A series of controversies around specific prediction markets highlight Augur’s governance issues.
The views expressed here are the author’s own and do not necessarily represent the views of Cointelegraph.com
Like many other ICO success stories and subsequent leaders in their respective market segments, Augur, the preeminent platform for decentralized predictions, faces constant public scrutiny. The latest episode that drew public attention is the allegation, voiced by cryptocurrency hedge fund Tetras Capital partner Alex Sunnarborg, that a developer group behind the platform significantly overstates the volume of trades that Augur processes.
While the real trade volume is an essential metric indispensable for understanding the scale and impact of a given project, it is still a quantitative, rather than qualitative measure. In this regard, a dispute around how much money is staked on Augur at a given moment of time is different from two major previous fracases, which sparked debates about fundamental aspects of decentralized prediction platforms’ usage and governance. Those two episodes concerned the so-called “gimmick markets,” and, earlier, Augur’s capacity to host death pools.
When betting money on the outcome of future events, like with any other contract, the rule of thumb is to make sure that you get all the details straight. What exactly is the outcome that liquidates your futures contract at a win or loss? When exactly does it occur? In most cases, these conditions are straightforward enough to go without saying. After all, when you wager on Real Madrid beating Liverpool in the Champions League final, 90 or 120 minutes after the kickoff time, everyone knows who won. And if something goes wrong, you can always appeal to the bookmaker.
This is not quite the case, it appears, when predictions go decentralized. Once everyone can set up a market, the terms of some contracts may become vague — either due to amateur bookmakers’ unintentional lack of phrasing precision or due to malice. And once the bets are in, users have no recourse if they suddenly realize that they were wagering on something different than the market is really about.
The latter likely describes the situation many people involved with the recent Augur political market have found themselves in. The question looked simple: “Which party will control the House after 2018 U.S. Midterm Election?” Anticipating that Democrats will have flipped the House as a result of the midterms, 95 percent of the bettors wagered on them. Indeed, the “blue wave” that pundits predicted yielded the Democratic-majority House post-election. However, the important caveat is that the newly elected members were not to come in until January 3, 2019; as of the market closing date, Dec. 11, the House remained exactly the same as it was before the midterms — that is, Republican-controlled.
The Augur community went abuzz: Those who thought they were betting on the election outcome demanded that the market be called for Democrats, while others — including the alleged creator and designated reporter for the market — insisted that the idea was to measure the state of the House on Dec. 10, which, to be fair, could hardly be different from what it was on the day the market opened. In a Reddit post, the self-avowed creator made this clear by referring to the deal as a “gimmick market” and declared his or her intention to call it for Republicans.
The fact that more than $1.3 million were at stake rendered this conundrum perhaps the toughest test for Augur’s on-chain governance system so far, and definitely made for the platform’s biggest publicity crisis since the summer hype around assassination markets. Despite the fact that these two controversies look quite distinct on the face, they are manifestations of the same deficiencies intrinsic to the nature of decentralized prediction markets.
Markets for death
In an episode of the British techno-dystopian series Black Mirror entitled “Hated in the Nation,” mysterious assassins begin to eliminate public figures, one by one, decided by whomever social media users post the most #deathto hashtags about. Once the bloodthirsty online mob realizes how the death pool works, they readily rush to bid on the next odious MP’s or obnoxious rapper’s demise in order to trigger the murder that mysterious assassins immediately carried out.
As Augur, a blockchain-powered — decentralized prediction market, went live in July 2018 — the media was quick to latch onto the minor yet captivating facet of its functionality: the capacity to enable the creation of so-called “assassination markets.” In the dark spirit of Black Mirror, albeit under a somewhat different mechanism, these arrangements could spell death for those in the public eye. Indeed, it did not take long after the platform’s launch for such markets to appear, with a number of prominent politicians, actors and entrepreneurs put on the spot.
Augur provides a decentralized infrastructure for users to set up bets on whether certain events will or will not take place. Taking advantage of blockchain’s anonymity and the absence of a centralized authority to censor the content on the platform, malicious users could potentially procure a tool for incentivizing other people to “help” certain outcomes occur. For instance, by creating a market on whether politician X dies before the end of their incumbency and staking a huge pot of money on a “no,” someone could effectively put a bounty on the person’s head. Wagering against the massive “no” market and then contributing — to put it gently — to a “yes” outcome, any villain could run off with the money.
Horrendous as it sounds, the scenario was not invented by the Augur community. The idea of a cryptographically anonymized death market has been present in the cypherpunk milieu for a while — at least since cryptographer Jim Bell had formally recorded it in his 1996 essay “Assassination Politics.” He envisioned a market that would predict the deaths of government officials as a means to punish those who indulge in corruption. The Augur subreddit has also been rife with various takes on the death market principle long before the protocol went live.
So, is this what blockchain is for: letting scoundrels ease the remiss bettors of their money or even anonymously order people dead and get away with it? The clamor over the dubious Augur developments jibes quite well with the broader, ongoing debate that concerns platforms’ responsibility for the content their users choose to publish on them. Think Facebook and fake news/streamed deaths, or Twitter and political botnets, or Youtube and videos of dead bodies on popular suicide sites. The centralized social media gatekeepers’ mantra of “we are not publishers, we are merely infrastructure providers” is sounding ever less convincing with each high-profile blunder, forcing corporations behind those platforms to haphazardly design new policies and interventions.
Critics often point out that, in the case of a decentralized, blockchain-powered marketplace for anything, there is no corporation or government to go to if the goods or ideas in question turn out to be immoral or otherwise unacceptable for the majority of users. Furthermore, immutability of distributed ledgers that carry information about transactions renders it impossible to take the content down. This takes us back to a more general problem of blockchains’ capacity to perpetuate the wrong — be it flawed land titles, unjust copyright claims or transferring a scam victim’s money to a con artist’s wallet. Does this mean that a responsible society should avoid using decentralized, permissionless systems to underpin any sensitive sphere of transactions? Not really.
The classic notion of the marketplace of ideas, as John Milton and J. S. Mill construed it, rests on the assumption that, once all ideas are allowed to clash freely in an open marketplace, the best of them will eventually prevail. Even though such reasoning might not seem indisputable, one need not take this leap of faith in order to be comfortable with blockchain-powered markets. Regardless of whether the natural tendency of good ideas to defeat bad ones is really a thing, there are still other mechanisms to fall back on — namely, governance systems’ design and a broader set of social norms that govern human behavior.
Prediction markets, as well as other blockchain-based idea marketplaces, may — and probably should — incorporate some on-chain mechanisms of community self-regulation. In the case of Augur, the community of REP token holders — who are also called “reporters” — are incentivized by the system’s design to document the correct outcomes of the events in question. The same people have the power to declare a certain bet “invalid,” in which case nobody gets paid after the outcome is decided.
This instrument of community self-policing looks like a relevant tool for stopping morally reprehensible bets from enriching those who might want to use the platform for malicious ends. The “gimmick market” case is a great way to test the system’s capacity to handle situations that are less unambiguously unacceptable than facilitating murder. Assassination pools constitute a marginal fraction of Augur’s overall trade volume, with just a handful of transactions. In contrast, gimmick markets on high-profile, highly bettable events may well become a feature of the Augur landscape, should the community set a precedent that lets the first one be.
In addition to on-chain community governance, there are things happening off-chain, too, that may serve as checks to potential abuse of prediction markets’ infrastructure. Even if we adopt the radical stance and accept that “code is law,” there is a larger ecosystem of constraints that influences human behavior. In the words of Lawrence Lessig, one of the preeminent legal thinkers of the Internet age, there are at least four discrete forces that shape people’s actions online: law, choice architecture, market and societal norms.
Even if the distributed ledgers’ architecture allows people to anonymously sponsor — and subscribe for — lawless action or con markets, and given the demand for them, social norms are still there. These norms suggest that murder is highly unethical, and fooling people into betting on the event that cannot possibly occur is not the best way to make them like you — even if you say sorry afterward. Also, there is a consideration of a perhaps more forceful effect: Both murder and fraud are criminal offences punishable within the legacy legal system, which still exerts a lot of influence over us all. On Augur, bets come in Ethereum, not REP, meaning that payments are very much traceable by law enforcement. And rest assured, the authorities are watching closely.
Most certainly, Augur already has regulators’ close attention, and recent developments are not going to make things better. Since markets that the platform hosts are essentially futures contracts, Augur and other decentralized prediction markets fall under the purview of the United States Commodity Futures Trading Commission (CFTC).
Reports emerged last summer that the agency was scrutinizing Augur for allegedly facilitating illicit gambling activities, since prediction markets as a form of gambling are illegal in the U.S. Those that manage to operate do so with multiple buffers and protections. For example, PredictIt, the largest non-blockchain platform that allows American citizens to wager on political events, is operated by a New Zealand-based, university-affiliated nonprofit and has strict limitations on the amount of money that users can stake.
In his October speech at a technology conference in Dubai, CFTC Commissioner Brian Quintenz raised a question of accountability on blockchain and sketched potential regulatory boundaries in the context of smart contract-powered futures products. In his remarks, Quintenz first demarcated the subset of smart contracts that potentially fall under the commission’s jurisdiction — the ones that manifest essential features of a swap, future or option — and then turned to the parties involved in their creation and operation: core blockchain developers, miners, developers of smart contract code and end users.
Quintenz suggested that it would be impractical to hold the first two categories accountable if some of smart contracts that operate on top of their ledger would be found to be in violation of the CFTC rules. Going after individual users of illegal decentralized futures, while normatively defensible, would likely be what Quintenz calls an “ineffective course of action,” given the pseudonymous and global nature of public blockchains. The only category left to directly target is, therefore, those who create and define the potentially illegal smart contracts.
Albeit Quintenz made sure to present his remarks as personal opinions, he is clearly not the only person on the commission who is pondering the ways to tackle these emergent challenges. Enter million-dollar gimmick markets springing on top of the largest political predictions platform available to U.S. citizens. Clearly, the whole deal looks primed for the regulator to step in and protect the investors — and if Brian Quintenz’s approach is the dominant one within the CFTC, it might be the right time for the Augur core development team to start getting concerned .
Governing with prediction markets
While regulators have yet to figure out how to deal with decentralized idea markets whose operations are apparently in conflict with the standing laws, it is unlikely that prediction platforms are going anywhere anytime soon. Since they essentially represent the pools of aggregate collective wisdom, such markets are often a feature of many projects aimed at creating systems of decentralized governance. Arguably, the most publicized of those is the futuristic form of government — called “futarchy” — that economist Robin Hanson proposed as a framework for enabling citizens to vote for optimal policies. The concept apparently gained traction with Ethereum’s Vitalik Buterin, who, in 2014, had dedicated a grant to support research on the topic.
There are projects that seek to build a versatile governance protocol around a pool of “collective intelligence”, where users determine visibility and prominence of policy suggestions by staking tokens on predicting whether they become a success or not. This way, a prediction market becomes a device for managing collective attention, stimulating members of the community to sift through policy proposals and evaluate their relative worth.
Meanwhile, blockchain-powered prediction markets are doing just fine in their primary capacity as platforms for betting on outcomes of future events. In November 2018, Augur’s trade volume in midterms-related contracts surpassed that of Predictit, the largest centralized competitor in the domain of political forecasting.
Humankind has had the habit of betting on the future for thousands of years, and the idea of doing it without a middleman for the first time is incredibly appealing. The CFTC seems to be up against an enormous task of wrapping the red tape around an ever-expanding infrastructure that facilitates an activity that many people enjoy.