I'll have to think about this one. Question is how to confirm real users of course.
Somebody controls 10000000 accounts. They can just upvote one account by another via scripts. Permutations.
Moderation of course doesn't scale, and moderators may dislike some real users but honest users (you like cats, he likes dogs, etc) and each other, leading to weird decisions.
Also new users still have to receive some delegation in this case to begin using the accounts.
This might be combined with second price auction bids to keep increasing delegations.
And bids get redistributed. Supporting valid delegations and accounts to flag phishing and spear phising and other clearly bad actors.
I was thinking they can bid for higher delegation, if they produced content that was upvoted. This would allow consensus to set the price for delegation, and rings without content would suffer attrition about delegated to accounts in the process. Can't as easily be gamed.
But increasing over time delegations on top of that, for a few months, except in certain flagging conditions might be a something to try, to experiment with.
What do you think?
This above is good. Was thinking about such things for academic publishing as well. Check out the last part of my post from a week ago, the channel update one for my subscribers. It describes a way to generate group specific tokens without further encryption / decryption and that can be used to give delegation within a group and by exchange rates between such subtokens for better or worse groups convert to tokens usable in other groups.
This can boost new accounts producing good content. Combined with some sort of second price auction mechanism for delegation this might do the trick.