One cuts the incoming audio to pieces of, maybe milliseconds, maybe dozensof minutes. Another builds them into rhythmic sequences. Third one tries to turn a piece of sound to a tune. The fourth imposes an occasional sound effect, like a weird endless echo.
300 random generators work at the same time, and most of the time they produce a little interesting noise, but sometimes the network seems to find inspiration, and it creates sound pattern s of amazing beauty and complexity.
We cut the human evolvement for as little as possible. The role of a person in this project about trying to capture the moment, isolate and save it to file.