2 ms·
> The stats would be different if people trusted strangers That's an interesting angle. Trust is the fine line between friend and stranger though. The stats wo
by jstarfish 3y ago
> The stats would be different if people trusted strangers
That's an interesting angle. Trust is the fine line between friend and stranger though. The stats would be different because "acquaintance" no longer has meaning.
> Overly connected is the issue, cause it incentivizes bad things for popularity.
Yeah, I thought about it some more and shit really went downhill once FB replaced the timeline with the algorithm. That surfaced the worst behavior for attention. You're right about extended networks.
- hot_gril 3y agohttps://ncase.me/crowds/ https://ncase.me/crowds/ is what got me thinking about this originally, supposedly based on prior research. It's a bit long and dramatic, so to summarize... They mathematically modeled complex ideas spreading only if they pass a threshold, % of people you know who hold them. Too many connections creates a noise problem for complex ideas, but allows simple ideas to thrive. Simple ideas aren't always bad, but they tend to be, like rage. Too few is also an issue, so there's a sweet spot. Obviously not proof of anything cause it's only a model, but I agree with the theory. Yes FB did go down the toilet with the algo feed, maybe Twitter too. So this is how I judge social networks now.