Smart contracts are an incredibly powerful invention, allowing for the elimination of many traditional middlemen, and creating trust by removing the need for it. However, smart contracts have some peculiar limitations, including an inability to collect data from the world outside the blockchain they live on. This of course is where oracle networks come in — by creating economic incentives for data providers to be honest, they allow smart contracts to trust data from the outside world. But these oracles are subject to economic incentives of their own, making for some interesting and frustrating bottlenecks. In this article we will explore these issues — and how ORAO solves them.
Economic Forces Incentivize Simple Data
Oracles, like every other kind of application that relies on on-chain transactions, are subject to transaction fees. We are in a time now where the transaction fees on Ethereum are truly prohibitive for retail investors, and it’s affecting the whole DApp landscape as well. But oracles have to print data in transactions, or smart contracts can’t use it. And the more data an oracle wants to print, the larger the transaction becomes, and the higher the fee. These fees are, of course, passed on to the data buyer, who is usually a smart contract. This means that data buyers have strong incentives to ask for only very simple data.
The most common data type being bought and sold is financial data, and especially price feeds. Even with simple price feeds, however, the pressure to keep data simple hurts the utility of that data. You want to know the price of Ethereum in U.S. dollars? Sure. Do you want the latest price a trade was executed at? On just Binance, or on multiple exchanges? Was that a big trade, with slippage? Was it a market buy or a market sell? Does the price vary between exchanges? It wouldn’t be uncommon, for example, for the price to be $2,502 on Binance and $2,488 on Coinbase. But adding all these details would be prohibitively expensive. Very few data buyers need data this detailed, so current oracles have mostly just… not provided it.
This leads to an unfortunate cycle. Data buyers overwhelmingly ask for simple, cheap data, so current oracles only offer simple data, so new smart contracts being developed can’t find any sources for more detailed data, so that functionality is scrapped, so nobody asks for the detailed data… And round and round it goes.
Escaping the Bottleneck With ORAO
The state of the fee economy is unfortunate. However, there is hope. With the explosive growth of DeFi in the last few years, many different solutions for the high fees are being worked on. Ignoring oracles for a second, the sky high transaction fees on Ethereum has led to the rise in popularity of a number of other blockchains, each with their own proposed scaling solution. Elrond, for example, uses Adaptive State Sharding, meaning the more traffic the network sees, the more chains are spun up, driving up max volume and lowering transaction costs. Polkadot, which exploded onto the scene last year, promises to bridge not only its own large number of parachains but also other popular blockchains like Ethereum and the Binance Smart Chain.
Ethereum is not just rolling over to competitors, of course. Optimistic and ZK-rollups offer good hopes for the future. By moving complex computations, many transactions and even decentralized exchanges to second layer solutions, the baselayer is about to be relieved of an enormous amount of pressure, starting July 2021.
ORAO is built with all these developments in mind. In particular, ORAO offers Predefined Data Protocols and is Blockchain Agnostic.
Our Predefined Data Protocols allow data buyers to fine tune exactly what data they want and don’t want. In our example of the Ethereum price above, you could ask for price and order depth, or price variance between exchanges, or price changes in the last minutes, or pretty much any other information you could want, in any combination. Importantly, our Predefined Data Protocols also are not limited to financial data, as the data offered by earlier oracles are. If you want to know who won the Superbowl or the last election, ORAO’s data providers can provide that. If you want to know how much rain there was in Hawaii this week for your decentralized vacation insurance DApp, that can be easily supported. Any data you want, from NFT trading to weather and climate patterns, with all the customization you need.
Now, importantly, this richness in detail comes with increased transaction size. We feel confident that Ethereum will scale well, but ORAO does not rely on that. In order to take advantage of the different scaling solutions offered by different chains, ORAO is Blockchain Agnostic. This means that data can be delivered on any blockchain capable of supporting smart contracts. Our approach is two pronged: we are building our own gateways to connect with chains we see demand for, such as Elrond and Ethereum. We are also building ORAO to be native to Polkadot, with all the inherent interconnectivity that offers. This means that, importantly, we can support any blockchain we see demand for, present and future.
Not only does this expand our reach and enable ORAO to support cross-chain data products, it means that as relative transaction costs change for different chains, data buyers and providers can move wherever transactions are cheapest at the moment, cutting down on costs for data buyers and increasing margins for data providers.
Proactive Data Rating
With such a broad offering of data products and customizability, we expect a very large data marketplace to grow using ORAO. This means a high number of data providers as well, and a need for strong protections to ensure the accuracy of the data. Here too our approach is two pronged. ORAO uses Validators, who ensure that all data being sold is accurate, and we also use neural nets trained with TensorFlow to judge the data as it comes in. This means that, if someone does try to sell malicious or stale data, the data is flagged before the buyer even receives it, and does not get delivered.
Our Validator program is supported by economic incentives. 20% of the total ORAO token supply is set aside for Validator rewards. Validator nodes can be operated by anyone, without need for programming knowledge or experience. These nodes, while not selling data themselves, are rewarded for double checking data products and making sure data being sold is fresh and accurate.
Our neural nets provide another layer of protections for many data types, where it makes sense. You can’t really apply neural nets to judge the accuracy of who won the Superbowl, of course. That’s a stand alone data point that doesn’t connect to anything. But for data products such as price feeds, or any other continuous stream of gradually changing data, neural nets are perfect for detecting sudden pattern breaks that could indicate issues with the data product.
Thank you for joining us for this deep dive into scaling issues for Oracles and our solutions. If you would like to learn more about ORAO, you can check out our website, dig into our whitepaper, or join our Telegram channel to speak directly to members of the team and our community.