a word devised and defined by the artificial intelligence network ‘Lexiconjure’, trained on the Merrian Webster English Dictionary and created by computational artist Ross Goodwin. Its definition is given as “a short piece of music or metal”.
With thanks to my collaborator Christine Payne (OpenAI), and her model MuseNet.
Save
Share
Tweet
