I remember reading something by, I think, either Asimov or Heinlein talking about how many people think that the utility of science fiction (beyond just transitory entertainment) is to predict what technological changes are coming.

The author disagreed with that, saying that predicting future technological changes is not really important. What is important is predicting the social effects of those changes. That's what a good science fiction writer should be figuring out.

They gave an example that went something like this. Imagine someone writing a science fiction story in say 1890 (although I suppose it might not have been called science fiction yet) set 100 years in the future in the United States.

Predicting that people in 1990 would have horseless carriages instead of horses, and they would be widespread enough for nearly everyone to have one is interesting, but according to author not really insightful.

The useful and insightful prediction would be for the writer to figure out, say, what affect that mobility would have on teen dating. In 1890, most of a teen's social life would involve other teens who lived within easy walking distance plus maybe some farther away that they met at school and either only socialize at school or at activities that are close enough to be in reasonably walking distance for both.

But in the future, when families have more than one horseless carriage and that teen has easy access to one, their social life can involve anyone in town, or in neighboring towns.

Also that 1890 teen that has to walk everywhere might have limited opportunity to get of parental supervision long enough to be alone with their boy/girlfriend. 1990 teens with their much greater mobility can much more easily get away from prying parental eyes and get up to things that would be scandalous in 1890.