Back on topic, some food for thought.
Yes, I had too much time recently and wanted to learn more about DT groups and how they function.I understand the idea of proposed changes to DT is for the benefit of decentralization of this group, and while I
don't believe hierarchical structures can become decentralized, I think they can become more
accountable. Arguably, this would help make the default trust group more decentralized to some extent, as it would be accountable from the ground up, as opposed to predominantly horizontally.
For the purposes of the following, the ratios are mere examples and not proportionate, this is more theoretical mathematics than practical.I'm wondering how a scenario would look if to be in DT1 you additionally require;
- A minimum of <1:2 ratio of inclusions from DT1 verses exclusions from DT2.
- This could go further and require <1:3 ratio of DT1 inclusions verses DT3 exclusions
- And so on, <1:4 ratio of DT1 inclusions verses DT4 exclusions
Or alternatively, continue with <X:X ratios, so that to be in any of the DT groups, you require less than X:X ratio of inclusions from your DT.X group verses exclusions from DT.X(-1). Example: to be in DT2 you require <1:2 inclusions from DT2 verses exclusions from DT3. Note how DT2-4 would be less determined by DT1 in this scenario, and therefore much more experimental and unpredictable, as it would empower an exponential amount of members in 2-4 trust groups. I'm not recommending this at the moment, but more as as a direction to follow for the future, in the spirit of decentralized accountability.
In summary: if someone is
proportionally un-trusted twice as a much by the trust group "below" them, as they are trusted by their own DT group, they would be demoted from that DT group. Although these ratios presumably would be better applied based on the ratios of numbers of members in DT1,2,3,etc. For example, if there is 4x more members in DT2 than DT1, then the ratio would be better set to 1:4 (proportionate % based ratios). I hope that makes some vague sense, without doing the calculations I have no idea what this would look like, but you get the idea:
lower DT groups make the higher DT groups accountable.
If the DT problem
isn't based on the majority of DT1 being untrustworthy but a handful of "bad apples", which I believe is probably the case, then this could remove these bad actors. I'm certain there would be some trusted DT1 members that would be maliciously removed by DT2 as a result, but it's a small sacrifice for the benefit of accountability in my opinion. Ideally, malicious DT2 actors can then be held accountable by DT3 members, and so on. I recognize there would be a need for enough members to have custom trust lists for this to work, and that probably these ideas are still too premature to implement, but there you go.
