Will it rain tomorrow? Who will challenge Donald Trump in the next US presidential election? Will humans ever live on Mars?
According to one notion from collective intelligence theory called the "wisdom of the crowd," consensus from a group of people is often better than any single expert opinion at casting accurate predictions, decision making, and problem solving. This idea has deep historical roots going all the way back to Aristotle, and is gaining renewed popularity thanks in part to blockchain technologies like Ethereum and the decentralized principals they instill.
One particularly interesting application of this theory pertains to predictions for an exchange or marketplace, aptly called a prediction market. In a prediction market, participants buy and sell shares under the presumption that they can surmise future events. At their core, prediction markets are simply a mechanism for placing bets.
This concept might seem counterintuitive, but it is being explored – or re-explored – by a new generation of researchers and technologists. Intelligence agencies, presidential campaigns, ICOs, and your local weather service all attempt to blur the temporal line between the present and the future.
Canadian Excellence Research Chair Erik Snowberg explained this for ETHNews: "This is because, as a market, it provides incentives for people to reveal their information about the probability of an event occurring. An algorithm using simple supply and demand brings all that information together into a single prediction. Put another way, if someone has information that an event is more likely than a prediction market price indicates, they will buy more of it, pushing the price up."
Although prediction markets might seem abstract, they are actually quite simple and straightforward. Rudimentary versions were used to assess presidential elections long before the digital age. History records Andrew Carnegie remarking to the American press upon returning from a trip to Scotland in 1904, "From what I see of the betting … I do not think that Mr. Roosevelt will need my vote. I am sure of his election."
Since Carnegie's time, there have been only a handful of true prediction markets. Hollywood has one for betting on the outcome of movies called the "Hollywood Stock Exchange," which utilizes virtual currency. There is a prediction market at the University of Iowa called Iowa Electronic Markets that has an ongoing track record for accuracy and serves mainly as a proof of concept for students.
IARPA's Geopolitical Forecasting Challenge
Like with so many other fields, prediction markets stand to be updated for the twenty-first century by blockchain technology, and we'll see how it might be applied in the Geopolitical Forecasting (GF) Challenge that's being launched by the Intelligence Advanced Research Projects Activity (IARPA) on March 7, 2018.
According to IARPA:
"This challenge gives you a chance to join a community of leading experts to advance your research, contribute to global security and humanitarian activities, and compete for cash prizes. This is your chance to test your forecasting skills and prove yourself against the state-of-the-art, and to demonstrate your superiority over political pundits."
Created by the Office of the Director of National Intelligence, IARPA lends support to the 16 agencies that make up the US Intelligence Community. "Our charter is to do advanced R&D," program manager Seth Goldstein told ETHNews. "We're to be looking over the horizon for … what's going to be relevant in five, ten, fifteen, or more years to sort of attempt to provide the US Intelligence Community with ground-breaking advantage over future adversaries."
The challenge will pit civilian teams against some of the best professional researchers and analysts in the intelligence community. IARPA explains that "GF Challenge solvers will develop solutions that produce probabilistic forecasts in response to numerous closed-ended forecasting questions that concern specific, objectively verifiable geopolitical events containing timeframes with deadlines and locations."
Goldstein went on to say, "In the Geopolitical Forecasting Challenge, what we are trying to do is say … 'OK, we want to know: Can you beat the existing state of the art?' IARPA has a process wherein we solicit proposals for research programs. We put out these documents called BAAs, broad agency announcements, and then there is full and open competition regarding who gets these awards. You have these large teams competing for research dollars and we negotiate contracts with them."
Prediction markets thus far have been relatively similar to the ones in Carnegie's time, in that they are all highly centralized. ETHNews asked University of Illinois at Urbana-Champaign's Andrew Miller why blockchain technology is reinvigorating the concept of a prediction market via decentralization. "First of all, prediction markets have been among those most closely regulated. Intrade, a popular centralized prediction market based in Ireland, famously shut down after being prohibited from serving US customers," he explained.
Miller continued, "Decentralized prediction markets are a good example of what can be enabled by disintermediation. A centralized prediction market like Intrade has to successfully provide multiple different functions. First, it has to receive customers' deposits and disburse their payouts. This mission requires a lot of trust, which is a barrier to entry for newcomers. A prediction market based on smart contracts instead, removes the need for a company to ever be in custody of users' funds. Second, it has to 'arbitrate,' i.e., determine the final outcome of the event. Even if the arbiter is a trusted central entity, the separation of this mission from executing payouts is valuable. However, this too can be implemented by a smart contract mechanism as in Augur."
It will be interesting to see how this field is shaped by bourgeoning blockchain development and application.
"In the future," Miller said, "we may see prediction markets as part of a company's decision-making process, alongside mechanisms like shareholder voting. Prediction markets are appealing because they use incentives to draw out unbiased information. Technology like smart contracts makes it possible to build more complex prediction markets than we have seen in the past."