The TikTok ban and Donald Trump's rise to power show how fragile our social media accounts are. We must normalize and invest in decentralized social media.
I don’t believe the transitive principle of trust that you cite is all that workable, unless it can be done at a finer granularity.
In my own case, I (A) trust B and C. But B doesn’t trust C, for reasons that have conditioned my relationships with both B and C so that I can still trust them. The reason for that is that trust is multifactorial: A can trust B for some things, not others. So what we’re trying to model is an ontological relation, not just a directed acyclic graph.
Based on that, the best we can probably achieve is being able to set the degrees of separation of delegated trust (maybe 2 hops and that’s all in my case), and add the ability to subclass or otherwise tweak someone else’s blocklist (say, B’s a fine person but habitually forwards Joe Rogan crap that I find to be nothing but vexatious noise), or C despises my favorite band but is otherwise quite sound, etc.
I don’t believe the transitive principle of trust that you cite is all that workable, unless it can be done at a finer granularity.
In my own case, I (A) trust B and C. But B doesn’t trust C, for reasons that have conditioned my relationships with both B and C so that I can still trust them. The reason for that is that trust is multifactorial: A can trust B for some things, not others. So what we’re trying to model is an ontological relation, not just a directed acyclic graph.
Based on that, the best we can probably achieve is being able to set the degrees of separation of delegated trust (maybe 2 hops and that’s all in my case), and add the ability to subclass or otherwise tweak someone else’s blocklist (say, B’s a fine person but habitually forwards Joe Rogan crap that I find to be nothing but vexatious noise), or C despises my favorite band but is otherwise quite sound, etc.