Meeting - 21st June 2022
Meeting 1 - Fund 8 - dRep White Paper Working Group Meeting
Last updated
Meeting 1 - Fund 8 - dRep White Paper Working Group Meeting
Last updated
Aharon Porath
Consenz
Frank Albanese
Kenric Nelson
Philip Lazos
Steph Macurdy
Stephen Whitenstall
Thorsten Pottebaum
George Ramayya
https://quality-assurance-dao.gitbook.io/community-governance-oversight/
Documentation funded by Fund 7 - QA-DAO Transcription Service
https://github.com/Catalyst-Auditing/Community-Governance-Oversight-Coordination/issues/71
F8 Proposal : https://cardano.ideascale.com/c/idea/398225
Duration - 1st June 2022 End - 30th September 2022
Brief summary of CGO F8 proposal, budget and white paper deliverable Tasked to examine changes in Project Catalyst Governance. Previous F7 Lit Review Editor roles Lead Authors
00:02 - F8 Community Governance Oversight proposal - Stephen Whitenstall
The F8 Community Governance Oversight proposal intent is to examine changes in Project Catalyst Governance. This includes Challenge Setting, Catalyst Circle Problem Sensing, Parameter Changes. And what we're doing here, oversight of D reps.
01:04 - White Paper Overview - Kenric Nelson
In the last Fund 7 part of this we completed a literature review and now we're moving forward with a white paper.
The basic way that the team is structured is around three roles. Stephen, myself (Kenric) and Thorston are going to act as editors of the white paper. Then there is four or five people who have stepped up to be a lead author for a section of the white paper.
Once we have progressed further I plan to have some community input to review the paper before we finalize it.
Updates from each participant - All
What have you been up to?
02:13 - Introductions
02:47 - Aharon Porath - Consenz
I am the founder of Consenz.
Which is another funded proposal from Project Catalyst. During Fund eight, we submitted and had approved a proposal for creating research in the accelerated digital identity challenge to build a different framework for governance and voting.
We named this proposal DID for Democracy on the Blockchain (Ideascale link).
We are also working with the CA and the vCA communities in order to provide them a tool for creating guidelines. And we are starting to research how we can use the digital identity technology to create a variety of voting mechanisms on our platform.
Right now the main role of voting in Cardano is on proposals and for the use of the Treasury. We hope that when the first version of Consenz platform is ready we can it to explore different kinds of influence in governance, like a creating a document that can act as the code of conduct or constitution.
By integrating principles of liquid democracy and other systems we can gain experience and really see how voting mechanisms work in real use before we implement them in the Catalyst process.
05:11 - Frank Albanese (Snapbrillia)
I am Frank Albanese currently CSO at Snapbrilla. We had two proposals in Fund 8. Around our Mentor-mentee bounty network (Mentor-mentee Bounties & eLearning) , similar to gitcoin but for Cardano, and then our quadratic voting and funding research work (OpenSource Quadratic Voting Funding). We have a Haskell Plutus team of around 25 engineers right now. And we're working through architecting systems that help with Sybil attack defense mechanisms.
With our with SSI (Self Sovereign Identity) research we will hopefully provide users with a proof of reputation platform. Starting with Hyperledger Aries integration (ATALA Prism ZKP + Hyperledger Aries) then with Atala prism working on the Roots wallet and RootsID teams and Tony Rose and the team from IOG around Prism.
What I am curious about is how to document the way we're looking at the UTXO model.
We were just talking about ways in which there are some attack vectors that are exposed through hyperlocal states, like with the UTXO model. And the way you can accomplish a local global state with either a Djed NFT where there's an issuer and holder model. Or there is this SSI approach with like a CL credential and forming a voting record, and how that can be expanded.
CL credential - CL Signatures for Anonymous Credentials
UTXO model - https://docs.cardano.org/plutus/eutxo-explainer
Hyperledger Aries - https://www.hyperledger.org/use/aries
gitcoin - https://gitcoin.co/
07:03 - Kenric Nelson (Photrek)
I am the founder of Photrek. We have been providing a variety of different insights about how complex systems can help understand governance processes within the Catalyst community. And Photrek has complementary efforts in machine intelligence and also in Environmental Science sensors. And I am really looking forward to working with you all. It's been a pleasure working with Stephen and I think this is going to be an exciting effort.
08:09 - Philip Lazos (IOG)
Hello, I'm a research fellow at IOG. I used to work in academia researching theoretical computer science, games, auctions, incentives and the like. In the last year or so I've been working full time in IOG on the research side on a variety of projects involving consensus or incentives. Thank you.
Joins the meeting later
08:58 - Stephen Whitenstall (QADAO)
QADAO was funded in Catalyst Fund 5. I have 30 years experience in quality assurance, test program management, project management and social enterprise. Most recently, I completed a degree in philosophy and became interested in Catalyst from that perspective.
QADAO provides a lot of project management for several governance projects in Catalyst. And we are also quite well known for documentation. Picking up on what Frank raised about UTXO documentation - QADAO is concerned with how to communicate technical subjects in a way that is easily understandable, to help onboard people and to understand the risks and the quality issues involved with that. Thank you. We have quite an exciting group of people here.
09:57 - Thorsten Pottebaum (adanamics)
I have been part of Catalyst since last year and have been involved in a number of topics. Lately working with Stephen W on audit and also on governance. I was also part of the team who did the DAO literature review (Fund 7 - Communty Governance Oversight). In my day to day life I work as an audit program manager for an international med tech company. And I have a postgraduate degree in Computer System Security.
Delegated Representation in Cardano’s Catalyst Governance
Zotero library - https://www.zotero.org/groups/4305236/dao_governance
11:39 - Kenric
As I mentioned the way we're organizing the team is that myself, Thorsten and Stephen will act as editors of the whole white paper. And then each of you lead a section.
There are some natural synergies so as we go forward we may change the arrangement a little bit. There also can be some overlap in terms of how we work together.
Presently the white paper has an introduction written during our literature review. For the moment I'm keeping that. but that's likely to change. It includes a little diagram of what liquid democracy is.
The reason we're focused on this is because it is foundation for what is proposed as the delegated representative process for Catalyst. We are viewing our mission to identify what measurements need to be done to understand the integrity, dynamics and quality of the community's representation.
A significant part of our work has been getting a deeper understanding of what's called the Banzhaf power index.
This is a way of measuring the probability that an individual in a voting system has the ability to change the outcome of a decision. It is a kind of combinatorial analysis about how often you are a decisive voter versus someone who's not influential in the outcome.
The two extremes are, if your probability is zero that means you're what's called a dummy vote, and you're not actually influencing any of the decisions.
And then the other extreme is if your probability is one, then you're a dictator and you are actually making the choice for the committee about all of the decisions.
Steph Macurdy has worked with me on understanding a paper about Banzchaf power index voting power in delegated Rep voting systems ("Power in Liquid Democracy"). And he has also identified some Wolfram tools that have already been built to do this measurement.
Philip Lazos is the co author of an important paper that IOG published at the beginning of the year on blockchain governance. An overview lots of different systems (see paper and analysis).
The the editors can also be co authors on some of the sections if they are planning to dig into a particular section and contribute writing to it.
A couple of you are already preparing contributions to a section and then also working together to build from that and put an independent white paper together on that topic. So Aharon Porath, is one case of that.
Finally, Frank and the Snapbrilla team are particularly in interested what identity security is required for blockchain governance. Particularly once you move beyond one coin, one vote and require security about the identity of who is voting. Where Snapbrilla are building a quadratic voting funding system the issue of preventing Sybil attacks is also quite important.
Discussion & commitments
18:03 - Kenric
Each section will have a topic - owner of the topic
Provisional commitments for each section.
Explain the pros and cons of delegated representation
Current literature; application to Catalyst
Show increases relative power of smaller stakeholders
18:21 - Philip Lazos
The length of the section will depend on available time. Given the title, "The role of Delegated Representation in Blockchain Governance" a couple pages explaining why delegated representation is useful, particularly for Project Catalyst. This is abstract enough to be applied to many settings that face similar challenges. So something of the length of a few pages and at a high level. I think that is possible but I don't know how much we want to expand and what the scope of it should be.
19:13 - Kenric
I do want to be sensitive to this being of limited scope. We can always do more in another funding round if there is community interest and support for that. But as you described, if you can write two to three pages.
If we go back to the original title, "Delegated representation in Cardano's Catalyst governance", and actually I think we're going to add a subtitle here to emphasize where the proper focus is, which is "What I aspects of delegated representation needs to be measured ?"
This is so the community has some understanding about how well the process is actually working. The introduction can be descriptive of delegated representation. But it would good to go beyond just being descriptive.
And also apart from voting power what other properties of delegated representation need to be measured in order to assure quality and integrity in the process ? This might come from your experience and knowledge in game theory, or some some other aspect of your work in governance systems.
21:12 - Philip Lazos
Something that could make sense as a section would be to explain the pros and cons of delegated voting and liquid democracy. There is a bunch of papers in literature review that paint a negative picture of it. But this is under a very specific lens. And I could go into the most current literature. Consider how this might apply to Project Catalyst and why it probably does not. Hopefully this would be a way to show that this [delegated voting] actually increases the relative power of smaller stakeholders rather than the opposite, which is the claimed result in some settings.
22:10 - Stephen Whitenstall
Just picking up on Philip's point, more generally if we take the whole thing as like a collection of different papers are we saying then two to three pages ? In terms of being clear about people's expectations here.
Practically, as we go through a series of meetings, do we return each meeting to review our progress on the papers ? Or do we review at the next meeting, for example, the level of detail that is required ?
23:01 - Kenric
Each section should be somewhere between two to five pages, depending on the scope that someone wants to dig into this. Each lead author is bringing a capability to the table, both from their personal experience and the company that they represent.
As we flesh these out it may not necessarily a good idea for just one person to write a section by themselves. They will have editorial support for sure and there will be reviewers at a later stage. But there might also be some overlap.
For instance, both Steph and Phillip are coming at this from the point of view of what Game Theory or Statistical Analysis can we bring to understanding how delegated representation works and what its pros and cons are. So they may decide to team up and be co authors. And it is fine to form teams. Some people may end up not having the scope for a whole chapter and we will have to adapt to that. But hopefully, we can have four to five sections.
Author(s): Aharon Porath, Kenric Nelson
Consenz is documented oriented
Which voting/governance method is relevant
Low / Medium / High level agreements -
Different voting mechanism as building blocks for versatile governance
25:17 - Introduction - Kenric
I would emphasize that this paper is focused on the delegated representation. You obviously have broader interests in terms of what consensus is trying to achieve. So within that context [of delegated representation] can you describe some of the things that are of interest to you ?
26:12 - Screenshare - Aharon
Because Consenz is an agreement building platform and agreement in this case is a document. So we can divide those governance documents to different different classifications.
And I try to think which voting or governance method is the most suitable for the different levels and classifications.
For the low level one coin, one vote is more suitable. But for the mid and high level, I consider direct democracy, liquid and representative democracy to be more applicable.
So I will focus on these kinds of agreements as documents in the chapter we would write for this white paper.
To give some very high level view of different kinds of governance needs or representative requirements
28:13 - Kenric
Thank you, Aaron, it very helpful to have a diagram like this. And if I can recap, what you are describing is you have kind of different levels of agreement that have to be made in your own process. And you're thinking that the high level agreements are the ones that have the most similarity with the Catalyst delegated representation process. Is that Is that fair ?
28:43 - Aharon
It is more like a suggested process for a DAO in Catalyst in this case, that Consenz can can provide.
29:04 - Stephen
If I could just ask a question about scope. Would this represent a paper on how dReps should be documented ? Or are you proposing to use this procedure to document our white paper ?
29:42 - Aharon
I think that is a third option of how directly will we [Consenz] participate in governance that is outside of the scope of funding
30:11 - Kenric
So the focus here could be applications of delegated representation beyond proposal funding. The constraint is to make sure that your section fits within the focus of how we are going to measure the properties of delegated representation.
For instance, one measurement could be how to facilitate the function of governance to produce reliable documents. That people trust, and feel they have a voice in forming.
Last time we had brainstorming about inputs and outputs. So I am going to add the ability to generate consensus documents. Such as a constitution, or guidelines to this as an output.
Author(s): Steph Macurdy
Banzoff Index - major index; Shapley index
Measuring voting power within Catalyst
Demo/Visualization within Wolfram Language
Author will adapt for Catalyst data
Query Catalyst data into Banzoff function
32:18 - Introduction - Kenric
32:53 - Steph
I have been doing my own research on the Banzhaf power index which is the major index. There is also a Shapley index as well, which I haven't looked much into. The Banzhaf power index is where we want to begin in terms of measuring voting power within Catalyst.
I came across a demonstration and a visualization for the Banzhaf index and calculating it inside Wolfram language.
And so I met with the author along with Kenric and talked it out. The author has been working on a function which can be fed Catalyst data and spit out Banzhaf power indices for specific entities.
So the work now is to figure out how to query and group what's happening in Catalyst. How to fit Catalyst data into this function in order to spit out what the rankings are and then make sense of it.
So I'm meeting with the Author again next week. And I'd be happy to explain the process of making the Banzhaf index a bit more hands on and workable. What I really need help on is how do we fit what's happening in Catalyst into a structure that we can query the Banzhaf index ? Because some of these things are created to look at a US presidential election. But we have different functionalities. And I don't know if we have all the data.
36:11 - Kenric
If you can do sort of a post analysis of Fund Four and Five data that is probably the most relevant. There is some data for Fund Three, but there was some unusual things that happened in that round.
If we focus on Fund Four and Five and you are able to provide analysis about the Banzhaf index for that round that would be a fantastic input to the kinds of things that are needed and we need to adapt in order to be able to measure this in a delegated representative process
This is relevant to managing scope. For this round, what we want to do is identify what measurements need to be take done, and then provide supporting evidence of why that would be a valuable measurement.
In terms of data the voting power of wallets is the most important thing. And have separate from that the actual outcomes. The Banzhaf index gives you the combinatorial analysis of all the possible outcomes. But as discussed, we're going to have to scope that down a little bit. So we can compare these combinations with what happened those funding rounds.
We have the wallet distributions in terms of the number, the amount of ADA that was in each wallet and we also have the outcomes of the elections. But we don't have what each wallet voted on.
Author(s): Frank Albanese, Kenric Nelson
Methods
Democratic - one person - one vote
Plutocratic - one coin - one vote
Meritocratic - reputation system
Attack vectors for each method
Reputations; identity;
SSI & DID technologies
40:18 - Introduction - Kenric
40:33 - Frank
I am really enjoying the ideas being shared here. I will add that if you need assistance with querying, you know, a test net. We're pretty proficient now in Blockfrost. And we're consuming the Dandelion solution through gimbalabs for querying this kind of data. Happy to collaborate on the technical side with the information gathering if need be.
The way I view this section is broken down into three categories. And Aaron earlier actually had a slide that mentioned some details on these three categories.
It's democratic, where it is one person per vote, plutocratic, where it is one coin per vote, and then meritocratic, where it is a reputation point per vote.
And within each of these categories, it might be helpful to view it through three different lenses from a prospective identity, reputation, and then technical systems design.
And also to include like quadratic voting systems, for instance, is one of those lines of thought.
And then within each of these lenses, per each three categories, we can break down the security measures and the different attack vectors that bad actors can pursue, and the defense mechanisms for each that we can document with recent research and going into the future.
Where systems are one person per vote, you get into forms of identity.
Verification, starts simply, as per Catalyst Circle in person election Zoom calls, all the way up the spectrum to like deeper forms of KYC (Know Your Customer) auth processes.
With plutocratic voting, you have one coin, one vote. There are many ways to game a system like this. And we can document this from a technical perspective through the lens of reputation and identity. We can get into the issuer holder model through Hyperledger aries and then Atala prism.
And when it comes to meritocratic voting, how do you game a reputation system ? We can talk about attack vectors for different SSI paradigms. That includes the reputation and verifiability of issuers themselves. And how do you defend against bad actors becoming issuers in a kind of a complex attack vector ? So I think these kind of three topics, through these three lenses, could provide some meaningful research.
Two to five pages will be a challenge. So we will condense our our most impactful thoughts into two to five pages. If that is the goal here ?
43:46 - Kenric
Yeah, I think that's best for this white paper given how it has to fit into the others. But then again, for each of you, this can be a launching point for an independent white paper that expands upon what you're discussing here.
44:06 - Stephen W
I think part of what we're doing here in a way is also governance, because this is part of the community governance oversight proposal. This is a proof of concept that people from different parts of the community can work together on this, and hopefully we can. It is very interesting material.
So it'd be like a taster. And hopefully we can then feed that into a Catalyst Town Hall presentation or something and just say, Look, this is what we've done. This is what we've covered, so it can also be a showcase for collaboration, and a showcase for each of our individual contributions.
Calendar invites Zoom and Google invite with Agenda links 21st June 2022 - 1700 UTC every two weeks until end of September 2022
After Town Hall to launch and recruit reviewers of White paper - July 2022
Please prepare Detailed Outline of your section for next meeting - All
Create a Different section in Zotero for F8 - Thorsten
Feedback Cooldown
adanamics ()