Have you listened to "music" created by non-specialized models that use WebAudio to create?

It seems like a good descriptor to me! ;)