one thing i find fascinating about generative systems is that while we have the tech now to do some pretty incredible things (AI based etc) you could get some pretty interesting results with something super simple. there is so much to explore (bytebeats, for instance, which you could see as generating both sound and composition).
this is pretty much the concept behind orca and make noise rene, among other things. both deal with a set of 16 preselected notes, it’s the order in which they are played that generates the actual sequences. rene does it by allowing you to control the x/y coordinates within a 4x4 block, orca combines 4 tracks to get a note number, and you control clocking parameters for those 4 tracks.
in both cases you have control over the rules but it doesn’t translate directly into what is played. what i found with orca is that i never really think in terms of “if i change the clock divider for this track to 4 it will change the sequence this way…”, i just change things randomly until i get something that sounds good. going back to the OP what is the degree of my participation here?
as i spent more time i did notice how certain changes had a certain effect on the melody produced, so i do have some control - but it feels more like trying to steer it a certain way without having full control. it’s an interesting way to interact with a sequencer, as you feel certain “stubbornness” that you learn to respect and play along with (as a side note, it can be super interesting to develop a feel for that particular character even if frustrating sometimes).
another interesting example is various conway’s game of life implementations. i did a version that allowed CV control over the rules. cross patching it to itself (and some noise on the inputs…) created some very interesting behaviours, and i noticed the same thing, you could “play” it but it felt like dealing with a living thing that has a mind of its own, so you don’t play it as much as steer it a certain way:
(another interesting aspect of the above for me is using automata but sorta in a granular synthesis way. you don’t generate sequences, you generate these sequence clusters)
another thing i learned from game of life and orca is that often it’s the meta rules that make for a more interesting results musically. apply some rules to the rules.
in a generative system using automata it’s the state itself that is the musical result, no? the string being parsed is the user input (or you could just have a system that transitions without any additional input).
