Oracles: How blockchains perceive the world
“Corgito, ergo sum (I think therefore I am)” — Descartes
Short of triggering an existential crisis, the issue of determining how blockchains perceive the real world and therefore its version of truth is a less talked about but indispensable part of the system. How do we go about bridging the real physical world with the ethereal digital world and translate probabilistic real-world data into deterministic on-chain data?
Those more familiar with the blockchain world might be quick to point out that truth on a blockchain is determined by miners continuously choosing the longest chain. This is true when keeping track of on-chain data and transactions. However, specifically in the context of smart contracts, many of the details in the contract can not be determined at the point of contract creation. Consequently, smart contracts still require a trusted bridge to receive data from outside the chain in order for the contract to be triggered on-chain. As such, the promise of smart contracts is deeply intertwined with its ability to access reliable real-world data.
Oracles: How blockchains sense the world
In the realm of cryptocurrencies, this link to the physical world takes the form of oracles. Though, unlike its namesake, oracles don’t provide prophecies about the future but rather report on a fact in the past which makes it significantly more accurate but drastically more boring. Essentially, oracles act as the interface for blockchain to receive feedback from the real world, much like how our senses enable us to make sense of the world around us. Hence, an oracle will have to inhabit both the software and hardware realms and this generally takes two forms: machines and users.
Machine oracles tend to fall under the all-encompassing umbrella of IoT as they are essentially smart devices which sends information to a blockchain rather than a traditional server. As reporting can be done via sensors, this opens up significant possibilities for automating much of the admin functions through smart contracts. Moreover, machine oracles enable such reporting to be carried out 24/7 without any delay in synchronizing with the database and triggering the contracts if needed. The biggest strength but also the Achilles heel of machine oracles is that such sensors ensure data objectivity as it can only measure if a condition is true or false.
Depending on the rules of the smart contract, this data objectivity can be supplemented when machine oracles work in tandem with user oracles. The most basic of this would be user oracles providing confirmation to the smart contract when the sensor is triggered which minimizes any false positives. Going forward, it is expected that smart contract triggers will increasingly rely on multiple data streams and even consensus rules. With this being said, humans have never been known to be reliable sources of truth and as such requires well thought out incentives to prevent any abuse of the system. This could be monetary or reputational but if the fake news scandal has shown us anything, it is that human psychology tends to precede any truth.
Nonetheless, one of the solutions that have popped up to solve this coordination problem are prediction markets whereby those who forecast the outcome correctly win money at the expense of those who forecasted incorrectly. In such markets, predefined options are created for each share value which sums up to $1 (100%). Upon the event coming to past, and the resulting market closing, the winning option’s share becomes redeemable for the full dollar minus any transaction costs while all other options becomes worthless. Given such a market, an individual will have the incentive to purchase shares that reflect their belief that they have superior information about the event in question. In essence, prediction markets collates information about the likelihood of a future event.
Those more familiar with trips to the bookmaker/bookie would be right (and probably hesitant) to point out that such prediction markets already exist. At the risk of sounding like an overly-enthusiastic preacher on how blockchains will decentralize the world, centralized prediction markets come with their own risks and limitations. Key among this is that centralized markets are significantly limited by the types of markets that can be created due to an upper bound on their scope (local regulation, population, etc). Moreover, as implied by their title, any funds will have to be held in their books with payout depending on the correct resolution of the market in accordance with their own rules.
Of course proper rules and enforcement have been implemented but a quick Google search of ‘bookie scams’ will highlight the frequency of bookmakers running away with the funds or withholding the winning payout. Prediction markets on the blockchain will enable trust to be based on a network where:
- value can be stored on smart contracts which are self-enforceable;
- enforcement and contract rules are transparent and bolstered by incentives that encourage participation and reporting;
- transaction costs are limited to compensating market creators and users that report the outcome.
It is important to note that prediction markets forces the oracle to share in the consequences of its determination. Essentially, whoever participates in the market will have their skin in the game which would theoretically lead to a more efficient market. In fact, prediction markets have been shown to outperform traditional forecasting and polling methods.
“Since markets excel at encouraging people to acquire information, share it via trades, and aggregate that information into consensus prices that convince wider audiences, they seem to be ideal information institutions” — Hanson
The benefits of a blockchain-based prediction market are plenty but it also comes with downsides. Key among this is that without a single point of authority to determine the outcome, economically-abled attackers will be able to overwhelm the system by pumping money into their desired result. This might also be exacerbated by the fact that winnings in a market essentially comes out of the losers pocket. Introducing such free market economics also brings in traders who would trade on noise rather than knowledge which might further steer the market away from the ‘objective reality’. As such, there is a dire need for well-thought out rules and mechanisms in the network in order to limit the influence of bad actors.
Exploring uncharted waters with oracles
Similar to oracles in the past, prediction market based oracles could actually result in the creation of new markets as it encourages the exploration of the great unknowns. With a greater incentive for information discovery, it is expected that more granular data will be made available which in theory leads to more efficient markets and better pricing of risks. Essentially:
“Prediction markets enable more nuanced and specific expressions about economic events, which in turn signal value more explicitly (along with risk) at both the macro and micro-economic level” — Gnosis Whitepaper
However, if we are to learn from the 2010 Flash Crash, we must be wary when correlating such signals with actual market value-add as markets have the tendency to be self-referential. If done right, creation of new markets will offer real-time access to all information relevant to the event in question. It will be interesting to see the effects of this on information capital controls as it has yet to be seen whether the incentives from prediction markets will outweigh the perceived security from hoarding superior data.
Another interesting use case for prediction markets is in incentivizing actions in the real world. This can be done through creating a market for the envisioned outcome and betting against it. If an individual sees the prediction market for the event and believes that he can make it happen, he/she will be incentivized to bet against the market creator in order to realize a profit upon resolving the market in their own favor. This essentially creates a performance-based marketplace for services where increased bidding competition (value streams are no longer limited to two participants) strongly encourages delivery by the most efficient provider.
An even more revolutionary and untried idea made possible by decentralized prediction markets is that of a futarchy, which is a market-based approach to governance envisioned by Robin Hanson. I plan to cover this in more detail in a future article but a futarchy is essentially a form of governance whereby voters will determine the policy outcomes desired but betting markets will determine how these outcomes are achieved. This is something to look out for because if a futarchy works, this participatory governance process will encourage focusing more on proposals rather than personalities ensuring that delivery of such policies are efficiently delivered and ideologically neutral.
Predicting the future
Oracles are just the mediums through which smart contracts perceive the real world but it is the underlying architecture and the resulting implications that makes it truly radical. Machine oracles ensure the objectivity of the data fed into smart contracts while also enabling 24/7 triggering of smart contracts. User oracles ensure that such data is in alignment with society’s agreed reality, be it through economic means or otherwise. These functions could be conducted on a centralized system but having it on the blockchain will eliminate the need for a trusted central coordinator while also encouraging greater transparency in digital contract clauses and data.
In addition to improved data integrity, it is expected that the information market ecosystem will see significant growth as it is no longer limited in scope. Consequently, market creation will only be limited to the market creator’s ability to define such a market. Additionally, having such a market effectively commits an individual’s action to their privately held belief aligning information gathering with their decisions. With increased data resolution and liquidity, the stage is also set for improved analytics to be carried out via data analytics and machine intelligence learning. The growth of this ecosystem will also likely bring about the creation of user-friendly oracle applications which will enable non-technical users to create different markets further growing the network and the ecosystem.
Prophesizing about the future is definitely exciting but there are significant hurdles, technical and non-technical ones, which will have to be overcome before the above can be realized. With this being said, I would be willing to bet that such technologies will play a central role in the future given the inherent advantage it has over traditional systems, just don’t ask me to define when.
Thanks for staying till the end. Would love to hear your thought/comments so do drop a comment. I’m active on twitter @AwKaiShin if you would like to receive more digestible tidbits of crypto-related info or visit my personal website if you would like my services :)
