It would be interesting to record your interactions with the models, and then construct secondary models based on those interactions - typically in AI that's where data starts to fall apart, but in art that's often exactly where we find the gems...
That's certainly an interesting take. What would you expect to find, maybe individual interaction patterns that differ from artist to artist? Do you have examples where this has been done, e.g. with symbolic data from behavioral patterns? Personally I somewhat appreciate that dialogue like process when interacting or performing with the models. Sparks more and more ideas while doing so - more like a jam session than deterministic composition.
woooooooooooooow!
I don't know what you are doing or talking about but the eery AI music sounds good.
It would be interesting to record your interactions with the models, and then construct secondary models based on those interactions - typically in AI that's where data starts to fall apart, but in art that's often exactly where we find the gems...
That's certainly an interesting take. What would you expect to find, maybe individual interaction patterns that differ from artist to artist? Do you have examples where this has been done, e.g. with symbolic data from behavioral patterns?
Personally I somewhat appreciate that dialogue like process when interacting or performing with the models. Sparks more and more ideas while doing so - more like a jam session than deterministic composition.
self made synth ?
In a way, yes. These are created by training neural nets with audio material.