The DeFi Download

Tellor: Decentralising data oracles to increase trust.

April 22, 2021 Radix DLT
The DeFi Download
Tellor: Decentralising data oracles to increase trust.
Show Notes Transcript

Join Piers Ridyard in this episode of the DeFi Download, where he interviews Nicholas Fett, co-founder, and CTO of Tellor.io, a decentralised Oracle Network. 

Tellor is a decentralised Oracle Network that facilitates bringing off-chain data on-chain. It accomplishes this by rewarding competing miners with the TRB token for providing and validating data.

Tellor allows data to be disputed for a fee, maintaining the system's integrity. The fee increases for each subsequent dispute, making malicious parties’ attempts to attack the network prohibitively expensive.


SPEAKERS: Nick, Piers 

Piers 00:04

Hello and welcome. I am Piers Ridyard CEO of the decentralised finance protocol Radix, a public ledger entirely focused on bringing DeFi into the mainstream. This is our podcast, the DeFi Download, a show about decentralised finance and all things crypto, where we dive into the details of the projects, assets and services that are powering the DeFi revolution. Today, I am joined by Nick Fett, co-founder and CTO of Tellor.io. Tellor is a decentralised oracle network. Nick, thank you so much for coming on the show. 

Nick 00:35

Yes, thanks for having me. 

Piers 00:37

So, I think everyone has an idea, or I think a lot of people think they know what a decentralised oracle network means. But could you just break that down a little bit and explain to us, what makes Tellor different? 

Nick 00:52

Sure, yes. It must be a special podcast when you assume that space’s knowledge. So Tellor, what we do is we enable smart contracts to get off-chain data. So, to tell you how Tellor began, before we were running Tellor, we were actually trying to do a decentralised derivatives protocol. We were going to do long and short tokens on Bitcoin and Ethereum. And basically, what all of these derivative contracts and mainly all of DeFi comes down to is you're basically betting on prices. The only issue is that Ethereum actually can't access pricing information. So, if you think about a traditional computer application, it can just go check the API on Coinbase.com, or on Binance. But Ethereum smart contracts actually can't read API data, it can't access information from the internet. Ethereum as a whole is this sort of self-contained virtual machine. So, if you want, say, the price of gold, or the price of bitcoin on Ethereum, you need somebody to place it there manually. As you might be aware, having one person with the ability to say what the price of bitcoin is, or the price of Ethereum in your DeFi app is a giant centralisation risk because whoever can place that price can basically determine who wins or loses in the smart contract. So, what we did here at Tellor, is we created a way to get that pricing information but without a central counterparty. The way that we do that is we have a network of miners. We call them miners because they actually do perform a proof-of-work computation, they compete at a proof-of-work competition. Then in addition to submitting the solution to the proof-of-work solution, they also submit some requested pricing information, say, the price of bitcoin or the price of gold, and they submit it all on-chain to an Ethereum smart contract. This is how we're different than maybe the other oracle designs out there. We like to say, “We're the truly decentralised ones, we have a way for your smart contract to get this information, you don't have to rely on some central third party to do it, or some off-chain person to actually sign the data and put it on there for you.” 

Piers 03:11

So, you're using a sort of Nakamoto leadership proof-of-work race to determine who's the next person who says this is what the data is, for this particular request? 

Nick 03:24

Yes, so, it's just the basics of crypto-economic security. Like rather than just having one person, why don't we just create a competition to see who gets to do it?  

Piers 03:32

Right. What happens if the person who is picked gives false information? Is there a way of making sure that the winner of the proof-of-work race is also honest about what they say is the price? 

Nick 03:48

Yes, definitely. So, as most people know, it's actually relatively cheap to break most proof-of-work protocols, especially for a short period of time. Even Ethereum or Bitcoin are probably much, much easier to break for an hour than you might think those multibillion-dollar networks are. But the way that Tellor works is, we don't have just one winner, we have five separate winners. So, the first five solutions to the proof-of-work challenge place the price on-chain. So, let's say we're going after the price of bitcoin. Five different miners will place the price of bitcoin on-chain and the median one will be the official value. In addition to taking the median, so you'd need to corrupt three of the five miners, we actually require all of our miners to stake as well. So, they're sort of staked miners, if that makes sense. But what they have to do is they have to place a deposit of 1000 TRB into the smart contract and then, if they lie, they can be disputed. Let's say one of the miners says, “Hey, the price of bitcoin is $10 million.” That's obviously false. So, you would open a dispute on him and then it would go to a vote of Tellor token holders as to whether he lied or not and should be slashed or if he should be able to continue mining and that was a good value. 

Piers 05:05

Got it. And I assume that when you— How long ago did you start building Tellor? 

Nick 05:14

Summer of 2018, I think was the first hack on something Tellor-related. 

Piers 05:22

And before Tellor came along [crosstalk], yes, well, because you said you started by building a futures and derivatives, you were trying to build a futures and derivatives platform and I assume as part of that you came across this problem of “Oh, I actually really need to bring some off-ledger data onto the ledger.” What was the thing about decentralised oracle services that already existed at that time that didn't match up to what you felt was important for making sure that this works in a way that it needed to, I suppose? 

Nick 05:57

Yes, so I left my previous job in 2017 to do the decentralised derivatives platform. We were part of an Ethereum Foundation grant in early 2018 and that was the funding that we had got initially and, namely, it was to do research on oracles. It was fun to be able to really do a deep dive because, especially in late 2017 and early 2018, there were no decentralised oracles or even people claiming to be decentralised oracles like—  

Piers 06:33

Wasn't Oraclize already around before that? 

Nick 06:37

Yes. So, you had Oraclize. Oraclize, they never claimed to be decentralised in any way. The way that Oraclize works, if you can think about the oracle problem, is you need somebody to place that on-chain, they're just a centralised service that will do that for you. They're a great service or they were a great service but if the Oraclize company shut down, then you were out of a price. That was really the big risk with that because there was no guarantee that they would give you the right information or more, so, we weren't concerned that they were actually monitoring the contracts. We were just concerned like, what happens if they're not around in a year? What's going to happen to it. So, the other people that may have been alive, my guess, coming alive in 2018 was Augur, which some people like to think of as a decentralised oracle. But that's not— Augur, as a prediction market isn't necessarily fit, especially the v1 Augur, it wasn't really fit for purpose for any sort of DeFi needs, if you will. So—  

Piers 06:54

Right, it's a betting marketplace that never had quite enough volume on it to make it actually good for using it as a price feed anyway, for anything. The values of the bets were so small that actually going in and if it was a valuable, you could attack the oracle relatively cheaply, essentially.  

Nick 08:06

Right, yes. Augur in general never really got the traction that we were hoping for it to get. But there were some other pieces as far as just— It wasn't necessarily straightforward like— Listen, we were looking for a once-a-day bitcoin price on-chain. Using the entire Augur setup was really cumbersome to do that. It wasn't necessarily built for that. I came from the regulatory environment myself and prediction markets for oracles— prediction markets, as you may know, are actually illegal in the US so it adds a lot of legal risks to a project that doesn't necessarily need to be there for just putting an end-of-day bitcoin price on-chain. 

Piers 08:52

Right. And so, you realise that there was this need for decentralised oracles both for the regulatory piece and also for the anti-fragility element that you wanted to achieve. 

Nick 09:07

Yes. I started building things on Ethereum and my philosophy on the space in general, if you're going to be building something on Ethereum, or on one of these blockchain networks, for the most part, it's because you value censorship resistance in some way, shape, or form. If you're okay having custody over people's funds, and you're okay with having central counterparties, you probably won't deal with the giant UX problems and speed issues that using a blockchain entails. And since that was always how we felt if we're going to build it, you should build it right. Because if you're just building something that has this decentralisation theatre, ultimately, the regulators will probably crack down on you, but even more so it just doesn't make sense. It doesn't fit with the ethos of why you'd be building it. So, we wanted to build something that actually made sense from a religious fit with our philosophy, and making sure that whatever it was, it was actually decentralised, or theoretically so and we didn't really see that in the space. Ultimately, whenever you looked at what the space of Ethereum looked like before oracles or anything like that, it's a different world. It's a lot harder because even things like most of DeFi, I'm sure as you know, rely on oracles. So whether it's, you know, Maker relies on their ETH:USD oracle or any of these derivatives contracts— 

Piers 10:35

[Crosstalk] And Maker's were an interesting one because they had to build their own, right? Because they did, it started so early in the space that a lot of the stuff that they had to build just didn't exist before they built it, like the auction model that they use for auctioning off the Collateralised Debt Position contracts, or the oracle contract itself. I think that they have explored moving to other providers, but still, I think it's just their oracle. Right? 

Nick 11:07

Yes. They haven't changed their oracle and they're okay with that. They don't really seem to be getting much pushback from community users on the decentralisation piece of their oracle. I think now, they're more focused on what the proper collateral on their system is, which is also a whole another discussion. But yes, they, like any of these early projects, you really just had to build all of the base, the base pieces. Now we're seeing protocols that are able to be an awesome DeFi protocol solely because there are stablecoins. Whenever you assume that the oracle is decentralised, or you assume that the stablecoin is actually stable, then you can do some fun things with those assumptions and you can create new products. But three years ago, everyone was still working on, “Oh, my God, how can we get a price on-chain?” 

Piers 11:59

Right, right, right. So, how have you seen the space develop around you? Obviously, the most famous or the most well-known oracle platform today at the moment is Chainlink. How have you seen entrants like that come into the market? What do you see as the opportunities and flaws in the way that they've approached this versus the way that Tellor approaches this? 

Nick 12:28

I think the biggest thing may just be like a philosophical way that we build things versus Chainlink. So Chainlink has actually been around for a long time, they just notoriously didn't have anything launched for a long time. They had just done some research and had some stuff up on testnet for you to use, but that was really about it. Now the only thing that's live with Chainlink, they have these price reference contracts that you can use, which namely, you can— So, if you want the price of Ethereum - US dollar, they have a contract on Ethereum that you can read the price of US dollar on and the way that it works is they just whitelist these data providers to provide an Eth - US dollar price that they push on-chain. So, they'll have a number of, say, 21 providers and all sign the prices and push them on-chain. This is a much different model than a Chainlink white paper, but also from Tellor. As you can tell, it's not really decentralised, it's more just sort of a majority consensus, even if you're ignoring say the whitelisting aspect of the central team. They still do have plans to decentralise and create the crypto-economic incentives to actually have it make sense, but we'll see. There's a lot of technical challenges around some of the design pieces, with having a really faster, a really robust oracle. I know they're still working on solving some of those pieces. We had the opposite approach. They built something that was very fast and usable right now, and they said, “We'll figure out the decentralisation piece later.” Whereas we said, “We're holding the decentralisation piece key, what can we build now?” So, our oracle might be a tad bit slower and more expensive than Chainlink's, but we prioritise one piece over the other. 

Piers 14:32

So, it's this idea of, again, coming back to the idea of anti-fragility, and making sure that the system itself in which you want to build a trustless ecosystem around is also inherently trustless itself. 

Nick 14:49

[Chuckles] Right. Yes. It makes sense. But a lot of DeFi, I guess, isn't really building in that way. Why are you building on Ethereum? Well, you're only really as decentralised as your least decentralised point. So, if everything in your protocol is decentralised except for your oracle, well, there's a giant hole there. That's one of the issues that DeFi really needs to address, as far as making sure there are no holes in the logic there that people can go after. And really just what we're focusing on is trying to be an oracle for things that actually do make sense to be built on Ethereum or to be built on some of these networks. We often have calls with people who are looking for oracles and they're trying to build things that just don't really make sense on these networks. The classic example is, if you want to build a high-frequency futures exchange on-chain, gas costs aside, it just doesn't make sense. There are giant front running problems on all of Ethereum and you have a lot of issues that you don't want this on-chain. Miners can extract a whole lot of value, and people can front-run your oracle, and it's semi-cheap to just censor your oracle or sensor people from calling settlement on your contracts. What do you do in these cases? People just aren't necessarily thinking through a lot of these. They're just seeing, “Oh, what are some things that are popular in traditional finance and how can I put them on-chain?” And that's probably the wrong approach to take. 

Piers 16:31

So, what are the things that you're excited about people building with Tellor?  

Nick 16:36

Oh, yes. I want people to build things that actually get usage and get usage in let's call the right way, [laughter] if it sounds cliche. But you know, whenever you see people using crypto, you want to see them using it, not just as these sort of scammy quick money-making opportunities. You want to see people building stablecoins off of Tellor. You want to see people building derivatives contracts off of Tellor that actually enable them to do something that they couldn't otherwise, or do it in a censorship-resistant manner, whether your government doesn't allow you to do this, or doesn't allow you to make these bets, getting around it by using one of these contracts, that's really what excites me. It's not one of these things where— I'm not really excited for the DeFi that is like, “Oh, we're going to have automated finance or have smart contracts that enable everything to go smoothly.” That doesn't really excite me. I think the pieces that excite me more, okay, these are outside of the reach of regulators, these are— You can build things that actually are competing with, say, the US monetary system or they're competing with products in a bank, and you can have things that are more community-based, you can have things that have different inflation protocols built in. And you can use Tellor to kick off some of the events or link to real-world data. Those are definitely some of the things that I'm excited about. 

Piers 16:41

Got it. And what kind of usage are you seeing on your network at the moment? 

Nick 18:16

So, we're still pretty new, as far as getting usage. We're just integrating with some users as far as some lending protocols and some stablecoins and it seems to be the traditional DeFi users or people starting to build on DeFi. It's the people mainly in DeFi now who are just looking for crypto prices. We're ultimately excited whenever we move beyond the need for just crypto prices. There's a whole world of different assets as far as real estate, gold, stocks and all of these things really have yet to be tapped by any of these smart contracts. 

Piers 18:56

How easy is it to add a new feed if someone comes along and is like, “Hey, man, I really need a feed that gives me the price of eels per pound against the price of gold”?

Nick 19:12

Yes, so, we actually specify all of the data off-chain, so you would get a new ID, and then you would have to tell the miners where to get the data. You can think we're a decentralised network of these data providers so you have to tell them where to get it. It's generally really easy to, like if you want to add a new crypto price, we can add that in a day or two, just basically for the time for them to all see the pull request and then run it. But if the issue comes in, we don't really do more of the manually entered data quite yet. We need to know like, “Hey, what API's can we pull this from that will actually give us the correct price?” 

Piers 19:58

What about API's that cost? What about feeds that there is no good, generally available public information to get it from but the feed is acquirable? I'm thinking things like bonds and bond pricing stuff that you can only get through Bloomberg or stuff like that. 

Nick 20:20

Yes, we can do those. The big issue that we try and— We try and avoid one or specific prices that only have a very limited subset of APIs. And the reason is so you can think about what data actually is. If you think about the price of bitcoin, there's the price of bitcoin, there's the price of bitcoin according to Coinbase, there's the price of bitcoin actually being traded on Coinbase, and then there's the price according to some other exchange or at some depth. The difference is that if you want the price of, say, a bond, according to the Bloomberg API, well, the problem is that now the Bloomberg API can just shut off and censor you. So that doesn't really fit with the idea of being decentralised. If you're reliant on one API, and if that API goes down, the whole thing shuts off. So, what we try and do is we try and say our ID number two is the bitcoin price. We actually don't care where you get the bitcoin price from. If you want to use a paid CoinMarketCap API, you are more than welcome to. If you want to use the free CoinGecko one, you can also do that. It's really up to you. Usually, for most of the miners, we give them a whole list of APIs that they can choose from, and they just take a median of it and submit it on-chain. Ultimately, it's this ambiguity that actually gives it strength and decentralisation because all you have to do is the bitcoin price has to be valid so that nobody disputes it. Somebody can challenge it and say that it's wrong so you just want to make sure that we're putting a price on-chain that nobody is going to dispute. The problem is that if you are reliant on, say, just one specific API, the API can show off or something like that and it really doesn't even make sense to use a Tellor. Just ask them to sign the price and put it on-chain. 

Piers 22:16

I mean, that can be difficult [laughs]. Just getting any company that has a price feed to sign it and put it on-chain or any kind of feed to sign it and put it on-chain that's often just as difficult. I have had plenty of conversations about it with identity companies, identity providers, and just being, “Can you just sign the identity once you validated it. You don't have to put it on-chain, you just have to sign it with a key and pass it across to us so that we can prove that you validated this identity.” Just getting people to do that can be amazingly difficult so I think things like Tellor are really important when there are huge numbers of data providers. But the universe of people who are data providers and also willing to interact directly with ledgers is still very, very small. 

Nick 23:07

Yes, and I think it's for good reason too. Like for exchanges, for instance, and even in the hypothetical Bloomberg signing bond prices example, if you create a derivative contract that settles to Coinbase's API, now— If you're a derivatives protocol it's completely decentralised now as Coinbase is liable for starting the derivatives contracts, from a regulatory perspective, or most of the time even Coinbase actually says, whenever you're using their API, you can use this to settle a derivatives contract. So, you don't want to be illegally pulling prices from Bloomberg and settling derivatives contracts on them, they'll probably sue you. So yes, you really want to make sure that— That's why we like to go with just more of the ambiguous approach to— You get to decide what is correct and, listen, if you want to go pull from them, fine. Please don't tell us about it. But you know [laughter], you're welcome to pull from any, any API source or any place you can possibly find the bitcoin price or look it up on your own or ask your friend. 

Piers 24:10

So, what's the vision of Tellor? Where do you want to be in two years, in five years? 

Nick 24:18

The vision was always just what can we actually create. We were building a decentralised oracle because we needed a decentralised oracle for ourselves [laughs]. That was ultimately something that we think that we've done a great job at, and now we're continuing to make it better. There's a whole lot of ways that we can continue to make it faster, support different kinds of data and also work on just scaling how easily it can be used by, say, smart contracts on Ethereum, but then also seeing it put on other networks. It's something that we're definitely looking into as well. So, within the next few years, I definitely see us working with other networks, making sure that Tellor can support chains that are not only Ethereum but also say layer 2 or other layer 1s. Then, from there going in, and just expanding things that we think need to be built. We built an oracle because nobody else was building a properly decentralised oracle so we felt the need there. The next piece is, what else? What do we think can be built in DeFi or in the crypto space that isn't being built? Where can we step in and lend our expertise? That's always been the vision as far as where we're going to be putting our resources and— 

Piers 25:37

[Crosstalk] So do you think the decentralised derivatives thing is going to be something you guys will still pursue? Or is it that was your path to oracles but you're sticking with oracles? 

Nick 25:50

I don't think we'd go down derivatives necessarily. There's a lot of other people tackling that problem at the moment. We'd love to partner with some of those people that are doing that and make sure that they're doing it correctly. Our vision for finance in general, it's at odds a little bit with the DeFi space right now. I've never been really big - I think I have already said it on this podcast - never, never a really big fan of the super-fast automated finance. The power of these networks, in general, is the fact that you can bring communities closer together and that you can enable more. Yes, you are trusting more in code but you can also work together to create a lot of things. Even if you have slower finance, if you have slower finance that nobody can shut down, that's okay. That's also really cool. 

Piers 26:40

Interesting. I can definitely see that for things like stablecoins, where you're like, “I want a way of creating a non-government backed, stable store of value and I need an oracle to give me a reference point for relative stability.” Is there anything outside of the stablecoin itself that you think is an important thing that hasn't been decentralised properly yet? That is waiting in the wings to be created or perfected? 

Nick 27:15

Sure, yes. I think the stablecoins are one that's definitely there. It's weird because stablecoins, and derivatives, and lending protocols, there are always ways to tweak it and make them better. But I think it's just going to come down to whether or not teams do as they're planning, or what happens is, as time goes on— With a lot of projects, a lot of times, what you'll see is, it looks like they're properly decentralising things, and then they don't. Then they go back on promises [chuckles] or move down somewhere out that you don't necessarily want them to. I think what we've seen, especially here with DeFi, is that a stablecoin is also a derivative but is also a lending platform. So, no matter what you call it, they're all doing roughly this similar thing, as far as you're betting on crypto prices in the long run. But what we really want to do is, what are you using this stablecoin for [chuckles]? How are people getting in and what are they ultimately buying with this stablecoin? I really hope that we can help crypto, in general, get back to that. You see a lot of focus around stablecoins and derivatives as far as, “Oh, come bet on this price with this amount of leverage,” or “You can make X amount of return.” It's like, “No, let's go back to buying pizza with Bitcoin.” Let's go back to, “Here's a stablecoin. Now, here's a bunch of things you can buy with a stablecoin. Here's a bunch of apps that you can use with this stablecoin.” Now you're back to crypto actually enabling something, not just this purely theoretical, nerdy finance loop that we've gotten ourselves stuck in. 

Piers 29:06

Would you mind if I just run a couple of numbers across the Tellor security paradigm? Because I'm just curious about [laughter] because I've been thinking about this just in the back of my head. 

Nick 29:21

Yes, I know what you mean. But let's go [laughs]. 

Piers 29:24

So, we've got a proof-of-work system. What kind of proof-of-work system is it? What's the proof-of-work algorithm? 

Nick 29:31

Yes, so since it's all validated in Solidity, it's going to be the three hashing functions in Solidity. So, you got Keccak, SHA-3 and RIPEMD. It's custom. 

Piers 29:42

Okay. So, it's— there aren't a64, but it is a6-able. 

Nick 29:44

It is an a6-able. Yes [laughs]. 

Piers 29:53

Right. And so, what we've got— Okay, so we have a standard hashing function and then we have a staking system where the value of the token is derived from the fact that 50% of the fee is burned, right? So, every time someone requests a price on Tellor, you are paying in the underlying token, and 50% of that is burned, and 50% of it goes to the miners a fifth each because there are five miners in total. Is that correct? 

Nick 30:31

Yes [crosstalk], for our listeners, yes, you can think of it like a gas fee on Ethereum. 

Piers 30:36

Right, right, right. So, the cost of attacking the network, on any given request, is the cost of achieving two-thirds of the hashing rate, well, slightly, three fifths, yes, right. So, three-fifths of the hashing rate plus— Well, no, actually, it's just three-fifths of the hashing rate, right? I need enough stake to stake, but if I am three-fifths of the hashing rate, then I can't be voted against so I can't actually get slashed. Is that correct? 

Nick 31:08

No, no. The way that it would work would be so you would need to be three of the five miners in the block. So, miner, one, two, and three, and then you can place a bad value on-chain. So yes, you could play someone on-chain, but then you would get disputed. People very quickly in our network, see that “Hey, he lied” and then they would open a dispute against you. That value would be taken off-chain. Then let's say the person who wanted the price of bitcoin would just request it again. This is an important feature in Tellor security that the actual cost to break and put a bad value on-chain would be as if you break the voting on the disputes. So, you end up with you own— If you own 51% of the tokens, then you're sure you officially broke Tellor, but actually, Tellor's security is measured in a cost-to-break-per-hour or per-block. So, it would cost you 1000 tokens for the stake three times. So, 3000 tokens plus the cost to break the proof-of-work every five minutes. So, it would cost you the 3000 tokens, and then the proof-of-work. So, if you assume the proof-of-work is negligible so 3000 tokens times 20 bucks. So, it's about 60 grand every five minutes. 

Piers 32:28

Well, no, because I don't lose the tokens. If I've broken the voting system, I don't get slashed. So, I—  

Nick 32:34

Yes. So, you would have to break the voting system.  

Piers 32:36

So, I break the voting system. 51%, I have to own 51% of the tokens, so as a single— How does the voting system work? Everyone has to vote on a dispute? 

Nick 32:46

No, it's a multiple dispute piece. If somebody, if you open up a dispute on the network, then there's a two-day voting period and as just a pure majority, whether or not it's good or bad [crosstalk] and you either get slashed or not slashed.  

Piers 33:05

So, it's an opt-in voting. So, you can have plenty of token holders not participate. 

Nick 33:12

Yes, no, I'm sure most people don't and that's fine. 

Piers 33:17

So, breaking the voting system would be less than 51% of the tokens then? 

Nick 33:23

Well, theoretically, so it's— That's what I was getting into with the multiple disputes. So, after the two-day vote period, there's a day where you can re-challenge, that you can open a dispute on the dispute [chuckles] to where, “Hey, that didn't work.” Now there's a four-day voting period, as far as whether or not it's correct. Then, if that one also breaks, you can dispute the dispute ad infinitum to where hopefully the network will, people will rally to vote around it and save the [laughs] vote with their TRB. 

Piers 33:58

Got it. Okay, so when I get the information back from the three, let's say I got majority right, and I go right, someone's requested the price of bitcoin to Piers coin, and I've been like, right, I issued a load of Piers coin. Now my Piers coin is worth one Piers coin to one bitcoin. It's been a very successful coin. So, I attack the Tellor network. Is it possible, from a smart contract point of view, is it possible from the entity that is using Tellor as the oracle to see whether or not something is disputed? How quickly does that get viewed? Let's say that it's some kind of oracle contract into a decentralised DEX and I've done a trade against that. As soon as someone spots that, say that someone's attacked the oracle and a dispute has been opened, is that something that can actually be seen as part of the feed that is coming into that smart contract or is that not part of the smart contract data in the oracle feed? 

Nick 35:00

You can view it, it's just— So once the piece of data is put on-chain, it's open for anyone to read. The way that we like to tell people to do it is, once a piece of data is put on-chain, you should probably wait a minute or two for a dispute, similar to like, “Why don't you wait 10 minutes before you read the price of bitcoin?” That way, there are 10 minutes for somebody to dispute it. This is similar to Bitcoin. So, Coinbase, if you buy something on Coinbase, you have to wait an hour or you deposit Bitcoin on Coinbase, you have to wait an hour before you can actually cash out. The reason is, you have to wait for block confirmations and make sure that the whole chain isn't going to revert. The same thing goes with Tellor. The more money you're dealing with, you might just want to wait to look at the Tellor price to make sure it's not going to get disputed. This comes into make sure it's slow, make sure it makes sense and can be secure.  

Piers 36:01

So, who can raise a dispute?  

Nick 36:04

Anybody. You just have to pay a fee. 

Piers 36:07

Oh, you have to pay a fee to raise a dispute? 

Nick 36:10

Yes, you have to pay a fee because we don't want you just disputing all miners [laughter]. 

Piers 36:17

Okay, so I pay a fee to raise a dispute, the dispute gets raised, and then it goes into a voting thing. Presumably, what would happen next is actually from the point of view of the DEX, they would just re-request the information because the information wouldn't be good. They'll just be, “Ah, it's disputed. I'll just request that again.” 

Nick 36:40

Yep. And then it could happen, theoretically, again, and again, and again, and so, obviously, we were saying it could be like 60 grand every five minutes, you'd have to be burning. So, as long as you waited a sufficiently large amount of time, given what piece you're putting on-chain, you'd be fine. 

Piers 36:59

So how much of the 60 grand is the power and how much of it is the buying of the tokens? 

Nick 37:03

Oh, this was actually just assuming that the power is free [chuckles]. 

Piers 37:07

Okay, so 60 grand worth of tokens. 

Nick 37:11

Yes, and so, if you look at what is that per hour, that's like 720 grand an hour. 

Piers 37:19

60 grand of tokens— Hang on, but 60 grand of tokens is 50%-51% of the total tokens, or is it, that's—?  

Nick 37:27

No, this is just a 3000 token for the stake.  

Piers 37:30

That's just the 3000 tokens. Got it. Got it. Got it. Okay. So, theoretically, quite an expensive network to break. But as you were saying, quite a slow network. The certainty comes from the way in which disputes and security is built and disputes and security is built over time in much the same way that Bitcoin security is built plus some extra clever twists like the staking aspect. So that suits it for things like settlement of long-term agreements, things like that. Okay. 

Nick 38:13

Yes, the one thing I always bring up that people don't pay attention to with oracles like I was saying, it would cost to break Tellor, for an hour it would be like 720 grand. But like with proof-of-work on Ethereum, so right now, it would probably cost you about 200 grand only to mine every block on Ethereum for an hour. This is a giant attack factor that most people don't pay attention to because if you have a derivatives contract that relies on a fast oracle, if somebody mines every block on Ethereum, they can just prevent contract calls to that oracle contract [chuckles]. So, you're never getting an oracle update for that hour that somebody is mining all of the blocks on Ethereum. So Tellor, I would say is sufficiently secure that it's actually cheaper to just mine every block on Ethereum than to try and break Tellor specifically, and just prevent the miners from submitting solutions to Tellor. You're not really breaking Tellor, you're breaking Ethereum at that point. 

Piers 39:10

Interesting. Very interesting. It's certainly a really cool system. I'm definitely going to spend a bit of time diving into it. If people want to find out more about Tellor and about your token and want to chat with your community, etc. where do they go and what do they need to read to find out more? 

Nick 39:31

Sure. So just Tellor.io, we have a white paper and then we obviously have a Telegram, a Discord. We have a very active community, same with, you can just go follow us on Twitter. I'm on Twitter as well. I'm on Telegram, Discord, just feel free to reach out to me. I'm very active in all of our groups too. 

Piers 39:49

Awesome. So that's Tellor, t e l l o r dot i o. 

Nick 39:55

T e l l o r. There's actually another Teller, t e l l e r, but that is not us [laughter]. We came first. 

Piers 40:05

And just before you go, why the name Tellor? 

Nick 40:08

We were thinking it was a mix between ‘tell’ and the word ‘oracle’. Originally, we were building this as an internal tool for our derivatives protocol and we didn't put much thought into the name [laughs] but it stuck and we've fallen for it. 

Piers 40:26

I like it. Tell-oracle. I like it. It's neat. It's been such a pleasure having you on. Thank you so much for coming on the show and really looking forward to seeing you guys grow. 

Nick 40:37

Yes, thanks. It was great to be here.