 A question for you, Pooja, as the co-author of the Decentralized Society paper, what was sort of some of the inspiration that led to wanting to write this in the first place? Thanks. So thank you first of all for having us here. Before I answer that question, I just want to acknowledge that the paper was actually a coordination of many individuals, including Divya and Leon, and members of the Radical Exchange community, and it was building off a lot of intuitions and questions that we've been asking for many years. So throughout this conversation, I really want to draw on Divya and Leon's expertise as well. So to answer your question, from my perspective, there was a really key, the key question we were trying to answer was, how do you build composable bottom-up networks of coordination for the 21st century in a decentralized way? And what I mean by decentralization is you can think about it sort of with three properties. One is where one level of social coordination isn't dominating, say, the level below it or say a level above it. Another way to think about it is where no particular actor or institution or nation state or set of actors can dominate another set of actors. And a key principle around this is consensus across difference or cooperation across difference. And the idea there is that we draw on a maximally diverse set of participants to draw out consensus in decision-making. I want to also draw on Leon and Divya and see what they have to say about this question. Okay. So what I find pretty valuable interest, like most interesting maybe around like the DSOC paradigm is that economics as an academic tradition has this very rationalist philosophy where we kind of, when you go to economics classes or business school, people tell you, okay, all our models kind of make sense based on the assumption that we are isolated from each other, rational individuals. And turns out like everybody is not like the most rational possible actor. And like we actually are not isolated. Like we care about our friends, our family, our like neighborhood, our like university, our like religious community, all like our soccer club or whatever. Like we have all these, we are embedded in like a network of communities. And the weird thing is that like even say like Gitcoins quadratic funding in its current form is kind of only like assuming like isolated individuals in its model, like where we don't really have any measure to account for the relationships between the people that are contributing to the project on the Gitcoin platform. And I think like, you know, the exchange community and then why specifically like, I'm not realizing we need a actually built formal institutions that can recognize like all these social culture groups that people care about. And like who cares about what social culture groups or as a member of them, and then feed that reality into the decision making models like Gitcong grants or quadratic funding. And I think the DSOC paper like, like incredible job and kind of bridging that social philosophy to Web three, as you might have recognized, like on social media, there was like a lot of like debate around like how it would particularly be implemented. And I think like it was super fair for the paper to like suggest that the Ethereum community or which like natively uses Ethereum tooling, right, would kind of use Ethereum native technology to to bring pluralism to its institutions. I think, you know, other institutions say like governments might use different data structures, maybe pen and paper, like usually or like, you know, like other technological communities, you know, Facebook could also do more pluralistic like content curation or like sortition, right. And their data obviously lives in different places and not on public blockchain. So I think, you know, whatever, like all kinds of like institutions, including web three, and I guess I was kind of the focus of the DSOC paper could like work on recognizing social culture groups and then like fostering cooperation between groups and not just between like supposedly isolated individuals. Yeah, I think for me, my background is in working on technology and democracy. I started out studying, you know, decision making processes and institutionalization of social movements, cooperative organizations, collective data ownership, things like this. And when you're in that space for a while, it sort of becomes clear that we need new primitives over time to allow for, you know, the subsidiarity principle that a lot of our democracies are built off of local autonomy as much as possible. And then you move to the next level up sort of questions to for those to flourish in an increasingly connected increasingly in need of coordination world, you need to constantly be pushing on what types of mechanisms can enable that and plural intelligence and collective intelligence to come together and have, you know, the three things that we want in a governance system, which to me are information processing. A lot of people have different information capacity and have different pieces of information. Everyone in the world has some small part of subjective information that not anyone else has. You want to process that, you want to coordinate people to take action, and you want that system to be legitimate, right? I think there's a significant normative component here where people should have that ability to affect their circumstances. And for those three things to be true, we need to constantly be iterating new primitives, identity and verification and ways of belonging in a community and deliberating are a big part of those primitives and always have been. So to me what's exciting about this paper, but more importantly as we've discussed this sort of like invitation to research and an invitation to create sort of more knowledge around this is what are those primitives going to be going forward while working with the structures that we've always had for this. And I think that's a big part of what we see here. We're not trying to replace, you know, wholesale any particular piece and really pluralism is about that organic growth forward and recognizing and correcting injustices that we currently have as well as building meaningful innovative positive visions for the future, which I think these primitives can lead us to. Yeah, thank you so much for providing sort of the baseline and thinking of why is this important from sort of philosophical and high level. And I guess coming down to more say a use case specific area, what are some actual application domains of this kind of thinking of soulbound NFTs or non-transferable NFTs, but of these SPTs? Yeah, what's an area that gets all of you excited? So, as I said, from my perspective, the goal of the paper was to think about how we build decentralized networks of coordination and one network that I think is really in need of disruption is actually scientific networks and the production of scientific knowledge. And as many of you who are researchers in this audience know, there is a particular rent extractor called journals who have placed themselves in the production of scientific knowledge that are, you know, very rent extractive and coercive towards researchers and peer reviewers. And I think non-transferable tokens or even verifiable credentials for that matter could be harnessed as a identity substrate to recreate these knowledge networks bottom up in a way where we can transition closed science to open science and also, you know, transform competitive science to cooperative science and creating research dows. And I gave a talk recently at DSI Berlin, which you can look up, which goes into this further. But for me in particular, I'm very excited about scientific disruption. And actually, the kind of flip side of that is also open source research, which sort of has the opposite problem where there's, you know, where scientific communities have a lot of status but low sort of closed information. Open source research communities have struggled to represent status and accomplishment, but have open information. And I think, you know, both of these communities could be very well served by non-transferable identity substrates. Yeah, Leon or David, what do you think? Yeah, sure. So I'm kind of, I think, like a cool area for like near term experimentation with this paradigm is like KitKoin grants, this like matching funding system that like KitKoin has been operating for a couple of years now. Basically, like, we'll usually have like a philanthropic organization or like wealth individuals like who provide matching funds, where you'll have like projects that kind of need like open source projects usually so far that need funding. And then based on community donations, which also kind of count as votes that matching fund will like, match the community contributions to these projects. And it's already kind of innovative in its current form because using this thing called quadratic funding where actually like, you'll look at the square roots of the donations, of the size of the donations. So kind of give proportion more weight to smaller donations and to larger ones. But even though like this already kind of strikes a better balance than like current mechanism designs like between like preventing tyranny of the majority and at the same time, tyranny of the minority. It's kind of like a cool trade off. It is still like assuming this very individualist notion and like doesn't, you know, it only considers kind of headcount and the square root of the size of the contributions. And like, obviously, this seems a bit ridiculously simple to like provide like very rich scalable social cooperation. I mean, think about like, think a very good example, like to just visualize these dynamics is like, consider like you would apply quadratic funding of that kind in the United States of America. Then like, although like we kind of allow minorities to kind of show how like much I care through like making, you know, larger donations, like that that anyway has a limit since like the costs are like growing serverly like with a square root and everything. Like it would likely be the case that like a large homogeneous group, let's say people in California or like people in New York or like people at the coasts would be able to dominate every decision, right? Because like, they care about like coastal issues or like New York or California issues. Well, like, the countryside might be less populated and just smaller headcount. So like to address it, we could and the system is kind of trying to address it already, right? By like giving local representation to the states, but that's like adding just one dimension of complexity to the system. We could also look at like where people work or like what universities they went to or like what high schools they went to, whatever. And I feel like basically we can map out like this like multi dimensional sociograph and like even one or two dimensions, I think would really help making making quadratic funding more pluralistic by then like putting increasingly like more matching to projects that not only have high headcount and like contribution sizes, but also diverse groups supporting it. So we'd add like diversity to the rewarding mechanism. And I think like this is something we could explore like other people who are building like quadratic voting or quadratic funding type stuff. I mean, optimism is also working on similar stuff. So I'd be pretty excited like about making these quadratic models more pluralistic. Yeah, I think for me, it's it's kind of a symbiotic process, right? It's not just use cases for Sylvain tokens, but really, why did we write the paper? Certain use cases for which they were created are the ones that I'm sort of excited about. And one of those I think it's around a network of how people handle and have voiceover and utilize data. And so a lot of my work has been around data as an input to AI governance and different types of technology governance. And one of the big problems that we face there is I actually came in in this space and was super excited about direct democracy over data, like people should have control over data, they should have an idea of the kinds of questions they want to solve, they should be able to target and basically donate their data to specific things. And quite quickly, which maybe you've all realized quicker than I did took me like a month or two direct democracy over data would be terrible, right? No one wants to have that kind of granular decision making power. You have no idea exactly where everything is going and where it's being stored. And that's not really the level of abstraction at which you want to operate. And yet there are major, there's rent seeking problems in the data space. There are so many ways that positive externalities and positive some outcomes are being curtailed, because a couple of organizations have access to so much more of this inherent problem solving input than others do. There's a lack of competition. And there's huge privacy problems. There is extraction from communities to build tools that will then go back and harm their communities. There are just a lot of issues here. And moving from direct democracy to then realizing, okay, we need something much closer to a network. We need liquid representation. We need new kinds of institutions. And we need ways to figure out what problems we can be solving. Suddenly you need these primitives like soulbound tokens to get a sense of what communities are people a part of, how do we track those values? What kinds of places would people be interested in delegating their values to so that they can be acted on their behalf? How do we have the appropriate inputs? Data lineage tracking is a big problem. I mean, how do we know what data is a part of which data set? We can imagine some use cases there. So to me, it is one of the inputs. It's not the only input that can build this kind of ecosystem. And of course, I'll speak to what I'm familiar with, which is the data space and the AI space. But this is true in creating a lot of different kinds of networks, I think. So I think that the networks that this can enable is what's most exciting to me. And it feels like that's kind of true for a lot of the space that's coming up around this paper. Right. Like, I just want to add on to that. So earlier in Juan's talk, he was talking about networked coordination. And so what we were trying to do in the DSOC paper is not just create like one big monolithic network, but actually recognize that we have different levels of coordination, right? There's local coordination, there's state, there's federal, there's global coordination problems like climate change that we need to solve, right? So how do we create this diverse, pluralistic network, composable network that is very dynamic at these different social layers and social levels, right? And as I said before, in a way where like one layer doesn't come to dominate the other or a set of actors within that layer doesn't dominate another set of actors. And, you know, IE is a truly decentralized network at different social scales. So the work that Divya does on AI and the work that Leanne does on quadratic funding, right, they're seeking to solve those problems at different social layers. That's great. And as we transition to questions, please raise your hands and I'll come run up to you. But a quick question for the panel as I start running around and catching anyone who does have questions is sort of what do you think is needs to happen in the next six months to really start making progress in these directions? Is that more of a tech problem or a culture problem or something in between? And yet if you do have questions, let me know and I'll start. So I would actually say first and foremost, this is a research problem. We need a lot of really creative thinkers in incentive design and a lot of small, you know, scale local experiments. We shouldn't be introducing any technology, in my opinion, whether it be soul ball tokens or whatever, unless they're actually solving like genuine coordination problems. So running small scale coordination experiments that can like scale up, right. And I would say I gave the initial use case around, is this still on? Yeah. Okay. I gave the initial use case around science and open source research. That's a great use case because there's already a high level of publicity. We want science to be open. We want research to be open. We want people to get recognized, right, for the contributions they make to a field. So that is a great place to run like experiments, but it is a research. There are a lot of open deep research questions. And if anybody is a researcher and wants to get involved, feel free to DM me. Leon and Divya, I'll let you also answer. Yeah, I'll shake up the order a little bit. Yeah, I think exactly like you were saying, a lot of it is around tight feedback loops and identifying what's going well and what's going badly, right. And so I think that, and I may have said this already, this is kind of like an invitation to enabling a network ecosystem. That means that there is a huge amount that hasn't been figured out. I mean, there's not a technical spec in the paper for a reason, right. And there's a lot of governance questions. There's a lot of questions at which level of communities this can be plugged into. And I think a lot of that comes around, you know, can this be piloted in small sorts of ways? Some of the questions we've talked about that are definitely still open are like, what are the ways that you have checks and balances on negative reputation? I mean, that is a, is a massive open question. But it's also been an open question for a very long time. It's not an open question that this kind of new primitive created. It is a problem that we contend with daily as we tried to live with each other in a pluralistic society. And so those types, you know, bringing together the research that has been answering those questions for a long time with some of the ways that we can use technology to enable this. And I think consistently being really humble about the role of the tech here, because it is purely an enabling mechanism. And the second that it stops enabling positive outcomes, we should not choose the technology over the outcome, right. And I think that that's something it's easy to be blind to when we are excited about a technology. And it's something we are really sort of invested in being careful and humble about as we determine whether experiments are going well, going forward. Cool. Yeah. I want to give space to the audience too. Yeah. So this is a question that's similar to what you just mentioned, Divya. So in a DSOC world that is doing a good job at producing good things for groups, for amplifying good things for groups, are you worried about the ability to amplify prejudice? And if so, what should we do about it? Yeah. So that's an excellent question. And we talk about this in the paper in dystopia as much, right. So one of the key principles is cooperation across difference or consensus across difference, for example, right. Elevating and surfacing consensus amongst the most diverse participants to signal, say, broader shared goods, pluralistic goods across, you know, broader networks. Like you could do the opposite, right. You could do consensus across similarity, right. And amplify, you know, the interest of the majority or say even the interest of a minority over a different set of groups, right. So ultimately in whatever technology we design and we make these sort of tacit decisions every day already, we need to sort of at the forefront, you know, declare sort of what our values are and what our goals are. And I would say like an existing technology like, you know, web 2, which I don't actually see as being very different from like Chinese surveillance state technology, right. We are making tacit value judgments when we use that technology as well. And we should just be aware of what are the values we're actually subscribing to and enabling. I think DSOC just makes it much more transparent that, hey, we are actually making value judgments. And like, what are those value judgments? But it's a concern, but I would say it's a concern not specific to the substrate, but is actually implicit in the substrates we use every day today. Yeah, I'll second that. I mean, if you think about it conceptually like the social credit system and like the people's republic of China and like kind of the attention driven like social networks like Facebook, Google are kind of that world where like we use a social graph, right. That's pretty rich, like much richer, I guess. And we can even imagine to like foster homogeneity to foster like attention bubbles like these like eco chambers, right. And like people start talking to each other in like social media world. And I guess the people's probably of China is like trying to like make their citizens more compliant and like homogeneous and like supporting the government that they have there. So I think basically we could, I mean, like all these systems for like pluralism will kind of have a similar architecture of like some social graph that you'll look at. And then the only question is like how do we define diversity instead of homogeneity? Obviously diversity is very like broad term and you need to make a lot of like normative judgment there, you know, like, it's not like that there'd be like a verifiable ground truth, you know, now this group is diverse or not. Like it depends on like what you consider diversity, right, or like valuable diversity. So, yeah, I think that the key will be like to at least have more accountability than like kind of the like homogeneous virtual societies that have been developed. And like people can understand, okay, why do I now get like lower matching in my good congruence program than like previously or like why do I get higher matching than somebody else? It should be like accountable like why why is it happening? I don't think like for instance, like that China or Facebook or Google are like very accountable in like how they do algorithmic recommendations or whatever. And on the other hand, it should be value aligned with the community that gets impacted by these algorithms, which is also not the case I would say on the like bad example side. Didn't mean to show you I mean like Facebook and just different. But so yeah, I think as long as it's accountable and like value aligned in the in terms of like value aligned with what the community that gets governed through this stuff thinks is like diversity, we are kind of going to the right direction, but it can like totally go wrong. And I think it will be like, like we will never like reach a perfect like diversity algorithm, right, will be like continuous iteration on like these algorithms and like aligning them closer to the community that it governs. Yeah, I think the short answer to your question is yes, we are worried about this. I think we're very worried about this. I mean, it's it's one of the major open points of research, right? And I think that as both Leon and Pujo were saying, the way to mitigate that is basically checks and balances is ensuring that we are creating a system that has as many different points of input and feedback and accountability as possible. And recognizing the ways that concentration of power means that that is cannot be done. And how many, you know, different power concentrating systems there are, whether it's at the nation state level or at sub levels, which is true in, you know, most of the systems that we see today, they're just incentivizing these types of power concentrations that don't allow for pluralism. So I think we are concerned about it. And building these countervailing forces is basically the best way we see to preserve that sense of non domination. Hey, a big fan of the paper, a lot of that provoking stuff in my community. So one of the big ones is this notion of like soul boundedness or non transferability. And I'm curious, basically the question is like whether either from your research or in circulating these ideas, like what seems to be the way that we can facilitate that? Like, for example, in the Linux Foundation public health working group where they're working on like COVID passport stuff, they have a whole task force on what they call the identity binding problem. How do you know the token or credential isn't isn't isn't is bound to a cryptographic identity, which we could share keys or something with versus like my soul or this human or something. So what's the, what's the, what, what seems to be a good solution or candidate solution from your y'all's perspective on that? Okay, so I hope I understand your question. Eugene, maybe you can go back so I can clarify. So I think your question was like how you were differentiating between like other identity systems and like soul boundedness. Yeah, like let's say I get a non transferable token from my academic credential, but it's bound to a cryptographic key. If I get to share my private key with someone else, all of a sudden they could impersonate me, but it seems like there's something with this notion of soul boundedness where we would want to potentially prevent that for the decentralized society. Right, right. So part of the key thing to understanding soul bounded tokens or non transferable tokens or community bound tokens, because ultimately we see individuals as a nexus of their social relationships, right? So they're different sides of the same coin. A key to understanding the idea around non transferable tokens is also this idea of community recovery where you cannot credibly sell your wallet because you could always just recover it through the communities in which you participate in. And we talk, we sort of gesture towards a mechanism which is very similar to the mechanism of consensus across difference. And the idea there would be that you would draw and say, for example, the most maximally diverse set of participants, ideally participants that maybe you know, but nobody else knows each other, right, in real time in a programmable way. So if somebody tried to steal your keys or you gave your keys, it would always sort of be recoverable. And so nobody would do it to begin with. And so this is like a very critical part of DSOC, right, is having these community recovery mechanism, which doesn't exist anywhere. And so this is a key, you know, thing to be innovative in like security research communities. And at the same time, though, like, we talk about it at the end of the paper, like, well, what comes first, right, these non transferable tokens or community tokens or community recovery. And actually both of the things are, you know, should instantiate relatively at the same time. And we kind of recommend, you know, these proto-solbound tokens, which are revocable by the issuer, right. So if you do need to, you know, take a new wallet, you can have those tokens like burned and then reissued in the new wallet and like carry on until we have community recovery in place. I hope that answers your question. Yeah, I think super briefly on the particular, like, what is it tied to? Is it tied to your wallet? Is it tied to your passport? I mean, what what is the mechanism by which it is tied to your quote unquote soul, which is not something you can really tie stuff to yet. To me, I think part of the way that we do that is also by attestation. So for example, if you shared your private key and shared your academic credential, well, everyone you went to school with would know that that person wasn't there, right, that whoever you shared your thing with, your school administration would know. And so we want to kind of maintain quote unquote soulboundedness by adding attestations on top of the kinds of credentials so that sharing in this fraudulent way is is basically made more difficult. Will there be slippage? Yes. And I think that's why we need to start out these experiments in, you know, the more environments in which we're okay with slippage basically where it's not like for death, right. And we're definitely not suggesting that that replacement happened early because there are all of these kinds of research questions that need to occur before we figure out, okay, is this really tied in the way that we wanted to or not? Much for joining. Unfortunately, we do have to cut off the panel here just because we're at time and we're already running slightly behind today. But big round of applause for our panelists. Thank you, Pooja, JV and Leon.