Interview With Augur: The Crowd Will Be The World’s Most Powerful Oracle
Augur’s crowdfunding campaign has been an incredible success by any measure. It has already surpassed three million USD, even considering the drop in Bitcoin’s price. The crowd-powered prediction market has definitely apparently touched a nerve in the cryptocurrency world and it seems everyone wants a part of it.
What makes Augur unique is the way it settles bets, rather than relying on some sort of central authority, users who participated in the crowdsale can use purchased Reputation tokens (REP) to vote on the outcome. In exchange for voting with the majority, users are rewarded with more REP, and therefore have an economic incentive to vote correctly.
You could almost call it “proof-of-truth” or at least “proof-of-contribution.” Not unlike how Bitcoin’s “true” block chain is based on the concept that there will be more honest miners than any one group of malicious miners working together and therefore the longest chain with the most proof-of-work will be the correct one. With Augur, the idea is that since the most prediction’s results will be public knowledge, more people will vote honestly than any one group attempting to throw off the results.
The difference between the two systems, astute observers might notice, is that in Bitcoin, malicious actors have untold numbers of ways to go with their attempted scam but many of the predictions on Augur will be binary. X or Y, something either happened or it didn’t. In those cases any malicious actors would have an incentive to work together, unlike in Bitcoin. X happened but they want Y. Everyone who wanted Y could vote for Y and theoretically throw off the results.
But, even considering that, it seems unlikely that malicious actors would be able to overcome the honest ones. Most REP holders won’t be betting on any particular outcome, just judging it, and so most of them will have no reason to lie. Furthermore, REP will hold some (likely significant) value and throwing off Augur’s results would destroy the faith in the system and therefore REP’s value. Just as mining pools with incredible amounts of hashing power theoretically have an incentive against taking advantage of their position, because doing so would tank Bitcoin’s value, the same thing applies here.
Following that logic, it seems that Augur will be safe from malicious actors, but this is uncharted territory so we will have to see how it plays out. Certainly, the more REP voters there are, the safer Augur should be and judging by their crowdsale, there will be a lot of REP holders. During my conversation with Director of Operations Jeremy Gardner and Core Developer Joey Krug, it became clear that they have full confidence in the knowledge of the crowd.
Which is what Augur is really about. I, as I suspect many others did, initially saw Augur as a novel way to gamble. Centralized prediction markets already exist, but they have issues with liquidity and from a gambler’s perspective, I’d much rather just deal with an established sports book.
But the potential of Augur goes far beyond that. They don’t see sports books as their primary competition, or even centralized prediction markets. What Augur offers that no other technology to date has been able to offer, is the insight it provides into the knowledge of the mob.
It is in vogue to paint the average person as stupid. With Donald Trump leading the polls for the Republican presidential primary, it can be a hard position to argue against. But the truth is that the crowd tends to actually be a fairly accurate predictor.
This is best illustrated in the guess the jelly beans in a jar game (or gum balls or any other small item) It is rare that anyone will actually guess the correct number, but if you take the average of a group of significant size, the result is surprisingly close to the answer.
But how will the knowledge of the crowd apply to more complex issues? The success or failure of a political campaign, the ability of a game company to ship their product on time, or which horse is going to win the Kentucky Derby. All these examples are significantly more complex than simply counting the number of jelly beans in a jar.
These are the kind of questions Augur is going to answer: exactly how powerful is the crowd? But as ambitious as that is, it isn’t the only thing Augur is hoping to accomplish. Gardener’s idea about it being used to give a financial incentive to whistle blowers is not something I had thought of.
We discuss that, how the system will work and everything else I could think to ask them.
Ian DeMartino: So, how did you guys like crowdfunding using Ethereum? Obviously, it has been a big success, but did you run into any technical or practical snags?
Joey Krug: Part of it is [built using Ethereum]. So there’s two halves, The bitcoin side just uses blockchain.info’s api to generate addresses which forward to our multisig, [that part is] pretty simple. Now the Ethereum side was a bit more complex.
We’re accepting Ether via a smart contract on the Ethereum chain. So, to make that a bit easier to use for people, we’ve created a helper within our sale site at sale.augur.net which has people input their Ethereum address and how many ether they want to buy in. Then, it gives them stuff to copy and paste on their local computer, which sends the funds to our contract (which records the time they bought and how much they bought + from what address) and then it forwards it to our ether address.
Ian DeMartino: cool, so, with most altcoins, which I know augur is not, just stay with me for a second, people generally invest in them for speculative reasons. They are interested in the features but usually only for the potential to increase the price. Even asset tokens or whatever you want to call them tend to fall into the same trap of being driven by speculation. Do you think REP will be an early example of the exception to that? Where, while some people are purchasing with speculation in mind, many more are purchasing REP tokens for entertainment or fun? (I know REP is used for voting not betting, I just think people will find that fun.)
Jeremy Gardner: Yes, I think it is the perfect exception because unlike with alt coins, there is are tangible incentives to hold the REP token
1. You help create and maintain the world’s first decentralized oracle/consensus system
2. You get a portion of the trading fees distributed at the end of each reporting cycle!
So, as Augur scales, and more people use it, with more trading fees being collected, the more you’re rewarded, simply for saying whether an event occurred or not!
It’s fun to be among this global community of reporters, maintaining this system, and for the first time turning “history” into “ourstory”, and you’re rewarded as a result.
Ian DeMartino: Are you concerned that the system could become overwhelmed if it becomes too popular? That it might annoy some REP users if they have to answer dozens of events a day?
Jeremy Gardner: So the way reporting works – and I’ll let Joey expand upon this – is that REP holders report on the outcome of events during reporting cycles, which occur every eight weeks, for a month long period. So really, REP holders only have to report for a couple of hours, once every two months
Joey Krug: The way we plan to scale things is there [will] be a minimum set of markets you have to report on – the ones with the most volume, every 2 months, then beyond that you can report on more events if you like, but you’re not required to. We want the minimum set to take roughly two hours every two months. To do so, it might take some tuning to get the number right.
Ian DeMartino: Gotcha. On the other end of scale though is not being popular enough. The crowdsale is a huge indication that you will start with a significant userbase. But, I have seen quasi-sportbooks that allow you to bet on any user-defined thing and the problem is that a lot of bets don’t get enough action on them to make it worth it for other betters to join. A sort of chicken or the egg kind of thing.
Do you think anything needs to be done to solve that problem, or do you think the early success of Augur should solve it? (or something else I’m not thinking of)
Joey Krug: So I think there’s a few things that help with the liquidity issues — one is that we use a market scoring rule as opposed to traditional order book, so there doesn’t always have to be a direct counterparty at the time you trade. Another is that we’re currently talking with market makers to help provide liquidity on augur.
Jeremy Gardner: Furthermore, by having this sale, what is created is a massive network of supporters of the project that will create and participate in markets and encourage others to use the platform
Ian DeMartino: Cool. So, how will betting work exactly? is it as simple as X amount is on this side vs X amount on that side? Do you think gamblers will enjoy that more than a sportsbook based on the Vegas consensus odds?
Joey Krug: It’s based on the number of outstanding shares of each side of the outcome
Ian DeMartino: Can there be multiple sides, like for horse racing?
Joey Krug: Yeah, we call those markets “categorical” where you can have a bunch of outcomes.
Ian DeMartino: Who do you guys see as our primary competition? Sportsbooks? Other prediction markets? Or are you sitting in a niche all your own?
Joey Krug: It’s hard to say exactly because on Augur anyone can create a market on whatever they want. Presumably, the most popular markets on augur will be the ones where it has the most competitive advantage, but the market itself will answer that question a lot better than me!
Jeremy Gardner: I don’t see any primary competition, I see secondary competition such as sports books, Vegas, sure. However, I think they stand to benefit from us more than compete with us. They can use the global odds-making engine that is Augur for their own profit. But to say that we are our own niche would not be inaccurate. I simply don’t believe there can be two of what we do successfully.
Ian DeMartino: Right, Vegas likes to get people betting as close to 50% on each side so that they win no matter what. It will be interesting to see if Augur becomes a predictor of where the Vegas odds will go. Any thoughts on that?
Jeremy Gardner: I think that’s exactly right, The odds made by Augur, assuming liquidity and volume, will force Vegas to conform to the predicted odds as set by the masses, But we’re far from experts on gambling. We really see Augur more as a forecasting engine, than as a gambling tool. I think that distinction has to be clear. We are giving probabilities of any event.
Ian DeMartino: Building off of that then, do you think that Augur will prove that the masses are correct in their predictions, perhaps more so than so-called experts?
Jeremy Gardner: Mostly, There are anomalies, always. Will American Pharaoh win the triple crown Odds said yes, but there’s a chance the car carrying the horse overturned, or it spontaneously combusts on the race track. But, when it comes to fairly predictable outcomes, such as elections and general trends, I think the masses will vastly outperform any given expert. Even product launch dates or climatological events. I just don’t think [prediction markets] can predict arbitrary, random events well. Everything else, they’ll do better.
Ian DeMartino: Well, there were lots of examples of horses that won the first two legs [of the Triple Crown] and then failed the third. They were always betting favorites. But I see your point.
Jeremy Gardner: I meant, the horse had a better chance than any other, but it still could have been abducted by aliens in the middle of the race, and there’s no way that could have been accounted for.
Ian DeMartino: So, just a few more Questions, where do you guys see this in 5 years?
Jeremy Gardner: In five years, when you go on Google and type “Will Jeremy Gardner be elected CEO of Bitcoin in 2020?”
Jeremy Gardner: Google will spit back, “Augur gives a .0001% probability of that occurring”
Joey Krug: Or you ask Siri that question and she pulls the latest augur estimate from wolfram alpha
Jeremy Gardner: Exactly.
Ian DeMartino: Anything else you’d like to talk about?
Jeremy Gardner: I just want to reemphasize that this is a tool for understanding the collective wisdom of humanity, and rewarding those make good predictions about the future.
Sure, you can talk about Vegas and sports books, but I really think that misses the point. This is an early warning system for the world, A way to tap into our collective consciousness, in a manner that has never before been possible. I think that’s a much more interesting story than gambling.
Ian DeMartino: The wisdom of the crowd, I think it’ll be interesting to see it square off against whatever AI they have in that time.
Normally I make that my last question, but I just thought of one more relating to what you just said.
So, a pretty clear problem I think people have these days is being able to tell which experts are credible. If you google “do cell phones cause cancer?” you’ll find seemingly credible experts on both sides. The same can be said about pretty much any controversial subject.
I’ve written in the past about how our education system needs to be focused more on giving people the tools to discern the more credible between competing “experts” rather than trying to shove an entire world’s worth of facts into their brains.
Do you think that an “early warning system” like Augur could be used to help people figure out which experts are credible and which aren’t by looking at the wisdom of the crowd?
Jeremy Gardner: Absolutely, because any PhD can be bought off to represent specific interests, [just] look at the climate change debate and the Koch brothers. However, when the PhD themselves has to put their money where their mouth is, things get a lot more interesting. Especially with topics where there isn’t as much vested financial interest (as obviously manipulation can always happen). Furthermore, Augur is great for creating a decentralized whistleblowing incentive system.
Ian DeMartino: I had not thought of that use case, can you explain more?
Jeremy Gardner: A market could be made for “Evidence that senior officials at the SEC have received bribes from Goldman Sachs will be exposed and lead to top level resignations in the next ten months” the market maker could bet a bunch of money that that won’t happen and then someone at the SEC or Goldman Sachs now has a massive financial incentive to bet that it does happen and anonymously give the evidence to the DOJ, knowing they can be rewarded without compromising their identity.
We want to thank Jeremy Gardner and Joey Krug for taking the time to talk to us. Augur’s crowdsale continues until October 1st 2015, you can find it here.