>
I don't think you can really decide whether something is a useful signal or not. I don't and can't assume anything about an individual because they follow Evil McBaddington.Maybe I misunderstood your comment, but isn’t that exactly the opposite of what you’re saying here?
> Who a person follows is a pretty strong signal of their values
If your position is closer to “on average, people who follow other people tend to be more aligned with the people they follow”, I’d probably agree. But this is a vastly different statement than “following someone strongly indicates endorsement”.
> I see this "scale invariant fallacy" all the time with programmers.
I think it’s problematic to bring too much of an engineering mindset to a social issue like this. Algorithms can be cleanly and rigidly defined. Social issues tend to be messy and complicated.
In terms of pure utility, it may be true that blocking people based on who they follow results in a clean feed at a low (effort) cost. But perpetuating the idea that following someone equals endorsement has longer term social costs and gets into dystopian territory pretty quickly, and depending on why someone is using social media, the impact of entirely excluding the group of people that “follow but disagree” is likely not small. If you’re on there for cat memes, probably not a big deal. If you’re trying to stay informed or be politically active, probably a bigger deal.
And regardless of use case, believing things about other people reflexively based on something like follow status is not a healthy place for society to be. Most people do not appreciate the nuance of what you’re describing as “scale invariant fallacy”.
My biggest issue here is the framing and the questionable/harmful conclusions that can be derived from it.