NN Ep:12 From politics to protocols to protocol politics with Santiago Siri

Transcript

So today we're exploring the question that is often times discussed in the context of blockchain or decentralized technologies and which is essentially can we escape politics with technological protocols or are we actually just recreating new political arena just within a different set of rules and to discuss this we're joined today by Santiago Siri who is the founder of democracy art democracy OS and proof of humanity And Santiago has spent the last 10 years navigating this tension between technology and politics and essentially journey from pure politics to pure protocol to end up in a situation of protocol politics. And so

back in 2012, Santiago has co-ounded Il Part which is Argentina's internet party uh which is a particular political party where the representative were committed to vote exclusively according to what the citizen will decide online. But after a little bit of time, politics actually proved a little bit frustrating and corrupted. And so Santiago pivoted in order to build infrastructure for democracy because the idea was well perhaps if we cannot fix politics, we can route around it. And so Santiago then created democracy OS and then democracy art and eventually proof of humanity which are all tools that let

communities govern themselves without having to rely on traditional existing institutions. Now of course those protocols that were designed to escape politics rapidly became filled with political struggle with governance disputes and personal conflict that simply could not be resolved in a technical manner. And so in this episode, we're going to discuss all those things with Santiago with a view to understand how if protocols cannot help us escape politics, then how shall we build protocols that can handle the political tensions that they inevitably create.

Thank you so much uh Francie for joining us. Um so with this episode we want to explore specifically um your journey actually uh as you navigated through um the various uh layers of politics and protocols and then the politics within the protocols uh and so forth. um and then try to see how this can help us as we are theorizing and but also as we're trying to implement the network nations. And so perhaps to begin with like looking uh straight into like your political work

um you started in 2012 a political party um where uh the goal at least uh if I understand it correctly was um not actually take power over the system but rather using this power in order to change the way in which power is being exercised. And so you created party de la um and uh uh this actually was relatively successful I guess uh in terms of getting at least 1% um of um of voting in Buon. Um and the idea was that the p the party will just be a proxy in

order to express the uh voice of the people that will vote via uh the internet infrastructure. So can you uh provide like a little bit of a overview of uh basically what was the um underlying motivation of uh creating this new party and uh why did you think that the the the system needed to be modified in the way in which the power is being uh exercised?

>> Yeah. Well, so I here have the the original ballot that we used for the election of Partio back in 2013. That was the election we participated in the city of Wenocidis. And during those days, 2012, 2013, uh the world was going through a big transformation by the rise of smartphones. uh smartphones being used uh in the in the activist arena around the world were becoming very relevant, very important. Uh the Arab Spring, for example, was going on uh in the Middle East. Um the Chilean uh student revolution was going uh very nearby where I live in in in Chile. And uh it

was interesting to see how smartphones, social media, these new networks enable new ways of connecting society and engaging people in political discourse, in political debate. And um with part, we tried to h put an innovation that had an an actual impact in the political system. We didn't want to just be noisy activists uh making you know rioting or making making loud loud manifestation of uh uh of you know how we felt about the political system in Argentina. We were facing back then you know rising inflation and um probably the sensation of being under very a very

corrupt political class. uh things haven't changed that much since but we felt that having a political party that was an innovation in itself in the way it organizes itself in the way that the participation is open to everyone that uses technology that technology is a is um tool for transformation. Uh uh you know one of the things that we used to say back then is that pol technology is changing pretty much every single field out there in the world and being is disrupted by digital innovations of some kind except politics which still runs on the 18th century playbook of how a political system should be run uh with a

few innovations every now and then. So we had this crazy idea of making like a proxy party like you mentioned where uh we would have a candidate that is committed to vote every single bill in Congress according to how people tell them online how to vote. So was was a way of making more a more direct democracy with more direct participation with this proxy candidate. Um, and we we had the tremendous challenge of creating the party from scratch, you know, getting enough signatures in order to get the party approved to run for an election, campaigning for that election, and making sure our proposal reaches everyone in society. That's the thing about democracy and about politics is that you have to be able to make a

proposal that that's exciting for everyone, not just people that look like yourself. Uh, so we needed to learn how to talk to older people. We needed to learn how to talk to people that were not interested in technology at all. The political system itself, which is also very biased to towards its own way of doing things. And um it was an amazing I always call it a kicksotesque adventure because it was a bit crazy. Uh we we did a huge Trojan horse to campaign in the streets of Wenocides. uh because our moto was uh everyone should get should get into the Congress and our proxy politician would be like a Trojan horse attack into the into the Congress. We

got 1% 1.2% of the votes back then. Um an interesting fact is that we actually beat like in our little uh fight of small parties uh we fought against the Libertarians and we won against them. Uh we got 1.2% 2% and they got 0.7%. Fast forward to the present, the libertarians are running the country. Now they have MLE. Uh back then MLE did not exist at all. But it was an interesting proposal of using technology for for for politics. The the the real challenge was on the software. You know, we were focused back this is prior to smart contracts to Ethereum to

everything that happened with crypto back then. We did talk about Bitcoin a lot in those days. Uh and um this is we we we had the challenge of making a software that would enable people to get informed, debate and vote. Uh so we did democracy OS and it was an open source solution. Uh that very quickly got uh forked and adapted to different context and different organizations. Uh we found in Tunisia for example an organization was using it to debate the new Tunisian constitution in the middle of the Arab Spring or in Spain uh PMOS which was a left-wing party that was getting started in the manifestations that were happening in Spain after the the mortgage crisis there uh they started

using it internally for decision- making of the party. uh in Mexico, in Colombia, in many places we saw the software being adapted and used by local activist communities and um it was you know an amazing experience. Uh we were a small party. Small parties tend to be eaten by larger parties in the long run. Uh and the political system is just way too strong for a for a new coming party out of nowhere. But sometimes you know I guess if you have more political capacity you can achieve things and disrupt things in the world of politics.

Uh but it was an an incredible experience and and and yeah we we learned a lot from that. Uh I I ended up focusing more on the technology side of things than on the political side. >> But so this was kind of like you know in the in the in the theory of activism you have like those three um different actors. You have like the insidider, the outsider and the beyonders all trying to change the system but with very different approach. So this was quite of a insider approach to trying to change the system and uh did you did you feel that it was successful? Did you feel that it was not at all successful? And what what what were the what were the highlight and the downlights in this uh

journey? Yeah, the I think you know it it was it was a very powerful experience that you know getting engaged in political activism uh just gives you some kind of very powerful spirit that is shared with like-minded people and thanks to the party I met so many people from my city that were genuinely interested and engaged on the idea of using technology to make democracy more democratic right Um, in that regard, you know, I I see the experience as one of the most memorable experiences in my life and and and I don't regret any single thing that we did. Uh, in hindsight, we didn't win.

Uh, we didn't get the part the party member elected. We need like 3% of the votes to get it elected. Uh, we only got 1.2%. I think that was for the better in the long run because I wonder that in in an alternate history where we get the candidate elected, uh I fear that that elected candidate would get too corrupted by the system and would end up betraying the party and somehow ending on a larger party. I I >> guarantee right it was like a trust based commitment.

>> It was a trust commitment. your political capital was tied to this idea of >> voting according to how people vote >> and uh just curiosity like how did you ensure the one person one vote uh via the online platform >> yeah that I think that's where we saw the limits our main approach was well let's make this software open source right that was the one tool that we had in order to somehow guarantee some kind of transparency of how the software worked to people but obviously is you that you just you can check the code but you cannot check the execution environment or how the database is being managed or all of or any of those things. Uh one of the problems we

identify uh throughout the process was that like for decisions that the party had to make internally we use our own software. We ate our own dog food. Um but it was very clear to us and or at least to me that whoever had control over the database uh over this the servers over the system could very in a very subtle ways uh tamper the any election happening in the software by allowing certain people and their family members to register to vote and delaying other people or making a much more a slightly slower process for other people to register to vote. So whoever had access to the centralized resource to the server to the database can very easily manipulate how uh on a c

you know talking about the client server architecture and this is prior to Ethereum and smart contracts. So it was very clear to me that whoever controls the database or the server and especially on high stakes decisions where you really want that you have two sides that you know they really are fighting each other. uh democracy is not polling you know or making a survey. It's a it's a decision. A democratic decision is something you do when you have to make a difficult decision, not an easy decision. And difficult decisions are where you know high stakes um you know two two factions or more fighting for for the for a resource or fighting for power. And we went through those types of decisions in the software

and it was clear that there were you know some players in any democracy will do whatever it takes in order to win and whatever it takes sometimes means bending the rules and we saw that. Uh so any kind of centralized solution to our pretention of making digital democracy would end up being corrupted in one way or another and it was like a dead end to build the architecture of the software that way if we were going to be serious about doing digital democracy. >> Yeah. And then I guess um you decided not to be an insider anymore and to become a beyond.

Um >> I like those definitions. >> Huh. I I found them very useful actually. Um in terms of like trying to and you know many people are a little bit of itch but uh um like I think the party that I read was clearly getting inside the system to try to change the system. But of course the usual criticis uh be that insiders get uh is that you just end up being corrupted because once you enter into the system you don't want to change it anymore because now you have the power right >> um >> so as a beyond then you started building another infrastructure which is

democracy art um which basically um in my in my understanding of it is really like a shift moving away from the traditional field of politics uh and moving into more of a protocol-based approach uh to governance. So do you want to give a little overview of uh of this uh this um system? >> Yeah. So, so, so what happened after the party experience is that with the technology we built, uh, we figure well where we can, you know, do our best to push the limits on what we're able to do in terms of technology. So among many things that we tried, we uh actually

applied to Y Combinator uh as a nonprofit as a nonprofit organization and uh we were very lucky because 2015 was the very first year that Y Combinator allowed Latin Americans to participate in the program. So we were like a pioneering generation of uh South Americans going to California. Um we were also lucky or maybe not so lucky. We we'll see about that. Sam Alman was the president of YC back then. So this is Sam Alman before Sam Alman became Sam Alman.

>> I have good memories of him. Uh he was always very interesting to talk to and very always very helpful. I have to acknowledge that. And um so we we got into YC and we figured you know let's focus on the technology side. I was kind of frustrated with how politics work. Uh the experience of doing politics in Argentina beyond just the mere fact that you're trying to run for a campaign or whatever, you know, trying to get elected, all of those things sound very nice, but the nasty side of it is that eventually you will go through smear campaigning against yourself. So, uh you have to have a thick skin in order to tolerate all kinds of trolls, haters, and you know, malicious attacks of all

kinds. That's just how the world of politics sadly works pretty much in most places around the world. So you have to have a thick skin for that. And then uh the you have to you know if you're trying to do a new political party in any country and probably in Argentina as well it's quite likely that you will get secret service agents trying to figure out who you are just to make sure that you are not a terrorist organization or someone. You know, in Argentina, we had terrorist attacks in the past and and we had a dictatorship in the past. Um, but we I had to tolerate some like really strange situations which now I I would definitely

regard the actors involved in those situations as Secret Service agents that were trying just they were just just trying to figure out if I was a a terrorist or something like that. I I guess they figured that that I wasn't. uh but it you you start facing like very very a very corrupt system or a very toxic environment and and it's it's I got just fed up of the world of politics back then back then you know I was young I was trying to I I'm a technologist at heart and and I figured well I focus on building technology for democracy more than anything else so thanks to y combinator with the democracy earth foundation it's a foundation we started in California in 2015 focused on building open-source free software uh

technology that would help democracies being deployed anywhere around the world at the lowest possible cost. Uh and during the first years we started doing democracy pilots around the world working with local activist organizations in different places. We work with the umbrella movement in Hong Kong. Uh we work with um uh the peace movement or the peace treaty uh that was being signed in Colombia between the government and the revolutionary forces. And through those experiences we always try to uh explore different types of democratic games and different types of technologies. Um and eventually um we we drafted this document called

the social smart contract that kind of tries to put all of the knowledge that was available about how to do digital democracy in a on a collaborative open paper that anyone can can check on GitHub and still to this day I'm getting translations of that paper into different languages and eventually uh the rise of decentralization finally uh met its promise of enabling smart contracts. thanks to Vitalic and Ethereum. Um, I was an OG Bitcoiner since 2011. Uh, I love Bitcoin. I tried to build stuff with Bitcoin. I did all kinds of things, but it was clearly that, you know, it was not meant to be used for for doing anything else other than moving coins around. Um and the

with when Ethereum appeared, it was a very exciting piece of technology uh that would finally kind of enable the focus on on the challenge that we face around identity. Identity is the mother of all battles. If we were able to somehow figure out how to decentralize identity, then you could definitely have like a digital democracy that could work uh you know using uh blockchain technology and uh and uh we we started focusing on that uh and uh to the point that we wrote a paper called who watches the watchmen that researches almost every single decentralized identity protocol that we were able to put our

hands on. This is around 2018 or 2019 and uh eventually all roads led to proof of humanity uh which is this project uh originally started by claros a French cooperative doing um decentralized justice or arbitration arbitration systems and um and proof of humanity I always you know tell colleagues and friends that proof of humanity was kind of like part 2.0 though uh because that's where we actually were able to implement the use of of blockchain decentralization to the problem on identity and and and in that way create a the very first uh DAO in in all of the Ethereum ecosystem that was

truly democratic uh and that was another amazing political experience in in my life. So democracy art before um before proof of humanity um was about enabling people to do like liquid democracy, right? Um so how did how did you >> uh ensure the uni unicity of personhood in the context of democracy heart before the advent of proof of humanity? We did different pilots uh but all of them failed miserably. Uh for instance, when we uh on the pilot that we did with the Hong Kong umbrella movement, we were

doing Telegram voting uh with uh people that would register on a telegram channel uh and we would persist that identity on the Bitcoin blockchain actually uh using this new technique called Merkel trees that would allow on a single transaction to compress the information of multiple identities and uh we got spam armed uh our Telegram channel by what we suspect it was accounts managed by the Chinese Communist Party and they kind of like took over the digital we were trying to do a shadow referendum to elect a major for the city of Hong Kong and they kind of like uh took over the the Telegram

channel and the Bitcoin transactions and it was an interesting case where we thought that maybe we could put a barrier of defense on the cost of uh registering an identity on the blockchain using Bitcoin and and it was exploring also what what Bitcoin could how Bitcoin could index a large list of identities at the lowest possible cost using the Merkel tree technique. Um and uh you know we we underestimated definitely the interest that the Chinese had in in whatever form of protest that was coming from the umbrella movement to be suffocated by the Chinese Communist Party and that was a failure. Um uh we we thought that telegram was safe, it

was not safe. Um and um we were exploring basically uh interesting ideas that people suggested to us and we try to implement them and we try to figure out if they work or they didn't work and in most cases they did not work. Uh in in the with with Colombia we work with local activist organizations uh that would help us verifying the identity of people by uh checking their national identity card. But obviously that's that's a risk on the privacy side of things. So uh it was that that pilot had around 12 12,000 participants if I recall correctly. Uh mostly were Colombian expats that were living abroad

and there we tried to do a liquid democracy approach uh that would allow people to vote on every single aspect of the peace accord between the government and and the Marxist gerilla. Uh and it was interesting in the sense that we figured that most people were in agreement with the peace treaty except for the one clause uh that was granting $15 million per year and 15 seats at the Senate to the to the guerilla fighters to the guerilla. That was kind of like the deal breaker. And it was interesting that in real life uh quote unquote uh the the no to the peace treaty actually won uh on the referendum on the official

referendum. Um we were doing this on like as an alternative referendum for Colombian experts to be able to share their voice on the on the debate that was happening around the peace treaty in Colombia. So all of those you know and there were some others that I I you know I can try to you know I need to remember my mind is not so fresh anymore but we did we did you know there was another one very interesting one that we did with an elected congressman in Argentina um Fakundo Garto he was a in Argentina we were debating the abortion the legalization of abortion uh and it was a very heated debate and this congressman

decided to that he would vote according to the software that we built for party dear according to democracy OS. And that was an interesting case because it ended up the vote ended up uh on the platform where the vast majority of the country voted in favor of legalizing abortion in Argentina, but the the constituency of this legislature in the province of Tukuman where he he was elected ended up voting against. So he as a congressman had to decide that whether he would obey the national result or he would obey the result of his province just to understand. So because when you say like voting according to what the what the people vote on the internet uh did you actually

get a quum like where was there enough people or like you know how how much people actually were using the platform to vote. I'm Neon. I'm remembering the numbers uh just by by by my own memory. But uh I think that this experience with this congressman had like it was a big number like 40,000 participants. Uh I mean it was not we're a country of 46 million. Uh it was clearly something that was niche to his uh the people that were paying attention to his own particular voice in in in Congress, right? and those who were like activists that were either fighting for one side or the other. So it was it was not like

fully representative of of the country but it was way larger than a survey uh or way larger than you know traditional polling or some other approaches to that >> uh how how did you um again how did you account for the people like what was the one person one >> it was mostly email verification uh it was not it was not very sophisticated this is all prior to the use of of proof of humanity or stuff that that these were all the experiences that actually led us to yeah we need to figure out you know strong identity or else anything we do is is is really can be gained or can be uh you know manipulated in very cheaply. But yeah

those were the some of the commulative experiences of trying to figure out what it meant to do democracy in a digital way. As you know very well, sometimes the challenge obviously the challenge there's always a technological aspect to it. But sometimes the challenge is is just finding a constituency or just finding like a political context willing to adopt a technology and sometimes we found the political context and we had to do something for them whether we had the technology ready or not. uh and and in those in those experiences we we were figuring out well you know all all roads led to the problem of identity uh and in we got a a grant to de further develop

decentralized identity and research on on that uh in 2018 or 2019 and that led to or joining forces with clos with a proof of humanity approach and that's I would say that the proof of humanity da is the very first digital democracy experience with strong identity even though still there we still found like uh you know exploits and and ways of gaming the system that were actually very interesting to see as well but uh at least the attacks now were way more expensive than you know creating multim multiple emails or m multiple phone phone lines or stuff like that >> you want to do you want to describe uh what was the solution that was uh

implemented by proof of humanity. >> So proof of humanity required uh video proof of yourself um which today I would argue that is no longer val you know valid because AI kind of kind of can can do it uh very easily. Uh but back then this is all prior to chat GPT and the transformer and the AI revolution. This is 2019, 2020, 2020, 2021. You had to make a video proof of yourself where you show a a banner that displays your Ethereum address and that Ethereum address has to coincide with the address that you will be using to submit your proof. And it was an optimist system.

The system uh you have to put a stake of uh some if and the system would assume that your proof is valid if in the process of three days no one challenges your proof. Uh if in those three days someone figures, hey, this person looks exactly like this other one or this is a civil, this is a duplicate, this is a deep fake, whatever. If your if your identity or your proof gets challenged, then you go to a Clos arbitration court and the Claros arbitration court uh would decide whether or not uh the the the proof of humanity is considered valid for the registry. Um and uh it was an amazing approach. I think it was a

like a very interesting use case for Clato. Claos was looking for that use case uh in order to increase the bandwidth of their or of their arbitration system. I was responsible for developing the UVI token. Uh so uh I did the a universal basic income cryptocurrency that would incentivize people to register on proof of humanity. So that was like the carrot in the system where uh people once they got validated on proof of humanity they would get a stream a real time stream. This is a cool feature that I actually wrote the code for that so that's why I mention it but you would get a real time stream of fractions of this token on your MetaMask wallet uh at the rate of one UBI token per hour. Uh so that was

the incentive for people to register and um the the registry grew uh to the point of almost uh 20,000 uh uh verified humans. Uh it was u a fascinating experience because the DAO that governed the proof of humanity protocol uh was uh one person one vote. uh we allowed for vote delegation so you could delegate your vote to any other valid member of the DAO. So it was a liquid democracy. We had over almost a hundred different proposals that had to go through three phases like the ediation phase, the quantum phase and the final binding phase. Uh and uh all of these proposals

were about protocol rules um uh parameters in the smart contract uh the treasury of the DAO uh or the rules of the DAO itself. And um we had some of the most interesting debates that range from game theory to pre cryptoeconomics uh that fought for power in this DAO. Um the DAO grew to the point where eventually he had two factions and because he had two factions he had two superdelegates that had most of the voting power uh that were Clement the founder of Clos uh and myself as founder of the UBI project. It was interesting

for a long time during I think the first 18 months. Both Clement and myself had like 80 votes or 80 delegations each. Um I had to say I had to brag a little bit that we were winning most of the decisions on our on our end because people were complaining. I mean Cleos amazing team. I I don't want to I don't feel like trashing. I think they are amazing and they're doing onto great stuff and they have improved many of the things I'm going to be talking about for sure over the years. But uh there was this suspicion that a lot of the people we being challenged on on on proof of humanity were being challenged by for unfair reasons or reasons that you know they lost their stake their if stake on

the on the on the on their submission because uh they did not take the picture properly or they were not given a second chance in order to improve their submission. And uh a lot of a lot of people were starting to suspect that Kos was you know just uh manipulating the system in order to trick people in order to make bad bad bad submissions to the system and trying to extract the stake uh uh in in in different ways. um mixing uh money and justice uh leads to you know very strange incentives that might put the the justice might not be it's not easy to align money and justice. I I'll just say that it seems seems wrong.

Uh it's like a positivist approach to justice that >> ignore to corruption. >> Exactly. And and we saw that and a lot of people started complaining. So we started voting changes on the protocol and those changes were not in favor of the interest of Claros. Uh and eventually uh Cleman who is an evil genius uh I think he has that that that element of him being extremely brilliant but sometimes playing on the edge of what's allowed or or what's ethical acceptable.

he started farming delegations for himself uh with uh with the with the users that were coming from Burma in Myanmar. Uh because he had a a girlfriend back then uh from Burma. Suddenly all of the people that were being registered in Proof of Humanity on the same transaction uh they were delegating power to Clement. uh and that was clearly farming delegations for him with people that had no idea that there was a DAO how the DAO worked that they were not given the chance to get informed in a legitimate way in an open way to the dynamics of the DAO. So suddenly we were very you know Clement and myself held most of the

power of the DAO. Suddenly Clement was farming delegations in Burma and that was like an interesting attack that eventually that led to uh the DAO proposing a new law uh that was that delegations were going to be accounted using the the quadratic voting method delegations uh quadratic delegations. So instead of counting delegations in a linear way we would count delegations with a square root. The game theoretic angle for this is that this would incentivize people to delegate power to those who have few delegations and not to super delegate super delegates or whales that had a lot of delegations uh and would keep the democratic spirit of

the DAO going on. That also meant that I would lose a lot of power myself. But that that the the delegation farming attack that Clement made in coming from Burma would also lose uh impact and we actually won that and we implemented quadratic delegations. Glenn Wild who is the author of a quadratic voting paper both Clementa and myself were trying to manipulate Glenn. Hey you should support us. Hey, you support uh and Glenn was like, you know, don't bother me. >> She says, don't get me mixed in your political troubles. Uh but it was great because quadratic delegations meant that it would really put power back into

minorities. It would break the problem of super majorities forming in a liquid democracy. Uh we we that got voted. the fun part for me at least and the I still remember the day I found the result of this and I was so excited home alone looking at telegram uh but eventually after voting the quadratic delegations we voted to kick out Cleman from the mission board of the DAO he helped to create and that shows you that is giving power to the people and not to the founders uh and we won that by a tempo difference like 565 to 55 and that tempo difference was a consequence of quadratic delegations and to me that was like like an exciting

political moment. >> If you didn't use the quadratic delegation uh it would have been the opposite decision >> I think. So yeah, >> I I was wild in in all in all to being truthful to to to Clement and to the Clato team, they always accepted the when they lost. You know, a democracy that works is a democracy where the losing side agrees with the result. And truth be told, even when we when you know we were fighting heavily on the internet like passionately debating about delegations and about the dynamics and game theory and truth be told, Quan was always very respectful of the results of the DAO as as we were all

but obviously you know himself being the founder of the project and himself being on the losing side of very critical decisions he was always uh you know uh he he took up very well the the the political attacks we we our side made on the on the whole proof of humanity debates but it was exciting and I think we everyone all of us learned learned a lot there were many battles in that DA but I'm I'm proud that we were the one of the very few if not the only democratic experiment in all of the Ethereum ecosystem um Vitalik for a moment he got he never got involved with the DAO uh but he made a proof of humanity for himself and he

on a given moment bought all of the circulating UI tokens on UNIS swap. Uh so that eventually led to the UEI experiment also having a tremendous economic impact in a lot of people and that was also like amazing to see because then I heard the stories of many people that got benefited with UBI and and you know some stories were very touching very powerful and and uh yeah it was like an exciting time because we were on the on the cutting edge of of UBI of democracy of of you know identity of all of the political aspects that led that that are important in the web three world and and you know most projects are on the financial side of things. Our

project I think it was on the more on the political end of things. >> So democracy art eventually started using this. >> Yes. Yes. Uh so we you know everything all roads led to to proof of you know we we were you know strongly committed to to to proof of humanity. Eventually what happened is that um artificial intelligence happened and uh Clos forkit the project. So they didn't want to spend more time with with the UBI team or the people that were on on our end. Uh so they fork it and they did a version two of the protocol on Nosis chain uh which is exactly works exactly the same way as as version one.

There are not really many relevant changes. >> Uh still a video proof. uh on our end we feel we felt that you know video proof kind of you know has a is a dead end in on itself because of uh artificial intelligence and generative AI. Uh it's like the AI apocalypse scenario that we sometimes used to when we were interviewing people to join the project. Uh we always asked them what would you do on an AI apocalypse scenario and well that scenario arrived with the the rise of generative AI. I actually made some tests making pe fake people do a proof of humanity video and it's you know impossible to to realize. Yeah, it it really works.

>> Um >> so that led to Yeah. And and then WorldCoin came uh Worldcoin kind of took over uh the identity aspect. warcoin in Argentina where you know became very big has almost 4 million users out of the 17 million verified humans they have they have the whole a different approach based on on the ISIS and biometrics uh also you know that's a complete different game to what we were trying to do with proof of humanity and they're a private company whereas proof of humanity was kind of like an open source project um but uh but yeah I I I I bet I guess that today I see this whole uh part of the game. You know, the world

coin experiments is gaining a lot of traction. I I wonder where where they they'll end up. I never scan my own iris even though they tried a couple of times to really put the the ball in front of me. >> Um but what but we'll see. It's an exciting experiment. It it genuinely has reached like uh very you know many parts of of society. >> Okay. So um but nonetheless I think like your uh your journey is very interesting because it shows how you cannot really escape politics even when you try to create a protocol.

>> There's no way out. >> Yeah. protocol. Protocol-based governance is also politics but but on top of the protocol but there is also this additional politics on who is in control of the protocol like who are the governor of the protocol uh which is a whole new layer of politics that needs to be accounted of uh which I think often times is a little bit forgotten right like when we like when people talk about like governance minimization and things like this of like we we want to minimize the people making decisions but then there is always the people making decision about the protocol um which is where the real politics come about right yeah from like uh so from all those uh

from all those experience that you had like if we if we bring back the conversation around the network nation uh which is what we're trying to build and theorize basically network nations they need like three separate things uh first First of all, they need like a shared space which in this context is like the the network, right? Is like the the connectivity between the people. Uh and then we need the population and the population obviously require a concept of membership and therefore a concept of identity and so forth. And then it needs like the institution which is basically the government structure for the this

population to govern itself within this shared space and infrastructure. And so you have actually been working with all of those things uh at different layers. Um so according to you like nowadays in in this particular time um when we're thinking about the network nation um structure what will be interesting I will say I wouldn't say like the perfect because there is no perfect but what would be like an interesting or a suitable uh design infrastructural design for membership identity and ideally also for like governance. Wow.

That that's that's the the open questions of every single day and uh sadly there's no formula or like closed answer to it. Um I think that the the elephant in the room these days is AI. uh you know there are amazing things to it that we can actually leverage and take advantage in order to um enable better governance or better human identification uh and you know it's a very powerful tool at the same time it's it's a tool that you know we we're still it feels like you know we're still in the very early days of what generative AI or these new

transformer models what can what they can do where you know whether AGI will happen or not or if is AGI real or not we'll we'll see but certainly it's extremely powerful and you know it it it brings back some of the nightmares doomers are asking you know warling people about on on the sense of you know that there are risks associated to the use of AI at the same time it's it's also a tool that's you know in cyber security Right now the sexy topic is AI jail raking and it seems to be very easy to do in if you have different creative approaches to how you prompt uh these systems. Uh so it feels that it's

not the safest or the most secure way to to do things uh these days. Um maybe it's a mix of blockchain and AI uh in some way or another. I am far I feel far away from having like a close concrete answer like there are some clearly loose ideas that I'd like to explore moving forward into the future. Um like for instance uh what if an AI can be like uh the you know uh the executive power of of an organization and has a autonomous control of a blockchain wallet that you know has the the funds or the resources of a community or an ecosystem of people and you have to

pitch that AI for for a grant and the AI will decide based on its understanding of uh the common common good or the or the shared values of the community you know then that's something I would like to see uh in order to sort out >> have you seen the recent experiment um from entropic they did this very um very interesting thing I created like this uh vending machine uh which was embedded with an AI and the AI was instructed to be a profit maximizer to like just trying to negotiate the price and like just try to make as much money as possible. And then during like the beta testing, they managed to convince the AI

to become a communist and to give everything for free. [laughter] >> And then and then they tried to create it again and they were like, "Okay, we are we're going to set up guard rails so that it's never going to be a communist again." And despite the guardrail, the journalist nonetheless managed to manipulate like just like it's just like psychological manipulation at this point. Uh and then the AI again decided to give everything for free. [laughter] >> Well, we have to be careful treasury of a to an AI.

My one of my favorite uh well you you showed me this this fun experiment called semocracy uh where you can create an agent uh and you write a manifesto for the agent and you put your agent to debate with other agents and they pre deliberate a decision and you get to simulate the congress and and decision making on any given issue. I I like I love this experiments. There's also this other experiment of an AI that has um is in control of a treasury and you have to pay money to prompt that AI and all of the money goes to the treasury and you have to trick the AI into giving you all of the money available in the treasury to yourself. And I think that this AI

went up to $50,000 and and and someone was able to trick trick the AI into giving the the money to the person. And there was this this recent thing that uh someone showed me that it was a video game where you are you have five minutes to negotiate with an AI uh that is acting as a terrorist uh and is about to you know trigger a bomb. So you have five minutes to negotiate with the AI and uh persuade her or persuade it to not triggering the the bomb. Um, I think all of these games and all of these >> uh exploration with with how we can use generative AI in in in real context

scenarios and with management of resources or or money are extremely relevant. I feel that it's still not safe. It's still not secure for large experiments. Um, my friend Matraisano who who recently passed away, but he he's an Argentine legend in in the in in our hacker community here. He was doing this VC fund uh called Angel AI and it's an AI that you would pitch on Zoom and you would it would just interview you. you would share your deck and all of your uh company files and the AI would decide in the next six hours uh whether or not it would give you a $5,000 check to your startup. I think that that kind

of idea would end up badly very fast because people are very smart when there's a bounty and they figure things out and AIs are still you know building guard raids around them to avoid jailbreaking is uh seems to be very hard prompting I think is still a very nent and very young discipline we uh it's very strange when when you write software for a or you write AI agents because you have to kind of like you're like uh play right you have to give the AI well if people talk to you this way or they ask you this way or they they are requesting this thing you should do this or that and it's sometimes not very clearly not very clear that the AI would act according to

what you expect the AI to do >> but maybe a model like a proof of humanity where you have the AI making decisions and then you have humans that can challenge those decisions you know you have you have a limited number place in which you can challenge the the decision that they are made so that if people are actually doing like advers adversarial prompting at least the humans can detect that uh but for every common and routine task then humans don't really have to intervene and they can do its own governance >> I like that humans can go crazy though they can find you know they can manipulate it's you know there's there there's also a project trying to do a

kos competitor using using AI uh to judge and and to read like uh contracts written in uh natural language instead of smart contracts um called the gen layer. They are very young but they raise a significant amount of money. We'll see how that goes. They're trying to really bring together the kind of consensus dynamics you find on blockchain uh and using LLM as nodes that would decide whether or not to execute uh a transaction based on the contents of the what they call in intelligent contract instead of smart.

>> Now the problem that I see with um all those solutions though especially like it to me it reminds me a little bit of the problem you mentioned for the party dear which is If there is if you use an AI, most likely you have a large data center of GPUs running somewhere. Um, which means that all of you are allowing whoever is controlling this server to change the prom to change whatever they want because they are doing the inference.

And so all of a sudden like all this corruption that you were mentioning that if you have a centralized server there is inherent corruption possibilities when you use an AI unless until maybe I should say until we get like the centralized compute um mechanism in which there is no single operator that is controlling the running of the AI. I will not trust running my network on on open AI. Right. >> Of course. uh you know so student you're like we're recreating the same problem mostly because AI as opposed to smart contract require a centralized server and operator.

>> Yeah. And and that's the reason I think most of the like AI tokens out there were not really taken very seriously by the market because yeah anyone can you know supposedly you have these tokens that would you allow you to make a node with an you know GPU and have an LLM in the GPU you know running on that on that node and you if you get paid tokens you know you would get give the the person an answer coming from the LLM but most people could actually, you know, just give you an answer that's cached on the DB instead of actually using the LLM and you would get like uh an answer that is not really spending the money to use the LLM. And that just and and that's why if

you you know I think the new generation of Nvidia cards are you you need to do at the hardware level you need to sign uh the fact that the computation was run on the GPU and not being just a database call on some random Amazon web service you know infrastructure call. So that's why AI decentralization tokens that most people see them as a fraud. Uh well well at least some smart people I know because you know you cannot prove unless you do it via hardware via you know and and I think that only the most recent generation of G Nvidia GPUs allow signing the computation with via hardware of in some way. So that would

give you at least a sensation that it was actually run by the GPU. I'm not a specialist on this topic, but I I can tell that, you know, if there was a successful AI token on the crypto world, we would realize by now in terms of market value, and still there isn't one. uh so I guess the market is still not very convinced on on on on how AI and blockchain can be brought together but um >> there's a lot of experiment but yeah I guess there is none that has reached a point of uh sufficient maturity to be trusted except for like the trusted uh execution environment system but it's still running on a server >> yeah TE that's another one that's

another interesting piece of technology out there for in terms for privacy and and uh yeah I looked into that some some time ago. All of the last dev dev connect was around this these topics right AI and blockchain you know how how they merge together how they can work together and I think you know still still many avenues to be explored and and still you know very little results to be shown. Um one one last question I wanted to ask uh concerning proof of humanity uh and network nation. Um would you would you say that to some extent uh the world effort around proof of humanity was somehow a proto network nation meaning like there is a global

population within a particular registry with its own governance system with its own economic system through the UBI like does it does it look like a network nation or uh like what was missing for it to actually be regarded as such? No, I I think it does like a proto network nation. No, not the full one. >> Yeah. >> Uh we were not singing the anthem of Proof of Humanity just yet, but it it felt like a a tight community and uh like most crypto projects out there, right? I think Ethereum is also like a protoet network nation or Bitcoin is like a protoet network nation. They're all like new levatans, new new bureaucracies uh that are being

constructed around economic incentives and uh certainly the identity element is is one of the of the biggest components uh that we could uh use for the construction of network nations. Um there's also like interesting efforts like CK passport which tries to leverage or piggy back on existing passport or existing legal documentation coming from governments and and you could make a ser knowledge proof on top of those in order to make sure that you are dealing with a unique human uh without revealing who that human actually is but you know you at least have this the proof that that human holds a valid legal document from a traditional nation state. Um but yeah it's it's I

think that you know these are all loose components very experimental uh that eventually will you know we'll you know we will see in our lifetime for sure the emergence of network nations that will have a a a lasting cultural strong cultural impact in the same way that Ethereum, Bitcoin and and and many projects in the ecosystem had uh you know you know Bitcoiners and Ethereums you know have very strong values, very strong principles. Uh we know, you know, where where, you know, how our communities think and operate and and even though we fight internally sometimes for one thing or another, you know, you you just can tell uh those

shared values in in in are expressed in the different community members. And I think that eventually as some a more complete solution that has the financial element that has the economic you know components and has the identity governance uh and hopefully democratic uh components as well will be like a much more complete and much more interesting uh in my view at least proposals to to what we have you know the strictly financial things that we have seen over the last 15 years since Bitcoin appeared. Yeah, I feel we already have a few uh networked nation meaning like we do have those

communities which have this common sense of belonging and identity and and to some degree kinship. Um but most of those do not yet uh organize themselves in this like true a self-s sovereign infrastructure. uh and usually many of them relies on like centralized corporate um entities or still rely on like the ideas of of existing government. So it's like I think like identifying the right mix of uh infrastructure like I think the institution of the network nation is a very hard one to nail down. Uh and I do I do agree we have like a lot of bricks

uh but we didn't yet fully understand what's the building uh that can be made that will actually uh empower those um those communities to self-govern to mutualize resources to like deal with all the administrative and governance stuff. >> Yeah. Yeah. the the road ahead is very long uh and and you know many many challenges. >> I think that the you know on a positive note that the >> the kind of experience know how knowledge that was accumulated uh I always look at the birth of Bitcoin like as as an interesting milestone that led to the a new generation of builders,

hackers, programmers, activists to start thinking in this direction. uh that the experience of the last 15 years has been significant. Um you know I I try from coming from where I come from in Argentina with democracy earth and everything we have done uh since party and everything I try to share as much as possible on this knowledge. So hopefully new generations will figure it out in a in in with new technology or with better approaches moving forward into the future but hopefully by not repeating some of the mistakes of the past. But I think that that we went a long way when I look at part in 2013 and then I look at the the um proof of humanity in 2022 and in that 10y year gap uh like I felt

a leap I really felt like a leap forward uh you know we had new kinds of challenges but definitely the the improvement of going from just an open source solution to decentralized smart contract identity uh on a blockchain erh It was exciting to see you know how far we we we were even though we were still like a niche thing of you know hardcore geeks trying to figure out the future of governance. Uh but you know those leaps are significant that you know hold good promise for the future.

>> I think to me what uh the challenge with all these uh decentralized uh solutions which eliminate the corruption at the administrative level. Um but then and they I think they are good because you can trade and like it's pretty good as a as a as a governance infrastructure but the problem is then um the question of who govern the governance infrastructure uh and this is like a different types of politics but that's a very challenging one as well. If you look at the Dows out there like Deententral just to give you an example of good friends of mine, >> they have some people have controls of

the multisc that controls the funds and some people and the pe and the users have vote on the DAO and many times what happens and it happened in proof of humanity as well. uh what gets voted finds itself on a gridlock with those who actually have the the the facto power of the multisig or the treasury of or you know in proof of humanity it actually did happen also and and you have this decentralization generates this gridlocks where you cannot enforce in a secure way or in a trustworthy way what's being voted on the on the actual execution of what happens in the smart contract. So maybe AI can do it. Uh but AI to this day is extremely easy to

manipulate. So you know uh it's a it's an open problem right now >> and AI has not yet a proper connectivity with those decentralized infrastructure. So that's yet another >> that's a big one. >> That's a very big one right now. >> Okay. So um just as a closing uh thought so because you have like a lot of those you have a big baggage of experience in this field uh and like our little network nation uh collective maybe uh has also experience but different ones.

Um if you were uh to like give like if there is one lesson that you've learned and that you would like to share to people interested in setting up new networks um what will be like the the key lesson that you got that you can share? >> What's the key lesson? Let me prompt my AI here. and see what she tells me. Um, you know, one of the the key things is is that you must make sure that you are like

legitimacy is not declared. It's it's enacted. You know, you you you need to to convince the world that you matter. Uh, and uh quietly becoming something people can't do without. um you know uh it's it's it's it's one of the you know at the end of the day it's like a cultural challenge. Uh networks don't win by by being recognized they they win by being used. Um and I think that's that's the thing you know if you provide to people around the world like uh an exit uh that's useful for them uh that uh you know really helps them you know you know

really gives them a place uh then you have something very powerful um but yeah uh well I think it boils down to to legitimacy >> so like acting already like practicing the things so that >> yes >> people come to you >> because they see you as being legitimate but but also I will say like legitimacy comes with this kind of guarantees right it's like >> I I think the decision is legitimate because I believe that the process was guaranteed to be in a particular manner so there is uh there is there is the trustbased type of legitimacy and then there is the confidencebased type of legitimacy

>> yes yeah You know networks become real when people use them, rely on them and and you know uh and and they make all structures optional. >> Okay. Amazing. Thank you, Santi. Uh where do people go in order to follow up on uh what you have up to? >> Anyone can follow me on Santi City. Uh I'm most almost on most my days I'm I'm on Twitter also known as X and on Instagram. Uh lately I've been mostly diving into the AI um arena of things but uh web 3 will always leave on my heart. So I always come back to web 3.

Uh but yeah you can find me on Sic City. I tweet nowadays uh mostly in Spanish. So you can easily translate with uh my content with the translate button. Uh and uh yeah looking forward to to to meet uh any idealist coming from this community. >> Amazing. Thanks a lot. Thank you, Prim.