“La Solitudine delle Moltitudini” (Immersive video installation and electroacoustic cantata for 16 female voices on fixed media (2023-24) - soprano and lyrics: Júlia Coelho) is an immersive room audio-visual installation that creates an immersive and emotional space for the visitors in a closed room. “La Solitudine delle Moltitudini” is a metaphor of our increasingly interconnected world, where data and information shape our perception of self. This ecosystem is represented both graphically and musically, by graphs and networks, that provide the main visual material and are at the foundation of the generative musical score. The score incorporates elements of AI generated sounds through a synthesis method mediated by a proprietary recurrent neural network (RNN) model based on the SampleRNN architecture of Mehri et al. (2017). The model was trained on the same samples recorded by the soprano Júlia Coelho for the installation, and thus does not use any data from third parties. The AI generated samples are integrated in the fabric of the composition exploiting the same geometrical structure of its pitch and rhythmic spaces to create a subliminal trace in the imaginary of the viewers.All the networks are generated using real data and are the results of scientific investigations in the BarabasiLab.