Beyls, PeterPerrotta, André2018-01-232018-01-2320169781450340823http://hdl.handle.net/10400.14/23940This paper describes a self-regulating artificial ecosystem in continuous exposure to human observers. Particles of variable morphology engage in local interaction and give rise to emergent overall audiovisual complexity. People only exercise influence over autonomous behavior developing in the artificial world. A machine-learning algorithm basically aims to maximize audiovisual diversity by tracking changes in systems behavior in relation to behavior in the artificial world. We suggest rewarding human-machine interaction to exist in the elaboration of dynamic relationships between spatial and cognitive human behavior and audiovisual performance in an artificial universe.engAesthetic human-machine interactionMachine learningParticle systemsAutonomyInfluenceEmergencejournal article10.1145/2851581.2891090