I think I finally came to a good explanation and/or theory as to why AI "music" sounds the way it does.
The vast amount of input data it has to have had in order to train their algorithms must've been with low bitrate, low quality samples in order to save on training time and storage space.
It just sounds so distinctly airy and robotic, I've never understood why all would have that je ne sais quoi but I think this is a good start