On last Friday night, Unitimes had the Tellor team at our Chinese comunities, talking about the importance of decentralized oracles, what’s the difference between Tellor and other oracles like Chainlink, Witnet and Band, and how two former government economists and a video producer made Tellor.
Tellor is a decentralized Oracle for bringing high value off-chain data onto Ethereum. The system utilizes a network of staked miners that compete to solve a PoW challenge to submit the official value for requested data.
Welcome to Unitimes AMA. Could you say hi to everyone and tell us what brings you to Ethereum and started to make oracle?
Brenda: I studied Economics, because I enjoyed it. It gave me a picture of how the world works in a different and fascinating way. Price dynamics, trade, game theory, it made it very clear that there is a lot of lack of free will and more social engineering than we care to admit. So, what really drew me in was Ethereum. I became fascinated (a little obsessed really) with the technology and the freedom it could bring. As a Statistician/Economist I had been coding for years so the transition to Solidity(Ethereum programming language) was not too bad. I started learning Solidity so I could build out some proof of concepts. But I finally dove into the crypto scene full time when Nick(CTO) started Daxia, a derivatives protocol on Ethereum. I had worked with Nick for a few years and he had been into all things crypto for years. I joined him with Daxia, and I was so excited to be part of this transition to a freer world and being able to be first witness to a transition of means of exchange (since I missed the barter to gold and gold to fiat ones 😉).
Our first endeavour was Daxia, a protocol to do long/short tokens (e.g. long/short BTC) (similar to a dy/dx or Market Protocol). The issue you quickly run into though is the oracle problem (how do you get the price to settle to). The companies now are just inputting it themselves, which is not secure and not in line with the ethos of decentralization. We wanted something better so we started building out Tellor internally. And it turned into a full time thing when other projects/investors became more interested in Tellor than Daxia.
Nick: Hi everyone, I’m Nick, I’m the CTO. I actually an economist by training but have built out our platform into something pretty novel and secure. I got into crypto back at an end the fed rally and then started building out trading bots for a few years before I joined the CFTC (US crypto regulator). I left that 2 years ago to do a derivatives startup and luckily pulled Brenda out of the government with me.
Mike: And I’m Mike, CSO. My role is strategic, creative and business dev focused. I was a former film/video producer and director turned Startup and small business consultant. Which lead me to working with a small crypto startup named Bitsie back in 2012/2013. They paid me in bitcoin, so I got interested pretty quickly. Later on met Nick and Brenda and joined them to build Daxia.
Why Ethereum need oracles?
Brenda:
Short answer: Blockchains are siloed systems that can’t natively access data from the internet or any other chain. But to truly enable blockchain’s use cases they need information from outside the network to execute transactions and provide services that people can really take advantage of (like hedging on assets not on the blockchain).
Longer answer:
So Ethereum is like a computer that runs programs. Unfortunately, it’s like a really old, slow computer that can’t access the internet. So right now, for those programs on Ethereum, if they want to know something about the outside world (e.g. the price of a stock, an interest rate, the weather, a sports score, or any other conceivable data point), they rely on someone typing that information into the computer. That person typing in the data is the oracle. Now having just one person is obviously bad, since that person can lie, be bribed, put in jail, or even die and your program on Ethereum gets wrong (or no) data. Tellor solves this problem. Think if instead of having just one person type data into the computer, you have a whole bunch of people competing to type that data in. This is Tellor. We have every person who wants to compete place 1000 tokens in a jar that they lose if they lie, and then they use Proof-of-Work to pick who the person is that gets to enter the data (just like Bitcoin). So now if your program wants to know the price of Bitcoin or petroleum, you can just ask the Tellor network, and our system will compete to give it to you.
There are many oracle projects on the market, and competition is intense. What are Tellor’s competitive advantages? How does Tellor survive and be stronger in the competition?
Brenda: I’d say our biggest competitive advantage is that we are a well-prepared team with experience delivering products. We went from design to mainnet in 1 year. We have made our smart contracts upgradeable and we plan to continue to upgrade and improve upon as scalability research is completed and tested. Us and the entire ecosystem is continuing scalability research we usually publish articles here, if you want to keep track: https://medium.com/@tellor
Some of our competitors have been working on their projects for much longer and still have no decentralized product on mainnet. And there aren’t that many significant oracle projects in the market, and especially not many that are currently live on mainnet and decentralized. Some of the other oracles in the market include Chainlink, Witnet, Band.
Chainlink does not have a decentralized product on mainnet. We have a decentralized product on mainnet, now. We took a very pragmatic approach to the oracle problem. We are similar to Chainlink in that we are trying to tackle the same problem and we both agree that data is a necessity for defi to continue to expand. We admire their ambitious goals for decentralization but instantaneous queries of all types of data is still a ways away, since many other computer science problems need to be solved to make that possible. We developed Tellor using technology that we already know works for decentralization. Staking and proof of work. However, we built it in a way that is upgradable and once the technology gets there we will also be ready to update our system.
Witnet operates an oracle on its own blockchain and it is a great design but they are not live yet and aside from the oracle problem they will also have to solve cross-chain interoperability, and that is also probably ways away.
Band is also a great design but their incentives are a bit off(charging per data request), since the data cannot really be kept private on Ethereum and freeloaders will happen. For this to be successful, blockchain privacy issues have to be solved and that is also a ways away.
I saw the news that Tellor received grants from the Binance and ConsenSys, congrats. Where are you now? What’s your next move?
Michael: We have some awesome partners and we’re super excited to have such a great network of connection in the space. As of now, we’re launched and focused on growing our security as we start to implement contracts utilizing Tellor. What this means is that we’re trying to grow our network of miners, increase Tellor’s security (through a higher token price (more staked)), and onboarding customers to drive demand. We have generally been really busy working on integrations. We’re a very technical team that loves to build things and get them into production, so we’re working non stop and now even some members in our community are also working with us for TRB, which puts our dev-share to good use.
How does Tellor provide off-chain data?
Nick: Similar to how blockchains reach consensus on every block. We expanded that to allow miners to reach consensus on the ‘correct’ data. Tellor utilizes staked miners, so miners have to deposit 1000 Tellor Tributes (our token) in order to do PoW mining. For each block, we take the first 5 miners to submit the PoW submission and the requested data (e.g. the price of BTC). The median of the first 5 becomes the official value and any of these can be disputed and miners can lose their 1000 tokens if Tellor token holders vote that they were malicious.
It quickly becomes very expensive to spam or “break” Tellor for any extended period of time since a user can just request the data again and will get it in 10 minutes but the miner loses 1000 tokens and it’s locked out of providing any more data once any of their values goes under dispute.
What would happen when a miner submit a wrong data? How do you deal with it?
Brenda: We take security very seriously and the first thing prevention from a bad value being the official value is that we take a median of 5 values, so even if you get in a bad value, it will likely not be the median. For all 5 miners though, we have a dispute period, so if they submit bad data they are locked out of providing data and earning rewards. The community then gets to vote for 7 days as to whether the miner was malicious, and if they were, they get their stake gets slashed.
What are the considerations for designing Tributes in the Tellor system?
Nick: The main consideration was creating the correct incentives for miners to provide correct data, for our users to hold their tokens, provide measurable security for our users. You want to design a system where tokens are both scarce and valuable, but also available for users in a way that is not prohibitively expensive. We worked hard to build our system properly, and definitely head over to our Medium articles, we have a lot of discussion around some of the design choices we made and the way it may evolve going forward. Overally, I’m just happy we’re economists building this and have some of the game theoretical pieces under our belts to make the system work as well as it does.
What are the use cases for Tellor and what are the possibilities in the future?
Brenda: We are currently focused on DeFi or high value data that need higher security, a premium oracle. For two main reasons: 1)”DeFi” is currently not decentralized. Most of these projects use centralized oracles, which makes it just the same as traditional finance but much slower. Our mission is to help companies become decentralized and secure and to ensure user’s funds are truly secured. 2) Technology limitations. Since at this point in decentralized technology you can either be fast or secure and decentralized. Our focus is to help create unstoppable applications, fast and centralized applications already exist in the traditional system. But we have made our smart contracts upgradeable and we plan to continue to upgrade and improve upon as scalability research is completed and tested.
Do you think the different attitudes among governments would limit the development of DeFi? Will Tellor be affected?
Brenda: Definitely. Government attitude toward the current “DeFi” can limit or help accelerate its growth just like with any industry. However, if a DeFi company is truly decentralized, that is, their smart contracts and the data feed they use to execute those contracts there is nothing any government can do to stop people from using them(short of cutting all internet access). A completely decentralized application that uses a decentralized oracle will be almost impossible to shut down. If governments started looking into DeFi right now and their centralized data feed practices, I think they will very quickly realize that the consumer is at a high risk of losing their funds to a company gone rogue and probably limit their growth, based on that alone, since once your funds are gone on the blockchain, you are not getting it back…and no court system will be able to help the consumer. DeFi with centralized feeds is one of the biggest threats to blockchains right now, since they are the biggest use cases and these companies hold quite large funds and will probably be the first to be scrutinized by governments.
I think Tellor will thrive if governments actually start scrutinizing current DeFi practices and will probably push them to implement decentralized oracles like Tellor. These networks (Bitcoin and Ethereum), we’re built to be censorship resistant. We’re happy to be one of the teams actually building things that make sense (i.e. can’t be done on a database) on these networks.
What do you think of eth2’s changes for the development of DeFi？
Nick: It should just make it better. The whole idea of Ethereum is a censorship resistant platform for computation and they’ve really built it. The only issue (that ETH2) is solving is making it faster and increasing the volume it can handle. As more speed and throughput happens, you’ll see defi applications able to handle smaller transactions (large ones only make sense when fees are high) and enable a better user experience as you won’t have to wait so long for confirmations. That said, we’re built on ETH 1 and we work now.
Unitalk is Unitimes online AMA on WheChat for Chinese Ethereum communities, once a week, sharing the lastest technical updates, powerful insights and significant improvements on Ethereum. Stay tuned for our next episode, or contact us at talk@unitimes.io if you’re willing to share your views with us too.
[The copyright of this article belongs to Unitimes. Please contact us at editor@unitimes.io if you want to repost the article. Opinions expressed by contributors belong to themselves.]
