Yuting Jiang on the Agora Citizen Network, Polis and bridging-based algorithms
Yuting Jiang is CEO and co-founder of the Agora Citizen Network. The Agora tool was inspired by the well-known Polis platform. Yuting understands how modern technologies can improve democratic processes and offers interesting insights into future developments.
If you enjoy our podcast, please consider making a donation. It really helps us to sustain the podcast.
Top Keywords
- agora 0.015
- algorithm 0.008
- different 0.008
- data 0.007
- community 0.006
- well 0.006
- citizen network 0.005
- space 0.004
- agora citizen 0.004
- interesting 0.004
- basically 0.004
- deliberative 0.004
Transcript
Welcome to another episode of the Democracy Innovator podcast, and our guest of today is Yuting Zhang from, Agora Citizen Network. So thank you for your time, Yuting.
Thank you for having me here.
And, as a as a first question, I would like to ask you what is, Agora Citizen Network, the project you're working on?
So Agora Citizen Network is a fully open source, deliberative tool, and, it's also a dialogue space, like a reimagined digital town square. We are a Paris based team, and we started a project about, two years ago in 2024. And what prompted us to start a project was when we witnessed that elections were happening around the world. Dialogue space were shrinking on the Internet. We witnessed a lot of, growing polarization. And although it was supposed to be a digital town square, so we wanted to reimagine of, how it could be and, how we could embed prosocial designs and algorithms in this, in this new space that we are co creating with, with the users.
And, I mean, I know that this software is sort of fork of policy. Right? How did you know? If you want to
Yeah. So we were very inspired by Polis. And and we we were looking basically, we were looking for different ways, to interact with each other. And we were very inspired, by Paulus. I think I saw somewhere on the Internet that, the founder of Paulus said Paul McGill, he said Twitter was, great for building a movement, but it was not great for writing a constitution. And, that really, sparked something in us that we need a different kind of, a way to interact with each other, a way that's more constructive. And, it's not about followers and influencers, but it's more about individuals and link among individuals. However, rather than forking another policy, we actually decided to, abstract the algorithm from policy. For us, that was what, mattered the most. It's a machine learning algorithm that is able to, distinguish different schools of thoughts, different opinion groups, purely by, looking at how people vote, agree, disagree, or unsure on each other's opinions. And and then this algorithm, they're able to basically put participants into different clusters, and then they're we are able to see which statements, divide scoops, which statements made them unique, representative, and, which opinions they unite the room, so which opinions they help people, build common grounds. And we we feel that this is extremely unique and valuable in today's world, and it's also what we call bridging based algorithms. So we decided, that in Agora, we are going to use this similar, kind of algorithm. And, eventually, we worked with the open source developer to rewrite, this algorithm in Python because the original Powless was built as a as a product. But we wanted to make, the algorithm into a library that, any team, who wants to, use this algorithm can build different products on top of it. That it doesn't have to be the same product. It doesn't have to be the same, deliberative tool, layout. And, in our context, we wanted to have a more button up approach, to add social features, to add, yeah, like a community vibe to it. So so, that's why we couldn't just we couldn't, directly fork on the. And, however, we on Adora, we have the same mechanism of, agree, disagree, onshore. And, also, on an, Dora, the analysis tab shows people, where the top and bottom lies, where the, division lies, and, what's the ping and mapping, ring mode time.
Okay. Very interesting this abstraction of the of the algorithm. And, I mean, first, you described the, why you think it is important, like, to have a software like this. And, and also very cool the the fact about the I wonder if, you were talking about a sort of mini app that maybe user can build related like a sort of plugins. And, yeah, I have a question. Like, when you first had the idea about the technology and politics that could be a a nice match?
It's interesting question because originally, before we started Algorand, we actually, were in the decentralized identity space, and we actually built a MVP, that is based on zero knowledge proof. And what is really interesting about zero knowledge proof is that it allows anyone to prove, certain aspects of your identity without revealing your personal data. For example, you can prove your age anonymously or your nationality anonymously. You can prove that you're a human, not a bot anonymously, or that you belong to a certain community. So the reason why we built the, belief that it had a lot of potential was also, what we witnessed, the is there that, the Internet was becoming completely faulted with bots. And, according to the statistics, from TOWIS group, I think, in 2024, one third of Internet traffic was generated by bots. And, that and there are also coordinated, astroturfing campaigns happening on all major social networks including Twitter, Facebook, and so on. So we decided to, do a small test, and, we built, MVP of a forum that uses, zero knowledge proof to allow students and, teachers, so people from the school, from the university to prove that they are members of this community, while posting anonymously. That's, how we started. Because we wanted to have both, anonymity, which we believe that is very important, for freedom of expression. But at the same time, we want verifiability. We want to know that these posts, they come from genuine, numbers that people from the same university, for example, that they really feel that they are chatting with others from their university. And, meanwhile, having the ability to express themselves, more freely. When we built this prototype, it was also, in a time of, increasing tension on university campuses, especially around censorship. And, there was a there's a nonprofit in The US called the Fire Foundation. They published a ranking of, freedom expression, a store per university. That and, and and so in that year, there was a lot of censorship going on in almost all the, all the campuses. Yeah, so it was a fun experiment when we decided to do it. However, what we witnessed was that although we allow people to speak freely, the classic form in is not design is not enough. The usual algorithm is not enough, because when people are anonymous and sometimes they, you know, they start to troll each other. And, and then that's when we started to look at, different kinds of, solutions, to help people have dialogues in a meaningful way. And that's when we discovered Powys and, decided to, oh, let's, try building using this, algorithm that perhaps, we can turn disagreements, controvert even controversy into, something that, that is useful, that is, constructive for the community.
If you have an example of, let's say, disagreement or other, or something that was discussed in, that could be also on some other platform, but also on Agora Citizen, something that, was particularly meaningful for you?
So about ten days ago, Agra was used at a, Web three community in Chiang Mai in Thailand called Four Seas. So it's a permanent node. And, we were there physically as well at a session where it was used. And, the topic of the session was about how can public, interest spaces and communities be sustainably funded, and, what kind of, commercialization could they explore and so on. It was very interesting because four seas itself is definitely is a public interest community. However, and and so they invited speakers who are building communities, basically, that are all, like, they're like intro intellectual hubs or, artists or sometimes they're just for students. And, so we were discussing how they can be sustainably founded. And, in the discussion, there was disagreement on whether if these kind of communities could only be founded by donations, or if, commercialization is viable, what kind of commercialization is viable. And there was disagreement. There were we mapped about five, different groups, during that session, and they were purists. They're so purists. There were people who who believe that, oh, if the if this community is for public interest, then they should not, seek monetization because it would pollute, the intention of these spaces. And, they were also pragmatists who who's who believe, for example, that, public good, builders, community builders, they can they should regard themselves as entrepreneurs. And, there was also controversy around if, a space is for public interest, if they should, sacrifice the depth of the discussions of the depths of the events, the intellectual depth, of the events, for more engaging topics and formats. So so that was a very, that was a very intriguing conversation. And there were four or five speakers, who who were all community builders and about, 50 to 60 participants who were visiting, this space. And some of them are are also residents of this community. So it was interesting to see, how they vote in real time and then also to see, to have the speakers to comment directly, give their opinions on the results of the conversation.
And, some sometimes I imagine, like, if in the in the future or next future, if people, could use this kind of tool to participate in in the political life. So how do you imagine it? Like, the future. If you thought about it.
We we choose the name Agora, and because in in ancient Greek, it was the place where citizens discussed politics and philosophies and, matters related to their city. So our vision for Agora is to become this dialogue space for citizens to discuss, matters that would help them to shape the policies collectively. And, we also named it citizen network because we think today, many social networks, they are they are the digital town squares. But a lot of them, they were not intentionally built, for this purpose. They were built for entertainment, keeping touch with friends or sharing fun cat videos. And so the so the design and the algorithm that increases engagement, sometimes it's counterproductive to issues that are complex and sometimes divisive. So that's why we decided to call it Agra Citizen Network. And what we wish is that more people would, discover alternative ways of governance that is not just about voting, and it's also about dialogue. Because, well, voting can take different shapes and forms. But, currently, the most, the most usual well, voting is is is out outdated, and it creates a lot of, conflicts, among among citizens. So we would hope, that by building a program, we would raise awareness of these alternative ways, to help people have more dialogues, to maximize consensus, to maximize their common ground before they move on to voting, and, to make, democracy processes more participatory in a in a constructive way. Yeah. So that's, our our vision.
Okay. Now yeah. I I also hope that, this will happen. So because there is a lot of I I see it also on the people around me, a lot of polarization. It's not easy to discuss about things where we have a lot of emotions and feeling. And so everything that is helpful to do that, or also to, yeah, to help a larger mass of people to decide about something. So I I really hope that in the future, we will see ever more like a like now, when you were talking I mean, before you were talking about the algorithm under, POLIS, and, I I was thinking or maybe in POLIS two, we're working on it. I don't know. Like, about, also possible, speech to text. So maybe people meet in a sort of physical Agora. And then, the technology, the algorithm in the background is helping in finding agreements and also point of conflicts. Yeah. Yeah. I'm really excited about the the future.
Yeah. Thank you. Yeah. I mean, transcribing transcribing physical conversations into text, is definitely something we are very interested in exploring more. And, also, we are currently only using colors, but the goal is to provide different kinds of algorithms that people can choose and compare the analysis results, to open a a space for, yeah, a plurality of, algorithms. I think that's, that's the the picture, the big picture that we have. And, also, just in terms of, transcribing real time conversations that happen in the physical space, we are also seeking to partner with other deliberative tools like, like, Danbury, for example. They they basically can transcribe, what people say in small groups into, insights that can be used to generate seed comments and so on. So we are also, working with them to experiment on this kind of projects together. This is great. Hope to be yeah. We hope to work with, more, civic tech tools, to bring values, to to reach others' users.
K. This is quite cool. I was also trying to imagine now in this moment, but, how to, because people talk, but then it's not very it's not really structured data while, in Agara network or policy. I mean, you have some struct more structured data. I need to think more about it, but it's something I really enjoyed. And, do you have any other functionalities in mind that you're working on, or you would like to work on it?
We have a lot of ideas all the time. Even when you just said about, structured versus unstructured data, we are also looking at ways to make sense of unstructured datas just, to have, different ways to to to fit into the algorithm. And, but our right now, our priority is, so there are three different things that we are developing. Well, the first is to, enable, to to to basically customize Agora for specific use cases. For example, if a community wants to use a certain type of, community credentials or if a city wants to, has a has its own database, for the digital identity system, We want to support that. We can customize it for them to allow, like, their residents, for example, to participate in anonymous but verifiable discussions. And, the second, project that we are exploring is to build the, API of Agora. So other civic tech tools can integrate Agora into their own products so their users will not need to juggle between different tools. And, we want to also integrate the APIs of other tools as well. And for the third part, we are now trying to cocreate a deliberative standard. So a decentralized deliberative standard, we call this, the DBS with, other teams. The idea is to really to allow interoperability among civic tech tools and by by making verifiable data interoperable. So these are the, three main projects that, we are working on this year.
Okay. And, is it, okay. The centralized liberty standard seems quite interesting to me. I mean, because it's thoughts about interoperability. And, I wonder if it because also inside MetaGo, they're working inside the inter about interoperability and the ontologies, and also outside the so there are I I see a lot of effort in in doing this interoperability things, and I I think it's, yeah, yeah, it's probably what we need because otherwise also now there is these things about data spaces, like a sort of place where we could put all the data from different deliberations, different softwares, and still have them. And this is also quite exciting topic.
Yeah. So, so on the deliberative standard, we are chatting with the team, from NetAdolf and and the on ontology that they made, I think it's a very interesting and, very complimentary, to what we want to build as well. And yeah, so, ideally, we've been cocreating this. And the reason why we chose AT protocol, is, is because so for now for now, on Agora, the, the data is hosted in our server. And this is this is, so this is interesting for, clients who want to keep their data private. However, for the use cases where the data is of public interest and we want to make it more censorship resistant, we want to share it on a decentralized web, then AT Pro Total becomes, like, the the perfect infrastructure for this kind of conversations, for civic consultation, for example, of a country or a city and so on. The main difference, I think, between this, approach to the, democracy data space, approach is the difference between open versus private data.
Yeah. Maybe now I was using the word data space, without meaning a specific one if private or or public, more as a concept where data can be, sort of collected and connected. Yeah. But agree that it should be very important to to see at, let's say, the technical level how that data is stored. And also so you you were saying that, now, everything is saved on the server. And, of course, but so what happen if someone else, let's say, install because then they is it open source or not? Yeah. It should be open source. So if I install, it on my server and I use it with my community or my, then the data is is exchanged through the AT protocol between the two servers or no?
So so for now, eyebrows, is basically like, like, power. So whoever hosts it, host the data. But of in the, for for users, who want to create a conversation that is hosted on a decentralized network, we will be able to allow that. So there if they post something on, these data will be directly broadcast post on the, AT protocol.
Okay. Also, this part is very interesting. Yeah. Let's say I mean, something about your background. If you
so I had a very different well, diverse backgrounds. I actually studied filmmaking, when I was a college student. But instead of, pursuing a career in filmmaking, I switched to tech. And so I worked in Microsoft in tech sales, before I decided to to to cofound, the startup with my cofounder. And, the reason the reason why I did that was was, also just, witnessing how, AI was changing the world in such a rapid way. But our societies or how the societies work, seem to be lagging behind. So we wanted to, leverage AI to, use, to build alternative kinds of interactions among us, amongst individuals. So yeah.
But so now in, in the project, how many people are working on?
So we we are two cofounders working full time right now on the project, and, we work with a freelance developer and a freelance designer.
K. Cool. And, so you have seen different places. I wonder, like, I mean, humans are humans everywhere and conflicts are conflicts everywhere. Yeah. I was wondering if you see differences between different countries where you were where you stayed, in terms of how people were dealing with, yeah, conflicts, polarization, or these kind of things?
So, actually, I I I moved around a bit before I settled down in Europe about eight years ago in in France, basically. And so, well, I never I actually never really voted. I couldn't really participate in voting. And and so I was an observer to, how the systems work in France. And there is, there is, I mean, I think the polarization in France mirrors what's happening, elsewhere in the world. And, so and we also we work in France. We work with a team of facilitators who are trying to raise awareness of, civic tattoos like Agora and Polis to counter this polarization trend in France. And, in September, we collaborated with them on, during the during a national strike, called, Block two. It basically means let's stop everything. Let's block everything. It was, response to the to a proposed, budget plan to cut public funding. However, so it was a little bit like a follow-up to the yellow vest movement from pre COVID times. But the movement did not really lead to to more concrete outcomes, although the prime minister sat down because of the because of the tension in the society. So what the facilitators wanted to find out was what united the, protesters. Although they were although they were demanding, let's block everything, although they were expressing anger, in in times of tension like that, how what kind of consensus could we build? What was, the baseline for their collective action? And what they found so they they drafted about, like, 20 seat comments like that. And, eventually, more than 200 people participated, and they contributed to more than a 100, opinions. So they were sharing what they were protesting for and what they were proposing instead. It was interesting to to see that there was basically a very strong consensus that, folks needed to touch the ultra rich. And, but the main divisive statements were around, if they need a deeper reorganization of the, of the government, or not, which is more which is more radical. So what they did was to show how even in protests, there can be different groups of opinions that, to that this would also allow the organizers to see, what kind of, what kind of, consensus they can build and, what kind of proposals would alienate, parts of the,
parts of the community. Yeah. I was saying it's very interesting, this part because, many times I thought what if people, could, let's say, protest using, arguments like, saying what they really want. Because often what happen is that, maybe there are thousand of thousand of people that want to demonstrate about something that they ask about something. But often, there is also, like, someone that take that fight and, instrumentalize it. And so I wonder, like, what if people could really, like, express themselves. And, and also it's very interesting how also some the radical things they were asking for. It's
Yeah. Basically. I think what happens, sometimes in movements or, protests is that the leaders, they would, say a message, broadcast a message, and they and then the members would follow. In these words, this is powerful. That's like how Twitter works. But a lot of the times, people have more nuanced, demands. They they do have different demands. Although they unite behind the same banner, even though they they are behind one banner, they they are still, different, groups of opinions. And, using a tool like, Artura or Polis would allow everybody to see and, to see the the different demands, from from their from their community, from their members.
This is in some way connected with the question about the future. Like, do you think that in the future could be in ten, twenty years or more, like traditional politics can be, maybe not exist anymore or, like, or maybe yes, but with a lot of of participation from citizens. So, like, because, you you know, some sometimes people were yeah. How do you imagine, like, the future in these terms? Like, people all people participate, not all people, but some people participate. How do you imagine, like, a sort of digital democracy or what?
Well, I I think, well, this is like a super big challenge. And, I think digital democracy, it definitely makes participation much easier for for people who know how to use the tools. But accessibility, digital literacy is definitely an issue. And, so I don't think Tishito democracy would replace completely replace, what it is today. And, but I do think it it's very important. It's a very, very important, complimentary, way to to understand what people think, to understand what people want. And, I I I think we are also in a very exciting time when more and more, well, people, but also governments around the world, they start to, to adopt more, innovative civic tech tools, and I think this is a great sign. But also, civic society organizations are super important to to show, the, institutions what's what's possible, what alternatives are possible. And, and then these by showing these different alternatives and proving that they do provide values, this would increase adoption by institutions as well. And, yeah, I'm I think I think this is, like, definitely it needs the entire we need to grow the ecosystem because it takes it takes more than a village to to make it work. And, we are also relative newcomers to the ecosystem. Now it's been three years. But, apart from building Agora, we are also trying to raise awareness of, digital democracy in general, through organizing community events, or partnering with, other organizations, to just talk about, what you can do with these tools and, working with students, for example, to show them, how AI could be used for deliberation and so on. And I think this is all extremely it's small depths, but these are all extremely important, to to to build, to build this movement.
Absolutely. And do you have, I mean, anything you're struggling at doing, I don't know, in, could be a a future that you're working on, or could be something else. And also a SwigTech field funding is not always, the best. Yeah. So something, for instance, where you could, need the help. Maybe someone listen to the podcast and say, I I will contact the meeting.
Yeah. Yeah. Thank you so much. So, yeah, funding is definitely a challenge from the beginning, but, we were we were lucky to receive two credits from the European Union through the Horizon Europe's next generation Internet next generation Internet, NGI, since 2024. That's what allow us to to start building Agora full time. And, yeah, it's, TV tech is definitely not the most, well, profitable, the most easy field. And, however, we believe that there is a way to achieve, sustainability, while remaining remaining ethical. And, we well, our product is completely open source. But right now, we provide services to customize it and to to help other teams if they want to self host Agora. And yeah. And we are we are also well, we are basically pursuing more paying clients, but also more more funding, either from grants or from value aligned angel in this, in this coming year.
Nice. And, as the last question, if you have any message for the people that are working, yeah, on similar softwares or, like, in in the field.
Yeah. I invite whoever that is interested in this space that is already building or that have ideas to please join us. We need, we need more people. We need more resources, and, knowledge is, one one of the is the most, valuable resource in today's world and, your attention as well. So we are I'm grateful to whoever that's listening to this, to this episode. And, please reach out if you like to, know more about, what we do or in this field in general or if you would like to contribute in any way.
Thank you a lot, Yuting. If you'd like to add anything else, otherwise, thank you.
Yeah. Thank you so much as well.
