i understand pnk is used so jurors can stake pnk so they can get slashed if they behave wrong and for sybil resistance. i am just wondering if its not more efficient if the cooperative just creates membership tokens and the cooperative becomes a membership dao where every member has one vote. this would be the kleros republic dao. all transparent decision making of the coop dao. users could stake any token to become a juror. all protocol decisions move to the kleros republic dao where members vote or delegate their power. to protect the court from an overtake of a big money whale the general court is only occupied by kleros republic citizens. isn’t this system much more robust against 51% attacks and also more fair?
1 Like
Kleros was made before we had Proof Of Humanity (https://v2.proofofhumanity.id/), so it wasn’t really a possibility.
@William is actively studying systems which would work as a mix of PNK staking and proof of personhood.
thats not my question. i don’t understand why you would need pnk if you have a certain set up like the one i mentioned where you dont need pnk as a sybil resistance mechanism for the protocol. my question is how could you create a more fair and robust system which doesn’t need pnk. i would like to understand why my set up is weaker compared to the current set up with pnk as the sybil resistant tool.
1 Like
We have done research that shows that a typically more secure approach than either using just PNK or using just Proof of Humanity for Sybil resistance is to use both, see the first half of this article. In the model discussed in that article, your odds of being drawn for any given case scale with the amount of PNK you stake, but you can only be selected for one vote.
Neither of these systems offers ideal, perfect Sybil resistance. Attacks that try to get disproportionate weight in a PoH-style proof-of-personhood system might look like trying to pay people that aren’t personally interested in being on PoH to record the submission videos for you so that you control their keys. How vulnerable your proposal would be to that type of attack might depend on exactly how you envision the dao granting membership tokens and how frequently you need to jump through Sybil-resistant hoops (like re-recording the PoH video) to maintain that status; however, it is generally difficult to eliminate this sort of issue. Such an attack has an economic cost for the attacker that has a somewhat different model compared to attacks on pure token-weighted voting. It is convenient that the two defenses seem to be complementary.
Using both PoH and token to defense against Sybil attacks on governance votes is more subtle because the advantages that we have found in how they interact for jury selection depend on how they fit into the random selection process for jurors. The existing research from juror voting would adapt more directly to governance if we had sortition-based goverance. (I am not aware of DAOs that use sortition based governance, but maybe it could be a workalbe idea. Sortition is an approach that has some support among people who advocate for the reform of democractic institutions in general.) I talked about the parallels and differences between how these tools interact in dispute voting versus governance voting in a talk a few months ago - unfortunately I don’t think it was recorded but here are the slides.
2 Likes
thank you very much for your in depth answer. As i understand now pnk provides a layer of security which you would not get without pnk. pnk has unique sybil resistance properties (need more then 50% of pnk to attack, ability to fork out, economic incentives not to dump pnk price). on top of that you can add additional sibyl resistance tools like membership tokens, proof of humanity, coop membership tokens or sbts. makes a lot of sense.
1 Like