This research project explores the impact of Machine Learning (ML) tools on compositional practices. Inspired by the film "Apollo 10 ½: A Space Age Childhood," the composition created for this project blends NASA footage with other sounds to evoke the protagonist’s blurred reality. ML tools like a text-to-audio model and Fluid Corpus Manipulation (FluCoMa) aid in generating and organizing sound materials, addressing challenges like limited accessibility, and finding unique relations as well as the interaction between different sounds. This project aims to contribute to computer music by displaying the potential of using ML tools in narrative-based compositions. By integrating these tools, it seeks to enhance the composer’s creative practices and offer new avenues for musical expression.