365: How Augur Protocol Is Bringing Truly Decentralized Prediction Markets Back

You know, auger's back, we're bringing a fully decentralized trustless information this protocol. That means the same face. There's a million dollar incentive for somebody to force this prediction market to resolve in the wrong direction. You have to think, okay, so there's a lot of money at stake in manipulating the actual prediction market resolution. So, what is the mechanism that's protecting us from that happening? So, auger is a really old school project. Your first launch in 2016, it was the first ERC20 token, the first ICO that managed to raise a lot of money. Hello, innovators, entrepreneurs and risk takers. Welcome to another episode of Web 3 with Sam Kumani podcast. And today on this podcast episode, I am interviewing auger from the auger network. And they are one of the OG chains or platforms or infra for blockchain. They have been in existence for a multiple number of years and today he's going to be sharing on what they are building when it comes to prediction market and how they are building the underlying layer that prediction markets can be built on top of. So, he's going to be sharing about his white paper that they have recently released. And as always, nothing that we talk about here should be taken as investment advice and please like, share, subscribe and follow with all that out of the way. Let's get into it. Welcome to the show. It is great to have you here and to be talking about everything that you guys are building. So, yeah, so, tell me a bit about how did you get involved with Ogre and also Ogre's journey because I understand it has been an existence for a while. Like in the pre ICO boom era in 2016-17. So, yeah, everybody. Yeah. August has been one of the first main projects in crypto. It was the first ICO, it was the first ERC-20 token, it was the first raise and they did well. So it ran for about four years. They launched the V2 afterwards. It really brought prediction markets and decentralized oracles to the forefront of crypto. And it was a reason that a lot of people got it to crypto in the first place. They've shackled a lot of people to go in and see the value of what was possible, right? Prediction markets are very important and it was great to bring that out. I got involved in Ogre personally in 2020. So this was about when it started getting a little bit weaker when DeFi summer happened. Ogre was on the Ethereum L1 chain. So really the fees kind of made everything difficult. So nobody wanted to pay $40 in gas fees to place a $20 wage, for example. So kind of the usage really slowed down after that. And people kind of forgot about it. They stopped using Ogre. The quality market came around. They were on the Matic network at the time. So now it's the Follygon network. And they kind of solved the scale problem. But what did they bring to the table was the initial kind of journey of Ogre. The initial anybody can create a market. Anyone can participate in market resolution. And that will kind of bring us to the next level, right? So I kind of saw that in a really like that concept of it. So I got really involved in the Ogre community for about four to five years. I decided that listen, this needed to exist again inside the crypto space. Yes. The crypto space in about the early 2020s was still kind of finding what it was trying to do. We had the tokens. We had the self yielding coins. If you remember all that stuff, indeed. But we kind of went a little bit away from the decentralization ethos, right? So all the projects coming out, we're kind of making those trade-offs. And they were saying we will decentralize at a later point. First we want to come in, first we want to build a product, we want to monetize, we want to attract the user base. And then we'll move into decentralizing things. And unfortunately what I saw in the space was that the latter half of it never actually happened. So they attracted a lot of money, a lot of capital, they built these products. But they never got to the decentralization. And really what Ogre brought to the table was no multi-six, no videos, no centralized points of failure. There was a peer open market approach to prediction markets and prediction markets resolution. And the design itself was really beautiful. And I said that this needs to exist inside the space. There was still money left over from the initial ICO days. And it was slated for future Ogre development. But it was kind of just sitting there, right? So I thought now was the time to bring this thing back to life. Now was the time to kind of remind the space what true ethos are, what true decentralization is, without all these backstuffs, without all these kind of wishy, washy things that kind of went up. And they agreed. In about 2025, with the purpose of reviving Ogre, it's continuing the experimentation of the Ogre Oracle. And ever since then we've kind of been working pretty hard. We have two development teams. And we just kind of put out a white paper about our recent design and how we're going to do things moving forward. So that's the short story of where we are today. Amazing, amazing. I do have a lot of questions around that white paper. But before we get into that, I would love to share my experience being in this industry all this years. And I absolutely remember back in 2017, 18 year olds, there was the cryptokitty's craze. I don't know if you remember. And because of that, you would have an NFT in that game that you might want to send someone or do something with it. The NFT itself would be $60. And then the cost of running it on the Ethereum network would be $120. Because it was so busy and consistent. Because the infrastructure wasn't scaling at this is pre-murge. And I remember the DeFi summer and all that. Even that was pre-murge and things were very challenging. And that did throttle a lot of progress. But now guest fees are becoming a lot more reasonable and things are a lot more scalable. So I do think that this is the right time for a certain prediction market. Tell me a bit about the white paper and everything that you guys, or not everything, but some of the things that you guys talk about and what so novel about it, kids. Yeah, for sure. So when people think about prediction markets, they think when they're betting on something with a prediction market, they're betting on what's actually going to happen. But that's not really the case, right? Because what they're actually betting on is how will the people running the market resolve this particular prediction market? The problem what happens is there's a lot of money at stake in manipulating the resolution source. For example, there's a prediction market with a sunrise tomorrow. Yes or no? You and I both know that it's going to happen. Yes. But on yes, because why would anyone bet on no for something that is not guaranteed? Of course, but I think we'd have bigger problems of the sun than a rise tomorrow. So that's what happens. So let's say there's a million dollars that goes on. Yes, the sun will rise. What this actually means is somebody can command and say, hey, I'm going to bet that the sun won't rise. And I'm going to try to take the million dollars of the people that bet that the sun will rise. Now the way I can do that is I have to make the prediction market lie. I have to make it resolve that the sun did not rise, even though that the sun did rise. So this is the question. The question is, what is the cost for me to make the prediction market lie? Right? And that cost has to be compared to the benefit of doing it. For example, in our example, there's a million dollars incentive for somebody to force this prediction market to resolve in the wrong direction. So now you have to start thinking. You have to think, okay. So there's a lot of money at stake in manipulating the actual prediction market resolution. So what is the mechanism that's protecting us from that happening? And there's a couple of different ways to do it. It goes from kind of the more naive ways, which is we're just going to trust the developers of the project not to misresolve it, right? Yeah. But the problem with that is that works when you got thousand million, maybe ten million on the line, but when you got billions of dollars, like these big election markets that we're talking about, you're putting a lot of trust into this resolution mechanism. So if it is just a multi-sake from the developers of the project, then you're kind of hoping that they, number one, don't decide to just walk off into the sun set with everybody's money. Or number two, nobody gets the idea to kind of force their hand to do something that they don't want to do. So that is the question of how do you do the resolution? And it's a very important thing to think about because not a lot of people see that as a problem because they don't understand that they're actually trusting that process. That's the only thing that matters in the prediction market. How will the resolution happen? So that is an Oracle. Yeah. A prediction market is a prediction market platform and to end, implicitly has an Oracle component to it. So the question is then, what is this Oracle component? Is it just a multi-sake? Is it some sort of a decentralized governance design with, for example, token voting? You can have people that want to vote with tokens and you can just say majority rules and that is a resolution. Or is it more of an open market design where you have a permissionless structure for people to come in and be able to put money behind what they believe the truth actually is? So what Algar brought to the table initially in 2016 is exactly that. It's a decentralized Oracle design that allows people to participate in Oracle Resolution. They do it in a couple of ways. So there are the escalation game component to it, and there's also the forking component to it, an algorithmic fork. This is what auger really kind of innovated, right? They brought the algorithm working in. So back to this kind of question, prediction markets, oracles, how do they resolve what is the cost of it? What we want to do in the industry is we want to say our design has the highest cost of manipulation. So if you want to manipulate the Oracle, it will cost you the most of any other Oracle system out there right now. And there's a couple of them. So the way we measure that is something called the security margin, the economic security margin. So, and we normalize it against the FDV of the token, for example, if you're using one token for your resolution and your token is worth 10 million in market cap, how much will it cost for me as a percentage of the market cap to manipulate the Oracle? For example, in majority rules voting, if we just have a majority rules token, right? Yeah. All costs is 51% of the market cap in order for me to buy 51% of the token and then resolve the prediction markets, however I want, right? Yes. So, if we're trying to defend a $1 million prediction market with a token voting design that has a 51% security margin, we need to have more than $2 million in FDV. Yes. Because the way we create economic security in crypto is we try to make the cost of an attack greater than the benefit of this. Exactly. And if you can do that and it's not an easy thing to do, it's a very difficult problem to solve. Yes. And there's many ways to do it. But if you can always keep the cost of an attack higher than the benefit of the attack, then no one's going to attack and we're fine and dandy and we're good, right? So, the question is, how do we do that? Right? So like I mentioned, there's shelling point voting, there's shelling point designs, there's token voting designs. There's vetoes, there's multi-sigs, there are security councils, right? It's trusted security councils. But then there's also the auger design. And what the auger design is, the auger design makes truth profitable. So, that's how we do it. We make truth the equilibrium that all honest participants, if they want to earn money, they should be betting on truth. And the way it works in practice, in reality, because I know I've said a lot of things. Even example, even example. Yes, yes, for sure. So, let's say we go back to our market, will the sun rise tomorrow, yes or no? Right? And you are an attacker and I am the honest defenders, basically. Yes. So, what happens is we start something called an escalation game. And what an escalation game does is we allow anyone to propose an outcome. Okay? So, you will say, I will bet $100, there is a stake, there's a bond that you put up saying I will bond or bet $100 that the answer is no, the sun did not rise. And now, what happens is there is a weekly, there is a timed round that starts. And anybody, anywhere, the whole open market has, let's say, 24 hours to say no, you're actually wrong, the sun will rise tomorrow. And for them to do this, they have to put up two times your bond. So, they have to say, you set $100 at the sun, won't rise, I'm going to put $200 at the sun will rise. And now, I become this tentative outcome. So now, the tentative outcome of the Oracle is at the sun did rise. And the round starts again. And now, anyone can come in and say, hey, actually, I'm going to say bet $400 at the sun did not rise. And so, this happens like this, right? So, it is escalated. So, anyone can come in and dispute somebody's proposed outcome. Now, if no one disputes an outcome, that becomes a truth, that becomes a canonical Oracle output, and that gets written to the chain. Now, what that actually means is that there is always an incentive, because whenever you win, you get the bonds of everybody that loses. And you get a 50% return on investment. And what that means is that you can crowdsource a defense. You can go out there in the space and say, hey, there is somebody saying that the sun will not rise tomorrow. And they put $1,000 as a bond that it won't rise. Do you want to take their money? You can take their money by putting $2,000 at the sun will rise tomorrow. And so, since you're crowdsourcing this rational outcome, the rational equilibrium here is that the sun will rise, because that is ultimately the truth. You are hitting the attackers against the entire rational market. So then you're saying it's a permissionless system. Anybody can come in and dispute and take the money of the attacker. So, this is the escalation game. Now, the escalation game has a threshold, because we can't just let the escalation game go forever, because somebody then with infinite money will always win, because then they just have to outspend wherever the defenders are. Yes. So, some say escalation games are the biggest whale wins games. So, the person with the most money does not take on a cost of winning the escalation game, even if they lie, even if they outspend and they are the dishonest resolution. So, we backstopped. So, now, what we say is, we have to somehow force this person that is being dishonest to lose money inside the protocol. How do you do that? So, we do that with a fork. This is the true utility of auger. This was their, in my eyes, their main innovation that they gave to the industry. It's an algorithmic fork. It's like Ethereum classic versus Ethereum, right? Yeah. It's like Bitcoin, cash versus Bitcoin. So, you do a social, those were social heart forks, right? So, they got everyone coordinated, they had a coordination system, and everyone agreed on what the new fork would be. Okay? So, what auger did is auger actually enshrined that into the code. So, it created an algorithmic fork. Now, the way it works is, we go back to my example of whether the sun rose or the sun did not rise. If the escalation game gets to a high enough threshold, we conclude that there is a very rich whale trying to attack the system. So now, instead of trying to outspend him or them because we cannot, we have to somehow force them to lose money. So we do this with a fork. The way a fork works is the universe. So this is one market with a sun rise or not. There are two outcomes, yes or no. What happens in a fork is that we now split the token into two separate universes. There is a universe where the people that the sun will not rise win and there is a universe of the people that bet that the sun will rise win. So now, what happens is every utility token holder, everyone that holds the token. When I was saying that we bet the escalation game, it's not in dollars, it's actually inside the rep token. And we use the rep tokens specifically so we can do this fork. Now, when we fork, we say, okay, so in one universe, the sun does not rise and the other universe, the sun does rise. And then what we ask everyone to do, every single rep token holder, we ask them to pick which universe they want to migrate their existing rep to. So their current rep token either converts to the sun that not rise rep token or it converts to the sun did rise rep token. Now, the attacker, the person that was trying to force the miss resolution, they get forced into the universe where the sun does not rise. All their rep tokens that they use for the escalation game automatically turn into rep tokens for that false universe. So what actually happens is then, since the rational rep token holders will choose to move forward with the universe that they deem will have future economic value, which means that the universe that will end up being used, they will pick to go, okay, where will all the future prediction markets exist? Will they exist in the universe where the sun does not rise? Will they exist in the universe where the sun will rise? So of course, they'll exist in the honest universe that becomes the equilibrium. I know this is kind of really confusing and kind of dealing with it. No, no, it makes sense. It makes sense. So then what they do is they say, okay, so I have two choices in front of it. I have a false universe and I have an honest universe. I know that the honest universe will have future prediction markets. It will have future fees. Therefore, it will have future economic value. So as a rational token holder, I will always pick to go to the universe that is the honest one because that's where the fees are going to be. Now, what that means is, since the attacker was forced into the universe where the sun does not rise because that's what they were betting on, their rep tokens become worth nothing. they lose their entire value of their SQC. the relation game that they chose on the false universe. - Yeah. - So that's how we force the attacker to take on a cost. We kind of leave them with the worthless token and we say, okay, sure, you won. Take your money. But you're token that you used to win with, that all the rep that you purchased you won't now won't be worth anything. So there's no way for you to recoup your costs of your attack. And then that becomes a question. The question becomes, how much value did the attacker lose? - Yeah. - And that is the economic security of the Oracle. We force the attacker to lose money within the protocol. - Yes indeed. - So whether they win or not, does a matter. All that matters is how much money do we force them to lose in order for them to force themselves to win? So that becomes the economic security. And that's really a design. So that's really what an algorithmic fork does. And that's how it's backstopps the escalation gain. So all rep token holders are service providers inside the protocol. They are not, it's not a passive asset because if you hold the token, there will potentially be some sort of an event in the future like a fork where you're gonna have to choose which direction to go. And it's in choosing the direction that creates the economic security because that's what makes the false universe worthless. So that's a migration. - Good question around this is, so how does over intend to make money? Is it from the infra fees for using its Oracle? - Yeah, for sure. Well the question is, the question of how does auger intend to make money? It's more about how does auger intend to pay these permissionless, trustless participants and how do we incentivize them to participate inside the protocol? Auger is not a company, it's not a business. The foundation does not make any money. The foundation just stewers the protocol. So the way auger makes money is we charge a fee to ask the Oracle a question. So if you ever want to resolve any sort of prediction market, you will have to pay a fee to auger that fee will be used to fund the labor cost of somebody doing the research to initially propose a resolution. In zone all the fees within the protocol, stay within the protocol. They don't actually ever go to some business entity that they were trying to make a revenue. There's a revenue in auger. All we're doing is we're saying, we will only charge the amount that we need to charge in order to make our incentives work. And for to make our incentives work, we need to number one, hey, somebody's labor costs to do research. You're gonna propose a question to the Oracle. Somebody needs to do research to answer that question, right? If somebody needs to do research, that means somebody needs to use their time where their time could have been used to make money. So this is the opportunity cost. So we have to pay for it. So we pay for it with the fees. The fees that we levied to ask questions to the Oracle go to the people that are answering. - Absolutely, yes. - Answering. - Yes. Now, a percentage of those fees get burned. So there is a lot of burning within the protocol. There's a lot of rep that gets sent to a zero with address and this drives value to every single rep tokenholder because the supply gets reduced. So there's burns and then there's fees, but there's no revenue generated because we're just trying to keep everything at a minimum. We're trying to maximize. - Equipment, yes. - Yes, exactly. - Fantastic. And what does the next 12 months at Oracle look like? What's the key plan over next 12 months? - Right, so what we did with Oracle Littis, Oracle Littis is one of our development strengths. We have two development strengths. So when we rebooted the Oracle kind of ecosystem, I guess we decided to do two things. So the first one was we decided to separate the prediction markets from the Oracle because initially, Auger was both. It was prediction market on the Oracle. So we decided to separate them and just package the Oracle by itself and sell that as a B2B service for other prediction markets just to solve the resolution. So the next 12 months, what it looks like is we will approach these prediction market projects. And basically say, hey, you're running your prediction market. You're good at that. You're good at getting users. You're good at getting things. But we know you don't like to do resolution because doing resolution is actually, it's a hard job and it worries a lot of risk. Because at some point sooner or later, if you run a prediction market project, you're going to be put into a situation where half of your community thinks you should have done it this way. And half of your community thinks it should have done it another way. And you're going to have this confrontation and controversy. And especially the bigger you get, the harder you get. Exactly. So you don't want these PR disasters. You don't want this risk of resolution. You don't want to do the research and hire this big team just to resolve these markets and kind of do all this. So you want to outsource resolution. You want to abstract it away to somebody else that has a good design, a permissionless design. And let them take the heat. Let them handle the resolution. And we're just going to pay you guys, you guys do resolution. Give us the answer. We'll use that answer to resolve our own prediction markets. So that's what augulitis is. Augulitis is a generalized modular oracle. It's essentially oracle as a service. So we just allow anybody to ask the oracle a question. It's going to sit on a theorem and anyone can go in and pose a question to the oracle. Pay a fee and really get the answer. As long as they provide a valid set of parameters. And yeah, all these prediction market projects, they just we just make it easier for them to outsource resolution. Lovely. Awesome. Final question, just fun line. Quick answer. And what is your ask to achieve this? D are you looking for a round of fundraising? Are you looking at hiring? Are you looking at any key partnerships? Anything? Just let me know. Or let the community know. Yeah, for sure. So we're not looking to fundraise. There's nothing we can fundraise with. We are a 501C3. We are a charitable nonprofit foundation. We don't actually have a token of our own. The token is in the open market. We have enough money to go with. We're looking for good developers. We're looking for good talent and really aligned people to just kind of come back in the fold. Augur is back. We're bringing a fully decentralized trustless and permissionless protocol back in the space. And yeah, just get excited. And really just bring me the prediction market projects that their users want to kind of increase the security. Amazing. That's fantastic. So with that, it's been an absolute pleasure talking with you. And one of the ways you best off luck in building this, especially, it's not easy building something that is not for profit in this way. For industry that's very used to profit. Fortune it. Yes. Unfortunately. Yes. But I look forward to how it develops. And we should do this again in 12 months, six to 12 months and see where you guys have reached and how it's being adopted by prediction platforms. So thanks a lot for your time. Good stuff. Thank you so much for tuning in and listening or watching this episode of the Web Tree with Sam Khmani podcast. By now, you know the drill, leave a comment or share this episode with a friend and leave a review. I would love to hear from you. So that's why my DMs are open. Reach out to me, especially if you're a founded building a product in Web Tree. There, I would love to hear from you. What are your challenges? Is there anything that I can help you or my community can help you with? Thank you once again and wish you best of luck in building your startup or your project.