This is basic idea for “pseudo” reputation system allowing PoH verified humans to add/subtract reputation points to other address by burning small amount of UBI. Gas cost and UBI burn will make sure this mechanism isn’t misued. Kill 2 birds with 1 stone.
- I’m aware of worst case scenarios straight out of
Black Mirror, Nosedive and The Orville, Majority Rule… but only way to find out how this works in reality is to Buidl it!
// Pseudo code
Human {
uint red; // -ve in votes
uint green; // +ve in votes
uint blue; // +ve out votes
uint yellow; // -ve out votes
}
// vote a human with +ve UBI reputation
function Green(address human, uint _ubi, /*bytes memory _comment*/) external {
// require _ubi < 1% of human balance
// human : (balance + √_ubi) && (green+_ubi)
// msg.sender : (balance - _ubi) && (blue+_ubi)
// emit UpVote event (from, to, _ubi, _comment)
}
// vote a human with -ve UBI reputation
function Red(address human, uint _ubi, /*bytes memory _comment*/) external {
// require _ubi < 1% of human balance
// human : (balance - √_ubi) and gets (red+_ubi)
// msg.sender : (balance - _ubi) & gets (yellow+_ubi)
// emit DownVote event (from, to, _ubi, _comment)
}
-
1% limit can be changed to fixed 1 UBI per vote or a max UBI value % set by governance.
-
PoH address can punish or reward reputation points to other PoH by burning small amount of UBI and paying gas cost.
-
Each verified PoH address can trigger one reputation vote per day / based on timer set by governance.
-
It’s number goes up/down meme in fancy way. PoH list can sort humans based on this 4 colors value. It’s fun reputation game like twitter followers/likes, nothing serious…
-
If gas price goes down / L2 scenarios, 1% burn and √burn value should reflect that to stop possible gamification of reputation system.
2 Likes
I think it would work better if the reputation impact would be quadratic (like burn 1 UBI for +1, burn 4 for +2, etc). This way, a user a lot of persons have trusted (burned a few UBI for) would have a better rank than a user having one whale burn a lot of UBI for it.
To avoid dystopia I think we should only rate business entities and not “humans” themselves. In some cases the human and business entity can be the same like a ebay-like seller with his real name, but it shouldn’t be mandatory and we should separate contexts (maybe someone is a very poor service provider but a very good party beast ^^).
By only rating business entities, the system would also be opt-in which prevents again against nosedrive-like dystopia.
1 Like
It’s actually for PoH/verified human “reputation” backed by burning their accrued UBI. Any EOAs/DAOs can join in if they want to burn their UBI in pseudo reputation game.
Limiting 1 UBI vote per tx, 3 outgoing votes per day/per address and 24 incoming votes per day/per addres – worst case would be 8 senders burning 24 UBI to trigger √24 UBI burn from receiver balance.
That’s on top of tx gas cost for senders. I think it’ll be used more for positive burns than negative burns. Negative side isn’t really required but it serves as tiny √quadratic “threat” model in this reputation game.
Err, it’s better to wrap this game in crypto memes instead of positive/negative, red green…
…
IDK, if this is good enough crypto Zen meme.
PoH governance can settle final values/limits so it’s more +ve and fun for everyone to burn UBI.
1 Like
Well if we use a quadratic system, we have no need to place limits on amount per TX, amount of outgoing TX and incoming TX.
We would still need to have an aggregation period (maybe 1 month) such that if you burn 4 times 1 UBI at a short interval it only counts as 2 points.
1 Like
I think that’s doable, I’ll try to scratch around solidity for minimalist logic & gas optimization…
I’m not a member of the Proof of Humanity DAO, but I am curious if there was any development on this quadratic approach. I am developing such an approach in my current white paper within the context of group currencies and internal UBIs.