Secure Internet Voting (SIV)

Transcript

Uh thanks everyone for joining today. We have a learn shared with SIF which stands for secure internet voting. We got David here and he's been leading um this initiative of like democratizing uh better voting for everyone introducing digital voting in the US and yeah some success stories to share and also some of the requirements that um as some of us are very familiar with like blockchain voting you'll see how things differ once it comes to like uh local elections so I'm just going to leave it to David to enlighten us. Thanks, David.

Totally. Um, thank you very much for inviting me and thanks everybody for joining. Uh, just to introduce myself briefly, my name is David Ernst. I've been writing code for as long as I remember. I'm very much a technologist. Um I'm reluctantly bringing technology into government. It's sometimes rewarding and sometimes very frustrating, but I'm definitely much more comfortable in technical settings like around this crowd. I've been doing I've been working on democracy technology full-time for 10 years now in a lot of different ways. Uh my basic my basic feeling is that the way that our governments work is based on like 250y old constraints and ideas

and they're pretty good. The if you look around the world, it definitely could be a lot worse. It's better than it's ever been in a lot of ways, but I think it could be radically better in so many ways. And so, uh, specifically, I'm I'm working to to build a vision of free government that is much more inclusive, much more representative, much more accountable, and harder to corrupt, and also much smarter, and just much more responsive to all the the issues we face in our public worlds. And I think computers can play a big role in that in so many ways.

Just like it's computers and the internet have supercharged everything else in our world. Um or just to give you a little bit of more of a timeline, I started doing this stuff around 2016. From 2016 to 2020, I was kind of full-time focusing on um piloting and creating models around peer-to-peer representative democracy. Sometimes there's this buzz word liquid democracy. I've come to not really like that term, but it's I think the idea is extremely powerful and uh there's all sorts of reasons why it's super promising in a bunch of ways and almost any it's what's nice about is it's very inclusive of almost any good ideas for democracy. You can kind of fit into it. We can go into

more details about that later, but I don't want to get too stuck on that. since 2020, since the pandemic, I kind of took um a step backwards in order to uh kind of reassess. And one piece of feedback that we heard a lot was like, okay, we have all these really interesting ideas for uh for how better government could work, but we can't even vote in our existing electoral system from digitally from our phones or computers or using the internet. And so, shouldn't we start there? like isn't isn't that just kind of like the even more basic problem? And I think there's definitely some truth to that and there's a lot of opportunities that open up once we have our governments recognize us as citizens

in cyerspace on the internet. It's our basic fundamental right as free citizens. And right now we don't have that at all. And yeah, and so we in 2020 we started this project called CIV. Uh it's short for secure internet voting. It's meant to be a protocol sort of like SMTP or Bitcoin or Signal, but specifically designed around the requirements for governmentra voting, digital voting, remote voting.

And I'm excited to talk about the specifics of what's involved in CIV. But before I even get into that, I'm was also invited to just speak to the the differences from blockchain projects and political elections. And unfortunately, there are a lot. I I kind of hate to say it. Um I'm a I'm a big fan of blockchains. I don't want it to come across any other way. Um, I mean, I'm I'm not a big fan of some of the seedier, sketchier parts of of blockchain communities, but huge fan of technology and all the various ideas.

And uh, just to give you a little context, I I first purchased and used Bitcoin in 2011, something like that, for as just like, you know, buying things on the internet. And I participated in the original DAO, like the the OG DAO for Ethereum in 2016 in like May, April 2016, whenever that was with Homestead and Frontier that led to the hard fork and it was it was gamechanging. I mean it really opened my eyes and really honestly it was it was what Ethereum launched that really uh helped me to see how we could create verifiable collaborative computation at

such a at such a big scale. And with that said, I don't I think there's a lot of constraints in the current blockchain designs that doesn't make a lot of sense in public elections and and I'm going to go into that. But uh I just want to make it clear that huge fan of of so many of the ideas and I've been advocating tell anyone that'll listen about decentralized money and how great it is and how great it can be. And just one last thing on this note, since this is a PSSE call, I just can't help but but add that I personally think that privacy coins and priv and private money and encrypted money is super important. And like that's really where

one of one of the major things that I'd still like to see even more progress on. And I personally think that it's like just having a minimum basic level of privacy is fundamental for real widespread adoption. But setting aside all the finance and currency side of things going more into government and political side of things. Um yeah, how how is it different? What what are the differences? So, um, first maybe I could just even talk about like where voting is at just to kind of kind of make this visual. So, I I have some pictures to kind of illustrate this. So, this is what voting currently looks like for a lot of people

right now. You know, you have to travel somewhere even if it's raining and some oftentimes wait in line. Depends on where you live, how how many precincts you have. or uh maybe you can use vote by mail, but there's a lot of paper involved. I mean, millions of pieces of paper being shuffled back and forth and really tens of thousands of of humans that have to man all the booths and man all the stations and carry all the pieces of paper back and forth and make sure that there aren't extra pieces of paper getting stuffed in the ballot boxes or votes aren't getting lost or uh yeah, I don't know. There's there's there's

actually I don't know. It's a it's a really complicated subject. I'm not I'm not an election denier in any sort of way, at least in in the West for the most part. I've studied these things pretty closely. There's there's extremely widespread mistakes all over the place. Every single election that's ever been conducted has lots of mistakes taking place, but they're not they're very rarely systematic. They're usually just like random things happening here and there at very small scales, not enough to flip elections. But yeah, the if if people could see how it worked behind the scenes, some of it is not great. I mean, it's just it's a lot of paper. There's a lot of logistics involved and so a lot of things go

missing and stuff like that. And it's also the other problem is like there's so many different jurisdictions involved. So I'm I'm from the US. I live in San Francisco. And so in the US there's 3,000 different jurisdictions that all run their own voting infrastructure. And so you have some really good precincts that are really on top of things, and then you have a lot that are like just their their local governments are like tight on cash, don't have a lot of funding, don't have a lot of volunteers, and so they'll cut a lot of corners and stuff like that. And so, uh, yeah, it's it's it's a whole process, but this is what it looks like.

And it's it's pretty crazy. I mean, if you think about it, this is like everybody voting on like one day or one period. It's like one of the very few things that like half of half of our countries will do all together collectively. So, that's pretty awesome. It's like currently this synchronous process, but I think it could be a lot better. And so, the world that we want to get to looks more like this. And it's just super easy. It's the thing that is in our pocket that we all are already interacting with, you know, 50 times a day, 50 times an hour and just make it super easy, much more

accessible. And so that's sort of the that's sort of the vision. I think the pitch here is is to make it easy and fast and cheap and accessible and all those sorts of things. The problem of course is the security side of things. Uh this is why we don't currently use uh don't currently use digital tools for the most part for voting in government elections even though the entire stock market runs digitally and our entire news media runs digitally and people are even meeting their dating partners digitally and our entire entertainment and a lot of people's jobs are 100% digital and e-commerce and everything.

But voting of course has really hard security requirements. In particular, they're kind of uh you know, I like to think of as these are the three major security requirements for running a a secure election. And so it's like one person, one vote. That's sort of like the the first security requirement. You need to make sure that people are actually who they say they are and um that people can't vote a million times and only registered voters can vote and only people in your jurisdiction can vote. And so that's not the hardest problem in the world. There's lot, you know, every time you use a credit card, it's authenticated to your name. And there's lots of authenticated systems, but we still need we still need to get it right. And then privacy. I'm going to

just skip this one for now and come back to it, but really this is why comes why becomes so hard. But then the other thing that people privacy is kind of the last thing that people think about, but the next thing that people really think about is like obviously you need your election results to be accurate. you needed to be protected against compromise like integrity attacks whether that's the server, you know, if you're submitting all your votes from your phone or in paper, you need to make sure that votes aren't being changed, votes aren't going missing, that tally actually adds up to what everybody claims it adds up to.

Nothing pretty crazy there. But where it does get really hard, especially for a voting context, is the threat model is that you can't trust people's phones themselves. And so like this is this is the really one of the first departure points from most blockchain contexts is every device that every last voter is using is an untrusted device. [Music] and that makes things difficult. And then again, on top of that, we have these really hardcore privacy requirements in government elections. We have secret ballots. And so, nobody is supposed to learn how you

voted. The administrator is not supposed to learn how you voted. And so, this is a big departure from the way like our credit cards work and everything where, you know, Mastercard or Visa can see every transaction that you make or PayPal can see every transaction you make. And so you you kind of need something almost equivalent to like endto-end encryption, but it's unclear who the end is because it's not you don't you're not sending a message to only one person. You're sending it to the public to the public at large. And so that is unusually challenging and again the really hard part I mean there are lots of anonymous systems in the world. Tour has been around for 20 years at this point. It's people people like it. It's it's got a lot of respect. I

mean obviously in this PSSE call everyone or not everyone but I imagine a lot of people here know of all sorts of really cool cutting edge privacy techniques but the hard part of course is is doing all of these things simultaneously. Doing any one of them not so hard. Doing all of them simultaneously very hard because um you know when things are private then it becomes really hard to audit them and you can have authenticated people but then that gets in the way of privacy like tour doesn't have any notion of one person one vote. You can just be as many as many uh nodes as you want identities as you want. Same thing in like any blockchain like Bitcoin or base level

Ethereum or Monero or Zcash. You just create as many wallets as you want. And so that gets in the way of this. And so achieving all these goals simultaneously is is tricky. And yeah, let me just step back for a second. And the other the other thing is in the face of all of these challenges your threat model is I mean really we're talking about governments right like this we're electing our leaders for our governments and so it this is a matter of national security your threat model is every nation in the world if they have an ability to take advantage of your election they will try to

compromise your election all the hostile nations of the world that you can imagine even your allied nations of the world even your own governments um rogue insiders employees at all the various companies so like you know again I saying you can't trust the device like you can't trust the device manufacturers you can't trust the operating systems of the device you can't trust Apple or Google or Mozilla or yeah can't trust anybody because specifically because all the results are private. And so if all the submission, everybody's vote is being cast privately

and they can verify something on their own device that looks like, oh yeah, your your vote got counted correctly, but Apple can ship some back door to 10% of all the voters. They can now secretly change all the votes. And it doesn't have to be a whole corporatewide conspiracy. just has to be one rogue person that can corrupt the deployment pipeline. And the reality is there are lots of supply chain attacks. If you if you look at nation state level stuff, you know, stuckset had 23 zero days, I think, applied to it. And so, yeah, that that that's that makes things tough. So, anyway, let me just check out my notes again for a second.

um other big problems. So in blockchains uh as well just again big difference you know there's this notion of gas fees like in in Ethereum or in uh you know if you're going to interact with a smart contract users have to pay gas have to pay gas to to prevent against like DDS attacks and gas fees are a total non-starter in government elections that's you you can't it's a constitutional requirement at least in the US and as far as I know just about everywhere else in the world that you can't charge people money to be able to cast a vote called pole taxes. There's like there it used to be a thing like

150 years ago and it was seen as very discriminatory and so yeah that that's a total non-starter and uh okay so few more points I want to get to. So in terms of privacy the so like I was talking about you know there there's systems out there for privacy right now like tour and and PSSE there's semaphore and and ethereum there's tornado cache or uh I forget what the what the new one's called the new mixnet but all these

systems are based around the idea of anonymizing the users and that is not really the best approach in terms of government voting. There's a different approach where you like if if you think about how vote by mail works, you fill out a ballot on a piece of paper and you seal it inside an envelope and you sign the outside of the envelope with your name. I'm gonna turn off screen sharing for a second. You sign the outside of the envelope with your name and you send in your your envelope, your payload with your vote inside it. And so then the government gets a big pile of millions of pieces of paper and each submission is authenticated. And so that's I I believe really the way that uh political

voting really needs to work. And um there's a few reasons for that. So one reason is just very simple legal requirements. I I can't speak to every last country around the world, but in many jurisdictions, at least in the US, just whether someone voted or not, just true false, whether they cast a vote is actually public record. You can go look up every single person that voted in the most recent election. And so, a system that is anonymizing, like using a ring signature or nullifier or anything like that to try to anonymize who's casting a vote just immediately is out of compliance with that particular legal requirement.

Um, it also if you if you're anonymizing who's voting, that also means that you're you're now forcing digital voting to be the only option at least for for those people that are that are opting into it. You need because you need to make sure like the my vision our vision is that voting from your phone is an additional option or voting from your laptop or your computer. You can continue to vote from by paper if you want. You can continue to vote in person. You can vote by mail if your jurisdiction offers it. It's just for people that want it, they can opt into it, but you need to be sure that people aren't casting multiple votes using like both the digital system and the paper system. And so, uh, if you don't know who's used the digital system, you can't

do that. That that just totally breaks that expectation altogether. And so, that's another reason why anonymizing who is voting is makes things really tough. And then the last major reason is just that it's it's really nice if you know who voted because it makes it a lot easier to audit things. You can you can like see the progress of the election taking place. You can be like, "Okay, we see that this person, this person, this person, this person have cast a vote and maybe we want to like poke these other people to like get them to cast their votes in." And uh if everything is completely opaque and you have no idea who's even cast votes, that just makes things a lot harder to audit and like move things along during the

process and we we just don't know whether it's successful or not. So that's a that's a big key element of the privacy requirements that a lot of people kind of overlook. The good news is there are algorithms to to solve this. This is kind of why why we started Civ in the first place was because I was introduced to this algorithm that this person came up with. It's a it's a specific ZK proof for a cryptographic mixnet for a provable shuffle. And so the basic idea in civ is everybody sends in encrypted ballots. You get a big pile of encrypted ballots and then you shuffle them all up using this provable shuffle. And so uh every you get cryptographic proof that everything that

went into the shuffle came out of the shuffle. None of them got tampered. And on top of that we have al we have multiple layers of verification. So you don't have to trust any cryptography for uh you don't have to trust any cryptography for being sure that the integrity of your election is correct. You do need to trust cryptography for privacy because if the cryptography is broken then you lose confidentiality. But again it's opt-in. people don't have to if they're if they don't want to do that. Um, those are kind of most of the most of the main technical requirements that come top of mind. Just again, yeah, and just going back to the first thing about the O requirements because you have to have one person, one

vote, that means fundamentally you need to have somebody assigning a voter role and and defining a voter role. And that's in government elections. That's always your your election official, the election admin, election administrator. And so you you've necessarily introduced this point of centralization into your entire into your entire operation here is who gets to define and decide who belongs on the voter role. And so once you've introduced that point of centralization, a lot of certain blockchain ideas like um the the consensus algorithms like proof of work and proof of stake don't really make a lot of sense anymore because the

whole point of proof of work is about saying anyone can submit to the chain and anyone, you know, we we don't have an administrator. Um, you could use more like proof of authority style blockchains or like just git itself. I don't know if this is one of my favorite questions I ask people if they think that git is a blockchain because it's a cryptographically linked list. But there's no proof of work in it or proof of stake or anything like that. But the the issue with with those sorts of consensus algorithms and that sort of decentralization is just that it introduces all sorts of throughput issues and uh and scaling issues and so you know Facebook is interacting has 1.5

billion people whatever it is interact with their web properties every day and YouTube and Google and all the all the big things that that people are using are getting billions of transactions per sec or not billions per second but billions per day I believe and uh currently blockchains aren't equipped to do that because ma mainly because of the consensus algorithms and so you're you've already introduced this point of centralization through the o system and so then adopting the the like slower throughput and go gas fees and things like that I would argue doesn't make a lot of sense there are some nice things that you can get out of blockchains of It's nice to

be able to mirror the data and um give some integrity guarantees that once data gets onto once data gets accepted onto the public bulletin board to use the jargony term that it's not going to go missing. So there are some nice things there, but it's not really Yeah, it's the So this kind of brings me to the bigger picture or the bigger a bigger question, which is just like the setting aside the technical requirements, the cultures of um political spaces and government officials and election officials. It's just radically different from like cyberpunk people and uh and like crypto

spaces whether that's cryptocurrency or cryptography and just in like so many ways like where to even begin on the ways. So most most election officials are like the the closest analogy that I'd say if you if you haven't interacted with them much is that they're like accountants in a way which is that they're very good at following rules. They're very like buy the book sort of people. They're very nice. Most of I mean most election officials are very polite, very friendly, very nice.

They're not getting paid a ton of money. They're not like uh they don't I can't think of I've talked to a lot of election officials. Can't think of a single one that had that ever had an ego problem, which to be clear is very different from elected politicians. A lot of elected politicians do have huge egos and that's a separate issue. But election officials are not like that. Election, you know, these are people work in the department of elections and stuff like that. Super nice. Um but yeah, they're they in general they're being given instructions by they have their own regulators above them that say here's how you are supposed to conduct an election and they have you know set of rules and they just follow the book and follow the rules and very very few

of them are going to want to try innovative things or like stick their neck out or like do new things. It's just that's not the culture for so many reasons. you know, they're they're they're promoted for stability. That's that's what people are looking for out of these roles and that's really what the culture values and it's like um consistency and reliability and like being able to expect certain things. And so again, I love blockchain spaces, but I don't think that's quite how I would refer to a lot of cutting edge blockchain projects that are like, you know, launched overnight and blow up and crash because there are too many

CryptoKitties. And so, yeah, it's a it's a pretty different culture. Um, they're also not technologists. I I'm very sad to say. They I can't think of a single election official that I've ever met that struck me as having ever written a line of code of their in their life. They typically will have technologists that work in and around them. Um there are technologists that are like stakeholders in in the system, but they're not they don't have the final say really. And yeah, so they will ask me questions like like I sometimes I need to explain like how

uh just just like things that I take for granted like if I'm going to do a a network call like a post request or something like there's you know TCP has um like error correcting code and like acknowledgements and like check sums in these things And I don't know, maybe maybe that's too deep. I guess a lot of programmers don't know that either. But it it's like very they they don't Yeah, they struggle honestly. They struggle with like their email infrastructure. Like we make this civ software that makes it easy to run votes. And honestly, the feature that they love the most, it's not any of the cryptography.

It's the fact that it can send programmatic emails to thousands of people with a distinct link to each person and just like you know it's their custom link for people and it's really just like dealing with email infrastructure is like what they really appreciate whereas I like writing cool I don't know cool cutting edge technology so that that can be a little frustrating at times just culture clash there. Um, just to kind of drive this point home, I have a quote here I want to uh share to the screen. So, this quote is from this guy, Jake Braraw, is one of

the founders of the voting village at Defcon. It's big thing where they uh, you know, they buy old voting machines and people try to hack them. And then he he works in the Obama White House and the Biden White House as well. It's like he is a technologist that's that's connected to this stuff. So anyway, so he wrote this book a few years back and I just came across this quote yesterday and I thought it really illustrated this. So he says, "The other day I was on the phone with an election administrator, sorry. And after going around in circles trying to get him to accept the free cyber security assistance we were providing, he rejected everything I offered. Fine. And I just said, "Okay, how about I send you a list of industry standard cyber security best practices? You can look at

those and decide for yourself which of them you want to implement." The guy said no. He didn't want me to even email him a list of best practices. He said that he assumed whatever the best practices were, he would not have the time or money to implement any of them and so didn't want there to be any record that he had received such a list. The official thought that if the Russian hackers did breach his system and alter the results, he didn't want there to be a record that he had ever received a list of best practices because that could imply he was negligent in his duties and could be held culpable for the breach. He thought it was better to say that he never knew what cyber best practices were in the first place, so no one would say he is at fault for not

implementing something he didn't even know existed. Yeah. Um, just just for full caveat, this is an extreme example. That's why I made it in the book. This is not a widespread example, but it's not hard to believe. I mean I I've had a lot of meetings with a lot of powerful people uh in in government and yeah some of them can be very depressing from just like a wouldn't it be nice if we had secure digital infrastructure point of view I could tell some stories but

probably shouldn't at least when we're recorded. you can ping me when we're not recorded and I'll tell you lots of stories. Um anyway, so let me turn off that screen share. Uh I want to open up to questions. Um yeah. Yeah, I'll stop there. And oh, I didn't talk about civ at all. I didn't talk about how civ works, what civ is. I talk about how, you know, what our track record is. I can talk about those things. I don't know if people are interested in it. I'd rather open up to questions. Alandre said that there's like a Q&A tool in here that if you add a question there, then other people can

vote on it. Um, and so that could be cool, you know, voting and all that. Uh, yeah, but I want to open it up, let let you guys direct things. Thanks for listening to me ramble. I was also told by the way that I don't think people were saying that they cannot send comments in the normal GM comment box. So I can you try can someone try sending a Q&A question even just to test just to make sure that that one is working? Uh someone that's not an admin. Oh, we got a raised hand.

Nice. Okay, I think nice looks like viewers. Hello. Hey. Hello. Thank you so much for sharing your perspectives as well as stories um around digital voting. Has always been thinking like digital voting should be a yesterday thing rather than a future thing. Um but apparently yeah we do work with a lot of problems when it comes to u digital folding. itself. Curious about like how did you actually pitch digital voting to I guess governments in this case as it feels like the conflict or the tension will be super high when it actually puts for example the election officials or the administrators out of work which is potentially the person

that you are pitching to. So yeah would love to hear more about your experience on those fronts. Totally. Yeah. So um I'll take that sort of in reverse order. So you said it puts them out of work. It it does put a lot of the like volunteer pole workers out of work of the job of being a volunteer pole worker, which some of them have expressed some sadness about. Um they, you know, a lot of people say it's like fun to get to, you know, meet all their neighbors on election day. And so we definitely we're not trying to to, you know, that's that's a beautiful thing and we're we're not trying to discourage that. But the main election officials very much stay stay in their stay in their role, stay in their job. They the basic the basic way civ works is um it just it we

provide the tools to election administrators. They use the tools to administer the election and so they still have to manage the voter role, design the ballots, make sure the questions on the ballots are all wellformed and qualified and everything like that. Get all the ballots to people in the various ways that that that takes place. And at CIV, just for context, for government elections, we actually recommend that it's that you still send an invitation to each voter in snail mail. And so people scan a QR code. It's you receive your normal vote by mail packet. It has an extra page in it. I think I have a picture of it here that kind of shows you, but you uh yeah, here's I can share the screen,

but so the short answer is they still have lots of lots of roles to play. Um the thing about Yeah. So you you still receive like a invitation in snail mail and then you can scan it or it has a number that you can type in on your computer and then you can cast your vote on your phone and you get confirmation everything on your phone. So there is still a good amount of logistics, but the the simple thing is in a lot of elections that we've run, basically we've done a number of side by side of paper versus civ and basically you just need like at least 10 times as many human pole workers to run a paper election compared to a digital election.

And I think it actually scales even drastically higher than that. That's for like that was like a few thousand people side by side with both paper and civ. And the thing about elections the thing about like if you're if you're an election worker most of the year it's like a full-time job uh for government elections. And so most of the year you don't have that much work, but then on election day and election night, you have an insane amount of work. And you know, you're supposed to be collecting like a million people's votes and like overseeing like tens of thousands of poll workers and like hundreds of precinct locations and everything. And so, especially on election night, typically everybody, all the staff is is

still there all night long, like till 4 a.m., 5:00 a.m., 6:00 a.m. counting ballots. And like counting ballots, what that means at this point is taking paper ballots outside of envelopes and trying to feed them through scantron machines just like one after another after another times a million. And sometimes the machines reject them and it's like, you know, when you try to put a dollar bill into an old vending machine and it like spits it out and it's like really sad. It's like that sort of thing. And so they have irons, like hot irons next to them to iron the pieces of paper to get the creases out and folds out and try to feed them through. And that literally this fold issue is the reason

why like the last two presidential elections we didn't get the results for like weeks for like 11 days was because literal paper jams of like the ballots not fitting through the thing and so yeah all that gets eliminated by a digital system and so it makes their I would argue it makes their job and their life and their work much easier. It automates the boring, annoying, repetitive stuff and it lets them focus on the making sure the ballot's well qualified, answering questions from candidates and voters and reporters and and all that sort of stuff. The other thing, so this is like a really tricky issue because especially in the US it's

highly polarized, but civ civ provides end to end voter verifiability that election results are correct. And so what that means in practice is that it allows election administrators to prove that they are not cheating. And I again I I've never met an election administrator who struck me as a cheater out of the hundreds or something that I've met. Uh oh, having a low battery issue. Um might need to stop in a second. But uh anyway, so civ yeah there are you know in the US just to give an example there was some polling on this over the last few

years I think it's something like 70 million Americans believe that the 2020 election presidential election was rigged. 70 million Americans according to Gallup polling. It's like I think it's like 60% of all registered Republicans. And that is a big number of people. And basically um that's very sad and and unfortunate and um and so it's just nice to not to be able to if you're not cheating to be able to prove that you're not cheating.

And so they I think they like that. It also saves them um just one um just in the kitchen. Okay. Anyway, uh it also saves them some money. And the other thing is a lot of election workers um Yeah. Great. Now my power's back. A lot of election workers again they a lot of them they they like like their community and they want to let people vote. but they got into it for good reasons. And so, especially younger generations are asking them, hey, can I vote from my phone? Although it's not it's not strictly a younger thing. It's lots of people want to be able to vote

from their phone. And so, they would love to be able to make it easy for people to vote from their phones. And in our experience, most of them. And so, it's like a win-winwin in a lot of ways. The big the big detraction is they don't want to be first. It's risky. They don't want to. Their biggest fear is that the TV cameras show up the next day and stick a bunch of cameras in their face and they say, "How could you let this happen?" And yeah, like like that quote that I was reading from that guy. I mean, that was an extreme example, but it's like they don't want to be seen as liable.

And so the short answer, I mean, this has been a very long answer, but the part of sort of the the strategy, and I think someone else asked a question about this, is like how do we do this? Basically, we just have to like make this stuff a lot more um common and like accepted and normal. And so, we're working to make civ very easy for people to use just for like their friend groups and like you can drop it in a WhatsApp room or signal chat or iMessage drop. You can just drop a link and people can vote and everyone can have privacy but also have verifiability. We can use it in like I don't know some people have like social clubs or religious organizations or school clubs where they need to vote for like presidents and

officers and things like that. Just make it super easy for people to do that sort of thing. kind of specifically where we've had a lot of success with this is there's lots and lots of political parties and so like every city will have like the this city in the US there the Democrats and the Republicans and maybe the Greens and the Libertarians but and other countries of course there's all sorts of parties as well and so they'll have regular meetings like oftentimes once a month or maybe quarterly or they'll have big conventions like once a year or twice a And so they have to constantly vote on stuff again just like party treasurer and party secretary and and things like

that. And so what's that's that's where we've had a lot of success to date because they um they don't have legal restrictions. They they have tons of flexibility. They're very strapped for time and resources. It's all like volunteer driven. And so they love I mean it's just like super easy to use and we're trying to always make it even easier. And so they love that. And what's really nice about that is a lot of those people that are voting in those elections are the legislators, are the decision makers. And so we have to get this stuff legalized. Right now it's not for except for like some weird exceptions, it's not legal to be able to cast votes over the internet. And so even if a local election official wants to offer it, the state laws restricted

or the national laws restricted. And so doing through the parties works seems to be very promising just because it gets the decision makers a lot more familiar with this and then it's like okay if we're using this and we can see how we can actually all verify it and we can see why this short approach makes sense and we've done it a bunch of times successfully why don't we bring this to the rest of the million citizens and let them use it. That's sort of kind of like our high high level vision. And then simultaneously there are some really hard security problems that we've we we've solved all the like core easy things. The tricky thing about voting I was kind of going over those three problems but uh okay I'll get to that in a second. Uh there's like there's really

like 15 problems here. I have a cool slide over it actually. Um here it is. So I want to pull this up. So this is specific to civ. This is a little bit out of date. this document. We wrote this like six months ago. But again, just breaking down the like authentication requirements, verifiable results requirements, privacy requirements, and then accessibility. And like each of these are individual either like security properties or accessibility properties that Civ is offering. Um there's like some very minor typos in there. I hope you don't notice them.

This one is not there yet. This is we're this is work in progress. Aleandro is championing this. We're doing a partnership right now with Macy to bring something similar to Macy's style receipt freess into Civ so that we have really strong protection against bribery attacks. And then the other ones, we have these ones down at the bottom. We haven't spent a lot of effort yet focusing on like really strong denial service defense. Hasn't been a problem yet at all and there's a ton of good answers there. We're not too worried about it, but it's just not something we focused on yet. And yeah, like quantum resistance for secrecy, we haven't we haven't done a huge amount of work on. We have like

some very basic answers. We have quantum resistance for integrity. You can't change even if you have a quantum computer, you can't change vote results, but you Yeah, this is I don't know how much you guys know about harvest now to cry later type stuff. And then yeah, just like other availability attacks. So, and then there's like some Yeah, there's like a few hard really hard technical problems we're working on. And then we just did this. So, this we just did this big contest at Defcon like six months ago. And so, we invited there's like a prize like $10,000 and we invited everyone at Defcon. It's in Las Vegas, big hacker convention to try to break civ. These were the challenges that we that we encourage people to do. Um, the main ones weren't accomplished. One

person did an incredible spyware attack. I wish you guys could have seen it. It was wild. He It was in person, but he was able to get spyware on people's phones in like a third of a second if you you had to like hand your phone to him and it was just nuts. But anyway, we're we're going to improve around that. But anyway, this this has been really great just letting people bang against the thing and attack the thing and we're working on that. And so this has been really cool. A lot of a lot of cool um like election cyber experts have been starting to look into it and review it and take more seriously. Another like crazy thing about election space is all the voting machines are made by private companies

and in the US just because I happen to know a lot about US infrastructure 90 uh 92% of all the talent machines come from just three companies. It's like a huge consolidation. And these are like the Dominion voting systems. They were in these big liel lawsuits a few last year, a few years ago. And there's these two others, ESNS and Heart Inner Civic. But anyway, these companies are super locked down. They don't want everything's closed source. They do not want people examining the equipment. They cite all this like intellectual property stuff. They will like get the FBI involved. there. It's like a huge crazy thing, but that it's

like very much security by obscurity. And so the security posture just doesn't make a lot of sense. And as people and the places like the voting village, you can like if you have physical access to one of these voting machines, you can like use a pen and like pop open a little thing and like stick a USB stick in and like totally overwrite the firmware and change the vote count to whatever you want. and it takes like 17 seconds or something if you have physical access to the machine. And so that's pretty wild. Um there are some there are some hand recount statistical techniques that you can do to test against that which are really cool. But

anyway, the point is part this is all just trying to get to this larger vision that part of our whole strategy is invite people to try to attack the thing. Make it all open. You know, we have all the all the code is is online. uh we want to put money we we reward people for finding vulnerabilities for finding ways to attack it. Um so it's a work in progress. I mean people used to not do e-commerce online and now they do. Now we do tons of it. But yeah, I'll stop there. That was a super long answer. I apologize. That was super insightful. Thank you.

Yeah. Uh I'm gonna check some of these Q&As's. Okay. So, I'm going to sort by popular. Okay. So, John Gilding with four up votes says, "Can you talk a bit about how civ works technically? How does and that's question one. And then two, how does civ prevent collusion? Eg. I screen record my vote when voting and use that as proof to sell my vote." Okay, two for one question. So, yeah, I can talk about save technically. I will pull up some docs.

I'm going to try to keep it very high level because there's a ton of details but uh want to we can get into more details but I don't want to get too deep into it but at a high level here's a very very high level overview you have you you start by having a list of voters and like there there's basically five steps in a civil election like five key steps and so you have your list of voters and so you send out vote invitations to all the voters and so that could be snail mail invitations like I was talking about that can be emails if you have email addresses for people can be SMS if you have SMS for people done a lot of inerson elections where they actually hand people QR codes if you have a public key infrastructure like you have

a list of everyone that is supposed to be allowed you can use a public key registry and just cross verify against that but some way to to authenticate people as that's like delivering their voter off tokens and then on device people are making all their selections I'll go to this one which shows it a little bit better. So you have these prep steps and then yeah you have these invitations and then on device people are making their selections you know it kind of looks like I was saying the actual interface looks like this super simple you have as many questions as you want you can use different voting methods this is like choose only one but you can do rank choice voting and score voting and quad well quadratic voting is a little bit harder for some reasons but you can do

budget voting we've done some participary budgeting type stuff now but anyway you have some interface where people can make their selections And so when people are making selections, there's like a plain text of how they're voting. And so if they choose Angela, their plain text is Angela. They choose Jane, their plain text is Jane. And so then that all gets encrypted. And so this encryption is an Elgal encryption thing which is it's basically a diffy not to get too technical but it's basically a diffy helman handshake with one extra step where you use the shared secret to also encrypt and it's very well documented.

It's got the same security assumptions as an HTTPS handshake happening trillions of times a day. We're specifically doing it in the elliptic curve group. Okay, I'm getting way too into the weeds. Anyway, there's some encryption happening. You can read about it. I can go into more questions if you want. I won't get lost on into the details. Basically, you're preparing this encrypted vote. Sorry. And so then one other important detail with your encrypted vote is that your device on your own phone picks a random number. And this is like your verification code.

And every time you make a selection, your own device locally is picking a random number. And that also goes inside with your with your vote. And so at the end of the election, you're going to be allowed to verify that. So anyway, so you prepare your vote on your device. You this all stays on your device, all the private info that went into it. And then you just attach your O token, which came from the first step. And so you're sending in this payload that has your encrypted vote with all the field. You could have like one for mayor and one for president, one for Proposition six, you know, all the columns we call them, all the different ballot questions. And you also attach your O token. And so the election administrator just checks, okay, is this a valid OT token? You know, you you send that in. The election

administrator checks, is this a valid O token? And has it is it unused right now? And if it is, it shows up in a big old list, public ledger, of all the votes that have come in. And so it's just a big it's it's the digital equivalent of vote by mail. you sent in your vote inside this envelope, encryption envelope and on the outside is your name and so the election administrator can see who each one belongs to. So then you get a big pile of a lot of votes and you get some confirmation from the administrator.

Hey, we received your vote. We actually send you send you this email so that after the fact you can confirm the step. Each each of these steps has individual paper ver or ver verification methods so you can see if anything goes wrong after the fact. But anyway, then after you have all the votes come in, the voting period closes and this is when we do this cool cryptographic mixet, the shuffle. Step one is you just remove all the off tokens. All the votes get permuted. Now, this is my favorite part. This is like why I love civ more than anything else because it does this really fun operation called a homamorphic reenryption.

And I have been into cryptography since I was like I don't know 10 years old. And no one ever told me that you could do this, but this is like the coolest thing ever. What basically what's happening is you have an encrypted payload and you send in the encrypted payload to someone you know the administrator or anyone and the someone else can then change your encrypted payload to another encrypted payload without ever decryting it. And to me that's just crazy. I think that's very cool and thank you for the heart.

however sent in the heart and very simply I mean from a mathematical point of view it's like your your vote was X and when you encrypted it you were doing a elliptic curve scaler multiplication by A which is a one-way function and then the reenryption is someone else is taking your existing thing and adding their own B additional encryption and so the combined effect of that is as if it had been encrypted in the first place as A plus B just we can I I won't get too into the weeds But anyway, so you do you combine all these steps where you permute the things and you re-encrypt the things and then there's also a ZK proof of the shuffle.

This is the thing this guy Andrew Nef came up with. We have more docs about it. I can very quickly just in case anyone's interested, you can learn more about if you go to the civ homepage and research and it's this paper here. These are all the like academic papers that went into building civ. Each of these are like different threads. And so this paper built on this paper which built on this one which built on this one and that is yeah and so this is different thread anyway. So this is like the final key piece that we needed although we have now upgrade everything to elliptic curve cryptography. So that's some other cool new stuff but continuing on. So then you have all these shuffled votes and the really cool thing about the shuffling algorithm is multiple people can do the shuffling algorithm. So, it's

like I get to shuffle them up and you get to shuffle them up and someone else gets to shuffle them up and it's it's a multi-party computation type thing. Sort of like a powers of tow or trusted setup type ceremony. And it's got the same property and that as long as at least one person destroys their or you know doesn't doesn't collude doesn't share their their private randomization then the entire election's privacy is preserved and so you can just add more people. So that's pretty cool. But then at the end of the election after it's gone through all this these shuffles, you know, I give you a list to shuffle, someone else shuffles it, they give the next person, they pass it, they give it to the next person. Then the fully anonymized list, we do a shamir secret

sharing type thing to unlock all the votes. And this is it's it's cool. It's fun. And all what's also neat is all this happens in a browser. Nobody has to install anything. We have we have open source code if people want to install it, but you don't have to. It all works in the browser, which is fun getting that to work fast and and everything. And then you just have a big list of votes. And so everyone just sees the now anonymized list of votes and uh everyone can tally them up on their own device. And the nice thing is you had your own verification number. And so you can very simply just on your phone it shows you, hey, your verification number was 83921 or whatever. And so you can control F and find a list. And so it's super easy.

doesn't require any cryptography. We've done this with probably I don't know like 5,000 people at this point who are not technical, you know, every age group you can imagine, every level of sophistication and they're able to confirm for themselves that their vote is in the final tally. And so that's awesome. And so each this is your vote each voter is getting proof that they are uh that their vote is cast as intended. And that's sort of the high level of the election. Then there's all sorts of like post-election stuff like auditing which isn't on this page but we have a lot of cool stuff around which I get into. So I think I'm going to say that that answers your first question. I hope that answers your first question.

We go into more details. There's also a um there's like a technical spec if you want to read like the very detailed notes around everything. This is all on the docs website. You go to if you want to know how to find this stuff, you go to docs. But then you asked the second question about vote selling and collusion. So I'll get right into that. Um hm where to even begin on this? So the first answer is that we have to like calibrate where we expect vote selling is right now. Like the this is one of those this is another one of those like kind of myths that cryptographers have

convinced themselves that isn't quite true. There's an assumption and I don't know someone told me this was the first lesson they heard they were told in cryptography class that voting you can't prove how you voted. Well the reality is with vote by mail you can just take your ballot and hand it to somebody. You know anyone sitting in your house can already see it. So at least in person at least for in-person bribers or coercers that assumption has already gone out the window at this point. You can also I mean you talked in your question about um about uh like screen recording as you were voting. And so again I if I'm in a voting booth even in person I can uh I can record a video of

how I'm voting right now in a paper election and I can be on a live video call with a briber third party briber and they can watch me cast my vote. And so there is a mistaken assumption I would argue about uh vote selling that people seem to think that it's totally impossible right now. I would argue that it's not. I we have like these score rubrics. It's hard. It's you have to do some steps, but it's not impossible right now. that like first kind of just setting the baseline and then um yeah I mean we could get into all these reasons why it's like kind of like a overblown issue but let's not I don't I don't mean to to be dismissive. So we have all sorts of cool answers. We have this really

exciting thing that we developed called the vote sellers dilemma which is like this game theoretic thing because the key thing to understand about vote selling and collusion is that it's highly illegal. If you get caught trying to sell or buy your vote, you go to jail for two years and you can pay a fine up to $10,000. And so that has been on the law. It's a felony. You lose your right to vote for the rest of your life. Uh I mean it's a very serious thing. And the thing is when you are selling your vote, you are it's like no honor among thieves. You are trusting that the other person isn't going to rat you out. And so um basically we can exploit that. It's it's like a prisoner's dilemma type

thing type situation. If the two parties can cooperate maybe they can carry out this scheme. But the cool thing about prisoners dilemma is you can set up the Nash equilibrium such that for either party it's always better to defect. And so we can do the exact same thing. So basically what we say is if you get caught trying to sell or buy a vote you continue to get fined this huge amount. We have a slight tweak. Instead of saying $10,000, we suggest that you um you get fined 100 times whatever the agreed upon vote sale price was. And so that is that is your fine if you get caught. And then of that 100x fund, we actually reward the whistleblower, the defector with 5x. So if someone's

offering to buy your vote from you for a hundred bucks, they can turn around and actually they can be a honeypot. they can blow the whistle on you and they get paid $500 for that and you get fined and you potentially face jail time. It's like a felony. It's a very serious thing. And the point is obviously we don't want anyone doing this. We don't want anyone to get fined. But the point is we can use this sort of approach to disrupt trust between wouldbe buyers and sellers. And so this is a legal mechanism, a game theoretic mechanism. It's not a technical mechanism, but it's pretty cool. And Civ has some like specific things that can make this easier. and we can like write in the interface warn people about it and alert people about it and it has all these

interesting properties and we can go more into this. We got to try this out at Defcon this past year which was amazing um that we could talk all about that. We're hoping to do another big contest this this coming year at Defcon to even lean in even harder to vote selling concerns. But anyway, so this is the current current defense and current things. I think it's pretty good. we have this um page where we compare various voting methods. So this is inperson versus vote by mail versus civ. And this is across a bunch of different properties that I think are valuable in a voting system to have. And so each of these have like scores for like why why they get different things, but they're rated on a

scale of 1 to 10. And so if you don't have that boundary reward system, we readily admit that this is a weak point for digital voting is uh we just call this like collusion resistance. But anyway, yeah, it starts out weaker than in person, but it's not nothing. I mean, people most people don't really want to sell their vote anyway. But then if you we believe if you add these banner rewards, it becomes a lot harder. There are still some slight edge cases where the banner rewards seem to not work. Um, that was something we learned this past year at um, Defcon. And so now we're working on this new thing in collaboration with Macy. And so this is very cool. I'll pull that up silently. Canable votes.

It's still a work in progress, but this is using a circom circuit. The basic idea is that you can self-destruct your vote. And so you would you my brain fights so hard on defcon versus devcon. Yeah. Yep, I'm with you. Uh, this is Yeah. Anyway, so silently countable votes, the basic idea is that yeah, you have you still have the normal normal civ published vote total, but now we have an additional total of self-destructed votes. And so any voter can accept a bribe offer. You know, you I the Vladimir Putin says, "I'll pay you a $100 if you'll vote for my preferred

candidate." And so you can take his money, cast the vote the way that they want, send them cryptographic proof that you submitted a vote the way they want, and then you go in in person to the voting booth. Has to be in person because we need to protect against malware abusing this tool. These are the the key threat model that we're working around. Um, the coercer never learns if you've used this tool. The election administrator can't abuse this tool, which is a problem. There's Estonian digital voting and unfortunately it currently has this problem where the election administrator can can can cheat a little bit here. But then also malware on voters's devices. This is unfortunately a little bit of a problem for Macy right now. It's not a big deal

in a blockchain setting because if you have malware on your device, it would just steal your entire wallet anyway or just empty your entire wallet. It's not really like an additional trust assumption in a blockchain setting, but in our setting, it's not a trust assumption that we're making. And so, we have to be able to protect against this. And so, the basic idea is if you're facing a threat uh of bribery or coercion, you can secretly use this tool. You can come in in person and you say, "I'd like to cancel a vote." and you scan some private data off your device and then you cancel your vote and you can cast a revote in person in paper and there's no possible way for your briber to know whether you use this

ability or not. And so basically you can take their money but then trick them. And so the idea is again it just utterly destroys the incentive to want to offer people bribery or money. And you can use this for collusion too. So this is actually a stronger tool I mean coercion. There's a stronger tool against like there's there's been a lot of talk because of vote by mail around like um domestic violence type situations or like um like employers trying to force people to vote a particular way. And so this is an additional benefit that we can introduce in a digital setting that currently paper voting doesn't doesn't have all these strong defenses against. And so that's cool. Um just one last thing on this subject.

Um, this is, uh, this was the president of the United States in uh, I think this was 2020. And so, yeah, again, just the this is like one of those the realities versus the the imagined uh narratives that sometimes we tell ourselves in especially cryptographic settings and communities that it needs to be impossible to learn how anyone votes. But even in person, the president of the United States is doing it. And yeah, anyway, I'm going to stop on that question. This is boat selling is a huge

topic and it's been like yeah spent a lot of time on it. But let's go to some more questions. Let me turn off the screen share. Um so I also want to see Brian. Okay. Yeah. Brian on DevCon versus Devcon Q&A. Nicholas, in your experience, what would be the best roadmap to get digital internet voting implemented at a government level? I hope I I've already kind of answered that. Um, I think just make it easy, make it very common place for people. Do you work with countries outside the US? Who are countries that are more forward thinking? Cool question. Um, yeah. So, half the world are democracies. Half the countries kind

of weirdly, it's both half the number of countries and half the population, the global population. Kind of weird interesting coincidental reasons. Um, yeah, we just ran an election like four months ago in Madagascar. There was a lot. We were at Zuzalu. I don't know how many people here were have been involved with Zuzalu. And so there was a lot of talk. We we met with the now Montenegron president and they were super excited about it. They want to use it for like internal party stuff. Um yeah, we we were super excited to work with other countries. The slight detraction there is that there's there's like a lot of nationalism when it come and like weird feelings about that. like people I don't know people have like

really weird feelings where it's like oh I can't trust software if it came from another country and so there's like this home like people want to use like homegrown software especially for for voting like a matter of national sovereignty so that's a little bit of a tricky issue right now um we'd love if if you want to bring civ to your country we would love to talk to you there's there's been a lot of talk in South Korea they're really interested in it Um, Estonia has been doing digital voting since like 2006, I think. Again, they have their own homegrown system that has some cool things, but has some slight ways that it could be improved.

Um, Switzerland has been doing a lot of trials around some stuff and um, I mean, there's lots of countries that are trying they're trying all sorts of stuff. Taiwan has all sorts of really interesting, crazy stuff um, through the gov zero type stuff. But I don't know. I mean, my my overall view is like it's it's a it's it's an inevitable point of it's an inevitable thing trend that's happening. It's hard to find the first countries, but yeah, we're slowly but surely making progress. Um, I hope that answers your question. So, the next question, you said blockchains don't have optimal properties for government elections. Do

you think this will always be the case in the future? Could you see a government ever adopting a digital voting solution that uses a blockchain? Yeah totally. Um, I mean, I would never say never. Uh, the I mean, the big the big thing that blockchains need to figure out in order to be usable for for voting is just like scaling issues. Like the famous line was always like Bitcoin can only do seven transactions per second. The main Bitcoin mainet. Obviously layer twos and everything have totally different totally different situations, but just the reality is we you know you need to be able to support hundreds of millions of transactions per second to to to work.

And again the gas fees thing, but if you can do everything offchain and then you do uh like aggregation layers and and proofs that that's great. Then it's kind of like at a certain point like if you're doing everything offchain then does that count as a block? I don't know. This gets into like semantic questions and issues like what even what do we define as a blockchain? Again, this is my favorite question is is git a blockchain, not GitHub, but git itself because it's it's decentralized. You know, anyone can fork a repo. Anyone can make their own nar history of of reality. Every block is a commit and it's cryptographically hashed and it can be cryptographically signed and it's a

linked list pointing to the previous block. But there's no proof of work or proof of stake or anything like that. So, I don't know. There's no currency. I I don't know how much that that matters. But we have we I can show you if you're interested in this. We have a really cool design. If anyone wants to work on this, we would love to work with anyone that would like to build this. But we would love to right now uh whenever someone votes, all the votes just go into our centralized database and um then we post it. We have a public ledger and in some elections we'll actually also post mirrors of that to GitHub as like text files of the JSON blob and but we would love to also post it to

IPFS and uh other other things like that decentralized storage type things and we have this really cool design. I'm super into this, but basically the one of the big issues with CIV as it currently exists, you know, I was talking about the these steps. So, you submit your vote, let's say on day, you know, whatever day it is, but then you submit your vote then and you can confirm like, oh, the administrator said they received my vote and they give you this proof that they say you received your vote. But then when the final tally comes out, that's going to be at some later time after the voting period has actually closed. And so there's like a gap in the time and the time there. And so that's something that blockchains are pretty

good at like finality once once a transaction is accepted to the chain, it's final. The problem is like we have to, you know, we're not going to pay like huge gas fees or certainly not going to pay a gas fee for every vote. We could do some aggregation type stuff. And so that's really interesting. We have this really cool design that I'm super excited about. Basically, the idea is we would put a big pile of money into a pot. And so, imagine we say we put $10,000 into a pot. And the pot doesn't have to be. It can just be an analog pot, but it could be a smart contract specifically. And so then at the end of the election, you know, we when you submit your vote, we

sign your your transaction that we receive your vote. If your vote goes missing between when you submitted it and when the final results are posted, you can claim the money in the pot. And so if it's not, yeah, basically everyone in the world can see whether anyone's claimed it. And there's like this big concern that a lot of people ask is like, okay, you have all these cool technical things, voters can verify their vote, yada yada, but will they? A lot of people ask that like, I'll verify my vote because I'm responsible, but I don't trust all these other sheeple out there. This is like a very common argument that we hear. It's kind of I don't know. It's I don't want to get in I don't want to diagnose it. But anyway,

the point is that people have strong feelings about that. And so this is potentially a really cool answer because the point is that um like money is nice. If if if they actually think that something went wrong with their vote, they can just take the money and everyone can see whether the money has been claimed or not. And so we have the whole design for how the how the smart contract would work. Um Alessandro even showed us a way to make it drastically more efficient. So instead of having to submit all 100,000 votes, we can flip it to like a disincclusion proof and then you only need the logarith you know login through the Merkel tree to show which how many votes are missing. So if anyone wants to

work on this would please do reach out that would be awesome. And then the idea would the idea would be election administrators would opt into adding this as an additional guarantee on their election because it's an especially high stakes election. And so you can just put as much money into the pot as is appropriate. And if nobody claims the prize, the money gets returned back to the people that put in the pot at the beginning. So this is one way that we could get, you know, all the finality and 51% resistance of something like Ethereum backing our our civ uh ledger of votes. And so that would be awesome.

But the reality, I mean, again, I I I don't know if I hate to say it, but the reality is is like very very very few voters are asking for this and election administrators are almost anti-asking for it. Um, it's this is going to be a little depressing, but there's this paper. Can you guys Have I not been sharing my screen this whole time? Oh, no. I was like highlighting things. I'm sorry. Ah, let me pull out. So, I was going through this whole design of how how this smart contract works and like how the money gets sent and like the specific each of the

transactions along the way and what each transaction the smart contract does. And anyway, so this is on the civ docs website if you want to look into it. We have this thing called further research in progress which is like work in progress stuff that we're still working on. And so this one's called financially guaranteeing boat integrity. But anyway, uh I was gonna say just to this point about appetites for blockchains, the yeah, there there's like really bad blockchains have a very bad reputation among in political spaces and we could try to diagnose why that is and like what that's about and there's

like a lot of possible reasons, but just as like evidence from it, I'll just show you this pretty influential paper. It's called from bad to worse blockchain going from bad to worse. And this is like pretty influential [Music] and yeah, I don't know. You know, this is the R and RSA like the these are smart cryptographers. They they have lots of ideas and they're basically Yeah. I don't know. They're it. It's all because it's because this one company, Votes. I don't know how much you guys know about them. They really screwed things up. V O A TZ. They

they were pitching all this marketing around their blockchain. You know, they're election secure because it uses a blockchain. And it was all it was like vaporware. It just it was like there was no blockchain first of all, but they were saying there was. And then they eventually added eight AWS servers that they were in control of all eight of them. And it was like uh it's like um the Apache whatever that like Apache authority thing is. So it it was not like a proof ofstake blockchain or anything but the way I don't know maybe they've actually take it looks like they've actually taken off the word blockchain from their thing. I don't know here they still say it but anyway there was like a lot of

people and then there was this one called like Horizon Zero or something. Horizon Dawn blockchain voting and this thing they did like a Horizon State. Yeah. So they they did like a hundred million dollar IPO or ICO, whatever it is. And so they were just like throwing marketing dollars and it I don't know, basically a lot of people got really pissed and so now there's very bad vibes about blockchain in these spaces and so we have to like it like triggers people. It's it's not good. Anyway, that was a long-winded answer, but yeah.

Obviously like my attitude is like yes of course we need cryptography. Of course we want verifiability. Of course we want credible neutrality. Like all these things are great things. Let's not get like too hung up on I don't know. We don't need like a currency in in the middle of it like a I don't know. But yeah I mean it will definitely these things will definitely change. Anyway, I I'll I'll stop on that answer. I think I don't know if there are any more questions. Oh, what's one seen with the raised hand for a little bit? Oh, cool. Let's check that. Oops, I just raised my own hand.

Hey. Yeah, thanks for uh I'm coming through. My audio works great. Yeah, thanks for sharing all the thoughts. I find myself agreeing with you on almost all of it, which is a little concerning, but uh why concerning? That's a joke. Um yeah, I one I just wanted to say that I also agree that uh like decentralization and blockchains are often just like shoehorned into this stuff more than they need to be as an ends rather than a means and there's definitely a fixation on on that rather than the properties that they provide. But yeah, all this stuff is like incredibly interesting to me. It's a a big of my interest in this stuff is a big reason why I got working on the project that I am right now.

Um, I'm not totally convinced that like approaching governments first for this sort of stuff is like the right avenue for adoption. I'm really interested in like trying to figure out how to make democracy and and voting and polling an app and like productizing it and making it something that folks are actually interested in and and find fun to use. And you did mention that, you know, Civ does support, you know, just like smaller interest groups essentially getting together and doing this sort of stuff. I was wondering if you could just share some thoughts on Yeah. just like productizing this and and that uh strategy for like driving adoption of of new tech. Yeah, 100%. Um yeah, just to

give some examples. So like will like when I'm visiting my family, my extended family, it's like where should we order dinner tonight? Should we get Indian food or Chinese food or pizza or whatever? And so we'll create a little SIV vote just for that. And it's it's like surprisingly nice. It's the the whole idea is just like um preventing peer pressure and preventing group think and just letting people, you know, express their their desires in like a safe environment. And it's it's super fun even with like tiny groups of people. We've like I'm in a Burning Man group and so we'll use it for our like Bernie man camp. We had like 50 possible issues that we were trying to figure out like what do we prioritize among like 50 of us and so we just threw them all up and we put a score vote and so you just

decide you can say like not important to me, super not important to me, slightly not important to me, neutral, kind of important, most important and just it's we we had argued for like eight hours all night long. It's like a retreat we were on and instead we just threw this up and everyone could do it asynchronously in like five minutes. We're all at a house together, but we got I don't know whatever 50 time 50 is like 2,000 responses in like five minutes and it's automatically sorted and it's just like super easy and it just totally unblocked us and it turned out we had been spending way too much time on like little issues that people really didn't care about that much and it just gave us a much clearer sense of priorities. That was a fun one.

Occasionally we'll use it on like work meetings again for like lunch like just in these settings where like we don't you know we want to make a good impression but we do have preferences but it's like tricky of like how accommodating do you want to be? Yeah. Again just preventing group think. One thing I'm super interested in I use Signal all day long. I like live in Signal Messenger and so I'm really really interested to just make it super easy to just drop in a link and people can privately do a thing. I was in this one big signal group with like 250 people or so. And last night, these two people got into this insane argument for like 12 hours straight all about whether LLMs are useful or whether LLMs

are the worst thing ever created and destroying society and destroying the environment. And they just went at it for hours and hours and hours and hours. And I think there's some opportunities to like let other people weigh in in like a safe way about like hey maybe you guys should chill out or like let's you know should we do like moderation type things and group chats I think could be really interesting because I think a lot of group chats kind of struggle with that right now and so safe moderation tools I think is really interesting but anyway yeah I mean it's I 100% agree with you I think there's lots and lots of lower stake settings that make a lot of

sense. People talk about like homeowners associations again like the facult oh workplaces is another really big one. Um there's all sorts of really interesting ideas around like building consensus almost like um like labor union type stuff like organizing like I have some friends again doing some AI stuff and so they they happen to be very connected. They're in one of these groups, not not the pause AI, the like six months pause, but connected to one of the other ones doing similar stuff. And so they're really connected to like 95% of the top AI researchers in the world, which is only like 200 people. I mean, it's it's

a pretty small group of people actually. And so they right now all these researchers are spread across different companies. And there's like all these race dynamics between like, you know, who's getting ahead of who. And so there's a lot of fear about like race to the bottoms and all that sort of stuff. And so there's a lot of talk right now of making it easy for these people, especially across different companies to build coalitions either internally at their own workplace or across all the coalitions and say like this is a line in the sand. like this is something that I personally don't want to be like I want to set a hard line here but I'm afraid to speak up because if I'm just speak up individually I could just get

fired as like a troublemaker and lose my job but if it turns out that like oh 60 you know 70% of us all feel the same way it sure would be nice if we could find that out and so it's like it's always this problem of you know that the jargony term is preference falsification just like social group dynamics and like the difference between private knowledge and public knowledge and just like um uh coordination problems. And so I think there's tons and tons and tons of opportunities for much for getting out of some like weird coordination problem type situations where having an algorithm that makes it easy to say okay here's our group of people and here's how we all feel and making it really

easy but in a safe way um I think is very powerful social tool to to move that direction. And then also just like fun things like we were at a house party two weeks ago and we were playing this game um where it was it was like a getting to know people questions game. It's like this lady Esther Prell, she's like this podcaster where where should we begin? And it's like very um like vulnerable questions like what what have you never told your parents or what do you hope people wish that more people knew about you that they don't see when they first meet you? like it's like like a card like cards against humanity style thing where you're drawing these and they answer the questions and I personally

was really into it. People had been doing it for a while before I got there and I got in and when I got there it was like eight people or something and two of the other people had brought their own game like board game. It was like a dice game where you roll dice and everybody guesses how many it's called like liars dice. Everyone guesses how many dice are hidden under the cup. And I did thought that game sounded kind of boring. I really liked the the other game and so we were all in this like awkward moment because we're all we had all just met basically and it's like no one wants to be like rude and it's just tricky and so it would have been so great if I could just pull it out my phone and like written in like dice game

or vulnerable question game and just hold up a QR code and everyone can scan the QR code and click a button and hopefully it takes less than like 30 seconds and then we can have a very clear answer and you know hopefully not I don't know move forward without all the pure pressure and stuff. So, I think there's tons of opportunities like that. Yeah, I'll I'll stop there. Now, I need a link to that LM signal. It was wild, but I can invite you. Yeah. Um, do you have do you guys have my contact info? I can Yeah, I can send you. Um, I'm davidsiv.org.

SIB.org. Um, if you if you really want to join. It's it's really about cyber security. It's like it's like um hardcore cyber security people um like how to root devices and I don't know just got into a weird weird thing about that but yeah uh okay any other questions see Q&A. So the last one I'm seeing is what's one hopeful exciting message you can leave us with? Okay, that's a great that's a great one. I want to talk about I have so many tabs open.

So I want to talk about democratic peace theory. I think this is super cool and not enough people know about it. Um this is like in my opinion one of the most fascinating discoveries of the last thousand years. So basically these political scientists just compiled a list of all the wars that have taken place over the last 200 years and 50 years and then they just categorize them by of the two sides fighting in the war um was the side what you know is this side a democracy or a non-democracy and it turns out that democracies do not go to war against

other democracies like all the wars ers are like an authoritarian country versus another authoritarian country or an authoritarian country versus democracy and democracies just do not go to war against each other and this is like wars and then they did it as well where just any violent incident just like smallcale stuff and so I don't know I think this is like a really powerful idea because the point is like if we can just make democracy easier and work and be Cool.

And I think there's a right now democracy is kind of struggling in a lot of ways. I don't know if you guys have seen the news, but there's all sorts of there's really powerful things coming on the horizon or like there's there's all sorts of ways to make it easier, but the point is that we can um I don't know just even more eliminate war. You know, what's that? The world has 100 people graph. Um like the this one. Oh, I clicked wrong. This one.

So this is just over the last this goes back over the last 200 years. And so yeah, like if you look like poverty, it's like way down and vaccination is like way up and education's way up and literacy is way up. Democracy it's getting there. again it's at like 50% right now and child mortality is way down which is awesome. homicide is way down in a big way. And so the trend the the long-term trends are in a very positive direction, but I don't know my my personal view is like our government systems are really struggling. They like they just haven't kept up in the same way. Like our news system and our

education system and our financial system and all these other things have been supercharged by the internet and by computers and our governments are really struggling. But there are all sorts of really powerful models um around better democracy and better better representative democracy. Not just voting in the existing electoral system, but actually being heard and being represented. Um just to throw out a bunch of buzzwords, I was talking about liquid democracy. There's that's again very inclusive. There's all sorts of ideas around um like talk or in the blockchain space. There's a lot of cool things around like prediction market based stuff like futarchy which is some really powerful ideas and there's all

these like citizen assembly type things where you can use sort jury type things and there's some like LM powered big data recommendation engines and it turns out what's interesting is all these systems um all get supercharged if you have strong verifiable digital infrastructure makes it a lot easier to experiment with them. a lot of them require like one person one ID type stuff and so yeah there's there's like way better ways to organize and do things on the horizon.

Um yeah so I I can leave it at that. So, there's another hand raised. Ariana, I don't know how I find the Oh, by accident, she says. Um, okay. I mean, does anyone else have any other questions? We've gone for Oh, there was another one from Ceno in the Q&A. Oh, sorry. Thank you. Q&A. Long-term strategy for adoption. Civ as an app or CIV as a feature in existing apps. Hm. That's a cool question. Um, I mean, Civ is a lot of things. It's, you know, it's a protocol and it's a codebase and it's an

organization um of of people in the same I think of it as like signal like Signal is a protocol that is now in WhatsApp for example and it's an app and it's an organization of people. Um, one kind of interesting note is that the way CIV works right now is there it's actually it's all like web- based. There's nothing to install. I mean, you you can install like the web app to like administer elections if you want. And we have some ideas for some tools that would be easier if you install them locally, but it's kind of it's actually it's actually annoying to force your voters to have to install an app just to be able to uh to participate in a lot of elections.

It's like a, you know, it's like usually a once a year thing. But yeah, I think just dropping links. I mean, it's like YouTube, right? Like YouTube, like the internet spreads because of links. Like YouTube spreads because of links. Instagram spreads because links, but people do have their infinite feeds. I don't I don't Some people are interested in the idea of like an infinite feed of democratic questions. I'm less interested in that from an entertainment point of view. I think it would be awesome, really cool if I could post questions to my city or post questions to my neighborhood or post questions to my state and make it like really easy for anyone to post a question in a verifiable way to all of our neighbors almost like polling on

every issue but decentralized not having like top down polling agencies and have the responses be have like verified identities. So, we have some really interesting ideas around that. Um, but yeah, it's, you know, couch surfing.com had this famous thing many years ago, which was like their one of their KPIs was the they wanted to minimize the amount of time that people spent on the app because their whole idea is like we want to connect people in person and like have people have cool experiences. And so that's yeah, it's sort of my design philosophy like I I don't it's not meant to be like sticky or like uh be like engagement bait. It's like it's a it's a tool. You have a

problem, you want to pull people. You It's the easiest tool. It's we think of it as the easiest tool to do it and then it gets out of your way. I mean, it's almost like plumbing, right? It's like you don't think about plumbing until it breaks, but you just take it for granted. And so, I think we kind of want it to be like that or I don't know, lots of stuff like that. But anyway, I don't know if that answers your question. Um, as a feature in existing apps, yeah, some people have asked us if they could integrate it into other apps. you can embed it as an iframe. So, we did this really cool election like nine months ago with this group called the New Democratic primary.

That's actually how I got connected with PSSE and Macy. I don't know how much Alessandre knows that, but uh yeah, we it was in partnership with some other PSSE people, the ZK, the open passport team. And the basic idea was uh this was in the middle of the presidential primary process. It was right after Joe Biden had like a disastrous presidential debate and everyone was like what is this this we gota we do we we definitely need a better option and we saw one of the issues was my personal opinion was one of the issues was like everyone that I talked to was like obviously this we need a better option but it was very

unclear who that better option should be and so there's been tons and tons of really interesting work on better voting methods to get us out of these spoiler effect type situations and the grid lock and the there's like all these voting paradoxes. And so we basically set up this thing called the new Democratic primary. It's still up. I think it's new democratic primary.com or.org. I think they both go there. And so it you could uh attach a passport proof, an anonymous passport proof to it. You didn't have to. You could also verify via SMS and then we could later add passport proofs as as it became necessary. And then it gave you an interface for like 11 of the top options for a replacement Democratic nominee and you could rank them in a

really fun nice animated interface and yeah it got a little bit of attention on Twitter few hundred people but then like two days after we launched it Joe Biden announced he was stepping down and he was endorsing Kamal Harris and I think that was not great. I think primaries are a good thing. I don't think we should have non primary candidates for office or appointed candidates for office. But anyway, the reason I just brought it up was because one thing that was really nice about it is we just embedded civ as an iframe to make it work. And so now civ has support to be able to do that. You can just put it in in anything else. And so that's cool. Um yeah, thanks everybody for listening so much.

I was told again that that was an accidental hand press, the raising hand. Um, yeah, I I guess I want to do one more. Thank you for clapping. I I know it's been so long, but I want to do one last thing just to show you guys uh different ways of doing democracy. This is we were talking about like long-term vision. We did actually include it in the event notes. But so you know right now this is how our democracy works. We have a tiny number of elected officials and then they have lots and lots of citizens around them.

In the US it's 300 million citizens, 500 legislators like just about a million to one. It's like 800,000 to one. And so we made this person in the middle these are 25 people around the person in the middle. And so we made it 25 times bigger but in reality it should be a million times bigger to match our current system. And the reality is on we have like thousands and thousands of issues like policy questions, laws, civil right issues, foreign policy questions, tax issues, technology regulation issues, healthcare policy issues. I mean, there's so many issues and sometimes we're aligned with our representative, but I don't know about you guys, but I'm definitely not 100% aligned with my representative all the time. So, that's very frustrating

because they have 100% of the say. And so one of these other models is more of this peer-to-peer web-based model where people can weigh in how they feel, but you can pick your own representatives. And so you can start to do much more dynamic and uh bottoms up empowering of leadership. And so as people that you trust, you know, you want them to represent you. You can always override them. You can always replace them. But people can start to accumulate voting power. And we can have leadership without needing our current model of elections, which I think of as like a cage match. It's like four people enter the ring and they try to beat each other up as much as possible and one of them

exits kind of bruised and bloody and we all lose really because that's not a great way to do things. And so in this model you can um pick your own representative and you can even get more advanced. You can say like this is who I want to be my representative on foreign policy and there's a different person I want to be representative on environmental policy a different person when it comes to how do we regulate new uh pharmaceuticals and so you can get much more expert driven and these don't even have to be people they can be organizations they could be AI models but again you have accountability over them you could fire them and replace them as much as you want they can be all sorts of other democratic ideas like I was talking about with um like futarchy type stuff like prediction markets or

citizens assembly. There's a lot of different ideas for how to do citizens assembly. And so what's nice about this is you create lots of options. And so it's like anyone can set up a citizens assembly and if it's doing a good job, people can just back here with their voting power and say yes, this is doing a good job of representing me. And so I don't know, this is a fun thing. There's all sorts of really, really interesting implications of this. I could talk about this for many more hours. We've been going for so long already. So, I should probably stop for now. But I appreciate everybody taking the time to to join for this. It's it's very fun stuff. Um and I hope we can collaborate and hope we can continue to push things forward. Question, how to get involved

or contribute. Great question. Um well, the first thing that you should do, the simplest thing is go to civ.org and maybe like scroll to the bottom or something and put your email address in somewhere just so that when there are interesting things happen happening. We can keep you in the loop about it. I am very anti-PAM. We try we send out messages like once every two years or something like that. We're not very good about it. We could definitely do it more. So you don't have to worry too much about that. And of course you can always unsubscribe anytime. So that would be the very first thing I'd say. Um, the next thing I'd say is like it depends on what your what your skill sets and and things that you bring to the table are. If you're h if you're like a product person or technical

person, there's lots of cool product to be developed. Um, if you're like a security person, there's lots of cool security type stuff to try to attack and and analyze and find vulnerabilities in. If you're a politically connected person, there's lots of opportunities for using political connections in cool ways. If you're a very wealthy person, there's lots of funding opportunities. If you are into that sort of thing, um, I mean, it's going to be specific to you, but you can always reach out and we can we can find all sorts of fun things.

It's a big problem. There's a there's tons of different dimensions to it. So, I'm sure there's fun things that we could find that hopefully excite you. And yeah, um, just tell people about it. Ask people if they'd like to be able to vote from their phone and a safe way. Obviously, it has to be safe. We It would be a very stupid thing to enable digital voting that's unsafe and then it's just back doorable and the whole thing breaks. That would not be good at all. We got to do it in a safe way. But anyway, yeah, I hope that kind of answered your question.

anonymous person. Um, I think we've answered all the questions. Thanks for all the hearts and support everybody. Yeah, thank you so much, David. That was awesome. And thank you everyone for coming. Yeah, like this is great. I learned a lot and hopefully you as well. It'll be recorded so it'll be on YouTube, I believe, maybe a couple of days. Um, and yeah, thanks so much everyone. Thank you, David, again. And uh yeah, I'll see you guys soon.

This is superb. Thank you so much. Thank you, everybody.