BioDigitalMediafilia is a reactive and generative audio-visual art work that delves into the rhythms of nature through the manipulation of image and sound and experimental inquiry through the use of digital images created and processed with various Artificial Intelligence (AI) algorithms based on generative models that allow visuals to be produced from training data sets. BioDigitalMediafilia expands the premises of new digital media art, in that the “human-nature-technology” relationship which is presented through art is no longer focused only on representations created with material symbols, but is emphasized on the whole process and how the creative process is generated. Based on these premises, BioDigitalMediafilia exposes the biodiversity of ecosystems from a digital taxonomic perspective, operating on a series of electromagnetic visions and focusing on the microscopic scale of the micro-collision of dissected “astromes” and the optical and aural fluctuations built like the universe, where regardless of its dimensions everything is made up of nanoscopic pieces of geometry and substance. The BioDigitalMediafilia project also investigates the "aesthetics" of Artificial Intelligence (AI) errors and machine learning through an artistic experiment that imitates AI with the intention to test machine learning, as well as the human audience, by hypothesizing that there is another type of of unknown AI, one that employs a process of learning that consumes less time for tasks that involve particles of light, called photons, to generate the combinations of image and sound.