• Recordings
• Notes
• History and importance of price oracles
• How do people get whitelisted?
• What is Tellor’s validation process?
• What are the main uses for Tellor?
• Can Tellor help with security concerns with bridges?
• How does Tellor work?
• What would it take to break Tellor?
• Reverse price oracles
• What challenges does Tellor face?
• Transcript

Recordings

Notes

History and importance of price oracles

• The oracle problem was first referenced in a post on Reddit in 2014 in a post about prediction markets and the need for a source of truth. The question is how do we have decentralized truth?
• Early designs were based on Schelling coins, where people would use an auction based system to find truth. As long as the majority of the participants were honest, the correct responses would get all the money.
• Early systems were too slow. Other solutions were centralized. Even up to 2 or 3 years ago, not many people were thinking about the oracle problem.

How do people get whitelisted?

• Maker had a secret whitelist. More recently they use larger companies. Chainlink has a similar model where a lot of times they get big companies on board.
• Reputation based, from staking or having a big name.

What is Tellor’s validation process?

• People can stake Tellor tokens and then they are allowed to report prices. Users race to fulfill requests. Anybody can look at the result on-chain and dispute it by paying a small fee, sending it to a vote. If they fail the vote, somebody else can step up to report the value and get the fee.

What are the main uses for Tellor?

• Bridges and price feeds.

Can Tellor help with security concerns with bridges?

• One of the pain points of bridges is that they are slow, and in order to get fast bridges you need a centralized layer of trust.
• You could make things better by slowing things down, due to security issues around lagging finality and the ability to roll back the chain.

How does Tellor work?

• Reporters stake 100 tokens for the right to respond to requests.
• Disputers are usually other reporters trying to gain a competitive advantage. They pay 10 tokens to initiate a vote.

What would it take to break Tellor?

• Breaking the voting system, accepting wrong data even after a dispute.
• An endless chain of disputes followed by resubmissions.

Reverse price oracles

• Possible to assume the price stays the same until disputed.
• You could have Tellor act as a fallback and use Uniswap prices as the primary source.

What challenges does Tellor face?

• Getting users.
• People are used to really fast or on demand oracles.
• Gas fees can be expensive.

Transcript

[Music] welcome to the beanpod a podcast about decentralized finance and the beanstalk protocol i'm your host rex before we get started we always want to remind everyone that on this podcast we are very optimistic about decentralized finance in general and beanstalk in particular with that being said three things first always do your own research before you invest in anything especially what we talk about here on the show second while you're doing that research try to find as many well-developed opposing viewpoints as possible to get the best overall picture and third never ever invest money that you can't afford to lose or at least be without for a while and with that on with the show [Music] cryptocurrency and specifically decentralized finance are currently going through the trough side of a hype cycle as investors look for safe havens and liquidities moved out of the space so many of the pretty shiny things that were so alluring even a year ago are starting to look like empty shells plenty of style but not necessarily much substance and as the market's thinned it's become more and more apparent how important defy primitives really are these core functions projects and products are fundamental to how the ecosystem works and over time how to make it work better in our own project beanstalk we see a positive carry token as the starting point for an endless number of transaction systems our friends at teller a decentralized oracle protocol see their project as a way to provide data in a way that's transparent permissionless and censorship resistant their system of validators can be called on to obtain just about any type of information and their challenge system provides the incentives and penalties that ensure that only the right data reaches an end user for this episode mod and i were lucky enough to sit down with nick fett teller's cto we talk about what the protocol does why decentralized oracles are important and how tricky it can be at times to determine truth in a trustless environment nick great to have you hey great to be here and then we've got mod back in the studio as well mod always good to have you with us glad to be here wonderful nick just to kick us off um i'd like you to kind of set the stage a little bit if you would um could you kind of talk us through the the history of price oracles and and why they're important and then walk us right into how teller fits into that general idea of of decentralized price oracles sure yeah you know that's a huge question i think um but to kind of give you the background so the oracle problem so it actually the first reference we know this comes from there was like a reddit post on our bitcoin back in 2014 and it referenced the oracle problem so like how would you create like a decentralized prediction market um and somebody said well there's a problem of how do you actually find out the truth and and that sort of kick-started discussions about you know well how do we have sort of decentralized truth in a way um a lot of the earlier designs so there was like a shelling coins or what they would call and they use kind of what's called like the shelling point in these auction based systems to to find truth in prediction markets and that's what a lot of them were kind of focused on on solving like you know if you would ask it a question so for instance like um is joe biden the president of the united states then people would place bets on yes or no and then at the end of the period you would expect that as long as the majority of the participants are honest you know then the yeses would have all of the money and then if anybody did vote no they would get their tokens removed and given to the yeses and it was these kind of systems that like were the backbone or the start of the oracle problem but as you can guess those were sort of very slow systems and people have in a lot of ways been trying to find different solutions to the problem over the years and this was actually so how teller kind of got started in this um before i was i was doing teller i i was doing a decentralized derivative startup so we were trying to do long and short tokens so once you know you would have like a price fee so like eat us dollar and you could have one token that represented the the leveraged long and one that represented the short um this was this was actually a pretty novel idea back in 2017. um and then we were you know but i sort of ran headfirst into the oracle problem as far as well how do you get a price feed and how do you sort of get a price fee that's a little bit faster even than some of the other price fee oracles that were out there this was you know 2017 2018 so this was before chain link had even launched back then um the only price oracle it was really live on ethereum there there was auger which was a prediction market that was they used that shelling system and then there was uh what was called oracle eyes they've since rebranded and are doing something different now but um back then oracle eyes was you would pay them like a quarter and you would ask them to fetch an api value and then this this guy in london was running a server and he would go and send you back the result to your smart contract and that was like basically how smart contracts were grabbing oracle data it doesn't it doesn't sound doesn't sound very decentralized no but you know like we hadn't people didn't really think through the oracle problem yet you know now now most people know the oracle problem but even up to like two three years ago it was very um the oracle problem was pretty nuanced not many people were thinking about it especially in terms of price feeds you know you had like people were sort of okay with the reputational based systems you know that that's like one way that people are solving it currently and people have done in the past is you know i guess i can step a step back a bit like what is the oracle problem and that's like how do you find out the truth of something so you know in my derivatives contract example if if mod and i are say we wanted to do a smart contract that bets on the price of bitcoin so we each place some ether into a smart contract and if it goes up over this week he gets all the eth and if it goes down i get all these the question is well how do you resolve that who gets to say what the price of bitcoin is at the end of the week and you need a trusted oracle if you will to sort of provide that and we we could obviously choose you rex like you could be the trusted oracle um i often am yeah it would work but the problem is is that okay well now that's this you know centralized trusted party but you know that was that was the oracle eyes model if you will and then people went on well okay so what's like v2 of that and the answer is like well a multi-sig so we can just have some trusted group of people that we white list and and they all get to sort of agree on on what it is and that's like so like maker is a whitelisted group of people chain link is a whitelisted group of people that that v2 system is kind of like where a lot of the space is currently just these you know you have a multi-sig and controls who gets to say what the price is um and it can work but then the question is as well okay so what are the farther designs how can we actually make something that is actually decentralized so it has some liveness guarantees it has some some properties that we would sort of attribute to a traditional decentralized system and and that's where we were trying to create with teller so well it seems like something like a multi-sig has its limitations you were talking earlier about about speed and i would also say i'm gonna i'm gonna use the term bandwidth and i don't mean that in a technical sense i mean that more in the idea of you know if you've got if you've got individuals that are validating you know a value or some type of data you know there's there's only so much general bandwidth you know that you can you can put in terms of stress on a multi-sig before you you get past the point of something that's well it's not necessarily like a multi-sig in the sense of like all transactions are flowing through one multi-sig like usually like what it'll be is like one multi-sig or a group of people will like white list oracle reporters gotcha i see and then you know like we hit this problem actually all the time uh with users is you know i'm sure you guys see this in d5 a lot like everyone wants price feeds faster it's like how do you get super super up-to-date and super fast price feeds and the problem with oracle's is that usually once somebody puts something on chain you need a validation period so how do you check you know especially just in like a decentralized system how do you check once it's put on chain was it valid that's how teller works and we can say whether or not we need to push that off to a vote and determine whether or not we want to slash someone but if you have like this white listed model or just the trusted model there is no way to slash them or say that they were wrong so you can just use it right away so with it actually leads to a backwards sort of thinking about oracles and that a lot of times these these centralized systems can do things really really fast so they can just sign you know price feeds like you know a data feed every second and push it on chain and you can consume it every single second because there's no way for them to be wrong and even if there were there's nothing you can do about it so you just sort of consume it and it's a very very fast oracle but then you know you're not decentralized in any way and there's actually no way to decentralize that um so that's one of those things that you know centralized oracles are are better in some ways because i often like to compare it to like an l1 in that you know and now one database will almost always be more performant it basically will always be more performant than a blockchain um and there's just you know comp sci reasons for that and in this in the same way a centralized oracle provider will most likely always be faster and more efficient and in the whitelisted model how are typically you know people whitelisted sure well you know like maker maker originally had um they had what were like secret white lists because they that they had said like well you know they didn't want to give away who these individuals were because then you could go attack them in real life so they had like a secret white list of people and then recently they added i guess not recently but a few years ago now they added where it was you know some larger industry players so you know like other companies would have signers and that's what you see kind of with chain link too so chain link gets to choose who their white listeners are whether they usually make them stake a certain amount and then a lot of times they get big companies on board so larger people with reputation that they want to be whitelisted and that's sort of how you trust them so usually it's it's usually just reputation based whether that reputation comes from staking or from actually you know having a big name in the space that's kind of where it comes from and then uh yeah usually a lot of these reputational based companies too so chain link and probably most companies you know who else uses like a white-listed model so like api three does um where you know their data providers are known as well um it's they have like off-chain agreements so it's not like stuff that it doesn't have to be sort of handled on chain you can have like a you know go sign something that says you won't lie and if you do we're going to take you to court um that's like the traditional model for an oracle and yeah i mean there's something to say about that and so to to kind of junk suppose that to tell her as i as i read through your information what came to my mind was the almost i hate to use the word traditional because i feel like in cryptocurrency and d5 traditional seems like a little bit of an odd term but i think of the let's call it traditional um hash validation process where there's an opportunity to dispute want to walk us through teller's process and and is i mean is that kind of an accurate way of looking at at teller's validation process sure um yeah in a way i mean in a way we're kind of optimistic in that sense um but the the earlier version of teller so originally when we had launched tyler we had a proof of work component to it as well to where in order to submit you had to solve a proof of work challenge as well so then there was there were disputes but then there was even a proof of work component on top of it which we were trying to make more secure but we we've since moved past that um we had some issues with proof of work like a lot of people but the way that teller works now so anybody can come so rex mod you guys can stake some teller tokens and then you're allowed to report prices so somebody would come on and say like in our example what's the price of bitcoin and you guys would then basically race so you would see hey the person that requested that is going to pay five dollars to whoever puts the price of bitcoin on so you guys race to grab the price of bitcoin from some sources and submit it on chain everybody can look at that value that was put on chain and dispute it so you can say hey he lied pay a small fee and it goes to a vote gets taken off chain and then the vote determines whether or not they lied most of the time 99.9 of the time nobody lies so it's just the people can use that value whenever they deem they want to use it um obviously the caveat being they want to wait for people to check it and whether or not it's going to go to a dispute but yeah that's kind of the simple way that teller works if anyone gets disputed somebody else can step up and report the value and get the fee and then it's there's no white lists involved there's no way to sort of shut it down um it just runs and it sounds like this process could really be used for really just about any type of off-chain data right i mean it's it seems like the bounds are really only the needs of a potential user and the you know computing ability of of the different validators yeah and that's kind of what we're seeing you know originally kind of most defibril most of most of the crypto space right now it's basically just d5 and people want price feeds but you're starting to see different kinds of requests so whether it's you know people want information about other chains so you can think of bridges or basically oracles they are oracle so if you're a bridge on one chain you want to read information about some external chain you need somebody to pass the information on and that's why you know most bridges still are that v2 it's just like white lists and white lists for who gets to push the data over but you know you if you want like a secure decentralized bridge you would use a model similar to teller to where you're grabbing information from other chains but yeah even even things such as you you can do things like off-chain computation so if you had a problem you could you could ask it to an oracle and then the oracle providers would go do some off-chain computation and return it on chain that's something else that you know teller could do but you really don't see that kind of demand yet in most of the crypto space sure and that actually actually leads me to what i was going to ask you next and you know so teller's been around gone through a couple iterations or let's say improvements what do you see as the main use cases now in in your current situation i would say bridges and obviously price feeds still you know people are people still just want a lot of price feeds i think something that we're seeing a lot of people coming to us it's just that they want to spend price feeds up quickly or they want to spin information up quickly and um that that's sort of one of the problems with the white listed providers is they're relatively slow to sort of add a lot of new data feeds so so teller has a place kind of there for for stepping in and adding them really quickly in a permissionless fashion um similar to how you saw you know i think you saw like decentralized exchanges really take off whenever people knew okay you can list any token here the centralized exchanges had kind of shut shut down the listing process as far as it was very very expensive and it was hard to do so then decentralized exchanges started to blossom because there were no permissions there in that way i think as more demands for oracle's kind of pop up the centralized oracle providers kind of won't be able to keep up and that's whenever some of the decentralized ones will really start to take off now obviously bridges have been in the news quite a bit specifically because of security concerns i mean do you see do you see tools like teller being able to help help manage some of those security concerns or or is it you know or is his teller and decentralized artillery and decentralized oracles away from that problem yeah no i i i think it definitely solves a lot of the problems and you'll see it probably starts to pop up here over the next year um one of the things that you have to get people to start accepting when it comes to bridges is just that they're slower you know right now people want really really fast bridges and you know the only way to really do really really fast bridges is to have kind of that centralized layer of trust um if you would just like slow it down you could sort of make a lot of a lot of things a little bit better you know same with like if you're going to use teller it makes sense you want to make sure that you slow the chain down just because you know even vitalik had posted like there were some issues with regarding uh wormhole which was solana's you have a lot of issues with really fast bridges because ethereum for instance isn't final right away like you you can roll the chain back and it and it's not it's expensive to do that but it's not like unheard of expensive so like if you deposited 100 million dollars on solana traded it on a dex for some other coin and then it it is significantly less than 100 million dollars to roll back the ethereum chain for an hour and then you could simply do it say you never sent that money over to this whole salon bridge but you had already traded it away on the solana side so so what happens um that was like an unknown attack vector vitalik had kind of posted about that obviously that the wormhole bridge got hacked for another reason just a smart contract bug but yeah um you know that was something that people were saying like don't use these really really fast bridges because it it even makes the main chain of ethereum less safe whenever people are building these really fast bridges it seems like when it comes to bridges you've got and i'm i didn't come up with this but it came to mind as you as you were talking you've got that there's like three variables of speed security and centralization and you can like pick two yeah well i think oracles are the exact same way you know it's like the the blockchain challenge holds for like basically every app along the stack um and and yeah people are just starting to realize that so but you know you you want your bridges to be nice and fast so yeah well i mean that seems to be the the the way that we move on to more efficient systems than l2 is to figure out you know however we can manage to get l1 and l2 working together well in terms of bridging and you know that that oracle component it seems like it's it's a really unique opportunity for both that and a lot of potential you know legacy l1 solutions i feel like i've read some of your information you know use examples like insurance providers looking for weather feeds or you know i i think about beanstalk and the you know the future um operations and smart contracts we're looking to to have built on top of on top of beanstalk in the amount of outside information they'll use that that oracle component that is well trusted and has it and is is relatively fast is going to be extremely important for some of those you know more unusual use cases for sure yeah i mean as you start wanting to actually make smart contracts useful and touch outside world information you know like real world information and make it relevant to normal people like you need oracles and that's kind of why we see them as a big thing going forward and i know you guys do too you know what the cheesy weather example that everybody gives a basic smart contract class aside but yeah no it's definitely something that we're working on and i i think one of the bigger challenges for oracle's just going forward in this decentralized way is just how do you sort of source all that truth how do you actually know what what the weather feed is in certain areas of the world and i think that's something that we're so excited to kind of explore nick i wanted to go back and ask a little bit on how teller works so you have you have a group of validators and then um those who can also like check um or dispute uh if the price you know is different can you summarize that back again to us sure well anybody can be a validator so you just take right now it's 100 tokens so it's like 1500 on mainnet and then you're allowed to be a validator so you stake those tokens and anybody can do it you can you know there's no limit to the number of people and then anybody's allowed to be a disputer so you know if you're a user if you're just a casual token holder usually the usually the disputers are the other reporters themselves trying to gain a competitive advantage because they're the ones that sort of monitor the system the most um but yeah one anybody can submit a dispute so right now that the dispute fee is 10 trb so it's like one tenth stake amount and then they dispute it and it would go to a voting process so it gets kicked over to a voting process uh the vote uh is split so we have this is kind of how the dispute resolution piece works but it's split uh into four pieces so 25 of the vote is the token holders 25 percent is the reporters uh 25 is the users the people who are paying for requests and then uh the last 25 percent is the team so we give ourselves some and we can throw it away at some point okay and correct me if i'm wrong the dispute is if if i as a validator had reported a price and then you know for some reason my price wasn't the one picked i can dispute it to you know get back my state tokens is that correct wait what do you mean so who who disputes and what what are you disputing so you're just disputing the value so like if you push the value of btc usd and you know you said it was one million dollars i would come and dispute that and say it's not one million dollars right and and then you take the the staked uh tokens of whoever you know sent the one million dollars okay what what would it take to compromise such a system so if we have 100 validators what what would it take you know for eventually terror to give out the wrong the wrong price sure so there's a few ways to break dollars so the first way the way to really break teller so like if you actually want to break for good what you would need to do is you would need to break that voting system that we had so like i you submitted a price of a million dollars i dispute it and then you break the vote and say like no that's actually a good price and you get your stake back you know because the problem is is then it would sort of reduce trust in the voting system in general um the other way to break teller is to sort of just censor it so if you would basically i would i would dispute you that it was a million dollars then i would post the correct price so i don't know what's what's bitcoin at today uh whatever the whatever the correct price that bitcoin is yeah 23-ish 23-ish yeah and and so i would post that on chain and then the way that you could censor the teller system is you would just dispute that so you would say hey that's wrong even even though it's actually right you could pay a fee and take it off chain and then i would probably just put it on chain again because i know like i'm probably gonna i'm gonna get my stake passed back to me plus your dispute fee um it's just it's a two-day long vote for that to happen so i'll just submit it again probably restake and submit it again and then you would just dispute that too and then it would just turn into this system where hopefully like the game theory would say that more people would want to come in stake teller and start reporting a correct bitcoin price because in the event that you dispute them they're going to get their money back in two days plus a nice hefty return so that would be um how it could sort of censor it if you will but those are kind of the the only two ways that you can you can really break teller so when it comes to those those voters you said token holders validators users and the team so how i mean is that for for checking a a particular piece of data is that like a unanimous process so if you know if i'm a validator and i give an incorrect piece of data and it goes back for dispute and it's voted on by those four groups do they have to completely agree or is it some type of basic majority or it's a basic majority and there's multiple rounds too so like you can if you don't agree with the first round you can kick it off to another round um and and this was something we had some disagreements on or not disagreements but like discussion on early on because like what you can do is you can do like the shelling systems where like uh you slash people who vote not with the majority so like i'm gonna say that it was true you're gonna say it was false and if more people say it was true the other false people lose their tokens but what we found with especially price feeds and a lot of these things is that it's it's really nuanced um sometimes even saying like what a valid price is isn't necessarily clear um you know like in the price of the bitcoin you know if it's 20 let's say it's 23 000 um if i submit it twenty four thousand should i get slashed or disputed yeah um it's really really hard um sometimes yes but if the price actually just jumped on a few exchanges up to 24 000 is that good um so we're allowed to vote you can either vote um to where you slash the person to where you don't slash the person and give them the dispute fee or to where everybody just kind of gets their money back and and in the past we've we've had examples of just where like listen we're just going to give everybody their money back like let's just check our api feeds from now on like it wasn't an attack it was just like there there was some somebody was using certain apis and another person was using another api and they just disagreed sure and that's i mean that's a thought that passed through my mind as you were talking earlier is that there are times when especially coin prices yeah such a good example if you if you are using different systems and we've had this discussion inside of beanstalk a number of times you know what's the best what's the what's the best price feed for eth and you know you ask 10 people and you'll get 12 different answers and i'm sure that that's a that's a common issue especially when there's a lot of let's say like a sensitivity in a given piece of data and maybe be like high rate of change yeah yeah well it's so like with teller we actually let you specify the granularity level of the price feed so sometimes sometimes users will be okay with you know like you can say like i want the eth price that's actually different than i want the eth price according to the coinbase api um sure those are two completely different things because you know some people might want according to these two apis and then it's it's actually a whole lot easier to say what a valid dispute is you know were these valid api queries and and you can you can sort of prove that the problem comes in however that if you're only using a handful of apis now you're actually really easily censored because those api providers could easily shut you down um if they knew what was going on so if you wanted to say a more robust price feed you could say something like i just want a valid etus dollar price doesn't matter where you get it from and now obviously that that means the the reporters can get it from any api they want or they can even manually go look at a website but you're not going to get sort of as tight of a range so you know there probably has to be a little bit more you know variability in the accepted answer and you kind of leave it up to the dispute process to to handle what that valid range is so there's pros and cons to both um it's just kind of up to the user to determine what they can handle yeah it's really interesting um you know so we were talking about the tension between speed security and centralization before it almost seems like in this case there's a tension between like uh specificity yeah and sensor ability yeah no we we have it like so like we have a whole like series on subjectivity and price oracles sure yeah and how yeah how how ambiguous you allow your data to be is actually more censorship resistant yeah that's i never never even thought about that but that's it's a great point and yeah maybe it may make your data feed more vulnerable to dispute between validators but just like you said it makes a less censorship resist or it makes it more censorship resistant externally yep yeah i mean this is like some we always push back on chain link because they always like to brag about uh they allow trusted data feeds so you know like paid apis and things that they can use and we're like well that's actually a really negative thing you guys don't want to do that like and publish which trusted data feeds you're using so yeah incentives can be very much very much a two-way street or double-edged sword sure yeah it's um but anyway that that's uh definitely watch it whenever you're picking your price feeds and particular ones i think some of the other pieces come in you know even when you're knocking on not price feeds like when you're talking bridges there's the whole finality issue there's the whole you know are you actually running are your reporters running nodes over there um how do you know if that's not being attacked there's a whole bunch of issues too nick i'm thinking out loud here um i'm wondering if this is already you know uh done in in some way um and and we reverse the way you know that the price oracle is working so let's say that we're interested in the price of three curve um and what we would do is that we would start with the assumption that you know the price of three curves was always equal one dollar and and we would be a validated and always push that the price of free curve is one dollar and once it gets disputed this is when we know that the price has changed um and you know it's not it's not equal one anymore does that make sense because something like that be used in this way yeah for sure um you you would definitely just want like a lot of times people will use teller as like the fallback in that sense um so you could just have like a lot of times when people come to us and they're like hey we want really fast price updates you know basically with every block and we're like hey you can't afford that neither can we um to push that many times but what we tell them to do is like you can actually use like the uniswap price for a given pool and then just use the teller as a fallback to where the unispot price has to be within two percent of the teller value so then you you can use it uh but we update the teller value say we'll update it if there's a two if there's a one percent change in the price and then as long as the unisol price is relatively within there we know that it's not being attacked and you can use it all on chain um but yeah you can definitely do things like that where you would just assume that it's a dollar and then have teller push it on and able to like override that in some way that almost seems like it's a security feature could be used as a security feature for some type of unplanned change to any type of any type of data feed that may have you know your point may start as you know internal to a particular protocol or project if you use teller as your as your backup you might be able to to detect unwanted or unexpected changes right for sure yeah i mean that's like i tell people like the uniswap price feeds for instance are normally for most things they're really great um you know as long as you you don't have a whole lot of value in your system um and the unison pool has some liquidity it's it works just fine you know nobody's going to throw the unisop pool to to mess up your system but if the liquidity on the unit swap pool goes down or you know if you swap upgrades to a new contract and all the liquidity leaves then you're stuck without an oracle so then you might want to listen to to the teller price which isn't going to change sure so another thing that was spinning around in my mind getting ready to talk with you is zero knowledge proofs and you know when i think about teller's ability to tell me a thing using its own internal set of validators i mean i feel like it's a relatively quick jump to zero knowledge proofs i mean is that something that is that something that teller is considering has considered you know is is on your radar or what what do you think about zero knowledge proofs and teller yeah i mean we're still trying to find a way to use them for like every data push you know or if there is a way to use it for that um the question is always like why are you keeping it a secret do you need to keep it a secret um sure so you know i don't i don't know there's actually we're working with a cool system to do where you would do like zero knowledge commitments on on one chain and then you can pass into another chain um which isn't necessarily teller using zero knowledge proofs but um sort of teller involved in the process if you will yeah no i mean we haven't really found some people are doing things with zero knowledge proofs as far as like random numbers so you can sign it and then reveal it later um but it's definitely something that we want to explore more i i definitely i personally have just gotten into it probably over the last six months i've been building with a lot of certain knowledge libraries just for fun now i was going to ask nick like what challenges are stellar you know facing today and where do you see you know price articles in general going in the future or you know which direction do you think it is sure um well i think some of the challenges that we faced today it's you know kind of like i was talking a lot about trying to find the right users um you know getting actual users in the space is really tough um obviously just because a i think a lot of times people overestimate the total number of users in the space um there's like not thousands of them um and then the other thing too it's you know we we have people have gotten used to whether it's like chain link's fault or or just the faults of using centralized oracles it's like people are used to kind of really fast oracles or on demand oracles to where you just throw a price on chain and then you can can use it instantly well well that's you know kind of like we were talking about we want people to sort of slow down some of these applications um and that'll make it good and then also just having to pay for oracles so you know every time we're relatively cheap to push it's like three hundred thousand gas to push a teller value one train um but on main net aetherium especially during like high times like you know that can be ten twenty dollars a push so if you want price updates every you know five minutes and it's going to cost you 20 it can get relatively expensive pretty quick for your protocol and a lot of times people are just not used to paying at all for oracle data they're trying to teach people that you know oracle's cost money and then there's some of these things about just just being slow and using it whenever you in a smart fashion as far as where i think price oracles and and oracles in general are going to in the future i think we're just at this stage where we're starting to see real decentralized oracles come about you know people are starting to realize that okay we can actually have decentralized price feeds and even just people worrying about oracle's is huge compared to where we were three years ago you know the fact that i'm talking about price oracles on a podcast is awesome um but then i think you know getting rid of a lot of these these multi-sig pieces you know as people start realizing that those are bad ideas you know there will be things what i'm hoping to kind of make teller into is is that kind of like the protocol if you will for oracles um in the same way like uni-swap in a lot of ways it's it's on chain it's unchangeable and people can just use it that's kind of my vision and how i hope teller will work um there's not really a whole lot of kind of fluff involved and a lot of these other pieces you know it still seems like really sort of early on and there's like a lot of governance and a lot of changes and a lot of you know just wool over the eyes about how the system actually works but ideally we just like simplify it it works you know the limitations you know how to break it but this is it that's how you have to use it go deal with it i feel like that's um when we look beyond you know folks in the d5 space just trying to to look to make money or or whatever when we actually look at different groups and projects that are trying to build let's call them like d5 primitives you know i feel like that is a place where where teller and beanstalk are very similar and that like the goal is to come up with a system that is really a springboard for other systems you know for for beanstalk we want to build we want to build a financial system that's a springboard for all kinds of other systems that can leverage positive carry seems like for teller like you want to build a protocol that is a springboard for all kinds of other applications that need reliable decentralized information that they can trust yeah no i mean it's kind of like why we all got into it you know it's you're building these things to you know smart contracts and all these primitives and protocols are supposed to sort of remove the middlemen in the system so you know obviously like an oracle is like the definition of like a middleman in like the you know the original bitcoin contact we were talking about like that is the middleman but how do you remove the middleman and make it sort of non-extractive in as much way as possible you know how do you make financial contracts that don't have somebody actually making money off of it um yeah i mean that's it and there's almost a difference between saying you know when you even the term middleman you know if you can create a middle it's it's in the middle but it's not a man you know it's it's it may be an intermediary but it's decentralized and trustless and censorship resistant you know that's that's the goal yeah you know and that's like we've we've changed even a lot of our mindset over teller over the past few years you know originally like we've stripped away a lot of originally we were thinking dao everything everything has a dao and and we've stripped almost every voting piece out of the system because we saw it as a risk for other protocols you know the more the more pieces that we're allowed to vote on you know the more we could potentially screw up your price feed so like how can we strip that out and just make it as much of like an immutable smart contract obviously we have to vote on the disputes um but other than that like it's all gone and and that and that's something that we're proud of like you know we want it to to be there and it just sort of works there's not even a dow that's extracting rent and making money off of this it's just a smart contract beauty and the simplicity right well my you have anything else you you took my you took my good ending question about what do you see is the future what else he got mad that says from my side this has been very informative nick oh so so nick what i'll do is i'll i mean you know i'll just hand over to you do you have any any final thoughts anything you want to leave us with no you know i think if you guys are interested in learning more about teller there's teller.i o is our website happy to just kind of anytime talk over oracle designs you know just definitely it's great that you guys are thinking about it um just the more that we can educate people to actually think about the oracle problem and you know whether it's how do you you know wait for data in in a safe manner versus what should the data be you know that whole subjectivity discussion um there are a lot of hard questions without really right answers so you know happy to talk it over with any users in the oracle space or defy space perfect yeah this is great nick thank you so much no thanks for having me um yeah excited to hear it live so oh man thank you again close all right as nick said you can find out more about teller by heading to their website at teller.io that's t-e-l-l-o-r dot io you can also find them on twitter at teller and on youtube [Music] the beam pod is a production of beanstalk farms a decentralized autonomous organization you can find us on twitter instagram medium discord and our home on the web at bean.money you can also find me on twitter at rexthebeam and as a final reminder this podcast is not financial advice thanks again for listening [Music] you