API3 and The Future of Oracles
The last few years in crypto certainly have been interesting, haven’t they? We’ve seen the rise and fall of the ICO bubble, years of bearish sentiment, the near-permanent hiatus of the very site you’re reading now, and all the while we’ve had steady progress throughout. We’re not talking about the platitudes that CZ and company love to tout regarding BUIDL – yes, this has been happening, but on the surface level it would seem stagnant. After all, when prices are down over 90% across the board for months and then years, it really can seem like doomsday. In many cases it was – weak hands got washed out, lots of companies got sued, and many more went defunct. Yet despite all of this, innovations and development has truly continued on. And so it is that we come to API3 – what it’s building, why it’s doing it, and if we really do have the next Chainlink on our hands.
The Oracle Problem
The infamous oracle problem has been one of the largest and most well known issues facing smart contracts and blockchain development, and it’s been that way for years now. You have a smart contract on-chain with enforceable rules and functions, but they are really only useful with data that is available inside the Ethereum network itself.
You can’t make a contract on the price of gold if such an input has to come in from meatspace – and therein lies the oracle problem. Just how do you get this kind of data on-chain and in a decentralized manner? Moreover, how do you ensure that this data is verifiably true, and how do you secure against an attack against such a data source? Certainly this increases the attack surface of a product dependent on a) the smart contract and b) the oracle provider itself.
Since the heady days of crypto, we’ve tried to resolve this oracle problem in several different ways. The most circuitous of which include prediction markets, such as Augur or Gnosis, but the real money has always been with an oracle provider that can deliver this data in an anonymous way, without third party intervention, and doing so in a cost-effective manner.
It would be poor form to discuss oracles and current solutions without mentioning Chainlink and all of the progress they’ve brought to the crypto ecosystem. In fact, this very site was a strong proponent of Chainlink back when they had their ICO in 2017. Of course, investing in an ICO is one thing – money was easy to make back then. What shows your strength as an investor more, however, is the ability to hold an investment through a bear market and reap the rewards once Chainlink was truly appreciated for what it was:
Chainlink is great. It’s one of the only true oracle projects that delivers on its promises, has a vibrant community of holders (known as the LINKmarines) and is poised to be one of the eminent blue-chip crypto tokens to hold in the future.
…So why am I writing this article?
Because Chainlink has problems. Problems that API3 solves.
The API Problem
So, we’ve discussed what the oracle problem is. In reality, it’s really more of a problem that’s been created by thinking too small about how we actually want smart contracts to function on Ethereum. The goal was never really to solve the decentralization of nodes that deliver oracle data, or to overcomplicate things such that “anyone” can deliver this kind of data. Even talking about it here it is a tad complex, is it not?
Actually, we have a much simpler problem. Really what we want is the ability to hook onto off-chain data and use it in our contracts. Oracles as far as blockchain middleware is concerned have been compared to API’s of the web in the sense that they deliver this data to the consumer. Rather than thinking of oracles as an abstraction of APIs, why don’t we just apply the design philosophy of an API itself onto blockchain?
Wouldn’t it be cool if instead of making an oracle call that costs you three dollars (quite expensive in the long run), you could make an API call that delivers the same data?
Wouldn’t it be cool to know who is actually delivering that data, rather than having to trust an anonymous node?
Wouldn’t it be cool to avoid all of the attack surface that multiple node providers enable and simply deliver that sort of data in a seamless integration?
Exit Chainlink. Enter API3.
The API Solution
So how does API3 work and why are we so bullish on it? In short, it takes all of the value that Chainlink nodes are currently aggregating (you know, the ones who are only incentivized by that value) and delivers it to the providers of the data themselves. I mean this in a direct sense. You don’t need to have some intermediary set of Chainlink nodes to hook onto an API provider and transmit that data on-chain. Actually, you can just have the API provider themselves provide that data and reap the rewards. This solves several key problems that Chainlink will have to contend with over time, and why we think API3 is a very bullish product.
Firstly, you now have a reputation to uphold when you are directly providing API data to consumers. Report bad or wrong data on Chainlink? There’s a monetary repercussion, but that’s about it. No one really knows who controls the node that did the damage due to the anonymous nature. As the API provider yourself you have a direct investment in the verity of your data, which means you remove a lot of the “oracle bribing” situation you can get with Chainlink.
Which is a huge problem, which Chainlink has solved by the way – only, they’ve made that solution prohibitively expensive to do so. Chainlink accepts that oracles can be bribed, and part of its design to safeguard against that involves using multiple nodes to deliver the ground truth of whatever data it is that you’re calling up. Multiple nodes cost money. Lots of money.
It’s elegant. Extremely elegant.
That’s what it’s all ultimately about, right? In the end, we need to ask ourselves what the actual advantages are for the data provider here, as well as the consumer. Aside from the aforementioned ease of onboarding (try getting any legacy company to set up a blockchain node) API3 is just… cheaper. It’s cheaper to set up, cheaper to manage, and cheaper to make oracle calls on-chain. Nearly every aspect of API3 has been built with the data consumer in mind, and we think this is one area where API3 just beats out the competition.
API3 is mostly focused on creating decentralized data feeds like feeds.chain.link that are composed of multiple Airnodes and decentrally and transparently governed. While you can call Airnodes directly and it’s as robust as calling the API directly, most DeFi runs on decentralized feeds and the ethos of the space is such that single-source oracles are seen as suboptimal (however these will be useful for things like prediction markets).
Of course, we try to be unbiased here – in our personal opinion, we still feel that Chainlink will provide a solid level of security with their node architecture. However, we feel that with the inclusion of the reputational element each API provider will now have, we think API3 offers an alternative that can plug in to existing systems and reduce gas costs – to the tune of 50% or more per call for very little downside.
Solid savings, solid returns.
Every project these days has a governance mechanic built into their token and API3 is no different. There is value to be acquired here in dollars and cents when it comes to owning supply of API3. You get to vote in governance changes, fee structure updates, and channel many of those fees to you, the holder. For a burgeoning data marketplace we think this is extremely bullish and once again are willing to make the play that API3 is less of a bet on their specific project succeeding (it will) and more of a bet on blockchain itself becoming even larger and more mainstream.
There’s also a good staking mechanic involved with the token which provides rewards to those willing to put their tokens up and act as insurance against malfunctions/errors. We expect these to happen, but will be few and far between. They happen on other existing systems too, and we are glad that API3 is taking the initiative here and being realistic about things rather than pretending these problems don’t exist.
Plus, staked tokens is reduced supply, less sellers, you know the deal.
There’s lots more to API3 that you can check out from their whitepaper and website, so we urge any intelligent investor to go ahead and check it out. Make your own decisions on whether you see any value here (we do) and if you’re interested in their upcoming sale in October.
Token Metrics? They exist
API3 has some really great things going for it. They have a lot of pre-existing customers who are looking to use API3 right out of the gate. The team has worked very closely with Chainlink themselves for some time now, so they know exactly where the pain points are.
There’s no reason for us to include things like token metrics, sale numbers, and team sections as you are free to do your own research when it comes to this kind of thing. Nor do we really exist in that kind of market anymore – attractive terms and numbers are never enough these days to get real appreciation. This time around you need some actual value proposition and real long term view to make a splash in this market.
Is API3 the end of Chainlink? Are we bearish on Chainlink? Is this a hit piece on Chainlink? Certainly not. If you think so, remember that you’re accusing individuals who invested in Chainlink big time and have held it for years of…being bearish on Chainlink.
No, we think these two solutions are great for different things – it will ultimately be up to the data consumer if they want a cheaper hands-off solution or a more expensive, but more robust offering with Chainlink. There’s definitely room in the data market for more oracle options – and existing services can only be improved by quality projects in this sector.
We’re convinced API3 is one of them. And we’re investing. Heavily.
Watch this space.