The musebot framework is a specification, and collection of implementations for, music artificial intelligence softwares that play together. Designed for musical metacreation, it consists in many individual software agents by different programmers which each have their own open sound control API that they expose to each other. Each musebot's API shares syntactical features with the others, but each interprets the messages it receives differently. Thus, when they play together, the musebots often interact with each other the way humans do: as different individuals, with different personalities, speaking a language with intersubjective meaning.

Getting Started Using and Making Musebots

The musebot framework has several components which need to work together. They can each be implemented separately. Download a test suite of them here.

Using the Framework

• Software Requirements: most of the existing musebots run as Max/MSP patches or as standalone applications built in max, but any program capable of sending and receiving OSC messages over UDP can participate in the musebot framework.
• The folder structure: where to store the components of the framework on your hard drive, relative to each other.
• The Conductor: the application, script or patch that launches the bots.
• A simple Template Musebot: a musebot template in Max/MSP.
• Sending and Receiving Messages: the languages the musebots speak to each other.
• Connections between remote computers: rapidly networking between musebot conductors on different computers.