From App to Infra: Is This How DeFi Scales? | Sam MacPherson - Phoenix Labs | Paul Frambot

[Music] of deposits and uh we started as a nap optimizing Aven compound and providing better rates and now we have a NRA which is a morha blue a ling and Bing protocol on top of which you can rebuild any lending app essentially hi I'm Devon Walsh I'm the executive director of the Unis poop Foundation we are a Grant making nonprofit focused on supporting the growth and sustainability of the unisa protocol and ecosystem uh today I'm excited to dig into the next version of Unis swap which transforms Unis swap into a true uh protocol platform yes hello I'm Sam mcferson I'm a co-founder and CEO of Phoenix labs and uh we are active contributor to spark which is a lending Market uh attached to Makers liquidity to offer the uh best uh die borrow rates in Def awesome thank you everyone so to get started on the topic I think there are a lot of different ways that you could address this first so I think it would be good to start by defining what do we mean by uh an application or defi app becoming more like infrastructure so Paul would you like to kick off the conversation there yeah sure so so the way we we Define defi as apps and as infrastructures like defi apps are directly used by Def and users while def infrastructures are used by Builders by developers right and so like the the thing that triggered this talk essentially is that in the very like uh early days most dii protocols you know were directly usable by users like you think of unisoft V2 like maker uh a compound Etc and and the general like mental model for this talk is that we're moving from those apps to infrastructure where you have like more make a core with subd Dows being the apps you have unisoft before with like Hooks and entire dexes that can be built as apps on top of uniso before you have Moro blue and on top vaults that can be entire Landing protocols on top of this uh infrastructure yeah and just to build off of of Paul's point I think the the note around developers truly becoming a first class citizen in the in the protocol um is is truly what differentiates like an app protocol from a platform protocol and in the case of uh Unis swop V4 which is the next version of Unis swop coming out later this year you really see um the developers who have always been building on top of of Unis swap developers who maybe for v2 or V3 may have built an an integration um have much more flexibility in what they can build and at the same time um be able to focus much more narrowly and specifically on the Innovation they want to introduce uh worrying less about um you know the security of an entire entirely new amm that they have to build to try and innovate um without having to worry about business like bding doing business development to get users to come to use their specific interface they get the distribution just by virtue of building into uh uni V4 as a first class citizen in the in the protocol yeah you see this trend across uh all of defi I think um as Paul said with uh or you know back in the day it was more user focused um as one of the oldest dii protocols um maybe some of you remember the old interface opening up a CDP and minting uh single collateral die um this is very like user facing but um these days especially with the restructuring under the endgame and uh subd Dows which are kind of more like uh user facing uh mini versions of maker that are oriented towards uh growth and development um maker core self is more just becoming a wholesale uh bulk lender to these subd Dows and then these subd Dows uh spark being one of them uh that's probably the most wellknown uh can then focus more on uh the end user um it's it's it's similar I wouldn't say it's exactly the same but there's a lot of similarities between central bank and Commercial Bank relationships where the subd Dows are more the uh commercial and maker can benefit uh from uh the scale of this whole system okay that's great so what I'm hearing is you're taking these protocols and you're aiming them more at developers instead of retail and you're unlocking new possibilities what are the benefits of doing so and how does that ultimately help defi scale um and I guess to scale in multiple ways right there are different ways it could scale to more users uh unlocking more Capital things like that so anyone who wants to jump in here please feel free yeah for sure I I can uh take that I guess so like maker I would say previously was hitting a lot of uh Growing Pains um with the Dow and so this giant monolithic uh structure trying to do everything trying to do uh real world assets trying to do um you know uh defi lending um all this stuff in one there was political factions forming within the Dow and um so like it just this is not going to scale up so as part of the uh endgame restructuring with these subd Dows these subd Dows can specialize in different things so the spark subow is focused on defi integration and specifically that we are not tooled up for traditional Finance you know that's an entirely different set of people our the people within spark are more like smart contract developers Auditors that kind of stuff whereas in the uh real world asset subd Dows which are you know focused on mostly investing into T bills at this time um they are like more like lawyers traditional Finance people and so when you Silo these things out and then you have multiple uh redundant versions of these you get this sort of uh competitive special Iz nature and um maker as a whole can then scale up uh through this process yeah I think to to piggy piggy back off of that a bit I am extremely excited about unisp V4 from uh the perspective of just someone who exists within the crypto space I think the the design of V4 specifically will allow for a much faster path towards uh more Innovation it will allow for much um faster cheaper iteration upon different ways to scale and make defi more flexible and able to suit a number of different users and use cases and just to to throw out a few different examples um for for one as I I mentioned a bit before um to innovate upon amm design today you have to build a new amm and that that may entail you know just Fork forking Unis swap using some of that code but innovating depends upon tweaking um the curve whatever it may be um that is extremely expensive to audit um and introduces a massive risk area in terms of the amount of complexity being introduced Etc um it also is very expensive and timec consuming to bootstrap liquidity to understand how your Innovation actually works and and functions um to see if you know it's it even works how you think it might work um the distribution element is another one being able to Leverage The distribution embedded within a protocol which already exists and has connections with routers and aggregators and and lastly more in terms of the um flexibility that's unlocked uh hooks really enable developers to um enable a path to be composable with many other Primitives that exist within the space and two that we've seen that I'm particularly excited about are incorporating flash bot Suave uh order flow auctions directly into a hook at the point of a a swap or liquidity provision and um ZK uh co-processors which allow for really mind blowing in ability to customize swapper uh swapper um experience based on you know uh history and and things like that um for me I think moving from app to infra enables mainly four things which are security efficiency scalability and network effects so on the security side of things like as as Sam was mentioning on the maker side of things when you're trying to do everything at the same time well you know you introduce more smart contract RIS and more economic risk and that's been the case in like almost every single vertical of defi like and in our case in lending you know we've been you know building for two years on top of Landing pools right on top of Ain compound and you know we found that you know building on top of what was supposed to be like the infrastructure but that was like upgradeable like the code changes all the time that has more than 700 different risk parameters that are changed you know real time uh that that made it like very hard from a security perspective to build on top now in the efficiency side of things there is kind of the same idea when when you're an app and you're trying to do everything at the same time sort of like you have to lower your efficiency in order to be able to tackle all those use cases in like one specific instance um then on the scalability set of things like you know in the case of unisoft before there are so many many different things you could do with an amm and you can't Factor everything in one and so now you have this sort of like permissionless creation nature that lets you like anyone build and innovate whatever they want without limits uh and then most importantly so like the the infrastructure layer is like focusing on providing this just very simple very secure and very efficient you know like primitive whether it's table coins whether it's Dex whether it's lending uh and and and the most interesting side of this is the network effect and the the state that gets aggregated into the instance so in the case of make a core like the the network effects of D is like huge right and like it makes all the D was benefit from it in the case of unisoft before you have one smart contract that is going to aggregate all the state and all the liquidity you know they'll benefit from Integrations on top like you know maybe unisoft X from unisoft Labs uh or like Aris finance that is building a bunch of things on top and in the case of lending in the case of moral blue so we have those micro primitive lending markets on top of which you can rebuild any Landing pool whether it's like the avv V3 compound V3 like you know compound V2 like model uh but the interesting thing is that if you have multiple vaults building on top they all benefit from a shared state so shared liquidity in that case to the extent that they share risk and that's like uh I'd say the superpower of the infrastructure is that it's like this aggregation of the state unlocks levels of liquidity and efficiency that I think will be second to one when you compare to traditional Finance at scale yeah that's fantastic really good to hear about I think we've really explored in depth a lot of the stuff that's happening behind the scenes next I want to shift the conversation more towards what this means for the end user the people who are actually logging on and going to Unis swap or maybe going to spark protocol and interacting with those apps how will these improvements affect them how will it make their lives easier yeah I can I can take that one first I see two core benefits that come for end users from Unis spot V4 one is better execution for everyone um which is great right um I I think some of the hooks that we've already been um seeing developed our Focus for instance on either Mev mitigation or redistribution to to end users um in Istanbul at at devc connect we had a hackathon entirely focused upon Mev focused Hooks and saw some some really exciting examples for instance a hook that would auction off the first transaction in a block for for instance and redistribute the the proceeds from that to the swapper directly so hooks like that can be incorporated across all all pools as a hook um and the second category is more um specific uh specific experiences built into a hook for specific use cases users or jurisdictions so one kind of hook that we think will unlock growth to the more Trad world are um for instance ofac compliant hooks which may be controversial but for sure people are already kind of thinking about those kinds of of hooks um on top of that another example are hooks which boost LP profitability for instance lending out outof range liquidity um on a or or Moro um uh so so benefiting LP so I think those those two areas just you know better better experience for all users and then customization as well yeah and I would say like for defi we have a long way to go the average user is still using us uh centralized exchanges um just because the ux is like so much better so on this front we have to be pushing on like all fronts here like um you know wallets uh infra you know more user facing stuff in defi scaling you know uh the scaling of the um infrastructure of like the blockchain itself it's you know this is the big focus of ethereum recently L2 scaling and all that so we really got to push this on all sides and where the superpower of defi in is the composability so we as the backend infrastructur say maker and others um having that in place so that uh developers can come in and they have access to this great resource which is uh say di liquidity and then being able to build these like uh killer uh user experiences within the apps in defi we need this and the composability is the superpower of defi we get this sort of exponential uh improvement from all fronts as everybody's working together and it's not these siloed uh exchanges and this is um you know this is kind of the dream we want to have the user experience of the centralized exchange but without the downsides that we saw in 2022 most infamously with FTX and others right and even add on this that it's specifically the separation from the protocol and the app that enables apps to specialize and to become you know very good at targeting one specific use case and you know en enabling for example in the case of lending like you know very specific risk profiles or very specific like user experience like taking a leverage or getting financing to you know buy a house those are every single one of those are different user experiences and once once you separate the protocol from from the the product well you can build specialized product that you know finally reach parity with web to and like traditional Finance or centralized uh Finance while benefiting from like aggregated State and you know composability which makes it strong ER than the you know Web Two centralized uh Finance world so do we think that perhaps in the future there'll be like a bifurcation between protocols like Unis swap or madal where there will be on like a bottom layer and then there will be more protocols built on top of that like I know you mentioned arakas Finance before as one example do you think that there will be sort of like a definite separation between those two types of def5 protocols will that start to emerge so yeah yeah maybe I can like start on this one like the way it's uh it's done in in morphos so like for those that are familiar with with a and compound like the way lending has been working so far is like doing multiple things at the same time notably lending and borrowing but also risk management and aggregating different assets and a bunch of features in in one protocol and what moru does essentially is separating the the the risk management layer from the protocol layer and so in in this model yes things get specialized like you have the protocol Network layer which is like extremely simple on top of which you can have like Risk Managers or or lending protocols like build their own vaults uh to have their own like lending specific uh experience and in that case uh things get specialized and they're able to you know like generate their own fees at each layer and it's much more like in incentive line so I think yes in that case you'll have like clear separation but you know in each areas of Def I guess is going to be slightly different Paul I would love actually for you should talk through the gauntlet um announcement you just made I want to talk too but I think it's a really interesting yeah no that you're right like I I didn't mention it but essentially like the way in the case of lending like you know uh gaunlet used to operate gaunlet is a you know one of the best Risk Managers uh in in the space and they've been for a long time working as consultants for you know Dow like Landing protocols like a or or or compound and and you know like when we started talking to them about this new model well in in the Morpho paradigm like they become the app right they instead of being Consultants they become you know uh creating creators of the the risk of a vault and so they're now able to make fees directly from the usage of their Risk Management Services not as consultants and this has proven to be a much more you know economically efficient uh business model for them and you know you probably have seen it like in in in the different governance forums recently like they decided to depart from the Avid Dow to build their morph volts and uh yeah and so there you know perfectly on point for the topic of today which is like you have the Morpho infra on top of which you have Risk Managers building their own pool like gaunlet for example but others like block analytica for example uh and I think it's important to highlight the um opening this up to a Marketplace of risk providers is super important here because otherwise you kind of have these in uh entrenched uh sort of like you know maybe there could be improvements I'm not saying anybody's doing a bad job at the moment but you want these to be marketplaces such that there can be room for Innovation and Improvement on the uh risk uh management front so I I really uh do like the structure of morol blue how they've done this yeah just to chime in from the amm perspective I think that the the competition theme the ability to to open up to competition in terms of risk parameters I think for for UNIS swop you see that you'll see that you're already seeing it if you go to Unis swop hooks.com there's more than a 100 hooks which have already been built to get nice array of how people are thinking about this one thing I'll I'll call out is you able to actually innovate upon different curves different um demand curves for for assets that is like one very very clear um uh manner in which different hook developers can compete and um that competition will benefit all all users across all types of assets um on top of that I think the the original question was about protocol develop in are new kinds of protocols going to be developed on on top of um you know these protocol called platforms um the answer is definitely yes but my team a few months ago when we last counted were talking to more than 20 teams considering you know developing protocols embedded within hooks themselves and again the the kinds of protocols that folks are are thinking about go from orderflow auctions that are built into um things like Suave and and maybe outside of of swaave built on um other other kinds of uh incorporating other kinds of Primitives to um um uh well there's a there's a wide variety out there that that is going to be very exciting to see so it seems like um for example with makad Dow you've gone the route of setting up these subd Dows that will build protocols on top of the main protocol Unis swap doing things a bit differently working with separate teams so to take a more I guess like futuristic look longer in the future do you think there will ever be a point where protocols like make out and Unis Swap and mor as well will become abstracted away to the same extent that maybe stuff like SMTP which is backel email uh you know a lot of email Communications today people don't even realize they're using it um I guess the other point to that is unis Swap and everyone else they have a really strong brand identity as well so it's kind of I imagine difficult to balance those two things what do we think I think absolutely I mean I think we see this already so uh things like the DI savings rate um already we see uh L some l2s launching where they uh just they have this intrinsic yield um in their stable coin Supply and they just like plug into the DSR and like the end user doesn't really know the details and um so we see this like all over the place and I think this this trend is just going to continue where the backend infrastructure is providing um you know all the uh usability for the um protocol on top of it but uh the end user you know as long as they kind of uh Trust TR the Judgment of the intermediary protocol um about where to allocate assets and stuff like that um yeah you'll see more of this like abstracting away uh the underlying infrastructure yeah from Unis swaps perspective I'd argue that Unis swap in many ways is already abstracted away Unis spot pops up is already integrated into openc friends with benefits like tons of interfaces that allow for just easy easy swapping so definitely see that that Trend continuing um as as more users on board yeah and you know I even add that you mentioned SMTP and HTTP and I think that's absolutely right like the direction where all of this is going that we're rebuilding Finance exactly like the internet right with those you know base protocols that have all the network effect on top of which you have different layers like the risk management layers the product layers on top Etc and each layer has you know very competitive and open markets like we mentioned the the risk markets and by the way like you could tell it like right now on Moro like the fees charged by racek managers on vaults is so much lower than what is charged in like in current Landing Landing pools right uh and so you could already see the benefits of this but also like the fact that you could you know from a same base a opinionated protocol reexpress different you know apps with the same way you know on SMTP you have like different clients right that sort of like give you a certain expression uh of the of the uh of the mail the mailing basically and you know they even complement the protocol with extra features with encryption etc etc I think they are definitely a parallel to be made with internet except now we Transit not only just data but data that has value in that case you know because that's enabled by the blockchain and so that's why it's super suited for for Define general so yeah 100% so to focus on the second part of the question as well does this cause a conflict with the strong brands that each of the protocols that you're from have created and nurtured that resonate with a lot of you know def native people um but if defi is to scale to more and more users those Brands will get abstracted away is is that something that's a conflict or do you think it's something that isn't you know really worth yeah I do agree there's a tension like like if you think of Unis swop you know unop is very unopinionated you can list any asset right so some people are listing scams as tokens right and you know to some extent the Unis swap brand can suffer from it if you're not a user that you know that you had the front end that was properly you know abstracting this away for you and the same thing for Mor BL you know you you know the day we deployed Moro blue we you already had like scam markets being created right so uh I guess it's uh on the app player to make the difference and what's allowed what's not allowed and even for compliance right like if you think of emails you can't send whatever you want through emails right that some things are prohibited depending on your jurisdiction right and it's on the side of the clients to sort of like you know make sure like uh uh you know certain rules are respected or or people don't get scammed etc etc and um I think inevitably it will take a little bit of Education to basically tell the user oh you have to make the difference between like the protocol and the risk decision that you're taking uh and you know I think this is just like general education because so far you know protocols like having compan have you know like users are getting used to having very risk-free experience where they get to was it without thinking and and you know assuming that everything will go well forever but you know I think in in the Paradigm that we're talking about it will have to be abstracted away through different layers okay thank you very much that's all we have time for um and thank you everyone for coming tuning in um and thanks to e Denver for organizing an amazing event thank you