Přehled o publikaci

The lecture presents four aspects of the use of artificial intelligence in musical composition using examples from the author's work. The explanation is directed from compositional approaches standing at the border of the current understanding of artificial intelligence to tools based on explicitly data-driven models. Attention is paid to generativity and algorithmic composition, i.e. modeling of musical material through formal rules that enable its variable transformation and adaptation in various musical contexts. Subsequently, the polarity between determination and aleatoric is discussed, illustrated for example by the principles of L-systems and Markov chains, and its relationship to the tension between composer's intention and machine creativity. This issue is followed by the question of mimetism and spectral resynthesis, in which artificial intelligence expands the possibilities of instrumentation and work with sound similarity. Finally, the question of recursion is opened in the context of improvisation and interactive systems that analyze input material, generate their own reactions, and at the same time reflect on their previous outputs. The lecture offers not only an overview of specific compositional practices, but also a broader consideration of the nature and boundaries of artificial intelligence in contemporary music creation.