In this episode, Mike Townsend chats with Hugo Philion, Co-founder & CEO of Flare Network. They aim to be the interoperability chain connecting all smart contract chains, non-smart contract chains, & web2 APIs together.
As a former commodity derivatives portfolio manager & a Master's in Machine Learning, Hugo is uniquely qualified to speak about the past, present, & future of DeFi - both on a financial fundamentals basis, as well as the tech driving it.
Megaphone • YouTube • Quora • Medium • Twitter • Facebook • LinkedIn • Soundcloud • Apple Podcast • Stitcher Spotify • Google Podcast • Player FM
Mike: In this conversation Iinterview Hugo Philion, the co-founder and CEO of Flare Network Flare hasraised over 11 million with over 50 people, and they are aiming to be theinteroperability chain connecting all smart, smart contract chains, non-smartcontract chains and web two APIs. Together, we talked about bridging.
What bridging is, where the faults andhacks and the $3 billion that was lost this year due to bridging failureshappened, what the future may look like, what Hugo and his team are building.Then we kind of extrapolated on some other ideas, , the impacts of the FTXcollapse and other existential risks and opportunities in Web3.
So I hope you enjoy this conversation.Here is Hugo.
Mike: All right, Hugo,welcome to Around The Coin podcast. Thank you for hopping on today.
Hugo: Thank you for havingme.
Mike: Yeah. , let's startwith Flare and what Flare is and, and how it's evolved. I'd love to just hearthe overarching story of maybe where you guys started and what you saw as theinitial opportunity and where things are today.
Hugo: Absolutely. , so I'mco-founder and CEO of Flare and I started it with, our CTO Sean Row and ourChief Scientific Officer, Dr. Ma Usha. And really we started out in 2017 justlooking at what, what, what could be done in crypto, what's interesting, what'sinteresting in blockchain. , we really started looking at consensus algorithms,and then we started looking at, okay, there's a lot of, you know, interestingnetworks like Bitcoin, stellar .
also these non-smart contract tokens.How could we build a, a layer that essentially allowed people that are holdingthese kind of new forms of value, to, to do more with them? So, you know,Ethere came along and kind of invented the idea of a, a global computer. butthat's at the time it was, and still, you know, to a very large extent, it'spretty much restricted to people that hold e and tokens issued on E Sure, thereare some bridges, but they're not, not, you know, they haven't been fantastic,and performed well.
And people have lost three, close to 3billion this year. Wow. In bridges, which is a lot of money to lose. so we werelooking at how, how can we, how can we enable those networks that don't havesmart contract capabilities to To have smart contract capabilities, causeyou've got a large amount of people and a large amount of value just sittingthere.
and really we've, we've gone a, that,that's kind of where Flare came from, but where it's kind of going in the lasttwo years, we've been looking at, really, if you wanna sort of s up Flare, it'sa blockchain smart contract platform. , uses the evm uses EVM principallybecause that's the easiest thing for developers to use right now.
there aren't any competing virtualmachines, or let's call them, you know, the abilities to write smart contractsthat have been, you know, broadly popular. UVM is still, let's say probably 90%of the market if that changes we'll change with, so if there's m layer one, butit's really structured around two things.
Well, one overarching objective, gettingdata onto the chain from external sources. So that you can build richer smartcontracts. this is really important because ultimately I think a lot of whatwe've seen with the speculation over the last couple of years has been becauseat the moment block, the whole blockchain financial ecosystem is incrediblyself-referential.
I think if you look at blockchain,there's essentially, if you look at DeFi, there's the, the principle model ofDeFi is ultimately people borrowing US dollars in, in some form. Mm-hmm. . Andthat's, that's pretty great use case, but we can do a lot more. but we needdata in order for there to be actions. So Flare has two core protocols.
the first is essentially a native oraclefor probabilistic data. that sounds complex, but think of it as prices, stuffthat moves over time. And then it has a native protocol. And we really thinkit's the. Most robust possible way to do it currently for deterministic data.So what has happened on another system?
So for instance, a deterministic pieceof data might be that you have sent some money on Bitcoin, some, some Bitcoinor on dose Coin to another, another person or another address. from that datayou can do interesting things. So one tiny, small, little interesting thing youcan do that certainly not taking the future of Flare on, is you might be ableto buy an NFT on Flare with a payment directly on Bitcoin or a payment directlyon do because, so I can read that and then release the nft so I can essentiallycome to consensus over that and then read the MFT and then release the NFT atthe smart contract level.
So it's kind of the overarching thesisof what we're, we're building, building smart contracts with. Much richer data.we have some really interesting areas that we are looking into how that can beapplied.
Mike: And, and the reasonit's it's rich is because the reason you use that word is because you're ableto scrape or read other blockchains and then you can create a smart contractusing the data feed from other blockchains.
Is that the correct way to think about it?
Hugo: So, scraping andreading are kind of soft. can't be very certain of the data. You know, thereare lots of different ways to do this. We are, I suppose you might call itproving the data on flat. and we have, this is the riskiest part or one of theriskiest parts of blockchain, right?
Which is how do you understand anexternal system without being centralized? So, you know, you've had a lot oforacles that come along and ultimately they're centralized. If you really wanta decentralized kind of future, then the protocol you are using to understandwhat is happened on an external system has to be meaningfully decentralized.
and we have a lot of thoughts about, youknow, applying things like proof of stake to such a protocol. Like then you canonly ever be safe up to a certain amount of value. Mm-hmm. , so we, we've donea lot of research in this area, and that's kind of where we're, we'reinterested in sort of going forward.
Mike: And what, what is, howdo you define an Oracle? What is an Oracle?
Hugo: So an Oracle is asystem that allows you to get data from an external source, right? It'sbasically like we call Flare State Connector, like the i in the sky, right?there's lots of different ways you can construct them, with various differenttrade off.
So for a price oracle, that's kind oflike a, a soft proof. You can't prove something that isn't deterministic.Ultimately, like if you say, what is the market price of Bitcoin? Well, it'sdifferent on different exchanges. it's different by fractions and it'sdifferent as to when you look at it. So you can't really prove the priceanywhere.
But what you can do is you can createshelling point within a properly decentralized oracle, which essentially allowsa true decentralized system to come to essentially an agreement that this iskind of roughly where the price is in a shelling system.
Is that, is that referring conceptually?Is this reading. I would imagine you are looking at the different exchanges oryou're, what would be the access point to collect the data?
Hugo: So yeah, at the momenton our, Canary Network, which is called Songbird, it's been live since 2021,September, 2021. Great name. , we thank you. we have a, what we call the FlareTime Series article, and that takes the native token, so the native solo tokenand people delegate to what we call price providers.
And currently we have about, I thinkit's about 86 price providers that are live and we have about 120 that areregistered. So quite a sizeable nber of price providers relative to basicallyany other price Oracle in the space.
It updates every three minutes. And eachone of those price providers is essentially calling, the APIs of all thevarious exchanges and putting their best guess of the price into the network.And essentially, each price provider gets, gets pricing power from, from thepeople that delegate their, the native tokens, songbird to that oracle, to thatprice provider.
and then songbird aggregates thoseprices, takes the center of the dis cuts off the tails so you don't getoutliers and takes the, the weighted medi of that distribution. And that's,it's been, you know, it's been live for over a year. It's been really robust.Interesting. Been, frankly, we've done quite a lot of analytics on, on how it'sperformed. and it's been a really stable, really valuable, price sourcefor
Mike: Yeah. I'm curious, socurious about this. Okay, so you have 80 different, are these 80 differentpeople or organizations managing each and API to exchanges? And why not justhave the direct API connection yourself? Why go to all the work of building outthis songbird ecosystem and bringing on all these different, pricing Yeah.Scraper API process?
Hugo: So the problem withjust having that API on a blockchain is ultimately someone has to, from thatAPI infrastructure, and that person or entity becomes a, essentially acentralizing point in, in the system. Mm-hmm. , , which means that, you know,if they decide to change the price because you've got a DeFi protocol, whichthey can.
Liquidate a large participant, theymight change the price in order to, oh, essentially, you know, benefitthemselves. And so you, you really want to have, as decentralized system aspossible. So for instance, in our, Oracle system, if someone did try to put ina bad price, but the others are honest, they would be, their price would beoutside of the distribution that is captured.
Essentially, we capture the center ofthe distribution and their price would be in the tails of the distribution ifthey're putting in a false price to try and liquidate someone. so this, theOracle problem's a really hard problem.
Mike: When you say theOracle problem, are you referring specifically to the, the problem ofcollecting the price for different, cryptocurrencies is the hard, is the hardpart that makes easy.
Hugo: the hard part is howdo you then get that onto a blockchain without creating, you know, a series ofattack vectors for the, essentially the parties that are operating that Oracle,to, to, to be malicious and, and to essentially erroneously take money fromother people.
Mike:Got it. Okay. And whenyou first thought about this problem, did you look at it and think that thehard part is, and where the value creation is, is your ability to safely in adecentralized way, collect the pricing, accurate, up to date pricing ofdifferent, assets that are out there, currencies that are out there.
And if you could do that, then you couldfacilitate smart contracts and all sorts of other things. Was that kind of theway you thought about the challenge?
Hugo: Yeah, we actuallylooked at the challenges. How do we get Bitcoin, , do Coin, all those. Largesources of non-smart contract value onto the blockchain, that has smartcontracts so it can be used.
and really the first part of that forus, in order to do it in a decentralized way, , was to have their prices
Mike: and Right. And, andthat's, and that's where you came up with Songbird?
Hugo:Well, that's where wecame up with the fts O which is one half of what we're doing. Mm-hmm. , , or,you know, one, one component of what we're doing, the other protocol in manyways, but there are lots of price Oracles out there and we're not actuallytrying to compete with them.
Mm-hmm. , , we just wanted to buildsomething for our blockchain that we felt was truly decentralized and best inclass. and it's really nice to do this at the network level where you'reactually using the network token to secure a price Oracle. The price Oracle isonly a little bit of what we're doing the.
Almost, I would say the much biggeraspect of what we're doing is the state connector. Cause the state connector isa, we describe it internally and it may, may not be strictly accurate, but wedescribe it as a true consensus protocol for proving events. So non-price, sonon probabilistic events, onto a blockchain.
So for instance, that, that example Isaid earlier, which is, you know, Mike has sent some Bitcoin to Hugo on theBitcoin chain, prove that onto Flare so that that can trigger something in asmart contract. And from those two protocols, a huge amount of things can bebuilt. All types of different bridges, all types of different messagingservices across chain because you can then set up.
You know, an example of a protocol thatcan be built using Flare is a simple relay protocol whereby relay nodes areexternal to a blockchain and they sit between blockchain to relay data todifferent chains. But how do you secure them where you can secure them bystaking them on FLA and then using Flare to prove what happened on chain A,what message was delivered to chain B?
And then if the relay node has liedabout that message, you can slash that node because they're staked on Flare andyou can slash that node because essentially Flare can prove what has happenedon BE chains. And to us, you know, these two protocols, price and state. Soproving, you know, coming to an agreement over what the price of something isand coming to consensus.
Over what the state of anotherblockchain is. This, these are kind of like the primitives mm-hmm. to build areally wide array of very robust interoperability protocols, starting withbridges and relay, all types of different bridge, however you wanna structurethem. We have our own thoughts about how to do that.
but anyone can come and build a bridgeusing Flare. Anyone can come and build a relay building Flare can structure ithowever they like. and that's sort of our, our thesis Flare is a layer forenabling interoperability between lots of different chains. .
Mike: Gotcha. Okay. That's agood way to explain it.
So pricing and state, the ability to seewhat the price is and the ability to see what people are effectively doing onthose other chains. When you, when you use the words relay and bridge, what arethe, I, I approximately know, but I wanna make sure I'm I exactly understandingyou. What is the difference?
How do you define a relay and a So
Hugo: I guess like, for me,a bridge is very simple. It's, it's moving, a token across chains. Now that'swrapped, like it's wrapped version, right? So you would have Bitcoin on onechain, obviously on, on Bitcoin, and then a wrapped version of Bitcoin on Flare,or potentially through Flare, via, through Flare to any other chain.
Let's say Bitcoin on Ethere, Bitcoin oncosmos, Bitcoin on pka, dot using Flare's, you know, bridges that people havebuilt on Flare. Mm-hmm. . And how about
Mike: a relay?
Hugo: So relay isinformation between chains, so. In the same way that Flare comes to consensusover what has happened on a chain, A relay enables that same data to be movedbetween different chains, not just Flare.
Mike: And of the re, betweenthe relay and the bridge, which is more, which is more technically challenging.You mentioned 3 billion was stolen or hacked through bridges this year. does,do relays face any
Hugo: sort of So both. Both,both are absolutely, , minefield. Mm-hmm. of security threats. Right. So thereason why bridges have been, you know, in the news, it's because that's whatpeople have been using.
relay of data is something that isalmost certain to be a very large product in the future. as essentially lots ofapplications need to go multi chain will go multi chain. there. The, the thesisthat there's going to be one chain to rule them all I, I really feel ismisguided. it's just like, it's not what we've seen to date and it's probablynot how the world works in general, at least when it comes to blockchain.
and so I think relay will be a huge areaof development in the space, but it's equally littered with, you know,complexity. Cuz ultimately those messages that you're carrying across chainshave value because if you deliver the message and it's incorrect, , that mighttrigger something in a smart contract that causes a loss that would not havebeen, you know, that loss wouldn't have happened if, if, if that message hadbeen delivered correctly.
So I think both are equally, riskyareas. I've seen a few products come to market in the relay space, which I'vefound hard to understand the, the safety, criteria. but, yeah, so, you know,we, we obviously have our own thoughts on how to build these products and, and,and, and are researching, that internally, to make sure that, you know, we arevery, we're very interested in building on top of our own chain, let's put itthat way.
Yeah. we have some, some, some things wewe're definitely doing. but the state connector goes a lot further than that.potentially the state connector can prove lots of things from external systems,not just blockchains. And that's where things become much moreinteresting.
Mike: I want to go in thisdirection.
I wanna ask you just to, on the Bridgesidea, since you're so familiar with this. 3 billion was stolen or hacked out ofbridges this year. What did people believe was true? What, like, what did welearn as an ecosystem around the construction or the analysis of Bridges? Howdid these attacks consistently happen?
Like, what was the fallacy? So
Hugo: I, I think they, theyweren't all for the same reason, right? you know, for instance, one of them wasa really sophisticated attack on individuals who were essentially operating abridge. Okay. So it was a, it was an attack on the infrastructure. So one ofthe core reasons we built Flare is that broadly the largest bridges have beenwhat's known as a multisig.
They tried to obfuscate this by callingit other things and, and having, you know, an NPC layer in the middle. and soit's not strictly a multisig, you know, they're, they're essentially shoutingthe keys and, essentially, but, but ultimately when you, when you boil it down,it gets back to say eight to 12 people, signing for something on a crosschains.
and that's an inherently nondecentralized system. it's inherently weak. and it, it's, it's also reallycustodial.
Mike: is that because eightis a small nber, like if it were 80, it'd be better, or,
Hugo: I dunno if it would bebetter. I mean, you know, ultimately how many people does it take to get in atelegram chat and convince them to steal the money?
Like, I, I, I don't, I don't know how. .I don't know. That's just not the style of bridge that we like, right? It's like,for me, it doesn't matter how large the multisig is. Ultimately, I think therecent shenanigans in crypto have shown me that if something, if a bad actorcan do something, it will eventually happen.
Mm-hmm. . and so for me, I don't care ifyou make your multisig 800 people, if I only need 600 of them or 700 of them inorder to be able to steal the money, you know, it's just not that hard toachieve. Right. and by the way, when you get to 800 signatures, you, you know,you are starting to, reach a point of, you know, it's impossible to actuallysign that on chain.
So that's why people build things likeNPC and stuff like that. But, , it's still a very large message. So, I'm not,for me, that's the, the, the key problems we've had with Bridges are centralization.poor code. and poor code was probably one of the biggest ones. Like, you know,wormhole, there was an attack.
They, they, they spotted the attack.They were just about to push the update. And just before they pushed theupdate, , someone, someone, someone compromised the bridge, cuz they, they,they, they either spotted the update in, in, in GitHub, or, or it was, I don'tknow. There's rors. Mm-hmm. , , but I, I won't, I won't, , I won't, I'll trynot to li anyone anyways.
Not here. Neither. Hint nor there. What,what, what, what's really bothered me about Bridges has been essentiallythey've been constructed with, very little risk management in place. Mm-hmm. ,right? Like they've been constructed with this idea that. Oh, we should be ableto bridge all of the money across the bridge in one step.
Like, okay, but it's a financialproduct, right? You don't wanna have any risk management. You don't want thinkabout risk. You don't want to think about what happens when goes, whensomething goes wrong. So we've put forward, if you've read our papers, we putforward at Bridge Design, which will, you know, we're, we're very, you know,happy to be working on, called Layer Cake.
And what that does is it essentiallyreduces the risk surface. It reduces the risk surface by, allowing the bridgeto be collateralized. And what do I mean by that? So if you think about aphysical bridge, you have a roadway. Mm-hmm. , , that roadway can only handle acertain nber of cars. and if you think about the nber of cars that it canhandle in say, one hour, the maxim nber of cars, that's our collateral.
So we essentially have collateral forall the funds that are in transit. , right? So, let's say we have a milliondollars of collateral on a layer cake bridge. Mm-hmm. , that means a milliondollars can flow across the bridge in one time step. Okay. So we reduce ourrisk by essentially setting a bandwidth in a million dollars, rather than allof the value of the bridge in one coat.
So if it's attacked, you can, you canonly, you can only take a million dollars. Right? So, but even then, if, eventhen if it's attacked, we've got the collateral. So essentially our bridgeproviders would be underwriting the risk of the bridge, and in fact, they'reactually underwriting at risk of themselves being malicious.
Mm-hmm. . , that's, that's why we havethe collateral. So you have this set of providers that operate the bridge.Fine. Absolutely. Now, how do you make that trustless? You make that trustlessby collateralizing them and then making the amount that can cross the bridge.The amount of collateral they put up.
So they own fees on their collateral,but if they're malicious, if the bridge gets attacked, they obviously pay out.So it's like insurance for a bridge. If
Mike: you, that's kind ofinteresting. Would it be fair to say that the analogy is like, in order for meto drive my car over the bridge, I have to have another car that, I I put incollateral so in case my car falls off the bridge, then I have another car tostick on there.
Hugo: I think that's pushingthe analogy too far. . I guess it's more like the, the manufacturers of thebridge will give you a car if the bridge eats your car. Got it. Yeah. Right.that's, that's the point of, of it, which is if the bridge goes wrong,principally if we get the code right. The bridge will only go wrong because theoperators of the bridge became malicious.
Right. at which point their collateralsused to pay out whoever they stole from.
Mike: Yeah. And is that achange in the typical bridge design? Do typical bridge now have a, have like athrottle limit where they say like a bridge can be full?
Hugo: no, not none that Iknow of. There may be, there's, I'm sure there's lots of designs out there, butnone that, none, none of the major bridges have a throttle limit.
We have aro limit, but ultimately youcan still bridge back as much as you like it. It'll either, it'll just takelonger. Right.
Mike: And the idea would beyou have, you have, instead of increasing the size of one bridge to be huge,you just add multiple bridges. So you have collateral for each new bridge.
Hugo: I mean, we wouldgenerally have collateral on one bridge for one asset.
Right. Okay. So like, if it's a bridgedealing with E, the collateral would be in E and then we'd have another bridgefor a different token. Mm-hmm. . And, , it's gonna cost probably a bit moremm-hmm. than conventional bridging. But if you don't lose a hundred percent ofyour money, I think that's a worthwhile trade.
Mike: It seems like peopleare willing to take a, take a fee in, in, in exchange for security and that, Imean, it's like, I think if, if people learn anything, it's probably that, it'slike I'll pay a few percent, whatever. It's like, can guarantee it. It's also
Hugo: really important forthe downstream ecosystem, right?
Mm-hmm. . So let's say you're adeveloper and you wanna build on a chain and you want, you wanna build a DeFiproduct for something that's being bridged, that, that includes, say, an assetthat's being bridged from a Syrian, if that asset's not safe. The value of yourproduct, the value of the ecosystem you're building goes to zero or close tozero.
Mm-hmm. , because, you know, let's sayyou are a lending protocol and you're using rat teeth and you are letting peopleborrow something against that rat chief, if that rat teeth isn't safe,collateral in your lending protocol collapses. And then you, your, your, yourall about hard work you've done, you and your team and your community and yourcommunity managers and all the people that, you know, put tvl in to get thetokens.
You know, the whole, all of that'sundone because of an unsafe bridge. And we've seen that a lot. You know, we'veseen, I think it was, I think it was Moon Beam that really, really suffered onP dot, I think it was Moon Beam. I could be wrong. Really, really suffered. Oneof the, the chains really suffered from the bridge going wrong.
It basically, I mean, you know, they hadto lay off a bunch of people. Mm-hmm. ,
Mike: have you been caught eitherthrough Flare or personally in any of these hacks or, , I haven't. And, and doyou suspect we're in the tail end? Do you have any ideas to how long thisripple effect of, , you know, seemingly this idea of like collecting customerfunds, reinvesting it into the market and then the market goes down, you don'thave the money to pay back customer funds?
Hugo: so there's, let's beclear, there's absolutely no bridges that I know of that have done that.
Mike: Oh, right. Yeah. , no,not Bridges. That would be more like
Hugo: a Yeah. but what Iwould say is like, you know, a couple of the bridges bailed themselves outthrough parent company funds. Mm. I would be.
Very circspect that given the bearmarketing crypto, the parent companies can be prepared to stp up that cashagain. Mm-hmm. if, if anything goes wrong this time. Right. And so, you know,what we are trying to build with Layer cake, or at least what we, what we'veconceptualized what we've put down on paper, what we're, where we areresearching is how do you build a bridge where the insurance is native tocrypto?
Well, you get that from the bridgeoperators, and that it's self-regulating, again, bridge operators in theecosystem around it. Mm-hmm. , , and provided our code doesn't, you know,doesn't break or that doesn't cause some unexpected, you know, issue then, thenyou know the risk. Essentially what we're trying to do is, is, is de-risk theentire bridging structure and, and then cover.
The only risk that remains if we aresuccessful at de-risking it, which is the risk that the operators of the bridgebecome malicious. and that's why we built a canoe net. That's why we have twodifferent test nets. We, we, we, we do a lot of testing before we releaseanything. Mm-hmm. , , perhaps, perhaps ungodly amount of testing.
I would've liked to, I would've, as ceoI would've liked to have gone a bit faster, but I'm very glad we haven't. andyeah, so we do, we do a lot of testing and, and we work with the auditors thewhole way through from inception to the, to the end of the product with theidea that, you know, rather than having an audit at the end of the build where,, an auditor has to go through thousands of lines of code, maybe, maybethousands of lines of code, in say a couple of weeks, if you work.
With them from the beginning, thenyou're gonna end up with, a group of people, a group of external auditors whohave a substantially better understanding of what you've built. Mm-hmm. .
Mike: Yeah, it makes sense.Sounds like the right way to do it. I, I, I wanna ask you this question. Thisis something I've been wondering about for a while, and I feel like you, you'requalified to answer it, be it's, , it's quite abstract.
So if I think about traditional financeand the, the trade off of safety, security, and price, do you, there's acertain fee, there's a certain, cost for that system. You know, for me to use aVisa MasterCard, it's like 3% and virtually all merchants pay that roughly approximatenber. There's costs In order to use a a, a Fidelity brokerage account, there's,there's costs everywhere.
There may be baked in, you know, theymay not be obvious, but there's some, you could get to some percentage fee. Incrypto, you have a, you have a different set of trade-offs between cost,security, decentralization. What, how do you think about maybe the, theconvergence of cost specifically? Do you think that in the traditionalfinances?
Hugo: I think this is reallyinteresting cuz something that people kind of forget. The reason why your bankwell used to pay you interest and probably does now pay you interest againafter having, after the interest rates having gone up. Although I must admitI've not seen it. , that may seem to be a bit slow.
, the reason why your bank does thatversus you paying a fee to store your funds at a brokerage, your bank doesn'thave your, you know, the money you have at your bank isn't yours. Your bank'susing it, right? Mm-hmm. . So that's kind of like the exchange model that'sbeen, , right. Very much, ,
Mike: sorry to interruptyou, but that, that same exchange model collapsed was the post, that was thepost World War ii.
That was the Great Depression, collapse,the run on the banks, thousands of
Hugo: banks on bank forthat, or, you know, 2008, which inspired our whole industry, right? Mm-hmm. , ,you know, which is differences in crypto. There's no, there's no central bankthat's gonna come along and make everything right. , not to say that they didmake everything right.
Let's be very clear. I mean, , I stilldon't think we've actually seen the effects of what the central banks did. Wehaven't seen the full gam of effects of what the central banks did actuallyafter 2008. I think we're basically still in that cycle. Right? , but at leastthey stopped the banks from completely collapsing and everyone having no moneyone morning.
Mm-hmm. , right? , there's no lender oflast resort in crypto. There's no one that's gonna come along when Sam freedloses all of your money, , and say, Sam, don't worry about it. , you know,here's a, a credit line. You know, just give up your equity and, and one of thebig boys will buy you out. , you know, that just can't really happen in crypto,, is a good thing.
It can't ultimately, but unfortunatelythat means that we are exposed if we're going to use centralized models tostore our money rather than using a, you know, a device, you know, like aledger or a wallet. , if we're gonna use centralized models and try and getsome income on our money, then that's the risk we are exposed to.
Exposed to, , you know, people makingbad decisions about what to invest in. , and people, people's egos getting toolarge to admit failure.
Mike: And okay. So do you feelthat there is a, , equal? Are we, would we converge on an equal price? Maybelike 5% of all, every year there's a 5% tax to just have a financial system,regardless of it's centralized or decentralized, that will eventually convergeon some
Hugo: I think if you, atleast we, we've taken a lot of, I guess, inspiration from the traditionalfinancial system to try and make, , to, to conceptualize and, and buildproducts that are just like better than others. Mm-hmm. . Right? , so ourbridges and the bridging model I just described to you earlier, it takes thisidea that ultimately you pay a fee to the person operating the bridge, but theypay you out if they aren't corrupt.
Mm-hmm. , right? If they, if they, ifthey steal your money from the bridge, the bridge holds their money in a smartcontract can verify that they've stolen your money and you get bailed out. Withtheir money directly. Mm-hmm. . So this is exactly the model we're looking tokind of, I guess, put forward in crypto, which is you want to do somethingexotic, like move between chains when chains were never built to be, , movedbetween then, you know, you, you, you pay the toll, , but you get safety.
Mm-hmm. , and I think that's a, I thinkthat's a fairly natural model, right? Yeah. To, to, to be honest with you, thebanking industry may have collapsed many times, but I don't think the insuranceindustry has.
Mike: That's a good point.Yeah. And, and so it may is that because the insurance industry is effectivelyforced by the government to keep a large pool of collateral that, cuz there's acost in that, you know, if you have a million dollars in collateral that'ssitting there under that bridge, that in case something happens in the bridge,you can't loan that out.
So that's interest, you're not making onthat money effectively. , which is, which is the correct way to think about it,because if you look at it as a lost opportunity and you start lending out thatmillion, well, what if the hack on the bridge is the same reason that people,what if the, you know, the hack on the bridge affects the money that was l outand that's where it's like everything dries up all at once and that's why Well,exactly.
Hugo: that's why you gotta,that's why you gotta pay for it, right? Mm-hmm. , you, you're essentially, theinsurance is being provided by someone else's money. You gotta pay an interestrate on that. And ultimately everything has to Harmon. to whatever else isavailable in the world and just finance is available.
And if, if, if people aren't paying feesin crypto and they are paying fees in crypto, , then essentially people won'tdeploy capital to crypto unless they can make a equal return adjusted for riskrelative to traditional finance. Yeah. So, you know, if your, your volatilityin traditional finances X and your volatility and, , crypto is y then, youknow, adjusted for that volatility, you have to be able to make essentially thesame return, if not slightly higher in order to attract capital to crypto.
, and that's the model we're, we'relooking to really kind. Put forward into the system, certainly with lackingbridging. Yeah. It,
Mike: it seems to me that ifyou almost take this from a, a even higher abstract perspective, there'ssomething that I think people are resistant to or maybe don't completely havetheir mind wrapped around this idea that preserved value.
So money in this case that is not beinglent out, is effectively, , it's like, , it's, it's value that you can thencash in on in the future. Whereas the other way around, if I take on debt, I'mborrowing from the future. It's like if you hold money, you're giving to thefuture. If you borrow money, you are, you're taking from the future.
So if the future is a down slope, likethings aren't as productive, something happens in the world, there's chaos. Weneed to, we need to cash in on those reserves and you don't have thosereserves, then it's like, you know, I almost think of this in a, in waterbecause it's more. Physical, it's like for me to have 50 gallons in a storagetank is great.
For me to have 10 gallons is still fine,but for me to go from like zero gallon, five gallons to zero gallons, the costis, it's just, you know, infinitely worse. So it's like for me to, so it's likethere's some balance where I, I look at this and think of this from an ababstract perspective. What's the correct amount of, we say this word insurance,but it really, it's like it's allocation of value to the future reserves thatyou're not lending out that that's not in the market.
Maybe the water concept to me ringstrue, but you could just use money and that it's, it's maybe in the futurethere's a drought and we need to cash in on that, and then the value that'sstored gets compressed. But you still have something, it's like to have to have$0 in circulation to have zero water is just.
You know, the game's over at that point,.
Hugo: Well, you know, thething about insurance and all major insurance companies is they have massiveinvestment portfolios. , they manage their risk pretty well. Right? So, youknow, they keep a certain amount of cash on hand to payout. , but you know, ifthey're not putting that money to work, then they won't be able to cover therisks that they've insured in the future.
Cuz they're only insuring risks based ona premi and also an assption on how much their capital pool is can grow.Mm-hmm. . , and anything they, and, and, you know, if you look at say ShireHathaway, which is essentially a vast business conglomerate, conglomerate thathas largely been built on the back of a very large and successful insurancecompany.
, and essentially what what they did wasthat they insured risks that are sensible premi. Took that, , you know, took,took that income, , deployed it into things they thought were massivelyundervalued, , or a portion of that income, put it into things they thoughtwere massively undervalued. And, and, and over 50 years have built one of thelargest companies in the world.
Like it's a, it's actually a reallygreat business. Mm. , and what we're, I guess looking to do is we're looking toallow people to deploy their east, their s d c, their s d t as collateral intobridges for which they earn the bridging fee. , and so essentially they get anincome from their, the collateral they're, they're sitting on and the bridgeitself is providing a service mm-hmm.
and they can roll that income back intothe bridge or into other products or whatever. But, , you know, that's, it's,it is a very simple model. Yeah. But it's one that has been. , deeply
Mike: overlooked. Yeah. Andit seems like you could take this same idea of the, it's almost like crowdfunded,, maybe I won't use that word, but it's, , you know, an insurance countrycompany is centralized.
It, it, yeah. Collects the money. Thiswould be more decentralized. You could take the same idea and apply it outsideof bridges, right? I mean, you
Hugo: Oh, absolutely. Look,there are insurance, I guess Dows mm-hmm. on blockchains, right? It's, it'spretty nascent. , but I think it's a fantastic, , I think it's fantastic usecase in blockchain.
And I mean, one of the things we thinkabout, like show Flare here for a second, right, is if you're running aninsurance Dow you want your risk to be as diversified as possible, right? You,you, you, if you're running an insurance dow that is covering DeFi, you may notwant to have a hundred percent of your exposure to that dow to a single DeFiecosystem.
So, you know, if you are in insurance onblockchain, what would be an interesting place where would be an interestingplace for you to go? , well, maybe a chain that can prove events from otherchains, so can prove that there's an insurable event happened. Mm-hmm. and payout upon that insurable event. , but allows you to create a, an insurance poolthat is diversified because it doesn't just do it for one chain, it can proveevents from multiple chains.
So you're in, you can be simultaneously,, essentially insur. , many different chains. Mm-hmm. , , and thereforediversifying the risk of your insurance pool, which ultimately over time shouldreduce premis. Can
Mike: I ask you a questionkind of related to this, just continuing with this idea. So could you go evenfurther and say, we wanted to create, , say, , fire insurance for your house.
And we say we're gonna do it on chain,but the way that we'll gather, the way that we'll verify on chain whether ornot your house is, has been burned down is we'll have the people who stake themoney in the insurance pool will be presented with the evidence. And thenthey'll vote and say, do you agree that, you know, here's pictures, here's thestory, here's, here's all the offline off chain data.
Hugo: So, so you can do itthat way, like, , dowels is what you're talking about, governance. , Dows arepretty weak at the moment. Mm-hmm. in terms of like, you know, there's some nefariousevents that happen with Dows because like say one person owns one, one entityowns the majority of, right. Right. , that's the problem there is you may nothave insurance.
What we're looking at more, you know,candidly is can the state connector our protocol for understanding otherblockchains? Can it be reliably applied to other systems where we'd like to gowith that? If the research bears fruit, which I think it will, is, can weprove, you know, from a, from a system, like, I dunno how it works in the USbut you know, let's say there's some database somewhere that says House X hasin fact burnt down.
Mm-hmm. . And, and people are preparedto trust that database. Right? There has to be trust, cuz ultimately the datahas to come from somewhere. , but if we can prove that House X has burnt down,then there could be an insurance product or a blockchain mm-hmm. . , and that,that is interesting. In fact, the whole area of can you prove something from anon blockchain system, i e a database or a website, , that's very interestingto us.
Mike: Yeah. That does seemlike to me, whether it's the house burning down or the car crash or somebodydying, or somebody being born or somebody, whatever the, the off chain eventis, I mean, it goes, it
Hugo: goes to so manythings, right? Like, you know, if I can do stuff that, if I can prove somethingfrom a website, let's say Twitter, I can do micropayments against, likehashtags.
There's a whole advertising model there., you know, The whole, like, I think about this a lot, like micro influences.Mm-hmm. , you know, it's like I'm not Kim Kardashian, , thank God. , but , youknow, and no one's gonna pay me a million dollars for, for filling something.They might pay me like 2 cents. , okay.
I probably wouldn't do it, but otherpeople might. , you know, there's, there's a whole, if you can start provingthings from the, from web two onto Web3, you can break open entire, like modelnew, you can create entirely new models. Mm-hmm. of how web two is monetized.And actually that's scenario.
Mike: Yeah. And alsocomplimented by the fact that there's a, a decentralizing effect happening onattention in web two. It's like, , my brother started a company called Cameothat allows people to like connect with celebrities and they do shout outs. Butthat, that worked because like to your point, you went from a few majorcelebrities, a few large, , channels, cnbc, cnn, ABC, to now 10,000 andeveryone splinters.
So there's all these like littleclusters of people. So how does that affect the world? You know, just like atYouTube
Hugo: just getting more andmore people putting out actually genuinely more and more interesting content.Mm-hmm. . Mm-hmm. . Right. And, and, and, and the only way that's currentlymonetized is by someone else selling ads and giving them a cut of the revenue.
Mm-hmm. like there are. Potentiallyother monetization models if you can either build a platform. Although I thinkthat's, to me, that's, that's really hard, , to build a new platform andattract people to it. You know, you've seen people like Master Don try to dothat kind stuff and sort of doesn't work, hasn't worked great until now.
For me, it's like, can you use Web3 as amonetization layer on what top, what on top of what has already been built?Mm-hmm. . , because you can prove the data onto Web3. , that's kind of an anglethat, you know, myself and the team are. Heavily researching. How
Mike: has the collapse oflike, , the Three Arrows and the Celsius and S SPF is, or ftx, how has thataffected the product development or anything else related to how you guys areworking or how other people are working?
Hugo: so for us, it hasn'thad a direct effect. , we were never in ftx. We, no, there's been no directeffect for us. It's obviously had a chilling effect on the industry. Moregenerally. People are scared, people are conserving runway. , it just meansless is gonna be built. Mm-hmm. , it's also, you know, there's lessanticipation around the industry right now.
Check sizes are smaller. Valuations aresmaller. It's healthy. Mm-hmm. ultimately, I mean, to be honest with you, youknow, my. Few of the last couple of years is, and what's been funded and what'sbeen popular is pretty thin. Mm-hmm. , I think like it's been a lot of, like,little bit quick to market, not very well thought through.
Things have been funded. , a lot of thethings that got a lot of interest. , were basically a, you know, essentially asole where, you know, some large backer would dp in a bunch of money, ,essentially pp it up and try to pull out a bunch of money. , we've been at thissince 2017 perhaps thinking about how to do things well, , and build a, a genuineproduct in the space that, that works is useful.
Mm-hmm. , like for me, I'm like, I, Iget, I've been. I've been semi, you know, the last couple of years, been reallyhard watching people do things that have largely been not useful. And I thinkas an industry we really need to take a lot of blame because we haven'tattracted billions of people to the industry because we've built nonsenseactually broadly.
Mike: do you think the realconstruction and production is is happening now in maybe the next few years?Like, where will we cash, like even from Flare or other similar networks, whereare, where are people using this? How does it cash out to real life now andthen
Hugo: how do I think so? Ithink there's, there's a huge, there's huge move to scaling.
There's a huge move to sharding. Thesethings are practically necessary. Prerequisites. Mm-hmm. for a, for ablockchain industry. To be able to scale to the billions of users. Right? Solike what people like Ethere and Polygon and, you know, all the variousdifferent networks are doing is that they, they're essentially attacking thecore infrastructure challenge.
Right? And, and, and I, I think that'sincredibly important because core infrastructure, you know, at least at smartcontract level is not, not yet good enough. if we, if we had a, a product thatgot the usage of Instagram Yeah. the whole, we wouldn't be able to handle it.Like it wouldn't work.
Yeah. It would just be too expensive touse. And so, you know, the, that's where the real work is happening. And then,you know, for us, you know, we are obviously very consensus focused. we stillhave a active research thread internally with regards to how to build a, youknow, consensus, , algorithms. but we're, we are also looking at, well, okay,once you've got a consensus algorithm that's really good and a way scale that,and, you know, all the various other things that people are doing, how do youget sufficient things onto the chain?
Sufficient pieces of data onto the chainwithout soft, without using soft mechanisms. Right? So like you, you, youmentioned that fire example and I think it's, it's, it's interesting becauselike that that governance mechanism, just like so corruptible mm-hmm. , likethis is what's been built for the last couple of years in crypto.
It's like super corruptible, methodsbecause of a cheap and easy, quick to put up and. People would throw 20million, a hundred million dollars at them. And for us, we're thinking about,well, how do we prove things like genuinely onto a chain and actually getusage? then, you know, others are thinking about scaling and I think hopefullythose two things come together whereby you've got super scalable blockchainsand really, really good data.
And then people can actually startbuilding products that can hit a billion people.
Mike: Yeah. That's what itAnd where do you, where do you see it like happening first? Is it you thinkreal estate funding insurance, financial tools, social media networks? Likewhere will people say, this is it I'm using,
Hugo: so I think, I thinklike what blockchain been really good at is figuring out financial models.
Mm-hmm. , like DeFi, right? Absolutely.Excellent. But their financial models almost like, But with no market orminimal markets to serve. Right? As I said at the beginning, I think broadlylike it's been how do I borrow US dollars? Cool. Great. So that's like DeFitoday. Sure, there's lots of other things that happen in DeFi, but like I'd saybroadly the biggest use cases, how do I get, how do I get S D T, right?
How do I borrow U S D T so that maybe Ican buy another token or whatever, whatever you wanna do with that U S D T, howdo I get it? I think that's been the huge winning use case. It's beenfantastic. And actually DeFi is amazing. Like it's, it's, it's, it has createdit truly, really in some ways decentralized financing system.
But today with very little to finance,that's the problem. Yeah. and so then, If you can match what's been built,which is a really great financial system with lots of use cases, lots ofgenuine use cases, which, which aren't just like guys and girls with likeblockchain logos on their laptops, . But like, like, like, like lots and lotsof actual people.
you know, the, then I think you've got,then I think the fruits of DeFi will actually be, be born out, you know?Mm-hmm. . but that's kind of where I wanna see this go. And where do I think itgets applied first? I think, you know, gaming Yeah. Is sort of the most obviousone. And then social media. Mm-hmm.
because, you know, if you, if you lookat like social media, the vast lion's share of the income probably goes to avery small amount of people. Would be my guess. Yeah. And most people areconsers rather than producers or whatever term they may want to use forthemselves. my guess is that essentially trends, , sort of continue such thatmore people need to earn a living from something that is non-traditional, likebeing popular on a social network.
Mm-hmm. or being really good at a videogame mm-hmm. or moderately good at a video game or moderately popular on asocial network. Like those things have value as well as just being, you know,having a million followers, maybe 200,000 followers also has value. Yeah. , youknow, so the question is how do you, how do you give that person that value?
, that's kind of the question that weare looking at.
Mike: And, and why is it,what mechanisms, what decentralized mechanisms are better than centralized?Cause centralized is pretty, I mean, it's, it's very, very efficient at doingwhat they do.
Hugo: It is efficient,doesn't work for small amounts of money. Mm-hmm. . , so for instance, micropay,that doesn't work great.
, you have to trust the centralizedentity to pay, you have to trust the centralized entity to know to, to, toessentially say, oh, you, your tweet or your video or whatever, got this manyviews. That's like a, it's quite a big leap of trust actually when your, whenyour income is tied to it, , you can be booted off at will, , You know, it's,it's not, , that you are also getting a huge, you know, revenue share.
Mm-hmm. , the, the entity is gettingvery large revenue share. , it interests me on whether that can be made,whether that structure can be made flatter, fairer, and more democratic.
Mike: yeah, it is, it isfood for thought. Hugo, I wanna ask you, where are you finding yourinspiration? Are there certain people you're following on online blogs orTwitter somewhere, or books? where do you find yourself consistently turningto, to learn and enlighten yourself?
Hugo: I, I tell you, Imostly find we're now a team of 51 people, and I must admit, I've, we've.
We, we haven't spent very much timetogether cause of Covid. Mm-hmm. , but recently we've been spending a lot moretime together and it's absolutely awesome. really get a lot of inspiration fromworking with the team. And then I've been reading a lot of different books,like what, so, so its, recently been, I guess listening to some of the JamesClave saga.
It's quite interesting. Mm-hmm. . and Ididn't know much about the Iranian revolution. not sure that the, book I'mreading is a fantastically accurate historical source, but either way it'sentertaining, podcasts, to be honest with you, Twitter is probably my centralsource of news these days, especially in crypto.
Mike: Yeah. And where,what's your Twitter? Where are you available online personally? Oh, at,
Hugo: at Hugo Fillin. Nice.Oh, I'll make sure. I'll tell you what it's been. Absolutely amazing to watch.The, the amazing, tragic to watch the unfolding of FTX in real time on Twitter.It's like nothing I've ever seen before. No,
Mike: it's crazy.
And it feels like there's two storiesplaying out simultaneously in two different worlds. It almost feels like thefrack, the fracking of like the, the web two, , traditional medias, mass mediacompanies telling one story. Yeah. And then the actual unfolding of it onTwitter .
Hugo: I mean, that, that,that is actually insane.
Like how, how the traditional media haskind of soft bold, the FTX saga is, I don't know. I mean, I, I hope it spells,you know, I hope it spells a reckoning for traditional media because it's beena huge dereliction of duty.
Mike: Yeah, I think the, themost critical interpretation would be that the media companies areintentionally soft balling because they were funded by this guy.
They received donations. And the, mythought is it's more, it's more middle ground. It's more like they don't reallyknow how to be harder. They don't know. I mean, they look at this guy and heseems like a smart guy who made a bunch of mistakes and they're like, well, whydid you do that this way? Why did you do it that way?
And they're, they don't quite know howto be harder. Right. I, I don't know that mean that, I dunno,
Hugo: I watched the, Iwatched the Theranos docentary on Amazon the other day, and you know, they,they took her down pretty well. Yeah. Yeah. And she didn't lose anywhere nearas much. And it was all private investors money.
Yeah. Private investor money wasn't,wasn't public. What was private investor's money. They took their risk. Shedidn't like, you know, she didn't say, oh, your money's not at risk. Yeah. Youknow, whereas Sandbank and free. Deleted a tweet that said We've never let outyour money. Ouch. And that I find that I find very difficult to handle it.
Mike: I do think that he'sprobably taking a very strategic approach and saying, let's go all out. Let'stake the exact opposite approach to what a typical lawyer would recommend andjust try to win the appeal of the public and have that be influential in the,in the court. Right. Cuz otherwise, if you're quiet, then there can be anarrative that's created that you're a monster and that influences youroutcome.
Hugo: his approach appearsto, appears to be, , that he is an idiot. . Right, right. And then he knewnothing, nothing that was happening internally in that he completely misjudgedeverything. And, oh, I'm sorry, I'm just a young guy. I need another chance.And all that kind of stuff.
Mike: And the thing is likethat, it's a weird approach.
It's gonna come out like the, the Slacknotifications, the texts, the emails. You're not gonna hide those. I mean,they're gonna be revealed in court. Right. I maybe he did
Hugo: know nothing. I mean,the guy was probably super busy. Yeah. But that's not an excuse. Right? That'snot, that's not like, I don't think if, if, if, if JP Morgan went down tomorrowand, and Jamie Demon said, oh, sorry, I was too busy to know what was happeningat JP Morgan.
I don't think anyone would accept that.Yeah. so yeah. Anyway, I should probably not get too far on this subject as I,, I, I, I may say something I regret. Well,
Mike: it can't be anywherenear as bad as what he's doing. Plus, there's nothing wrong with side share,speculating and just observing, making sense of what's happening so far.
Cause sometimes it diverges from realitythat like there, there's a, there's a strong group on Twitter of smart peopleand crypto and elsewhere in tech that are like, this guy is obviously stealing.And I, I think that that draws a lot of attention. People kind of see throughthe facade pretty quickly if you follow it.
but either way it's gonna be indicative,it's gonna be influential regardless, so it's worth paying attention to. Hugowill put all these notes in the show notes and I really appreciate your timetoday. Congrats on all the progress so far. I hope you guys significantly, ,influence the direction of crypto for the better.
Continue to do so and thanks so much,man. Appreciate it.
Hugo: Thank you for yourtime.
Mike: All right. Take care.