CoinGecko Podcast - Bitcoin & Cryptocurrency Insights

Decentralized Oracles with Nick Fett and Michael Zemrose of Tellor - Ep. 16

August 07, 2020 Bobby Ong Season 1 Episode 16
CoinGecko Podcast - Bitcoin & Cryptocurrency Insights
Decentralized Oracles with Nick Fett and Michael Zemrose of Tellor - Ep. 16
Chapters
CoinGecko Podcast - Bitcoin & Cryptocurrency Insights
Decentralized Oracles with Nick Fett and Michael Zemrose of Tellor - Ep. 16
Aug 07, 2020 Season 1 Episode 16
Bobby Ong

In this episode, Bobby Ong, co-founder of CoinGecko is joined by Nick Fett, CTO at Tellor and Michael Zemrose, CSO of Tellor. Bobby interviewed Nick and Michael on the Tellor oracle, how it works, the utility behind Tellor Tributes, as well as Tellor’s upgrades in V2.

[00:00:02] Intro
[00:01:20] Background on Tellor
[00:03:59] The main difference between Tellor and Chainlink
[00:05:19] How does Tellor work?
[00:12:44] Tellor’s upgrades in V2
[00:18:55] Who’s using the Tellor oracles?
[00:21:03] The main utility behind Tellor Tributes (TRB)
[00:25:40] Attack risk on Tellor
[00:33:42] Where to follow Tellor

Quotes from Episode:

“I don't think Chainlink is going to be able to achieve that type of quality going about things the way they are using white-listed data providers like they do. And so that's a huge difference. Like with Tellor, we focus on that Bitcoin style quality where decentralized is our number one priority and censorship-resistance is our number one priority. So you know, in a nutshell we feel that we're the decentralized option between the two.” [00:04:49]

“The problem is, is that, you know, at Coinbase for instance, they just started an oracle service where they sign the price every so often and you can grab it and say, 'Hey, this was the price at Coinbase.' The problem with that is Coinbase can at any point stop doing that. There's no guarantee that they will continue to do that.” [00:07:42]

“Now you have to stake a 1,000 TRB and you can only mine one value per hour. And the reason that we're moving to that is, any sort of minor centralization can sort of be, if you want to own 90% of our hash power, you're going to have to own a whole lot of our token as well.” [00:15:25]


Links

Tellor - https://tellor.io/
CoinGecko - https://www.coingecko.com/
Tellor (TRB) on CoinGecko - https://www.coingecko.com/en/coins/tellor

Social Media

Tellor:
https://twitter.com/wearetellor
https://t.me/tellor
https://discord.gg/n7drGjh

CoinGecko:
https://twitter.com/coingecko
https://t.me/coingecko

Show Notes Transcript

In this episode, Bobby Ong, co-founder of CoinGecko is joined by Nick Fett, CTO at Tellor and Michael Zemrose, CSO of Tellor. Bobby interviewed Nick and Michael on the Tellor oracle, how it works, the utility behind Tellor Tributes, as well as Tellor’s upgrades in V2.

[00:00:02] Intro
[00:01:20] Background on Tellor
[00:03:59] The main difference between Tellor and Chainlink
[00:05:19] How does Tellor work?
[00:12:44] Tellor’s upgrades in V2
[00:18:55] Who’s using the Tellor oracles?
[00:21:03] The main utility behind Tellor Tributes (TRB)
[00:25:40] Attack risk on Tellor
[00:33:42] Where to follow Tellor

Quotes from Episode:

“I don't think Chainlink is going to be able to achieve that type of quality going about things the way they are using white-listed data providers like they do. And so that's a huge difference. Like with Tellor, we focus on that Bitcoin style quality where decentralized is our number one priority and censorship-resistance is our number one priority. So you know, in a nutshell we feel that we're the decentralized option between the two.” [00:04:49]

“The problem is, is that, you know, at Coinbase for instance, they just started an oracle service where they sign the price every so often and you can grab it and say, 'Hey, this was the price at Coinbase.' The problem with that is Coinbase can at any point stop doing that. There's no guarantee that they will continue to do that.” [00:07:42]

“Now you have to stake a 1,000 TRB and you can only mine one value per hour. And the reason that we're moving to that is, any sort of minor centralization can sort of be, if you want to own 90% of our hash power, you're going to have to own a whole lot of our token as well.” [00:15:25]


Links

Tellor - https://tellor.io/
CoinGecko - https://www.coingecko.com/
Tellor (TRB) on CoinGecko - https://www.coingecko.com/en/coins/tellor

Social Media

Tellor:
https://twitter.com/wearetellor
https://t.me/tellor
https://discord.gg/n7drGjh

CoinGecko:
https://twitter.com/coingecko
https://t.me/coingecko

Bobby Ong (00:02):
Welcome to the CoinGecko podcast. I'm your host Bobby Ong. Each week we will be interviewing someone from the blockchain industry to learn more about this fast moving crypto currency economy. If this is your first time listening, then thanks for coming. The CoinGecko podcast is produced each week to help you stay ahead of the curve. Show notes can be found at podcast.coingecko.com. I highly encourage you to join our newsletter where we send out top news in the crypto industry every Monday to Friday. Come back often and feel free to add the podcast to your favorite RSS feed or iTunes. You can also follow us on Twitter and Telegram at CoinGecko.

Bobby Ong (00:39):
Hi everyone. Welcome to the CoinGecko Podcast. For today’s episode, we would love to welcome Nick Fett, CTO at Tellor and Michael Zemrose, Co-founder and CSO of Tellor. Nick left his job at U.S. regulator, Commodity Futures Trading Commission, (CFTC) about 3 years ago to build an Ethereum based derivatives startup. Now he's the co-founder and CTO of Tellor, a decentralized oracle and an evangelist for censorship resistance and proper crypto-economics. Michael Zemrose is the co-founder and CSO of Tellor, whose entrepreneurial journey zigzagged all over the map until it landed on something that matched his passion for emerging technology, freedom, and trolling scammers on Telegram. Welcome to the show Nick and Michael .

Nick Fett (01:20):
Thanks for having us, Bobby.

Michael Zemrose (01:20):
Thanks for having us, Bobby.

Bobby Ong (01:20):
So, for the first question, can you explain to us what Tellor is in a simple, non-technical manner to our audience here at CoinGecko?

Michael Zemrose (01:20):
Okay. So Tellor is a decentralized oracle, which in blockchain terms is any entity that brings off-chain data on-chain. Because blockchain are siloed from other network. So with Tellor our unique way of doing this is through competition and crypto economic incentives to bring that off chain data on chain. That's like a super high level, super simple way to explain it.

Bobby Ong (01:59):
And I remember reading in your FAQ that Tellor was born out of the need from your previous start up Daxia, which rely on a centralized Oracle. Which centralized Oracle was it and why was the architecture flawed, causing you to start Tellor?

Nick Fett (02:12):
So we were just kind of using, we built our own centralized Oracle. So that's like, you know, if you're running a derivative startup, whenever you are just sort of, you need somebody to enter in the price. You just start-date and end-date the price off of it. And this was back in 2017 and actually nothing was launched. Oracle [inaudible] You had Provable, which we were using, you know like oracle-ized at the time, and that was, we had experimented and played around with that, but that's just a completely centralized Oracle service. So yeah. And then we realized, you know, whenever you have a derivatives contract, you need to decentralize the Oracle. So we set off to kind of build our own since we didn't really like any other models being built.

Bobby Ong (02:55):
And your derivative startup was either a centralized derivative exchange like BitMEX or was it more like a decentralized derivative exchange, that's why you need a decentralized oracle?

Nick Fett (03:06):
Yeah, it was a really decentralized one. So it was, you know, I like to say similar to like a dYdX. So we wanted to do a tokenized derivatives, which was the kind of the goal of what we were, that was kind of our tagline. So you could have like a long Bitcoin token or a short Bitcoin token. Kind of a synthetic cryptocurrency token that was backed by a week long or month long derivatives contract.

Bobby Ong (03:31):
Which is kind of what the leverage tokens are doing around these days, right?

Nick Fett (03:35):
Yeah. So dYdX is pretty similar. Synthetix is similar except, you know, they use the pooling mechanism versus the individual contracts. And yeah, there's, you know, it's definitely something that a lot of people kind of realize the need for. And we're glad to see other projects can take it head on, we just wanted to build the critical piece of infrastructure, which was the Oracle.

Bobby Ong (03:59):
And I'm sure like everyone in the space is quite familiar with your competitor, Chainlink when it comes to the decentralized Oracles. I'm sure you get this question a lot as well. What's the main difference between Tellor and Chainlink?

Michael Zemrose (04:12):
So I think the main thing is that Chainlink is focused on what they call quality of data, which they get by using known reputable nodes who have access to private permission API data. And so that's how they define this quality term, which is kind of like how we think of quality in a more centralized world. But the thing is Bitcoin came along and sort of enabled a whole new type of data quality. Data quality that would be defined by it being decentralized and permissionless and censorship resistant. And I don't think Chainlink is going to be able to achieve that type of quality going about things the way they are using white-listed data providers like they do. And so that's a huge difference. Like with Tellor, we focus on that Bitcoin style quality where decentralized is our number one priority and censorship resistance is our number one priority. So you know, in a nutshell we feel that we're the decentralized option between the two.

Bobby Ong (05:19):
All right. So Tellor basically, I mean, for example, CoinGecko has an API and we are one of the trusted providers on Chainlink that feeds data into the Chainlink Oracle. And they use the same, for example, to get Bitcoin price they also use data from some other API providers as well. So let's look at Tellor, right? So how does Tellor work? Like how do you get data on to your Oracle, for example?

Nick Fett (05:42):
Yeah. So the way that Tellor works is we have a network of stake miners. So anybody who grabs a thousand TRB tokens and stake their tokens in our smart contracts and then they can compete at proof-of-work mining. So you, if you running, say a derivative's decentralized, derivative's contract, and you needed the price of Bitcoin, you would come to the Tellor contract and asking for the price of Bitcoin. Then those network of miners would compete at a meaningless proof of work calculation, and they would submit the solution along with the price of Bitcoin to an Ethereum smart contract. We take five of the submissions and the median is the official price of Bitcoin. And the important piece is that any of those five submissions can be disputed. So Tellor open holder would say, "Hey, he lied" and then you can potentially lose your stake if you did maliciously lie.

Nick Fett (06:39):
But the difference in kind of how, whenever you're talking about what data they're putting on chain, so you have to, it's almost a separate question versus how it gets on chain. The idea of what piece of data it is, you know, like there's the price of Bitcoin, there's the price of Bitcoin according to CoinGecko, there's the price of Bitcoin according to Coinbase, there's the price of Bitcoin that you can trade a thousand Bitcoin at on Coinbase and then there's the price of Bitcoin at 4:00 PM on this specific exchange, you know, that's going to be physically delivered on the CME exchange. So there's a whole bunch of different ways to define what data that you want on chain. And what Tellor's focused on, you know, if you actually want censorship resistance in any of these DeFi contracts, which is what we think that people are ultimately using Ethereum for, you probably don't want to just use one service provider or one kind of, we're going to use Coinbase's price.

Nick Fett (07:42):
The problem is, is that, you know, at Coinbase for instance, they just started an Oracle service where they sign the price every so often and you can grab it and say, "Hey, this was the price at Coinbase." The problem with that is Coinbase can at any point stop doing that. There's no guarantee that they will continue to do that. And that's kind of the life experience piece, but there was still always be a Bitcoin price. And that's what Tellor focuses more on. It's more of that abstract concept of, you have to post a Bitcoin price and it is the Bitcoin price according to the Tellor network, not according to this specific API. So it's a little bit different of a concept but it's one that we think can actually provide censorship resistance. For instance, you know, if the goal of the API is what's the price of Bitcoin according to CoinGecko and you guys are ordered by some regulator to not give the price, you know, that's too bad for that contract if the API shuts down. So we would have to, you don't want to be kind of be hold to the API service provider.

Bobby Ong (08:47):
So, and the miners get to decide like what would be the best Bitcoin price because they have, at that point in time when they submit the data. And they could like for instance, take the average Bitcoin price or the Bitcoin price on Coinbase or the average Bitcoin price across different exchanges. It's up to the miners to decide what to submit to the chain.

Nick Fett (09:06):
Yeah. So usually, you know, whenever the Bitcoin price gets updated quite often with our Oracles and we give them a whole list of free APIs that they query from and should take the median of, CoinGecko is one of them. You guys are great. And then, you know, but if they want to throw in paid APIs, they're more than welcome to, you know, we just, we just need the price of Bitcoin. If you want to manually type it in, by all means, go for it. It's just basically all that we see is like what you push to Ethereum. So you enter the price of Bitcoin on Ethereum. And as long as the Tellor network agrees that that is a valid price of Bitcoin at the time, you're good to go.

Michael Zemrose (09:45):
Yeah. We feel like we incentivize accuracy. And so that'll be there, but it's really not about that in and of itself. It's about the sort of the principle behind how that data is delivered. And that goes back to why is anybody doing anything on a blockchain? You know, why are we all here? Like, what is this really about? Is it about accurate data? You know, I would say, I think we would all agree, it's about more than that. It's about censorship resistance. It's about unstoppable applications. And so like, that's the important thing. It's like, you're getting data that's unstoppable. Whether or not it's as reputable as of a piece of data that come from one of our miners versus Coinbase, is just not what we think of. You know, we don't think that that's the big consideration that you should have.

Bobby Ong (10:35):
And how often does this data get updated? Are blocks mined every 10 minutes, like the Bitcoin blockchain, for example? So if that's the case, then Bitcoin price only gets updated once every 10 minutes or so on the Tellor Oracle. And I guess my next question is how many submissions are there, like for the Bitcoin price for every block? Will there only be one submission because they will only be one block winner for each block or will there be a multiple submission? And how do we decide which one do we choose?

Nick Fett (11:03):
Sure. So I think it's like two questions in here. So yeah. Tellor block times are currently 10 minutes long. So we get one data point every 10 minutes. We are upgrading this for V2. So this is coming out next month. I'm sure you're, I know you're going to ask questions about that but we're going to be updating five data points every five minutes. So we're like 10X our speed. But even that, you know, it's not necessarily the fastest of Oracle's. Uh, so, you know, you can think you can get a Bitcoin update every five minutes as far as what the price of Bitcoin is, but we think that this is the sort of fits with where the Ethereum space is and what the DeFi apps actually need. So yeah, that's why we built it that way.

Michael Zemrose (11:47):
Yeah. We don't think you should build your [inaudible] to rely on really, really fast to Oracle. And so like that again, like we were saying with the having high quality data or whatever, is the same sort of thing as is trying to have super fast data. Like when you're talking about building on Ethereum and building on a blockchain, those should be lower down the totem pole in terms of what's important. And I think we've seen recently what happens when you try to rely on fast updates and the network of blocks and all of sudden you can't get those fasts updates and things. Bad things happen. So we designed our block times and just the sort of throughput of our Oracle. We were very conservative with it and sort of wanted to make sure that we could make, the assumptions we were making on a security level where we could meet. And we did that for the last year. And now with V2 coming up, we're pushing the envelope as far as we can with those things.

Bobby Ong (12:44):
Let's talk about V2 right? So what are the upgrades coming on in V2 and what are the main differences between V1 and V2?

Nick Fett (12:52):
We talked about it. So we just sent it off to our auditor today. So we're going to get a beautiful security audit done on it. And assuming we pass that with flying colors, it should be live, you know, a few weeks after they're finished and we can properly test it. So probably sometime in mid August, I'm guessing. But the bigger thing, so as I said, we're moving from 10 minute block times down to five minute block times. And instead of one data point, we're moving to five data points. So from a user's perspective, those will be the big things. Other big changes, so before what we had had was whenever, let's say a miner would submit a value and it was a bad value, you could get disputed. Now what would happen would be that would then take, it would go to a week long vote of Tellor token holders, whether or not that was a correct token value. We actually didn't like that.

Nick Fett (13:42):
We thought that it was, it was too final. And so we moved to, it was A, too final and B, too long. We move to the, first it's a two day long vote. So, and then, but the result of that vote can then be disputed and then it moves to a four day long vote and that can also be disputed. And then it moves to an eight day long vote and it can sort of be disputed and you can go to multiple and multiple vote rounds with increasing A, cost to dispute and B, time for voting. But we think that that model of, you know, the security of Tellor and a lot of these networks that comes from kind of that lack of finality in the systems and just the ability to, you know, make sure that you keep iterating to make sure that it corrects itself.

Nick Fett (14:25):
And so we're moving to that voting mechanism. As far as, I don't have like the document in front of me. So I apologize if I miss any of the small ones, but we're also moving to a, similar to kind of like at the Ethereum with the token fee burning. So whenever miners pay for tips, there's a big piece of our monetary policy. There'll be whenever you pay for a tip. So let's say you wanted the price of gold. You would tip the price of gold and the miners would then be more incentivized to go submit the price of gold. Just like, you know, they're more incentivized to include your transaction in block on Ethereum, but now we're going to be burning half of the tip. So it helped decrease the supply.

Michael Zemrose (15:06):
There's also we're changing the amount of rewards per miner address that each miner address can win in a day. So what is it? What are we moving to?

Nick Fett (15:16):
Yeah, so right now we were at, so every miner had to stake a 1000 TRB and they could mine as many values as they wanted. Now you have to stake a 1000 TRB and you can only mine one value per hour. And the reason that we're moving to that is, you know, any sort of minor centralization can sort of be, if you want to own 90% of our hash power, you're going to have to own, you're going to have to own a whole lot of our token as well. You know, like we want you to be really incentivized to keep this thing going in the right way. And it's sort of linking up the two as far as just incentive mechanisms.

Bobby Ong (15:54):
So you mentioned that in V1, you basically have one block mined every 10 minutes and in that one block they can only do one value provided in the Oracle. So this could be the Bitcoin USD price or the ETH USD price. But in V2, there'll be blocks mined every five minutes, in each block there'll be five data points in each block. Is that right? So I guess every five minutes we can have like Bitcoin, Ether, Litecoin, 0X and some other coins like price. Yeah, I guess my question is, if they are more demand for data points, say in that five minute period, someone requests for Bitcoin USD, Bitcoin GBP, Ether USD, Ether EUR and like gold price and the oil price. Which one is five out of the six queries to be included in the block?

Nick Fett (16:40):
Whichever one has the highest tip. So it's similar, it's similar like gas prices on Ethereum. You know, whichever one has the most, you would assume the miners are gonna do.

Bobby Ong (16:51):
And then if I pay less then I don't get the data in the chain, so I can't really use the Oracle that Tellor is giving, right?

Nick Fett (16:59):
It's just, it's a time thing. So you would just, just wait for the next block. You would just be on the next block.

Bobby Ong (17:04):
Okay. Okay. Unless someone is constantly paying the highest to get like BTC USD or ETH USD to get like included in every single block, then like it's constantly being filled out, right?

Nick Fett (17:18):
Yeah. I mean, well, this would be like in Etherum, you know, like if somebody is willing to pay like, you know, a hundred dollars for a, to fill up every block, every 15 seconds, you know, like theoretically you could, I think it's actually much lower of a cost than people think it is to do that. So if the gas prices, but yeah, no, it's, that's like sort of a briefing mechanism.

Bobby Ong (17:39):
But I guess, I guess if that's the case, then if there's a lot of demand for using Tellor, then it's going to be hard trying to provide all sorts of data points on chain, right? Because like there's, like 10 people are requiring data every five minutes or like this?

Nick Fett (17:52):
That's the thing. So like we, you know, kind of right now, so we can increase it. So you can, you know, right now we're including five values and you can move it up to some farther amount of values or reduce the block time. But it's always like a balance between kind of the block size. And you want to make sure that A, there's scarcity for the tokenomics to work and B, the other thing is like, miners have to pay gas fees on Ethereum in order to submit these things, these solutions. And at certain times the gas fees can actually be, you know, like black Thursday for instance, the gas fees is gonna actually be higher than the rewards and your TRB token and then your whole network freezes. And you, you need users to tip way more. So if there's a whole lot of demand and tips are really high, then you would assume that the rewards are way higher than gas fees. So then you can sort of expand the block size on Tellor. So like it's a flexible system, but you have to like, there's a lot of variables that almost are more dependent upon Ethereum scaling and upon like the demand base versus just purely "Can we do it?". You know what I mean?

Bobby Ong (18:55):
Yep, I guess my next question is who are the people demanding for Oracle data on Tellor right now? Like who's using the Tellor Oracles?

Nick Fett (19:03):
Yeah. So we were just working with our first partners on getting them on. So we should have some exciting news over the next few months. So we're not really in the habit of telling. We just want people to kind of start seeing data being requested on chain. And that's one of the big differentiators we want to make between, you know, we've always just riffed on the space a lot. As far as, we hate whenever people make announcements about future partnership and nonsense. You know, yeah, like an announcement of an announcement. We'll make an announcement whenever if someone's actually using us.

Bobby Ong (19:42):
Looking forward to hearing the announcement then, when they actually...

Nick Fett (19:43):
Yeah, exactly.

Bobby Ong (19:46):
Don't want to spoil the announcement later on.

Nick Fett (19:49):
Announcement of an announcement coming.

Bobby Ong (19:52):
Well, we sort of made these right. There's an announcement coming up in a few weeks time.

Nick Fett (19:56):
No, I know. That's how we tell people, like, just don't, just stay away. Like if you're looking for like constant reassurance and you need announcements, like you're too fragile to be in crypto. Take a step back!

Bobby Ong (20:09):
And all these mining right, it's sort of happening, like on the Tellor chain itself or is it a side chain or is it a SHA-256 and then you use a CPU, GPU, ASIC to mine these blocks?

Nick Fett (20:21):
So it's all off chain. Everything is sort of on Ethereum. So the, you know, the valid blocks are all stored on Ethereum. So it's like, they actually used a, kind of the initial design actually came from, if you remember 0X Bitcoin. Somebody built like the whole Bitcoin proof of work protocol on top of Ethereum. And I was like, that's so dumb and pointless, but then I was like, wait a minute, you can do something with this, but it's a combination of the different algorithms. So there's SHA-256, Sha-3 and then the RipeMD160. So it's kind of a Tellor customed algo. And right now it's kind of FPGA mixed with a whole lot of GPUs.

Bobby Ong (21:03):
Let's talk about the TRB, the Tellor Tributes, right? TRB. What's the main utility behind it? I know there's a constant growth amount for TRB to infinity, I believe. Every block, like X number of TRBs minted and then a percentage of them is given to miners. And in the new V2, a percentage of them will be burned, right?

Michael Zemrose (21:23):
Yes. Correct. But the basic utility behind TRB is they're used to pay for requests. They're used as the base rewards for the miners, as you mentioned. And they're also used for those tips that we mentioned. And so the way the requesting works is you pay some amount of TRB that you want, which is added on top of the base reward for the miner to further incentivize that they go and fetch your requests. So that's one of the main uses. The other, perhaps even just as important or more important, it's used to secure the network.

Michael Zemrose (21:56):
So the miners have to stake a bond of a thousand Tributes in order to participate. And so, and that can be slashed and in case of a miner submitting a bad value or a malicious value, and there's a dispute that a miner would have their stake slashed. So it's the disincentivization to sort of be a bad player. It's also used to initiate the dispute. So if you see a bad value, come through on a disputes.tellerscan.com, this is our dispute center. You see a bad value come through, you can pay the dispute fee and initiate a dispute. And then that goes to that vote that Nick was talking about, and then all TRB holders can participate and vote using their token weighted vote power. And so, you know, in a situation of a dispute being resolved, if the miner was deemed bad, his stake goes to the disputer, if it's the reverse and the disputer was wrong and the value was voted as good, that dispute fee goes to the miner. So, yeah, that's the TRB token in a nutshell.

Bobby Ong (23:00):
All right. So that's a good explanation. So I'm curious, right. Have you guys encountered like a miner submitting like fraudulent information onto Oracle, and then have you guys gone through this dispute resolution before? What was the experience like? Like to hear it from you?

Nick Fett (23:13):
So we've gone through it 38 times. Well, so the first, I think quite a few of the, probably the first half was like in the first month of Tellor being live. So this was like, you know, we had a CPU miner. It was like, you know, people running like their own custom Python miners. It was pretty early rough rowdy days. And, you know, everyone was, it was like free for all, because the token price back then it was basically worthless. So, you know, everyone was just trying to hack each other and come up with new gimmicks.

Nick Fett (23:50):
So those ones don't count. I like to think of them, it's just almost trials. But after that, you know, we've had several disputes. There's been, most of them, so I don't think we've actually had anybody being malicious per se, but we have had quite a few bad values. A lot of them arise from sometimes people build kind of custom miners and they build them like most of them it's like these PGA miners and they're not built properly in the sense that like the Binance API will go down and then they'll start submitting zero values. So like, there'll be submitting that the price of Bitcoin is zero. And it's pretty obvious that the price of Bitcoin is not zero. So they get disputed.

Bobby Ong (24:35):
So what happens then? Do they get their stake amount slashed?

Nick Fett (24:38):
They do. So I think in probably about half the cases, they've gotten their dispute amount slashed and the other half of the cases, we actually catch it. So like, if the team, if we catch it ourselves, like we'll dispute them and then we'll just kind of like, figure out what happened and then give them their tokens back. So we're nice. You know, cause we don't want people to think like we're making money off of these. So we try and like work with them. Like a slap on the wrist, like what were you doing? What were you thinking? Like you should check if it's a zero value before you submit it. That's been kind of the case so far and we're working on just to continuously making it better.

Michael Zemrose (25:20):
It's an open dispute. It's an open network, so anybody can initiate a dispute. And if somebody else catches that and disputes there's little, we can do. So you know, the important thing for us is that to maintain the integrity of getting those accurate prices on chain or those valid prices on chain.

Bobby Ong (25:40):
Do you think that the 1000 Tributes or requirement for staking it is a bit low? I mean, at this point in time, I mean, it's probably around $7,200, but I think you pointed out a few months ago when TRB was worth very little money. I mean, I think a few months ago it was even $1 or $2. So it could be like $2,000 worth. And some people, for example, may think that, well let's talk about someone who has malicious intent. If someone uses the Tellor data for some derivative contracts and that derivative contracts worth like, let's say half a million dollars, like they don't mind like losing this 1000TRB stake just to earn some of the contract on the derivative side that can earn them half a million dollars, for example. Do you think that will be a bit of a risk?

Nick Fett (26:23):
So the way the Tellor works is you can't necessarily just based it on that. So like the cost of to break Tellor, so like, let's assume so, you know, we take five submissions and you would need three of them to get the median. So you would need to lose 3000 tokens to submit a bad value. So you submit a bad value and then you get disputed. So that value actually comes off chain. So like, let's say you're the person in the derivative contract, you requested the price. Now that value got disputed. What you would do is you wouldn't wait for the week of the vote to end. You would just request it again. So, and then somebody else would bind it, say the next block and the next day. So if you wanted to continuously grief the other person, you would need to do it again.

Nick Fett (27:07):
You would need to lose 3000 more tokens. So similar to like, you know, in Bitcoin or Ethereum and proof of work networks, it's calculated like the cost to break Bitcoin is measured in like the cost of break Bitcoin for an hour or for a day, in terms of a 51% attack. Tellor is the same way. So like the cost of break Tellor for an hour is the cost of the 3000 tokens plus the cost of the proof of work plus time six. So 10 minutes blocks for an hour.

Michael Zemrose (27:40):
If you do the Math there, it's probably pretty close to what would it cost to attack Ethereum for an hour, which right now it's $152,000, which is probably way lower than you would expect.

Bobby Ong (27:56):
Okay. That's pretty low to attack at $150,000 per hour.

Nick Fett (27:59):
I think like Bitcoin is only like 500 grand. It's a constant 51% attack, like once you have the hash power, isn't necessarily astronomical. It's you know, the security of a lot of these networks is a lot more in the, it's in the lack of finality. So everybody knows it's eventually going to revert back to the truth. So, yeah.

Michael Zemrose (28:22):
So it's not like, I don't know how I, sometimes I think about it more in terms of not the cost to break, cause that definitely is there, but more like the futility of it. And that's what he was saying about the lack of finale in these things. Like it's expensive and ultimately it will be futile like, you would have just ended up griefing the network and they would have to just deal with it for a little bit. But eventually,

Nick Fett (28:46):
That's why like, so we want to like, that's why it's such a big education piece. You know, whenever you're talking to about like what people are building with these networks, like you want to teach them like, this is how you want to build a proper derivatives protocol to last an attack. So if somebody attacks the Ethereum network, how are you going to handle it? If someone's going to attack the Tellor network, how are you going to handle it? And we're working with the projects to let them build like these resilient networks. You know, same with like, you know, Bitcoin, like if you have 51% attack, Bitcoin, like you want to teach centralized exchanges. Like you wait an hour for confirmations to go through before you accepted. You wait, you know, and if it's an order of greater than 500 grand, you want to wait a day before you accept the blocks. You know, and like you just have to, it's like an education piece. Like there's nothing wrong with Bitcoin being that it's only 500 grand to break. You just have to, listen if you want to do a really high value transfer, you got to wait for a bit for confirmation.

Michael Zemrose (29:45):
We always say build things that makes sense. This is kind of what we mean.

Bobby Ong (29:52):
So I'm guessing if I run a derivative contract that expires on say the 31st of July at 4:00 PM EST. At that point in time, I have to query the Tellor blockchain, like maybe every hour for the next 24 hours. And if I get consistent numbers from the chain throughout the next 24 hours then I can sort of say that this number, the Bitcoin USD price that is given in at 4:00 PM on a 30th July is sort of correct and I can finalize my, I can set up my contract.

Nick Fett (30:21):
Well, you wouldn't have to keep quering it. You could just query it once. And then like, it's going to get taken off chain if it's, if it's bad. So you can just sort of wait and like, you can check it yourself, let other people check it, make sure it's good. If it's bad, people are going to dispute it like right away because you know, [inaudible] a thousand tokens for disputing it. So they're going to want to try and catch it. So if it's usually on Tellor, like if you wait an hour, like it's going to get disputed. So it's just one of those comfort levels, you know, do you want to, do you want to check it? Or how long do you want it to wait?

Bobby Ong (30:53):
It's interesting that you guys use proof of work in this way. Like, do you see any other projects doing something similar, like using proof of work system in this Eth stake tokens and on Ethereum for example?

Nick Fett (31:06):
I guess no. Yeah, no. Um, there's like cool benefits to using proof of work. You know, like we could easily do this with a stake method, kind of a purely stake method and you know, maybe we'll move there at some point in the future. But I think proof of work has some interesting concepts about having people actually do something in your protocol, like it's really nice having people, you know, active members of the community, especially when you're early on. And then also token distribution proof of work. You're doing something you have some costs to getting these. Whenever you have purely proof of stake networks, you know, we started with a zero token supply and minted our way forward. It's like, we didn't have some big ICO. Like usually what happens if you, if you start with a really low token supply and then just have proof of stake, like the rich, just get richer. It's like that kind of mechanism. And you know, that's not good for any network. So you really to have a good proof of stake, you sort of have to assume are really good distribution initially, which yeah. Which doesn't happen whenever you start out right away.

Bobby Ong (32:08):
Yeah. I completely agree with you. Uh, it's one of those things that a lot of people don't really appreciate about. Some of the OGs in crypto really appreciate that mining and proof of work is more the fairest way to distribute tokens to the community members. Everyone's got a chance of getting the tokens fairly and it's not disputed. but yeah, as you said, but proof of stake like, all these risk, especially also on the risk of an attack where there's zero cost, right. Because it doesn't cost you anything to get fixed, but then easy to attack based on the zero cost factor as well. Yeah. I've asked most of my questions here. Is there anything that I should've asked that I haven't really asked you on Tellor yet?

Nick Fett (32:45):
I don't know. Uhm...

Bobby Ong (32:46):
I guess, I guess there's one question here, right? So have you seen any interesting data that have been requested or made available on the Tellor Oracle? So I get most people would be requesting for crypto data Bitcoin, the price, the price, but are there any non-crypto related data that has been made available or caught by developers right now?

Nick Fett (33:07):
Yeah. So we're starting, um, the price of gold is being requested. So hopefully we'll be working with a project that's using that one here and then another one, somebody requesting the U S inflation rate.

Bobby Ong (33:21):
Oh, interesting.

Nick Fett (33:23):
So, you know, which is it's actually, so our CEO, Brenda of which you've met, so she, she used to work on the inflation rate survey in the U S government. So it's a, it's like a weird back around for her to where she's back providing that data to people. So that's cool.

Bobby Ong (33:42):
All right. Last question, before we end these rights. So if someone is interested in learning more about the best place to follow?

Nick Fett (33:47):
Well, our website, basically, so www.tellor.io. So there, you can find a white paper or documentation, um, our Medium blog, and we're really, really proud of the content that we've put out there. So definitely check out the blog and read all our stuff. Uh, if you really want to figure out how we tick and, and get some more details on sort of how we designed Tellor and why also our social links are going to be on the, on our website. And so we think we have a really fantastic and welcoming and helpful community on Telegram and Discord. And so we love it when people show up there and I think that's the best place to interact with us directly and, and get questions answered by our community.

Bobby Ong (34:32):
Alright, sounds good. Thank you very much for coming on the CoinGecko Podcast, Nick and Mike. Definitely learnt a lot about Tellor and I'm confident that the CoinGecko community will learn a lot as well.

Michael Zemrose (34:33):
Yeah, don't worry. We're happy to be a part of it. I go to your website everyday.

Bobby Ong (34:33):
Awesome stuff, guys. Thank you!

Nick Fett (34:33):
Bye!

Bobby Ong (34:34):
All right, that wraps up the show. Thank you for listening to the CoinGecko podcast with Bobby. If you like our show and want to know more, check out podcasts.coingecko.com or please leave us a review on iTunes. Do you have any feedback? Do drop us an email at hello@coingecko.com. Join us for more next week. See ya.

Disclaimer (35:20):
This podcast is provided as part of the overall information on cryptocurrency content on our website is for your general information only and does not howsoever constitute any endorsement, financial, or investment advice, nor any solicitation or offer of securities, other financial instruments. CoinGecko and the podcast presenter makes no warranties implied or express of any kind in relation to this podcast, including without limitation, the accuracy and updatedness of its content. All opinions and recommendations there in the podcast are based on the personal opinion of the presenter. Please conduct your own research and procure professional advice should you, at your own risk, decide to howsoever invest or trade in relation to the content contained in the podcast.