Also look into the current state of play to earn games [1] and some NFT auction methods where people are selling their participation quite openly [2]
It’s no longer a Sybil attack if it is real people who value their moral “vote” less than the market rate of that vote. It’s just an open market encouraging wealthy participants to on-board buyable users.
[1] https://restofworld.org/2021/axie-infinity/ [2] https://www.coindesk.com/layer2/2022/04/29/please-dont-buy-a...
As for people selling their vote, this occurs even without costs in the form of vote brigading, and seems pretty controllable on e.g. reddit.
Right now the cheapest option is a bot. A perfect anti-bot solution shifts this so that the cheapest option is the global poor.
Zoom out, is the problem is manipulation not bots. Bots are just the best weapon today. Does this new scheme present a sustainable solution to manipulation?
I don’t believe it does. over time, power and wealth will be more able to pay the per vote cost and raising it will disproportionately deny real peoples voice.
Looping back, this is why I asked if any of it was needed once we assume the presence of a working moderation scheme (which you credit as the protection today from vote buying).
I think the key is to have individual sets where each user has particular people whose upvotes and downvotes count for more and use that to filter content. This would be transparent at least to the users involved so there’s no black box algorithm.
It kind of reminds me of overlaying vectors and for any given user there would be particular vectors to overlay to rate every thing according to its value to the user.