{
  "metadata": {
    "transaction_key": null,
    "request_id": "metagov:on-neural-quorum-governance-liu",
    "sha256": null,
    "created": "2025-10-27T23:38:02.154400+00:00",
    "duration": null,
    "channels": 1,
    "models": [
      "metagov-manual"
    ],
    "model_info": {
      "metagov-manual": {
        "name": "metagov-manual",
        "version": "2025-10-01",
        "arch": "manual"
      }
    },
    "warnings": null,
    "summary_info": null
  },
  "results": {
    "channels": [],
    "utterances": [
      {
        "speaker": "Speaker 1",
        "start": 0.0,
        "end": 0.0,
        "transcript": "Welcome everyone to weekly MediGov seminar. My name is Val. I take she or they pronouns, and I'm one of, two MediGov community leads, Senti here with me as well. And today, I'm excited to be introducing our speaker for today's session. So today, we have Anka Liu, who's gonna be joining us or who is here to talk about neural quorum governance. And so I could give you more information, but I'm gonna let Anka take it away and start her presentation. So welcome. Thank you all for being here, and let's enjoy. We'll have about 15, twenty maybe of presentation, and then we'll do the rest q and a and discussion. So buckle up. Exciting stuff. Let's get into it."
      },
      {
        "speaker": "Speaker 2",
        "start": 15.0,
        "end": 15.0,
        "transcript": "I gotta gotta mute in order to say some things. Yeah. I might actually, like, like, do a little bit more presentation, but anyone, is actually, like, able to, ask any questions at any time. I'll try to, like, keep a good eye on the chat. So, it's a little bit of a a meaty meaty topic. So, like, if anyone has any questions, feel free to just, like, send something in a chat or just unmute yourself, and I'm happy to kinda answer on the spot. So don't feel shy. Great. I'll get started. Well, thanks so much, Sante and, Sante and, Val for that awesome introduction. Excited to see so many folks, people I know, but also some new folks here, so definitely don't be shy to ask questions. My name is Anka Lou. I'm the ecosystem growth lead at the Stellar Development Foundation, which is a nonprofit organization that supports, development and growth of Stellar. At the foundation, I, lead our Ecosystem Growth team, which manages the Stellar Community Fund, which is actually the first use case that is implemented in your own governance. So neural chrome governance, in short, it's aimed to set the new governance standard that prioritizes community alignment, participation, adaptability, and aims to effectively address common hard to solve issues in that web three and beyond mechanisms are experiencing. So today, I'd like to take you on a bit of a journey. So I'll start a bit with the overall introduction to Stellar in the Stellar Community Fund followed by an understanding of the entire, like, design process and reasoning behind creating the mechanism. Then we'll dive a bit into what the definition is of neurocoron governance and how the mechanism works and how it's also implemented in the community fund, and finally, dive into some of our learnings and next steps and how you can get involved. Alright. Let's start with the basics. So is anyone here familiar with Stellar by chance? Some people. Alright. Well, I'll I'll I'll go over it briefly. So Stellar is an open source decentralized layer one blockchain network. It's been around since 02/2015, actually before Ethereum, and it's specifically designed to power applications and services creating equitable access to the global financial system. Stellar provides infrastructure developers can use to issue assets, build applications, and deploy decentralized financial protocols. Developers can choose to either build a dedicated set of functions available by SDKs in several languages, or we now also have a native Rust based smart contracts platform, both of which allow direct access to the Stellar network. So I haven't updated these operations from about these stats from last year, so pardon, they are a lot higher now. But it's pretty powerful, right, since apps can actually tap into the Stellar extensive on an offer and provide a network to get access to the, world's currency. So for instance, Stellar is actually the largest cash out provider network through collaboration with MoneyGram, which Stellar apps can integrate with to offer end to end solutions to real world problems. Alright. So that was a little bit about Stellar. Now let's switch to the community fund, which actually powers more of the ecosystem growth side of the Stellar ecosystem. Neurocrone governance has also been first implemented in Stellar Community Fund, and this is a, fast paced award program that leverages community input to fund and support teams building on Stellar from, ideation to successful launch and growth. We've, like, it the fund has been launched in 02/2019, so it's been around for a while and awarded over 31,000,000 in dollars in excellent over 400 projects. Most of the these funds have actually been distributed since Neurocorner governance was added to the, the, stack. We have a number like, SCF actually number of, project types. So we have, like, financial protocols, education community, infrastructure services, developer tools, and applications. So anything that an ecosystem is necessary to to grow. It's also got a fly one community that votes to determine award allocation, supports projects, and participates in SCS governance to make sure it meets the ecosystem needs. You can see a few of the tiers here. I will mention them later after when I actually talk about our implementation. Alright. Well, let's actually dive into some of the design journey. SCF, has been active, as I mentioned, SAI for five years now. And, in those five years, we've really gotten a lot of community input and also, like, shifting, ecosystem needs, a lot of feedback, and and trying to, like, navigate the issues that are associated with effective award allocation by community input. And so last year, we collaborated with Block Science to I I believe everyone here should probably be, like, introduced to Block Science already. But they are kind of a firm that really, like, does a lot of research to solve complex issues. And we've collaborated with them to design a governance framework that would answer to not only some of the issues experienced with the community fund, but also in the broader web three governance space. And some of these, like, common issues include, like, inaccurate community sentiment. What we see also is a lot of, like, centralization and collusion, especially by whales if there's token based governance involved. Voter fatigue and apathy are still really large, kind of issues in web three governance as well as civil attacks and manipulation. We have a low information voter problem as well and, like, as well as a lack of scalability and adaptability. Oh, Brian, are you not able to see my screen?"
      },
      {
        "speaker": "Speaker 3",
        "start": 30.0,
        "end": 30.0,
        "transcript": "I'm sorry. I can't. I don't know why it wasn't showing me your screen, but now I see it. So my my apologies for that message."
      },
      {
        "speaker": "Speaker 2",
        "start": 45.0,
        "end": 45.0,
        "transcript": "Oh, okay. Good. Okay. I was like, oh, it got a little heart attack there because I thought that people weren't able to see my screen and just looked at my face while I was presenting. Great. Well, good thing that the screen is now it's showing for you as well. So So kind of going back from, I guess, the common problems that we experienced, we started to, like, map out kind of the systems requirements and also desirables. Block science actually uses this following, like, three phased process to design solutions for complex problems, starting with the desirables mapping, followed by the specification design and integration of it, and finally, the the validation of the system by thoroughly testing it and simulating its performance. Starting with phase one, we fully analyzed kind of following, like, desirables of how we wanted funding agent allocation and governance to look like in SCF. Like, it should be effective in determining projects that should receive funding. It should also be fair and reflect community sentiment, reward beneficial behavior. It has to be scalable to avoid overwhelming to, voters, transparent in a way that, that information is publicly available, but also remain individual proto privacy, which are, two very difficult things to combine. I will I will I had, like, explain more about that later in the presentation. But the also, the system should be flex flexible and to be able to adapt. So in in the, like, in five years, the SCF has been around. Like, we have changed the mechanism a lot, and each time we have to kind of update it, in order to, reflect the current kind of community feedback and ecosystem, needs. And finally, we also wanted this mechanism to be a bit innovative and also use Stellar tech Stellar's technical capabilities wherever, possible. This is how I'm able to fund the research. So while exploring some of the different types of, like, voting mechanisms across the space, we found that most governance mechanisms, whether it's for Web three or not, rely on either direct or representative governance. And most of them are weighted by token based governance, which are, by nature, introduced some of the undesired issues that I mentioned before. And as a Stellar community, like, it's primarily actually not a token based community. And also we had the desire to keep the governance process semi open rather than totally open or totally closed. We divided kind of our, like, we actually dived into two areas of interest. One is reputation based governance and liquid democracy. So I'll start with the first. For anyone who is familiar with reputation based governance, it it involves evaluating one's trustworthiness, expertise, contribution history, and also alignment with the initiative's value to influence voting power and as or assign a membership. So the reputation based governance could actually lead to, like, more informed decision making as there's more voting power to experts. It could also be more inclusive as, like, the commute as the governance doesn't need to be limited to token holders or validated, but could include a plethora of different user types. It could drive intrinsic motivation rather than often misaligned extrinsic motives such as monetary gain. And another, another, benefit is that it's quite versatile and can be layered with other mechanisms as well, as well as being kind of, like, intuitive. We kind of make choices based on reputation in our everyday lives without even realizing it. It's definitely, like, challenging to, translate it this intuitive process that we use every day in our brains to actually, in a, kinda smart contract based governance system. So it should require structural manufacturers over might be overwhelmingly complex. And another another issue is that we have to be wary of the influence of certain members who recruit reputation from the start so that they don't remain dominant and unchallenged over time. So another area of interest was the liquid democracy. So instead of, like, having either direct democracy direct voting or representative voting, you actually have a combination of the two to offer, like, the best of both worlds where where people can either choose whether they like to vote directly or delegate their votes. So and as voters are able to delegate to experts, it can also lead to more informed decision making. Also lead to more, like, inclusivity by flexible participation based on availability, expertise, and interest into a particular issue. There's also an availability to to re delegate and in some cases, override the delegates' votes, which hold delegates accountable. I'll dive a bit quicker because I see that I am talking for quite a while. So let's actually dive into then how we arrived at Neurocorp governance. So out of these design processes with block science came Neurocorp governance or NQG for short. It takes up their form of liquid democracy combining both modular voting power attribution, neurogovernance, with often delegation of voting power to trusted groups, quorum delegation. As an open source framework, anyone can use it to design governance systems that prioritize community alignment, participation, and adaptability. So it offers also more sophisticated alternatives to token based governance and has the ability to measure things like reputation, expertise, and trust, in rather than simply relying on token holdings. We with NQG, we really wanted to, create a governance system that resonated with stellar's ethos of creating equitable access to the global financial system, where someone didn't need to buy their way in or be one of the early adopters or always have the time to to contribute. So, yeah, at the end of the day, like, the main goal of any governance system is to get ample and diverse participation for effective decision making. And NQG offers cons customizable design where members can work their way up, gain trust, or also reputation in the ecosystem to have more influence and decision making with flexible options for participation. So let's dive into the individual pieces of it a little bit. And, again, like, feel free to, like, pause if you have any questions at any time. So neuro neurogovernance, which is, like, one of the pieces of neurocognitive governance, is the modeler kind of transparent, flexible architecture for dividing divide defining one's voting power. And this is inspired by neural networks, as the system mimics our brain's kind of information processes through interconnected voting neurons. More of a simplified version of that, but that's the idea. Each of these voting neurons consists of both an oracle function and a weighting function. The first one encodes the core logic of data input, and the second one encodes how it should be weighted. So these outputs from these voting neurons can be aggregated together in layers and then also pass as input to the next layer until a final value is reached. Some of the, kind of, benefits of neurogovernance are that it kinda simplify the structures of evaluations or reputations, break down this intricate process into smaller, more manageable parts, and offer kind of a plug and play experience for adaptive governance. It can also be used to have more inclusive participation in governance as it can also be used to determine membership. In terms of, like, neurons, there's a lot of different types. And depending on, like, if you're doing an on chain or an off chain, reputation rating, anything from, from, like, expertise or trust or even, like, token holdings. Anything can kind of be added as a a neuron, basically. Then quorum delegation is also, like, is one of the more interesting parts, in my opinion, is, like, the neurogovernance has is kind of also based on some of the other types of Oracle waiting functions. With quorum delegation, it allows users to passively vote. Oh, Lydia, I see that you have a question. What model of reputation do you use? I'm going to actually go into that in the implementation section. So wait just a few minutes. As far as quorum delegation, it allows users to passively vote by delegating their voting power to a quorum and, or group of other users. It's inspired with a Stellar consensus protocol, which is a proof of agreement type of consensus mechanisms, where validators actually define their trusted quorum of other validators in order to validate a transaction on the network. So Stellar's consensus protocol consensus mechanism works is based on trust rather than, proof of work or proof of stake. I am and that it's kind of mimicked by this mem mechanism of quorum delegation. So voting power actually goes, to the majority's decision or whatever kind of quorum, like, minimum quorum that you have sent of the chosen delegates that can be changed each voting round. And you can also choose delegate purse decision, so it's quite modular in terms of the implementation here. Without having to kind of reveal information about the quorum, the user can now automatically vote. So this can also reduce attention costs for users while still allowing high flexibility and privacy. Those were very mechanism for a settled community fund. Also, as these quorums are self selected and can be altered with each decision or topic, it could avoid potential collusion issues, that you see with other mechanisms. And as quorum results are determined from the consensus of a self selected group of voters to trust, it allows for a bit of a buffer to protect against nonvoting or rogue voters. Alright. So some of the use cases of of kinda like a Norcom governance can include, like, public goods funding up allocation as NQG has the ability to incorporate eval evaluation criteria beyond profit potential, such as prior network participation or voting history. It can allows the community to easily adjust the weights in those signals. It also has the main expertise representation to improve maybe the quality or, like, kind of, like, the reviews of the voting as well as allowing for adaptive governance. Actually, in in several different ways where neurocovernance can actually accept inputs at each layer, transforming them in weighted outputs of the next layer, and this allows for cross round adaptations. It can also, like, take in results of prior voting action such as project success and outcomes to be an important signal for future voting power allocation, something that we're very interesting in exploring and and integrating into the Stellar Community Fund. So where can you find all of this information? Because, obviously, maybe I just talked for a lot and it kind of all goes one ear in, one ear out. We actually have something called the governance model library that contain the specification, implementation, also tuning guides, and simulations of neurogovernance and quorum delegation. This is created by Block Science and defines all the requirements and inputs and outputs. Probably, like, hover over this quickly. It's just if you go into all this message it in the chat here, it's the and has all the different kind of specifications as you can see here. Alright. I won't go into do into too much depth because I want to finish at some point today. Then, also, like, you can like, it takes quite a bit of time and effort to set up some of these systems, like, to figure out, like, what what are some of the neurons that you'd like to create, figure out the weighting, figure out the parameters on those systems. That's the one thing with neurochrome governance. It can be a bit more complex. So to help anyone else with experimenting with the system, like, LogScience has created open source tool called CADCAD, for anyone to do research and simulate different outcomes and counterfactional analysis, sanity checking on the simulations, and these implementations are readily available today. If you go to this is the other one. Okay. Yeah. Blocksaint's here. This is the the repo where kind of all of the the CADCAD demo is and also, like, some instructions on how you can get it up and running, and these instructions can also be found in the governance modules library. So, yeah, that allows you for doing a number of different things and simulating. And I'll go into, like, what our pilot implementation has done, in a bit. So, Lydia, I'll get to your question. So, yeah, Neurocorp governance has been first implemented in the Stellar Community Fund, which I shared earlier. Little bit on the road map of things, we have started early twenty twenty three last year. And, like, this year, we're we have been really focused on simulating and calibrating, also, like, deploying to Mainnet and getting the contract live and getting the get the contracts also on open sourced. And later this year, I think we're the timeline has a bit shifted since we have been working on the the actual award structure structure a bit more instead of the Erichrome governance. But the, later this year, we'll focus a lot more on ecosystem collaboration and trying to support any other, implementations with, yeah, getting used to the mechanism. So a little bit of an overview here of what the pilot implementation included. So the NQG's implementation of SCF was done through some Stellar smart contracts. There's actually an open source repo if anyone is interested in in exploring more, basically, to balance the user kind of privacy, transparency, and also computational efficiency within the current, constraints of our smart contract system. We currently have employed a hybrid model and, for, for the computational hybrid computational model. And currently, the contract actually, like, a a Gergai is saying results from each neuron to a certain, like, the final voting power for submissions, where computations are performed partly, like, off chain and the result are, uploaded to the contractor by an admin. It's not an ideal solution as we have it right now, but we hope with the coming of a g k features to Estela smart contracts that we can have a fully verifiable and also a system that actually still provides the voter privacy that the community has expressed to require. So if you're interested in kind of diving in deeper, this is a really great blog that I wrote that kind of goes into, like, how the votes are uploaded on the on chain contract, how then the delegations work. Delegation are currently resolved on chain because we are actually connecting the Discord username. So that's the identifiable part of a user to the public keys on Stellar, which are then, you know, publicly viewable. So the connection between those two should remain private. And right now, that's the only way to do it was to keep it off chain. And similar with the neurons, all the details on the contract, how we share it has been there. But but yeah. Okay. Eugene, I see that. I will answer it later. And then and then also, like, you can see some of the insights from from, like, building on on the new smart contract system and then some of the issues that we that we ran into building it on, on it, to actually be able to protect further privacy, enable delegation reputation, and allow trustless verification to have the ideal version, we might need a more of a z k implementation. But let me go back to, to some of the other implementation details. So as I mentioned, as the community is mainly based on merit and reputation rather than token holdings, our initial choice of neurons included number of users that trusted the voter, also voter history and discord tiers. So these are relatively minimum, kind of like neurons. We are actually like, there's a company called a project called Trustful, I think, by Danny, and who is currently working actually on an expertise neuron based on some of the stellar stellar badges that people can earn to as they complete exciting challenges on the Stellar network, and so I'm very excited to see that happening. Currently, we just have these three. These neurons were relatively, like, low weighted to keep the results in control. So individuals couldn't have more than two or 10 times the voting power of of others. Okay. Next slide. Also, like, the current minimum quorum that we required was three active users. We We actually loosened this from some of the initial settings to prevent delegate votes from resulting in abstain. So you can see in one of the rounds recent rounds, we actually realized that that 30% of the votes were delegated, right, of which about 40% of the votes resulted in a stain as the minimum quorum could not be made. And so this is why also, like, the minimum quorum was loosened. But we realized that this was also not enough. So in the next, actually, system, we are introducing category specific delegates panels, which actually allows kind of the nomination of these delegates. And so if people, you know, nominate to be a delegate, they're still like, well, I I am definitely going to be available for these three rounds. And so people should be can then select their quorums from the nominated delegates, which will help us a lot in setting up and making sure that the yeah, that, like, the people who have been delegated to that they actually that those go through and note resolve and abstain. Also, like, I think maybe some of the interesting learnings here were that the NKG kind of the outcome didn't change the clear yes or no votes as opposed to, like, one person, one one vote. You can kind of see in this example of the two two award rounds where you can see here where it's just NKG outcome and then there's one person, one vote. But it did actually offer some alternatives to the the gray areas, the maybes. And so after each round, like the, kind of, the voters and the verified members, we always discuss the voting outcomes before we decide on final award allocation to prevent any anything, like, on uncertain from happening, that people didn't, like, realize before that was, like, not captured. And so far, this the NQG kind of NeuroQom governance outcome has consistently been favored by the community and so these individual findings may suggest that this type of governance may be from better and gray area decision making but further research is needed to prove this out. Yeah. As I mentioned, we also were able to like, because, like, SCF leverages criteria based tiers pretty heavily in governance as I shared earlier. These offer the ability especially the the higher two tiers, you can either vote actively or you can you can, vote you can delegate for your votes. Then we also have, added kind of a lower tier that was, like, more, like, easily attainable that they're only able to delegate and actually disallowed us to double the voter count without negative income, impact on the outcome. So some of the future features will include dynamic voting power based on log logistical curve rather than linear. Right now, like, someone is able to kind of, like, if they just, like, let's say, vote more often or if they get more people to trust them, like, their voting power just goes up in a linear curve. We actually want this to look much more like a logistical curve so that, like, new members to the ecosystem, like, once they kind of reach a certain inflection point, add, like, a more more or less a consistent, I guess, like, voting power as as some of the, some of the longer term members. And this is to kind of in in include be more offer more inclusive kind of governance. We also want to add the ability to override some of the quorum results and that this would make it, like, a true, like, liquid democracy type of system where you can let's say, your your BFR around, you'll be like, okay. I'm going to delegate my votes now. This is my quorum for, let's say, financial protocols. This is my quorum for applications, blah blah blah. And then, actually, you realize that later on, like, oh, so these this was these were the results from these different quorums that I selected. And, actually, like, I don't agree with those results. So you could actually, like, choose to override them. And so that would be the ability to override the"
      },
      {
        "speaker": "Speaker 4",
        "start": 60.0,
        "end": 60.0,
        "transcript": "most essential function of liquid democracy that exists."
      },
      {
        "speaker": "Speaker 2",
        "start": 75.0,
        "end": 75.0,
        "transcript": "Yes. Exactly. So, sorry. I was, like, a little bit shaken when you said that because, like, I was just, like, going on my brain and then, like, I heard someone say something. So, yeah, thank you for for bringing that in. Exactly. Right? And so, like, so, ideally, like, this is the system that should be in place for liquid democracy to, like, truly be added to, like, some system like this. Sure. Then also Sorry."
      },
      {
        "speaker": "Speaker 1",
        "start": 90.0,
        "end": 90.0,
        "transcript": "Can you share the last two presentation so we can kind of follow along? I would love to bounce back now that you've kind of gone into more of the implementation. I wanna now bounce to the beginning of the presentation and kind of get my full head around everything. But, yeah, folks, please feel free to keep sharing questions in the chat. And, yeah, there's no such thing as a dumb question. I have lots of dumb questions coming to mind. So feel free to ask whatever is on your mind. I'll let Anka finish, though, so we can move to the q and a. But how much longer do you think, Anka, do you wanna present?"
      },
      {
        "speaker": "Speaker 2",
        "start": 105.0,
        "end": 105.0,
        "transcript": "A minute two minutes? Yeah."
      },
      {
        "speaker": "Speaker 1",
        "start": 120.0,
        "end": 120.0,
        "transcript": "Cool. Alright. Get ready. Let's move to discussion in two minutes."
      },
      {
        "speaker": "Speaker 2",
        "start": 135.0,
        "end": 135.0,
        "transcript": "Alright. Cool. So, yeah, some of the also experimenting with additional neuron weighting and additional neurons as well as enhancing some of the verifiability and transparency in the system as I mentioned earlier. Great. So, like, we recommend folks to get involved. You can learn more in the SCF handbook, which is kinda like our, like, overall official rules, guidelines about everything about the fund as well as kind of if you look here, I have a specific area about neurocognitive governance that kind of sums everything up and also links to the blogs and updates and also reports. Can can for instance, go into one of the reporting voting reports that I shared earlier as well as updates. Then, also, we recommend people to contribute to the governance models library. There is a it's an open source. So if you actually go to governance, GitHub. Carbonon's. I'm sure I have it somewhere here. Okay. Well, I'll I'll find it exactly in a GitHub link. It should be somewhere in the in kind of this module overview, but I'll find it and then add it here in the chat. But we recommend to, like, add more kind of, like, features and also, like, different types. It doesn't need to be so this is definitely not Stellar specific. This is kind of, like, overall trying to add a new implementations to Neurocrone governance and even some other mechanisms as well. If the folks are interested in getting involved, we encourage you to become verified. You can become verified on the if you go to our site. And, yeah, and then learn more. And then, yeah, if you have any questions, now is the time to ask. I'm also on the Medica Slack. I'm on, Telegram, Twitter, and I hear also some other links to to dive more deeper into. Cool. Well, thanks so much. It was a bit long. So thank you again for your patience, and, yeah, happy to answer any questions now."
      },
      {
        "speaker": "Speaker 1",
        "start": 150.0,
        "end": 150.0,
        "transcript": "Cool. Let's do it. And, yeah, if you don't mind sending the link to the presentation in the chat, that would be that would be great."
      },
      {
        "speaker": "Speaker 2",
        "start": 165.0,
        "end": 165.0,
        "transcript": "Yes. Let's see. Wow."
      },
      {
        "speaker": "Speaker 1",
        "start": 180.0,
        "end": 180.0,
        "transcript": "Yeah. So we had Lydia ask her question about what model of reputation. Lydia, do you feel like you got your question answered? Do you want Anka to go any more into the reputation side of things?"
      },
      {
        "speaker": "Speaker 2",
        "start": 195.0,
        "end": 195.0,
        "transcript": "Okay. Yeah. I mean, I can it's it's mainly just, I guess, the, right now, we have three neurons. So I think if also, you will be able to view it in the in the contracts section as well if you dive into our yeah. Kind of, like, the assigned reputation neuron. We have the prior history and also, like, the trust graph Neuron. And then, also, some of these, some of these details can be found in the, governance models library as well, especially, like, on the trust graph bonus, which was actually added as a separate module."
      },
      {
        "speaker": "Speaker 1",
        "start": 210.0,
        "end": 210.0,
        "transcript": "So those are kind of the three models of reputation, you would say. There's can you go over them again? The trust graph? What was the other"
      },
      {
        "speaker": "Speaker 2",
        "start": 225.0,
        "end": 225.0,
        "transcript": "Here we go. Like, go over so we have assigned reputation. So this actually takes based on, like, the the peers. Right? And so if you could, like, look at let me see if I can go this one. And we actually have, like, four of four different tiers right now and three that are able to participate in in some kind of voting or delegation. So if you have, like, for instance, like, Pathfinder, that's, like, our lowest one. You can attain the Pathfinder role by, gaining expertise. So there's actually an automated role. We have something called Stellar Quest where you can complete, challenges on the Stellar network, and it proves that you have somewhat of a Stellar expertise. And so once you have kind of, like, proven that, you can get the Pathfinder role and you can start to delegate your votes. And so that also, like, gives you a bit of low voting power as well that you can then attach to your delegate vote. But then navigator and pilot, that's where things really start to kinda get interesting. These two roles are currently voted in by other navigators, and there's also an automated process, that happens on our Discord. And so navigators are able to, like, both delegate it, but also actually vote directly to delegate awards. And then the pilots are able to now become delegates that, let's say, other pathfinders are never gonna be even pilots could delegate to in the case that either they don't have the role like path pathfinder does, or they don't have the availability or expertise, or maybe they have bias, and they feel like, actually, I don't feel comfortable voting directly for this project now. Yeah. And navigators"
      },
      {
        "speaker": "Speaker 1",
        "start": 240.0,
        "end": 240.0,
        "transcript": "and pilots are both voted in?"
      },
      {
        "speaker": "Speaker 2",
        "start": 255.0,
        "end": 255.0,
        "transcript": "What did you say?"
      },
      {
        "speaker": "Speaker 1",
        "start": 270.0,
        "end": 270.0,
        "transcript": "Navigators and pilots are both voted roles. Like, people are elected to the roles?"
      },
      {
        "speaker": "Speaker 2",
        "start": 285.0,
        "end": 285.0,
        "transcript": "That's correct. Yes. They have to, like, meet, I guess, like, someone can nominate them, and then they have to kind of get three. For navigator, they have to get at least, like, I think, like, five, yeah, delegated or navigators to, like, say, like, okay. Yes. I think this is a this has been a good, delegation. And for pilot, there's also, like, a voting history kind of involved as well. Yeah. Well, so there's rules."
      },
      {
        "speaker": "Speaker 1",
        "start": 300.0,
        "end": 300.0,
        "transcript": "And then the trust graph is different?"
      },
      {
        "speaker": "Speaker 2",
        "start": 315.0,
        "end": 315.0,
        "transcript": "Yes. Yeah. So there's a prior history voting, like, on history it's basically on how many times have you voted historically. Like, the more times right now, like, how it goes, it's like linear voting app attribution, which is not ideal. But it's like the more times that you have voted historically, like, the the higher that particular kind of neuron, weighting score is. And then the last one is trust graph neuron, which is a bit different than the delegates. It was designed for, to mimic trust. And I think I I think maybe, like, the neurocromal governance might have some more Let me see if I can find here. This is trust graph. So, yeah, I mean, that doesn't really say that much. But the idea is that it use, like, a page rank algorithm from Google where, like, you can then select as a user. You can select, oh, I trust, let's say, Val. I trust, I trust, Brian, you know, like Steve here, David, Ben. And then it's a and it actually, like, adds, like, to those who trust you, it adds more voting power. So the more people that trust you, the higher that neuron's waiting function is. Eventually, it could be really cool if we actually take in the the the voting power of the individual users that trust you. So, like, if some, like, a high trusted user trust trust you, that gives you more power, but we haven't gotten that complex yet as you can kinda see. Yeah. It's like a big trust graph indeed."
      },
      {
        "speaker": "Speaker 1",
        "start": 330.0,
        "end": 330.0,
        "transcript": "Oh, yeah. Sorry. What were you gonna say?"
      },
      {
        "speaker": "Speaker 2",
        "start": 345.0,
        "end": 345.0,
        "transcript": "No. Like, it it just yeah. So it it gets a kinda you can go as complicated as you want it to, basically. Right now, like, we have a relatively simple interpretation of the neuro chrome governance."
      },
      {
        "speaker": "Speaker 1",
        "start": 360.0,
        "end": 360.0,
        "transcript": "Eugene's question, I feel like, is a perfect next question because it it speaks to the it does like, of course, the system does have a lot of possibilities. Like, it does it seems to open up and experiments and play with a lot of different types of modular governance, you know, experiments. But in the sense that it's all encompassing, like, it can be a little complex and even just in this presentation has been hard to navigate. So Eugene's question around, like, how let's see. I wanna word it how he has. How has the experience been sharing neural quorum governance with the community? It's a really interesting idea. When I've shared it with other folks, the consistent feedback I get is that it's too complicated. Have you found easier ways of presenting this to community members and other operators to get more folks using the mechanism?"
      },
      {
        "speaker": "Speaker 2",
        "start": 375.0,
        "end": 375.0,
        "transcript": "Yeah. I think that's an excellent question. I mean and because, like, it's such a, like, I guess, like, all encompassing, like you said. Like, there's literally, like, so many ways that you can kinda configure how you want that system to look like. It's a modular structure, and there's way of, like, waiting functions that are relatively, like, customizable. And then there's this, like, idea that you delegate to a group of people instead of instead of just one person. Right? I think in terms of, like, as we, as we have been, like, working with our community, they have been finding it. I mean, I think neurogovernance is one of the the more complex ones. Right? Because it actually, is very community specific. And so these types of mechanisms, like the reputation types of mechanisms, will always have a harder time of onboarding, because not only, like, there is some some way, you know, there's a number of different factors that can potentially influence your voting power. There's also those factors individually are weighted, and they can even be weighted differently depending on, like, how the attribution graph is. And so that, like, obviously, like, can mimic, you know, a little bit of, like, how you actually, you know, determine reputation for a particular person or for many different types of people or an entity. But then, like, consequently, like, brain's kind of complicated. Right? So, like, trying to mimic that, like, similarly, like, these systems are relatively complicated. I think in terms of, like, you know, like, receptance to the community, it has been interesting. I think, like, in terms, like, how to, like, better do this in the future, it's, like, doing a lot of simulations and, really, like, explaining well. It's like, well, how does this actually affect your voting power? Like, and how can you effectively, like, also, like, add your voting power? But so that's on the neuro governance side. Now I think on in terms, like, when we think about a quorum delegation, I think that system actually has, like, a pretty good actually set up to get more onboarded within the wider kind of our three ecosystem, and that's what I'm also, like, trying to, like, do in the next few next few months, maybe even, like, early next year. Since, essentially, like, it's a a system that doesn't actually, I mean, delegate, like, your voting power to someone else, but it delegates to the consensus of a group that you chose. Right? Which works quite interestingly in a way that then it, you know, as I mentioned, it protects against a number of different things. But that's a like, it's a bit of a different and that system to think about in terms of delegation, but but can effectively, be, adopted in any type of community that uses any type of delegation system. So, yeah, I hope that, like, answered your question a bit, in terms of, receptiveness and kind of future implementation of explaining the systems."
      },
      {
        "speaker": "Speaker 1",
        "start": 390.0,
        "end": 390.0,
        "transcript": "Yeah. I I have a quick follow-up. I wanna give Brian's question a chance, and then I wanna follow-up as well. But Brian says, is there an iterative way of adopting such a system? What would the smallest change to a typical public voting system look like to step in this direction? I'm thinking about ways of adopting as little risk as possible from complexity."
      },
      {
        "speaker": "Speaker 2",
        "start": 405.0,
        "end": 405.0,
        "transcript": "Yeah. I I think then that a good next step could also be I mean, if you want to, like, actually divide both, like, is it because a neurocrome governance is a juxtaposition of neurogovernance and quorum delegation. So, yeah, you could choose, like, if you wanted to add, like, reputation based, like, even, like, a you know, let's say you still have you don't have a you still use direct governance, so you still use, like, oh, everyone can there's no delegative component here, but we will choose to actually take in someone's I mean, even if you currently have a token holding system, you can add that as a neuron in, and then you can weight it with, let's say, well, maybe I want to actually weigh in voting history. And and so, like, making the system, the neurons relatively simple, making them being able to, measure pretty much against, like, any type of on chain activity or any other type of activity that you can easily measure could be a great way to get to get started. And trying to see, like, well I mean, obviously, if that is something that your ecosystem desires, if that you want to actually be a bit more inclusive or put in different types of features and have them influence the the in which and the voting power and see and see what it does. Right? I think the second, if you're more like, oh, actually, like, I, you know, want to keep my voting power pretty much in terms of, like, token holdings because I believe that reflects the community sentiment. Another way of do but I want to, like, experiment with different types of delegation. Then you could try to see, like, well, actually, could I in include a system such as quorum delegation that, you know, instead of delegating, it takes actually the consensus of the delegated vote. So but yeah. In any case, I will, in the next few months, I would love to, like, collaborate with people as well if they're interested in, in kind of, like, contributing or, like, saying, like, hey. Actually, I'm I might be interested in, having the system, being added or, like, at least part of the systems or, like, at least, like, something that might, you know, sound, like, relatively similar, I'm happy to, like, support as well. I mean, what I'm trying to really do, I mean, in terms of, like, to I also foster some of the adoption is to create, tooling, that is just, even just, like, spreadsheets, even things that aren't off chain that people can try with, that people can experiment with, to be able to, like, at least, like, simulate, well, what would the results be if we do a, like, or, like, our current situation, and what would it actually look like if we implemented b or c? So, yeah, I definitely like I I see the the issues as well as we have trying have been trying to, like, focus on even implementing it just in the Stellar Community Fund. So that has been the idea so far, like, trying to see, like, validate that this this mechanism even works within the Stellar Community Fund. Quorum the quorum delegation delegate panels will be really interesting. Just something that I want to share about more later on once we actually implement it, and tried and, see the results of those. But then after that, I'm like, okay. Like, how can we actually find more adoption? So right now, it's more like, hey. Whole research, still a bit complex to adopt it in t TLDR, but finding ways to simplify and adding tooling and resources accordingly. Totally. Answered the question a bit."
      },
      {
        "speaker": "Speaker 1",
        "start": 420.0,
        "end": 420.0,
        "transcript": "Yeah. And as you know, I mean, I think, like, Steve has done a simulation during a community call. Like, definitely, you know, if you wanted to do something like that, and we could help you as a community get together a group to run some kind of simulation that worked well in the past."
      },
      {
        "speaker": "Speaker 4",
        "start": 435.0,
        "end": 435.0,
        "transcript": "And He also wants to use your thing that starts with a c to do simulations of his thing on your"
      },
      {
        "speaker": "Speaker 2",
        "start": 450.0,
        "end": 450.0,
        "transcript": "Well, it's it's not mine. It's block sciences. It's called Yeah."
      },
      {
        "speaker": "Speaker 5",
        "start": 465.0,
        "end": 465.0,
        "transcript": "Whatever that is. Yeah. I Yeah."
      },
      {
        "speaker": "Speaker 4",
        "start": 480.0,
        "end": 480.0,
        "transcript": "Yeah. Yeah. It was a little much. It was going by a little fast. So"
      },
      {
        "speaker": "Speaker 2",
        "start": 495.0,
        "end": 495.0,
        "transcript": "I also just shared the the the link in case anyone wants to look back. Yeah. CADCAD's been really interesting. I mean, I I definitely it's you know, the the more customizable a system could be, right, the it it that in turn introduces a lot more complexity. So, yes. But I I think, like and I think also, like, let me see. There's some folks here from there were some folks here from Block Science here who could definitely help. I mean, also, like, we have, like, many folks in the from Block Science in the medical chat, who'd be happy to, like, also introduce. Yeah. I'm I'm working also with Jacob a lot, and, I'm just, like, blanking out on his name. Like anyway, I'll I'll find your name very soon, and there's many folks in the Medigolf Slack that can, support. Zargum. Yeah. Oh my god. How can I forget? Yeah. Of course."
      },
      {
        "speaker": "Speaker 1",
        "start": 510.0,
        "end": 510.0,
        "transcript": "Cool. Yeah. Yeah. This is great. And definitely, let's do another simulation soon in a community call. That would be really fun. I saw a question I wanted to give. Mel? Yeah. Mel, you had a comment back on Brian's question. Did you wanna ask anything or bring up any points?"
      },
      {
        "speaker": "Speaker 5",
        "start": 525.0,
        "end": 525.0,
        "transcript": "No. I just I'll I'll double tap on that, though. I think that that trust graph is incredible. Getting participants to engage in in that kind of activity, gets you to a place where you can use that data, I think, in in really, beneficial ways. And we we've talked at length. I I don't know if you're going to the, to the multi gov event tonight. There's, like, so much going on with, with Tali and all the like, get, like, MediGov, Tali, all the organizations that support. But, yeah, I think a big, a big aspect of what I've been working on relates to a trust graph and very directly using that as a delegation, mechanism and, very explicitly. And I think that, kind of where my mind went is almost a a neuron that would essentially look in on itself, right, and say, okay. In process, in a vote process, what are the people that I trust doing if I don't think I know the answer? And so I think that that feature, as I see it, would be would be really cool because it's it takes somewhat of the the mechanized process out of it. It's really very p two p in a way that I think is very smart. And so yeah. Kudos on getting to that trust graph because I think that's getting anybody to engage in a process of, self sortition, is is a bit tricky. But, I think now that you have it, it's really powerful."
      },
      {
        "speaker": "Speaker 2",
        "start": 540.0,
        "end": 540.0,
        "transcript": "Yeah. And, I mean, yeah, I mean, you're I am actually going to the event. I wasn't sure how long I was gonna stay because I have some other things too. But but yeah. I mean, like, in terms of, like, the the possibilities are really endless. Right? If you have, like, infinite, I guess, engineering power and time and resources, there's a lot that you can do, which I would love to have, by the way. In terms of, like, the trust graph, like, I mean, we decided to to be, like, separate also from the delegation because you might trust someone. Like, this was actually a reasoning of, like, how can we design certain systems that are a little bit more able to onboard newer users rather than kind of close off, like, a community of, like, highly, I guess, informed people and then, I guess, making it impossible for newer users to kinda come on board. And so the trust graph was really like, the trust norm was really meant for someone to say, like, well, you know, like, see some new people on Discord. They look kinda like they're doing some good stuff. Right? I kinda want them to friend them or like them or give them some kind of credit. I wouldn't, like, delegate my votes to them yet, but I definitely wanna see them succeed. Right? And so that's the idea of, like, then someone can actually build up their trust graph. And then in that form, then also build up reputation. And if you then combine that with something like I guess, like, if you're like yeah. Like I mentioned earlier, like, a highly trusted user then trusts you, then you can actually, like, get, like, a lot higher voting power. But then subsequently, let's say, like, if you were to, like, lose a let's say, like, if you give certain or, like, expose, like, certain, like, behavior, like, that is, I guess, like, against, like, the best interest of the community in, let's say, the Discord or any other online platforms. People can be like, hey. I don't trust this person anymore. I can go untrust him. Right? And let's say if you have, like, an interesting voter attribution curve that then is, like, logistical that actually like, if, like, you know, like, a certain amount of people in short amount of time were to, like, untrust you, you can significantly, like, lower, let's say, your total voting power. So, anyway, I could go on about these things as you can see. But but that's the idea. Right? Because then you have, like, somewhat of an automatic process for reputation based governance that doesn't require, yeah, any token holdings or the like. Yeah. Let's continue the conversation tonight."
      },
      {
        "speaker": "Speaker 1",
        "start": 555.0,
        "end": 555.0,
        "transcript": "Yeah. And please, we'll start a thread in the Medigov Slack too for continued discussion. We might not get to all these questions and points, but we'll open up a thread so we can continue there. And please share the link to whatever the event is you're going to tonight so we can add it to our calendars for us, not as plugged into web three space, but it does sound like an interesting event. So please share that. But with only three minutes left, I wanna Mel, you had another question around best IRL use cases, meaning non Web three orgs. I'm really curious, Anka. Yeah. If you see opportunities I know you live in New York. I live in New York. New York has started doing participatory budgeting experimentation, and I'm curious, yeah, what you see there. Like, what where can we bring this outside of the Web three space?"
      },
      {
        "speaker": "Speaker 2",
        "start": 570.0,
        "end": 570.0,
        "transcript": "Yeah. I mean, like, honestly, like, like, pretty much any anywhere. Right? Because like I said, like, it you can, like as we actually have, like, a hybrid, like, on chain or off chain, addition into, like, neurons. But, actually, like, these kind of systems, you could, transfer also to, yeah, things like community, like, local community projects, like or, like, even, like, that's, that's, where members can actually rise to the top based on reputation rather than, I guess, like, maybe appointed or, yeah, kind of or, like, just, like, normal delegation. Right? And, and especially as, like, you can then also build contribution, kind of models or incentive models to these, like, point like, reputation points. So let's say if someone can get a lot of reputation points for I mean, like, getting, like, get getting out. I mean, there's actually who is on the call today? I forgot. I forgot the name. Okay. He left. Great. But they're building some you know, so, like, a system that is well, somewhat also similar based on reputation on Slack where, like, a community of people, like, living in the same house, they can, like, do tasks and then actually get a, you know, a score based on that. Right? And so, like, you could also use neurochrome governance for that where you can actually grade. Yeah. Yeah. Yes. Exactly. I just, like, keep, like, you know, forgetting things as I talk. But but yes. Exactly. So, you know, you can get, different, like, I guess, like, based on your contribution history, based on your, like, expertise, based on your, you configure, like, who does certain roles, like, how people yeah. Then it'll create chronicle. Yeah. Yes. I met him several times and yeah. So I just keep forgetting people. But yeah. So yeah. Just trying to, like, actually, like, like, measure those scores. So I can measure that reputation and measure that that contribution history. And then, like, then rewarding that with either, like, roles, or, like, maybe even, like, monetary incentives. And so, yeah, I would love to see, I guess, like, maybe I will will do that after my start work at Stellar trying to see, like, how we can actually make those systems a lot more efficient to help the real world. Right? Yeah. But, yeah, let's see. If anyone's ever interested in doing anything collaborative, I'm happy to also, like, support. Very interested in these systems and, yeah, wanting to make systems more impactful and efficient and inclusive so more so more people can participate. Thank you."
      },
      {
        "speaker": "Speaker 1",
        "start": 585.0,
        "end": 585.0,
        "transcript": "Awesome. What a rich presentation, discussion so much here. Yanis, I'm gonna kick off our thread with your question that we didn't get to, but I really love. So come to the MediGov seminar channel on Slack, and let's continue discussions there. Thank you, Anka, for your presentation. Everyone, please unmute, give our speaker a big round of applause, and have a beautiful rest of your day. Thank you, Anka."
      },
      {
        "speaker": "Speaker 2",
        "start": 600.0,
        "end": 600.0,
        "transcript": "Bye. Have a good one."
      },
      {
        "speaker": "Speaker 1",
        "start": 615.0,
        "end": 615.0,
        "transcript": "Thanks, y'all. See you. See you next week."
      },
      {
        "speaker": "Speaker 2",
        "start": 630.0,
        "end": 630.0,
        "transcript": "Bye. Bye."
      }
    ],
    "summary": null
  }
}