Why not? It seems like some kind of categorization system (even a really naïve keyword-based one) is already used for things like advertisements. If such a system was surfaced, and you can filter out specific categories, you are then able to see what the hypothetical racist $relation posts about the family, without having to see their hot political takes.
Remember that we are talking about words on the screen 99% of the time. I would be willing to bet most of us are reading on a post by post basis and don't know the full spectrum of every given persons beliefs (in fact, I'm not sure this is healthy or desirable).
Remember that we are talking about words on the screen 99% of the time. I would be willing to bet most of us are reading on a post by post basis and don't know the full spectrum of every given persons beliefs (in fact, I'm not sure this is healthy or desirable).