Defi Panel with @MariiaYatskovs2, @NourHaridy, @0xmikko_eth, @sarangparikh22 and @clemlak

Transcript

Um, hey everyone. Um, this is the D5 panel. I'll let you guys introduce yourself. So, please go ahead. Hi, I'm Nor, founder of Inverse Finance and Dola Stablecoin. So, yeah. Hi, I'm co-founder and CTO of Gearbox Protocol. We reimagine how lending could be. Hi, I'm Clement, solidity developer at VFAT and also Wetstone Research. I'm Maria. Hi. Hi again. And I'm leading people operations and recruitment at Cow DAO. Cow DAO is decentralized organization of developers, market makers, and community members. Thank you.

Um, and you know before we get into the panel and get those spicy questions, uh, just give me a one word answer, not more than that. Do you think it's better now to build in D5 than last two years? So like was it better to build in D5 2 years ago or is it better to build in D5 now? Almost start. No no no. Um well I mean I can say you want more work. Yeah. So let's say like is it better to build now? Oh it's it's horrible now. Yeah. That's my answer. Um I mean I've been I've been in I've been building we've been building at inverse for 5 years and it just gets I guess more competitive but also people are

getting further and further away from like cypher punk values and getting more and more centralized. I think that's the market is more money chasing than tech chasing. You know I think people chose between you want to be people used to say do you want to be rich you want to be right I think they chose rich uh yeah so good question uh of course it's it was much easier to start a few years ago but in terms of technology it's uh we made a big shift in terms how to build projects thanks for security auditors was uh unfortunately hackers we learned a lot. So what I can see now

people could build more robust things taking the previous experience. So it's a little bit easier from the technical standpoint but it's much much more harder in terms of markets than trucks and capital. Um yeah my opinion is a bit similar actually. I would say it's harder now because in the past there was uh less competition. uh but the tech stack was a bit harder and now we have also more regulatory constraints but if you're successful you can be way more successful now than before but uh so I would say that now it's easier

because we have better infra we have more clarity and uh everything is again like just smoother and faster however I would I I still miss lots of security uh lots of um lots of UX and uh but however in like if you talk about the two years ago also it was more structured it was uh more I would say we took it more serious we were following more on the principles of the decentralization we were really caring about the products we build the protocols we build and how do we build them yeah um and actually that's a good segue to My next question uh do you think like the trust assumptions on D5

apps has been increasing or decreasing? Um previously what what I personally felt like there were like apps which were like more reliant and built for a purpose that tomorrow if it's a nuclear war or something the apps still perform. Is it the same case now or do you think like where where are we heading towards in in that fashion? I can I can pick this up if you don't mind. I would say it's decreasing. Mhm. We have more trust into defy. We have more protocols which are solving the real use cases and we are one step closer to the mass adoption which I would love to see in a few years. However, um as as I mentioned, I would love to see more I would love to see more

security. I would love to see more trusted products and protocols. uh better composibility where protocols can just build on top of each other and trusting that that block is safe without the malicious actions and so on. So, uh answering your question, yes, it's decreasing. Um but we wish all for more. Yeah. So, basically like the trust is decreasing but there is uh inherent borrowed trust from different protocols which are safer and you you you're using it from there. I would say no trust into sorry maybe I need to explain myself but trust in defy is increasing okay but the risk is decreasing okay makes sense makes sense thank you

um very quickly I would say if your question is are protocols making more trust assumptions as in they're trusting um they're making more assumptions about things being working for example in the in our you know sub industry of stable coins oh yeah I I mean it's wor it always gets keeps getting worse especially with the RWA trend you know like for example when DIV1 was just ETH you know collateral I remember when it upgraded to multiolateral and that was a mess everybody was so angry everybody was complaining like he could not like put more like other collaterals we could not trust other collaterals this is the end of D and now D is adding censorship I mean USDS for example including

censorship or you know or just adding RWAS which are custodians etc. or not just die to be honest. I'm not picking on D but like every other stable coin project now using RWA and everybody is okay with it. I guess the trust assumptions have shifted dramatically honestly it's uh no one at this point no one really cares about what the trust assumptions are and everybody's maximizing them in stable coins at least. Yeah. And in terms of lensing market and maybe it's regards into the whole industry the initial values were totally removed because like maybe three four years ago it was significant to make things

unstoppable to make things uh like bank and banking people know nothing. So we really try to make a machines which works without any possibility to be stopped and uh really people evaluate crypto as an alternative financial system and now there are a lot of conversation how we can really develop all these barriers and things like legislation on the smart contract side.

So people do not valuate that the system works independently. They really just focus maybe on yield or even accept pretty risky solutions. So some of our competitors they just hardcode one to one from one stable coin to USD which was a pretty risky solution because the price is never guaranteed and at the same time we we in gearbox even have a pricing for USDC. So we didn't write even USDC but it's one of the biggest tokens on the market and people say okay it's totally fine why not and these

steps really remove the base of perception and at my point if we really do not evaluate the value of uh decentralization the value of unstoppable finance etc. It's really move the price and value of the whole defy further and further to the lowest things because people initially came to crypto as alternative system and if you try to build the things closer to existing financial system we really start to compete and of course this competition could not be really won because we have a very major huge

banking system. we can't compete in any kind of money or size or even technology or person involved. So yeah, unfortunately this happens. Yeah, I guess also trust is uh increasing but uh in very specific actors you can see like people will trust like OG protocols or protocols that have been there for a long time even as you said like with D things like that people will trust it even if they are changing things but when you are like launching new protocols it's harder now I guess to get liquidity to get traction because people in the

past uh they were willing to like put money in some unedited projects that got uh launched just because they were willing to try DeFi, try new stuff like with alchemics project like that. They were really willing to try new things and experience DeFi and new financial uh yeah like a new financial system I guess. Um so yeah like and I feel like with that like with the trust assumptions we do have actually a lot of still a lot of hacks going on right like I I feel like okay at the start there were a lot of hacks now still like you know the the

only people that get benefit from that is like probably North Koreans right uh they've probably made the most amount of money in crypto at this point um but we still see a lot of hacks and some of them are like smart contract but a lot of them are also more on the web two side Right. Uh and still the space uh I I don't really find a lot of monitoring uh data and tooling. A lot of recent hacks could have been avoided with a very simple monitoring scripts, right? So how do we try to actually improve it and you know like force protocol people and devs to focus more on it, right? Because now it's been like oh if if you most likely get hacked. I think like previously the like protocols did take more responsibility versus now if they got hacked or like you know probably

just take our tokens or whatever you know. So what should we do like you know who should be the on maybe like Ethereum or like you know L2s because they directly get impacted with any hack right so what are your thoughts to improve the situation? Um I I would say at inverse we for example we um over the time we we have an analytics working group over the time it's really grown to like a almost like you know an onchain intelligence agency like we monitor every single user above a certain threshold. We monitor the assets and the markets outside. We actually uh and and you know like you basically we have watch lists of users

and assets and markets and and all of that to make sure there's no risk that could spread to us uh and if there's any um unusual behavior. We get instant alerts and we also have charts of past past data etc. We've done some work on open sourcing that. Think in the web two industry a lot of times when there are problems like like this that are shared by you know people like projects um that are competitors the solution is usually open source we've we've done some work on that we've open source we have a platform a pretty sophisticated platform called inverse watch which is a an analytics platform we've open sourced it but I think we we and other projects could do better on like creating some kind of shared monitoring tools that we could use especially lending and civil coin protocols because they have the

same problems they have the same stuff that they're monitoring the monitoring their users, their markets, etc. Um um so I think some open source work could work there doesn't have to be like because all I see is service providers and providing as a service but I think we do a lot of redundant monitoring work which could just be collaborated on as open source and it doesn't have to be commercial. I think open source is a solution to that but um because I know what we're doing at inverse is probably being done at gearbox and it's probably being done in a morphone etc etc. So it's just better to learn from each other and do and and create kind of like a shared infrastructure. That's just my idea to be honest more open source.

uh I don't know maybe in uh terms of protocol it's a pretty hard to explain because uh in some kind of ways protocols consider from different people like something like pretty complex inside them and people more think about products not protocols. But on the protocol side we built like a very very uh strong infrastructure regarding security things. So we have many monitorings and we already open sort the whole supply chain. I believe if we talk about

uh protocols we should think a little bit wider because the last hack for example of by bit it was not related to a protocol hack. It was related to the frontier when people sign in correct transaction and in protocol uh should really cover the whole supply chain. For example, we built a lot of uh interesting stuff related to processing front end. So uh before going to production, we launch some tests and we check that it works as defined. So we get some user inputs and we have some predicted transactions etc etc. But it's not on couldn't be limited on the protocol. My oldtime pitch was related for example DNS building DNS checking

board. We also outsource it. You can find it in gearbox protocol GitHub. It's a very simple solution but you should really as a product owner check it first because there are multiple layers when where potential attacker could really make any bad things for you. So you should monitor your DNS name, your front end. Then we check and monitor liquidity using special optimistic liquidations at the system. We check that everything works fine and we have enough liquidity to really liquidate if the problem happens on the market and so on and so on. So multiple things to evaluate and I think it should be done maybe openly. We

pitch a lot of this stuff, but other protocols should also help and really build like a community to at least make a value of this security measures in eyes of potential user because sometimes when people could see wow it's 216 person APY it doesn't matter they have a security outit how to deposit and it is here so we should really talk a lot and maybe these discussions could help users to distinguish what should be done what's done by different protocol and how security is significant because 100% loss is not really people expected um yeah I think like over the years we

are taking security for granted and like it's the same thing like in the daily life you won't believe something will happen to you until it will happen to you won't believe you will be hacked until you get hacked. And I think as you say like it's pretty clear that now uh maybe protocols are getting better at like testing and auditing the smart contracts. Maybe not all of them, but I feel like we are in a good uh in a good way about this. But like many aspects like we've seen that as you said like the hack with the safe wallet or many other hacks that happen even with like

centralized exchange, they are not happening like on um like a protocol or smart contract issue. They are either happening with like fishing attacks like social engineering. So people are not like following security uh practices like the the good practices or also like um you're going to build a front end and you will trust like some random uh modules. Why? I don't know. But uh I mean this is something pretty common and you will be building a front end and you will be like using the latest version of this JavaScript library built by someone you don't even know and you will never check what's inside. Maybe there is like a huge possibility of getting wrecked at

some point. Like I feel like audits are good for the contracts but no one is going to audit the front end. No one is even going to do like proper testing of the the front end or checking the origin of the packages. So I mean now we are in a I feel also in a small uh how can I say like yeah as you said compared to trafi we are very small and I think it's okay now because like all the security practices are not like uh here but if D5 wants to grow even bigger like everything should be checked otherwise people will get wrecked all the time. Yeah.

So many things has been said already. What I can add that I believe uh personally that security is one of the biggest issues and I believe it's there is no one or two solutions. There are many many many things which should be improved and we should take it as a complex thing. And me personally I would love to see chains taking more security measures. I would like to abs um protocols to take care more of the observability and even maybe we have some opensource monitoring stack like primitive use who can you know control everything and to ensure that security is taken seriously. we do

improvements and this is also one of the huge steps would be for the defy to reach the the mass adoption because the reason why we are growing but slowly growing not rapidly is because um people need to trust more people need to trust that if they put the funds if they do I don't know swaps or anything on device they are secured and there is less and less we would see situations like this by bit or even smaller attacks. So we have security and users who are not even

native DeFi users can trust us and can step in and say hey yes that's exactly what I was waiting for for so many years and I actually like defy more than uh traditional finance and um yeah I can start using that on daily basis and I don't need to think about what's going to happen my funds are secured the UX uh the UX X is good, UI is easy to understand and um and so on. Yeah, I think we're making efforts with like C91 is like a good initiative like that helps people. Maybe we just need to give them the pausing role or even like PE shield a pausing role because half of them they monitor everybody's contract.

So yeah um moving on what are the extent we are making protocols reliant on off-chain systems such as intense right let's say in the case of cow swap and 1 in and also in in in in lending space as well there have been some RWI assets which can only be liquidated by a very uh off-chain actor right in those cases that and most of the lending protocols that we do have today have shared risk right with other assets so that puts like with adding just that one RWA which can build build give you like a hund00 million makes your uh lending and borrowing at a billion dollar risk versus also in in swaps like the trade liess risk right um and the data shows at least from what I saw from the 1 in

data most of the swaps goes to not naming the the solvers but one of the top two market makers in most cases right and at least for like the the major payers right so uh how do you think the situation would look for onchain lping or activities and you know is it is is this model good or are we like degracing back where we shouldn't maybe if you want to start from that side yeah I would love to start thank you so much we already discussed that with you uh before coming here so it's not black and white there are pros and cons and let me talk about the pros I believe um personally that

intense based mechanism like batch option is great. The reason why because it's indeed can provide for the users the best possible price. The solvers who are solving uh those intents and in the batch auction they can really do their best and get rewarded for proceeding with uh best possible price and surplus. On the other hand, it's true that there are market makers who are big and uh they are solvers as well and most of the times they are winning the let's say the solver competition and batching. I would I don't see it as a

big problem. I think it's has place to be. What I want to see is that the smaller solvers or those solvers who are having less liquidity than market makers. Let's put this way. They also have fair competition and condition to be able to provide their solution and to be able also to get rewarded and again like be huge part of the um of the of the competition basically.

um in uh on in the batch auction. So I'm not sure if you heard my previous talk in case of the cow swap we in the end of the May we actually introd introducing more fair auction where not only one solver can win but there actually can multiple solvers win at the same time. So we believe that this is also a gamecher and that's the way how it should move forward and it will involve include smaller solvers to be um beneficial out of the solver competition and the most important that the users will benefit out of it and uh see again like multiple solutions I'm talking

right now here peer-to-peer ring trades not only single trades when it's connected with the market makers I can talk a lot but I think I I will stop here. I um again like it's not black and white. I'm huge believer. I'm I I think that have the intent despite that it's taking a slightly longer than just executing the trade on chain in one transaction. It has lots of benefits. It still needs to be improved but that's exactly the reason why we here and that's exactly what what we are working on. Yeah.

Do you guys have any questions? Uh I think so the first part regards to market we should distinguish uh like a major coins and long tail because solvers could be only beneficial for major coins. Nobody want to be a solver of a shitcoin to be honest. uh and we really go deeper to this market when we talk about liquidations the expectation of yield of people who sold it's a very very high so it's a big problem when you launch for example the new chain ask someone to put money on it just to use there so it's

like a such kind of barriers and of course it's create a very sustainable business for solvers so they really get and squeeze a lot of money from chains uh could solve a lot of problem regarding to starting your own first token because you can provide liquidity by yourself. Community could help itself. So I believe AMM will exist forever. So no one could really cover all this stuff. But another very risky point and nobody cares about that.

Imagine that this solvers it sounds a pretty pretty we have a solvers all over the world. It's a really 100 computers, not more. Even it's a billions, it's 100 computers. Nobody cares about sustainability. Yesterday, we got a blackout in European Union. If we will have like a two major solvers who own billions there and it was a price drop, everything could be got because in EMM you can really compute how much money you have. You can at least make any expectations in terms of price drop. When liquidation started, you can exchange money and buy something back. In terms of offchain liquidity,

nobody cares. So, and it could be totally different. Of course blackout is a very strong case but in many stuff I was experienced people just was panic about like a market turmoil and they say hey guys let's keep a pulse and in this case it create a worse situation with liquidity because the market players expect that hey people would be willing to buy Ethereum for example like at the price of 100 1,900s but people say oh it's a very very huge turmoil we are really not satisfied with news maybe war happens or

anything else so we stop any operation and in this case a lot of capital could be disconnected as this such event so I really appreciate that you have a multiple small sovers because if you have a very big party we rely of a huge pe of capital which could be disconnected by one decision if we have a decentralized solvers who really have small part of capital it could be more reliable so I believe it's like a things which could really reduce the stuff more decentralization makes more sustainable stuff but of course capital want to do the opposite stuff because capital is all time about centralization more

capital more yield and of course banks also compete by their sites no services so it's a tricky point even now we previously seen with like even like with stock apps you know like there had been like liquidity crunch when like big market makers you know did not provide any codes um however those apps which were actually smaller than the bigger ones they had like a lot of small market makers which actually did during those times as well so yeah it's like it's a very interesting point um I think we're way above the time so I'm just going to skip to the last question um so imagine if you know like you you are head of DeFi at Ethereum. I'm not sure if that's an official post. Uh so what what is one

change you would do? Like is it something that you would do with like EIPs to put push certain EIPs or probably get a table at the discussion where they actually discuss EIPs or what what what would be a grand change and you have to just do it in one day, you know? So it has to be like a very atomic change what it would be if there is something. I think what I would do is what they're doing right now actually and they've been the the Ethereum Foundation just yesterday I think last week or so has been making a huge shift to um L1 scalability. I'm honestly sick and tired personally of like multiple L2 deployments and I wish we could just I mean we on in inverse we're our lending protocols on L1 on Ethereum Dola is

mainly on L1 but we're basically we have some activity in LL2s but it's just so much work to bridge and and create the infrastructure to support those L2s and and fragment the liquidity. I think we need L1 scalability to be honest. I think what they're doing now is that they're basically moving on from the EVM towards uh risk 5 which is an alternative that is more scalable significantly more scalable there just the architecture itself allows the L1 to scale and I think that is the right thing for D5 D5 does not want does not work well with liquidity fragmentation it works well with composibility we want to be always

composible with every learning protocol with every stable coin with every AMM uh we inverse and we we integrate with gearbox we integrate with with a we integrate with Pendle you know and gearbox you know just the same we integrate with all the steel coins with fracks and and uh you know all of us we're really all tightly knit even if we're competitors some sometimes but if we have this like we have hundreds of L2s and bridging uh also like third party bridges and first party bridges but the point is it's such a mess to be honest and people who are you know in the Solana community who are you know just they have a point like it's for for us as builders it is annoying I mean

probably even more than the users because it just increases our work by 10x for no clear reason um I don't think L2 is going to go away but the answer to your question is I would focus on L1 scalability which is what everybody's been saying and which is what I think Vitalic and the rest of the EF are doing. Anyway, I don't think on the other hand the EF or the Ethereum foundation should be interventionist, you know, and try and and I think, you know, just create some specific changes that would help DeFi as an application.

It should stay credibly neutral as it is and as they've always been advocating for. Just give us scalability. Um, please. That's all. uh I believe uh maybe the most significant part because it's like I totally against any market intervention to be honest. So we have like a two things which could be done on this level. The first one is standardization and some stuff could be dramatically improved if we can really build something uh more standardized.

Luckily, we have for example ERC 4626 is a good example which really solves a lot of stuff and make the life easier. However, I believe the biggest part should be focused in terms of values because uh the same things happens with capital there. So we have a really like a market leaders for example we have 30 30 plus billion dollars and uh we have a big uh stakers on a serum who also control like a huge amount of ease etc etc and the problem here if we can say hey guys we are for

decentralization come to a serum and this is top protocols this is top capital held It's not about decentralization. Decentralization should be on each step of the stuff. So give attention not incentive but give attention for different protocols for different stakers. It's Vitalik also made uh supporting solo stakers etc. Because if we have a lot of different solutions and we have a less like less difference between the market leader and others it's create better way at any terms what I said if you have a huge market maker who control 90% of swaps is

serum totally depends on it if you have a only one deck which really has the major values it's also not the best solution if you have only one lensing protocol so costs everyone Everybody want to be a market leader. We also depends on this and in case of hack you should really ask the community hey can we make a step back if you have different competition but not like a big levels it's create more sustainable way. So um yeah I guess yeah I'm in favor of decentralization on many aspects but I guess something interesting would be more leadership uh in the development of

Ethereum and I guess this is why also Solana was able to maybe grow fast is because feel like it's maybe less centralized on the development side. I don't think we necessarily need to be just like a centralized or someone deciding but maybe someone leading the development and saying okay this EIP we can spend this time trying to see if it works or not and uh maybe someone also trying to organize the development um of the IP. We can see that with EOF how it went and I feel like there are many like lessons to learn here. But uh personally I would say scaling L1 is good because too many L2s is just too many L2s. Uh

there's like fragmentation. Users just don't know what to do. If the UX was just like completely abstracted that would be a huge win for the user experience but it's not the case. So right now it sucks for developers. It sucks also. It's very bad. Maybe just scale L1, increase the gas limit, increase the contract limit, make the ERC20 uh native, so no more contract, just add it to the EVM, adopt new up codes like I like 7702. So EOS can be can do like multi go things like that. There are so many good EIPs that are have been there for like maybe five years and no one is trying to

push them anymore because the author just like kind of abandoned and I feel like uh this should be like the first step trying to make Ethereum uh simpler for developers or better and then once they're building something promote the apps because I feel like many people are building stuff but everyone is just like promoting their own stuff. I don't know exactly how to explain this but I feel like there is like uh things are not like correlated between the Ethereum foundation how apps or or dabs are promoted things like that so I feel like there is no marketing from like user point of view like we're not like trying to attract new users

it's not easy I don't have solutions but uh I guess this is something we could be working on and taking like lessons from other chains that are more hyped at the the moment. I partially join everything with you as said guys. Join everything with you said guys. I'll try to be short just in mindful time. So there are a few things I think are important and I would like to see them in the future. So first I would like to see the layer one as a stronger piece again for the visibility reason for the security reason and for a few others. The second one for the two what I want to see is that protocols take more seriously what and how the

yield to avoid monopoly to ensure that it can be used. The second one I want to like come closer closer in the corporation definition finance because for some protocols for some products it will be impossible just to do you know read the defy. So they need to come closer and to curate the FI to ensure that actually DeFi is moving forward that there are more users coming on the market and it's more well known and trusted by um also users who are coming from the digital market. So I say more globally I don't want to focus only on

the what we have right now but I'm thinking how can we scale what can we improve and how can we become better. Thank you. That's all. Yeah, this is a lovely brand. I really enjoyed it and I think we're out of time. So, if anybody has questions, you can find them and you know, you can ask them personally. So, yeah. Thanks for doing this, guys. It was really, really fun. Thank you.