We are not dealing with trivial machines, which react according to the always same transformation function, but with self-referential machines, which determine themselves by their own operations, what they start from in the following operation, thus becoming always new machines from moment to moment.
# Thinking of Stories as “Self-Referential Machines”
[…] to deny the idea that works of fiction “represent” life in any sort of straightforward way. They are little self-referential machines that give off energy—but the relation of that energy to the reader’s actual life is mysterious. We feel that a good story “improves” us, or “makes us less lonely”—but the actual pleasure comes from the machine, and that machine is ornery. So in this model, a “dystopian” story is not saying that the author believes the universe is dystopian, but just that she believes she will find out the most interesting things about life by modeling it as being a dystopia. Kind of like: does a tango-composer believe life is a tango? No. He believes that immersion in that form will allow him to produce the most and/or more undeniable energy. That produced energy has something to do with life, and something to do with his vision of it—but that relation isn’t linear.
~
WEINSTEIN, Alexander, 2017. “Oomph-Making Machines”: An Interview with George Saunders. Pleiades: Literature in Context. 2017. Vol. 37, no. 1, p. 66–70. doi 