I think what makes a story “science” fiction as opposed to the other kinds is that it includes something that doesn’t exist but could. As in, the idea invented by the author only has to be the tiniest bit plausible on the spectrum of things that we believe, to the best of our knowledge and positivism, to be possible. Science fiction is stories about everything that we as a society haven’t yet ruled out as definitively “not real.” That’s why straight-up magic doesn’t count, but wormholes and hive minds do. That’s why dragons from Earth are pure fantasy (until someone discovers one) and resurrected dinosaurs are essentially just bad science (highly disappointing), but aliens from space are the uncontested charismatic megafauna of the genre.

Since there’s no such thing as literary phylogenetics, though, we’ve settled for a taxonomy based mostly on morphology – that is, content. But it seems to me like the category of science fiction generously expands to include any story that is labeled as such, which makes the definition hard to pin down. If box didn’t start with strict inclusion criteria, and the label is a semi-arbitrary designation anyway, then we really only care about the box and the label when we’re arguing about it.

Science fiction is still just stories, though, and stories are about human emotion, and ethics, and aesthetics. We don’t want to do logical thought experiments with the author; we want to feel something. If science fiction stories explore possibilities, the generative potential is in the differences between the world of the story and reality as we understand it. In the space of that artificial mismatch, we get to explore how we relate to one another, how we relate to the unfamiliar, the things we find amusing or brilliant or terrifying. We find new ways to tell each other about what matters to us.