ZKPodcast: Exploring Uniswap V3 and a Multi-L2 Future with Noah and Moody

Transcript

[Music] welcome to xero knowledge i'm your host anna rose in this podcast we will be exploring the latest in zero knowledge research and the decentralized web as well as new paradigms that promise to change the way we interact and transact online in this week's episode tarou and i chat with noah and moody from uniswap we talk about the history of uniswap from v1 to v2 the learnings along the way and how v3 aims to add more flexibility and options for liquidity providers we touch on sushi swap what operating on

multiple rollups simultaneously might look like and more but before we start in i want to let you know once again about the upcoming zk sessions event happening next week in this session we will be looking at dows and nfts and how we may be able to incorporate zk or privacy into these new systems i will also be hosting an interactive community discussion about if and how the zk community could potentially start using some of these tools as well event happens on june 23rd and i've added the link in the show notes hope to see you there i also want to thank this week's sponsor ave ave is an open source decentralized non-custodial liquidity protocol on ethereum with ave users can participate as depositors

meaning they can provide liquidity to earn a passive income or they can act as borrowers to borrow in an over collateralized way or an undercollateralized way think one block liquidity flash loans the ave community has unanimously voted to introduce an ave grant dao to grow the ave ecosystem and if anyone's interested in building on ave or learning more be sure to go to avegrants.org we've added the link in the show notes so thank you again ave now here is our interview with noah and moody from uniswap so today tarou and i are chatting with noah and moody from uniswap welcome to the show guys hey thanks for

having us anna yeah thanks for the invite so primarily we're gonna be looking at uniswap v3 but actually this is the first time i have anyone from uniswap on the show and so i think it's worthwhile for us to do like a little bit of backstory about unit swap i should say though we did last year do an episode about sushi swap which i can link to in the show notes where we definitely talked a lot about you swap but yeah this is the first time i actually get to talk to you guys about it so let's do that let's start what is the history of uniswap where did this project come from how did it become what it is today sure happy to be a primary source here so unisop started way back when uh before i was even around actually with

hayden of course really just riffing off of an idea that um had been sort of floating around in various forms there's you know we want to make sure that we credit the appropriate people and i think that there that's been done in twitter threads and things like that but you know the the most i think relevant motivation for uniswap was a blog post about automated market makers that vitalik made on each research you know two or three years ago at this point and hayden sort of ran with that with a grant from the ethereum foundation of about 100k which i think mostly just paid for an audit and maybe some out-of-pocket living expenses well while hayden was doing the development and that was how v1 worked you know i sort of had read about that after it had launched without having any insider knowledge of the process or

really anything that had led up to it and i think that's how most people actually um who weren't involved in those very very early days sort of heard about it they maybe read a blog post and they thought it was some sort of wacky sort of fun idea you know they thought it was like bancor right which had which was sort of a historical precedent and you know i didn't think too much of it until i actually met hayden in person at a meet-up and he was you know just like someone who you'd expect to have phil you swap right he was super in the weeds very passionate really really excited um and just like brimming with ideas of what he wanted to bring to the table next um and what he wanted uniform to become in the future and so i think i was very very excited by that

and you know that that was really those were again the very very early days you know d5 wasn't even really a term back then or it was just very it was in its nascency and so you know i was sort of in this circle and you know he sort of needed some help had some ideas about what he wanted to do next and i thought you know this is this is maybe a great opportunity so um along with cal who is the sort of design guru and and genius behind the sort of uni user experience of uniswap you know we were the really early team members and we sort of worked on just these really fun early ideas including unisocks and of course v2 which introduced oracles and flash loans and all kinds of things but you know without getting too much into the weeds really that was that was the

genesis it was an ethereum foundation grant um which transitioned into series a which transitioned into sort of a full fully fledged company which is which is what it is today and so um you know that's a bit told from my lens but but that's that's how i saw it what were you doing before uniswap you said you were sort of around but what were you working on yeah so i actually started my sort of career quote-unquote at the new york fed so i was doing economics research in preparation for a phd which i very quickly realized i wasn't cut out for and so bailed which is the best decision i've ever made um and i actually worked with uh andy turlian who who was a guest just a couple weeks ago i think on on your podcast and yeah um yeah we were we were we were very close

friends um at a small crypto startup in in new york and we had a great great team and um you know that just sort of didn't work out for a variety of reasons and i was you know at that time i was sort of going to meet ups and you know things were a little frothy so you know it was always it was always sort of fun to see what people were building and um try to cut cut you know cut through the the chaff and find the wheat um and things like that that was 2017 i guess when you say frothy yeah but one thing that's funny is i think that uh that company has produced a lot of founders in d5 in the last uh three months i feel like it's uh yeah it's actually kind of crazy that you guys worked at like probably the

best little like d5 dev incubator that i can think of in spite of the fact that was it called hydrogen hydrogen did it ever launch i guess you had a token that's right yeah it was just sort of at this it was at this sort of awkward in between stage where it was trying to appeal to the sort of open source what we now call d5 ethos and um sort of a more corporate version of of you know a fintech app and it just never quite hit the right balance i think and ultimately i think there was just you know it just didn't really work out in that sense and so um it is funny i agree that like there's you know there was this big exodus and you know we've since gone on to some very interesting projects in the space so cool all right moody where do you enter

this story at what point did you join unit swap yeah so i i joined about i think a year after noah mike was actually trying to recruit me for rainbow so he invited me to a tweet up and i met hayden there and i was also trying to get hayden to come speak at google about uniswap with the blockchain group so i heard hayden was hiring at uniswap and i kind of forced my way into the office for an interview and uh yeah wait who is mike in this story uh mike is the founder of rainbow the the wallet okay and he uh he he's very social very very fun guy um so he throws these

tweet ups in new york with all these crypto people and all these these twitter people which i never experienced before so that was a lot of fun did he buy your ens domain does he have moody dottie no i actually i front run him on that so he might have sun moods now which i'd be a little sad about a little a little background on this and as mike's entire growth marketing strategy for his startup is to buy both engineers and famous celebrities.eth domain names and squat on them until they use his app which has worked crazy that he releases it if they use the app like mark cuban.eth for instance he bought that and he got mark cuban to use

his app and now he's investing amazing well it's genius it's supply side and demand side too right because he's got the celebs who have the clout and then he's got the devs who are like well this is what is this you know and then they get hooked amazing that's so good moody what were you doing before what were you doing before this meeting yeah i was at google doing privacy and ads and sort of in the meantime i was like studying crypto and i have no idea how mike found me i had like 30 twitter twitter followers and and yeah he's just he's a good scout i guess um i had been going to like hackathons like east new york uh east denver east waterloo and just getting to know people yeah building a bunch of silly projects like i'd say cool we have now your two start dates so

2017 was noah 2018 2017 i think 2018 was the official date i believe but okay something something around then and 2019 i guess was moody well i think i yeah i started a year ago so 2020. okay 2020. yeah got it let's let's kind of go back to the story of uniswap we've done v1 that very kind of simple mechanism this this idea that had been originally proposed by vitalik and maybe also by martin we had martin from gnosis on the show recently and i think it was like a conversation between them um what happened after that like let's let's continue through that story of this this company in this product well and if i'm being honest we spent four weeks

uh four weeks four months building unisocks um that's exactly what that's what you did okay you so you literally paused everything and built unisocks um which you know i'm look i'm sort of somewhat self-trained as an engineer i don't have a cs degree and you know i had been working on smart contracts almost exclusively and wow i get hired by hayden and the first thing he's like hey what if we build some sort of like web ui to interact with you know these tokenized socks on unison v1 i was like well sure i don't really i haven't done that before but that seems fine we built this whole redemption scheme and you know for those of you who don't know unisocks are if you don't know unisocks you know you're probably not listening to the show but um you know we tokenized 500 socks and put them in a user pool and let them

be bought and sold freely and you know you could redeem them on our website for for an actual physical pair of socks which we are still to this day safeguarding in the unisop office so it was it was a lot of fun it was actually great because it really showed people what unisop could do and those are the kind of things that hayden saw in his head um but needed to actually bring into reality for others to be able to buy into to what you know to that essence of what uni swap made uniform interesting to him unisocks this is like were they nfts were they the 721s kind of nfts sort of so there's there's an nft element to unisox but the really the core unisock is it's just an erc20 contract of with a total supply of 500 and so they're fully fractionalizable

i see and you know it just takes one full you know denomination of a unisoc to redeem it for for an actual sock uh two you know a pair of course a left and a right uh oh yeah but once you've once you've redeemed your erc20 which is fully freely tradable you do get in addition to the physical stocks a non-physical non-fungible nft token and so those exist as well and they're exactly corresponding to how many nf nft socks have been sort of burned what was the connection though like it sounds like a super fun project and i get it's good for branding and stuff but like could you use these in uniswap as it stood were you able to actually like trade yours i guess if they're erc 20s you could yeah absolutely and that was you know the sort of algorithmic pricing was what

was very very interesting to us about about this experiment and i mean they're the most valuable pair of socks i think ever created which is sort of just a funny little meme and you know it really just i think it it it showed both the sort of mass market appeal the uniform could have while demonstrating the underlying mechanism in an extremely simple way um and also you know frankly touching on some of the pain points of of that you know first v1 design and you know we've seen i guess we'll get there to v3 but you know there's this capital efficiency that v3 brings to the table um and we just didn't have that in in v1 and what we did is we actually had to lock up a bunch of eth alongside our socks to sort of um seed this liquidity pool you know to sort of support trading and

so there's there's there's very interesting tie-ins to unisox that go much deeper than sort of like a one-off little project very cool this is still with v1 was this still like being built in tandem with v1 well this was post v1 launch and so we were it was just we were using the viper contracts um it was sort of we were still v2 isn't you know in sort of ideation phase and we were just sort of riffing on on some fun things cool so when was v2 and what did it do so v2 was about a year after that um so i think you know late 2019 maybe um early 2020 and the big there were three big ideas the first was oracles the second was uh non-eth specific pools and the third was flash loans so

unicef v1 worked only with eth so v2 opened up the floodgates and allowed any token to be paired with any other which is you know sort of incredible to think about because that's just taken for for granted these days right yeah um but yeah that that was that was a that was one sort of high level feature um oracles which which are very interesting to me i think still underexplored in the d fight and broader crypto ecosystem um introduced a way to sort of trustlessly and with certain known risk parameters obtain the price of assets that are trading over uniswap which can be used as a primitive to to input you know values and data into other systems which rely on on these you know relative prices and then the third of course is flashlights which

uh it doesn't always have the best reputation but flashlights are this feature that unisof v2 pools have where you know you can come to the table and say i'd like to borrow all of your assets you know uniform pool and then i'll return them you know in the same block so it's totally safe for you but um you know don't worry about what i'm doing with them and basically you know that that can be used for sort of ill effect if the the people borrowing assets are willing to do something sort of unsavory but um i think there's a really positive narrative to flash lens and the reason why we introduced it really is because it allows uh no capital arbitrage and so really unisop is about democratizing trading and finance and lping and so um i think being able to arbitrage you know without

any upfront investment is extremely powerful and basically how that works just post-it no version is um you borrow a bunch of assets that are relatively overvalued from uni-swap you sell them on an external you know venue it doesn't have to be off-chain on chain it could be off-chain um and then you you return the now correctly appropriately priced assets you know that process is normally only available to sort of professional market makers but this flashlight feature made it available to everyday people which which flashlight came first aves or uh uniswap i think ave we certainly didn't originate the idea i can't we can't claim credit for that but it does work quite nicely within the unisoft ecosystem of you know like making a swap and sort of simultaneously returning assets and things like that i think the the flash swap was that originally from uniswap

the flash swap yes that's that's sort of our coinage of it and yeah i think that was a that was a novel thing because you know with ave it's i think it's you know exclusively single asset but but unisof has this interesting dual asset property where yeah you can flash swap optimistically and receive your output before you're actually required to pay your input were there like these lp tokens on v1 and v2 or was that sort of an introduction with v2 yes the lp tokens were existed on both and they are okay yes they were just a you know a way to track um relative asset values and so you know we needed that for accounting purposes got it so moody you were joining 2020 was v2 already out i think i think we're about a month out from v2 launch or a month before v2

launch when i joined a lot of the first work i did was just like getting into the interface and modernizing the code base i did a lot of conversion to typescript in the early days and some unit testing and yeah just building i actually wrote the first naive routing algorithm which is still unfortunately a little naive in the interface uh for for the v2 swap since we had not just e pairs but we also had token token pairs so it got a little bit more complicated than v1 cool okay so i think we've set it up nicely to start talking about v3 which just hit mainnet on may 5th a month ago i guess so this is pretty fresh i'm excited

to actually dig in on this you know v2 was out at what point did you start planning for this v3 and what were kind of the initial kind of challenges or things that you wanted to fix with this new version things you wanted to change right well moody maybe you can speak to the technical aspect but i think just to tie this back to unisox i think something that's pretty illustrative to people is basically what we wanted to do is is to say you know we wanted to support a unisox style token issuance uh without having to lock up this east that we knew was never going to be sold below the the starting price of stocks right because we control the total supply we have this guarantee that um there's sort of a floor price a reserve price for socks and that they're only going to

be sold on the way up and so um you know unisof the 1v2 just didn't allow that kind of use case because they were constraining liquidity providers to a single shared position which was you know behaved exactly identically economically in terms of risk in terms of required capital in terms of fees earned and so the over you know the underlying motivation of v3 is to introduce a level of flexibility in lping and allow users to customize their risk preferences and then their expectations over future prices and so that's really the core idea in innovation we can get into how that that works but that that's what that's what it is doing and i think the motivation for this really began much much earlier on than v3 at all but v3 specifically began to form as a coherent idea and

as something that could actually be implemented basically right after v2 unfortunately and so we didn't have too too much of a rest period after v2 launch but you know that was when sort of dan robinson from paradigm began to really flesh out the math here and and and figure out that this could this could work and hayden sort of began pushing and and prodding and seeing you know what what strings could we pull together how you know how could we start staffing this things like that tell me a little bit about like what did happen under the hood what changed it feels like it's pretty major well yeah from a technical perspective uh you don't just have like these two token balances from which you can determine everything you now have to have this price curve broken down into these things we call ticks so like

all the different uh discrete prices are sort of like bucketed into these ranges for which there's different amounts of liquidity that are staked in supporting trades in those ranges and this like timing wise for this i'm trying to picture like a lot of this work was happening i guess before sushi swap even was on the map right it was certainly in the works um but before that yes and and frankly we were sort of not focused too too much on v2 because we felt that the core reasons why v2 had come about were sort of thriving and and it was its own self-sufficient system and we we had sort of bigger plans for what we wanted to work on next i am curious now though like what happened the day that sushi swamp emerged in the company how are you guys feeling what happened

was there like a what the hell moment like the music executives in 1999 when they discovered napster or what it was you know i mean i think that we you know there were mixed feelings because we felt that you know uniswap was an important thing to exist and so we were happy in some sense that right that you know the code was being put to good use and you know we open source unisoft for a reason right and we we have gpl code for a reason and so you know from that perspective we didn't you know we didn't sort of really i think notice too much until you know until the numbers started cranking up and um you know ultimately we were actually quite busy and so we sort of didn't want to devote any time to this what sort of felt like a distraction and

you know there's an important element um at play here which is right that like in an unpermissioned fully sort of adversarial ecosystem you know anything can happen at any time and so i think we we did react as a company right and we we sort of had to you know adjust our expectations about um about what people wanted from unisop or from something like unisop and i think it actually taught us a lot about you know the community and and these there's really important like foundational you know things at play here like you know venture capital versus crowdfunding versus um governance and you know all these things are very core to what crypto and and and what it means to build a crypto company is and so i think that basically it taught us a lot it forced us to introspect a bit more about why we're building

unisop and what we want unisoft to mean to different groups of people and so i actually think from that perspective and i you know i don't want to go into really details here and you know i think the history is sort of you can you can look up what happened and all that but i think i think what we took away from it is we need to be intentional and involved and engaged and and you know just make sure that we're building things that people feel involved and engaged in that's what i got from that experience cool if you do if you are interested in the details i would say there definitely was a re-prioritization under the hood um when sushi sort of got big and gained a lot of traction we sort of shifted our engineering team just a little bit i think towards like governance and community

work did the uh like so i'm just thinking all of this history is on the way to the launch of v3 but was originally the plan to actually include because there was the airdrop that happened in september but i'm curious like was the idea originally include the token in v3 and then that was moved up or was there always a plan for tokens or was that something that actually everyone was like oh yes we should because like from the outside it wasn't completely clear there had always been sort of the no token ethos right it was actually always cited as the example of the project that doesn't have a token actually the episode that we did we did this three-part series on sushi swap and in it we did the whole episode the day before you launched the token

so then we had to record sort of a third part because it changed the entire narrative in a way once you guys had dropped one too so yeah i want to hear a little bit about that like was that always in the works was it sort of pushed up yeah i'm really curious what was happening yeah well i can't say too much and i i don't want to speak for for hayden in this sort of official company position i think it certainly moved the timeline up i think that there was always you know we always had the v2 fee switch right which which was known to to people and i think we were pretty transparent about what that did and you know who's going to control it you know we never wanted to sort of have this as a company um we just felt at the time of launching v2 we weren't

ready to you know concretize what it meant to you know sort of govern you know this aspect of uniswap and so um we sort of punted a bit right and i think you know obviously it's hard to replay things right but i think yeah in retrospect if if sushi hadn't gained the sort of traction that it did i think we certainly would have been a bit more thoughtful not not thoughtful about what we wanted to do but thoughtful about the the timeline and the the way that we prioritize from an engineering perspective um the launch of what what is now governance right and the uni token and so um we we potentially could have coincided that with the v3 launcher or just done things in a slightly less frenzied uh timeline from from an

engineering perspective right there's a lot of a lot of code to get written when you're launching these kinds of things and a lot of tests to be written i think that we we don't feel that we've sort of betrayed anything fundamental about unicef biological token you know there is this no token narrative and i think that like the original meme was that you know hey uniswap was just bank or without bnt little did people know it's it's actually it was much better and more gas efficient but uh but you know no one sees uni as like an intermediary like routing asset for the protocol the protocol is self-sufficient it stands on its own i think that there is this interesting conversation to be had about um the meta level of governance where you know you're thinking about incentives and incentive alignment and moats and there's a point at which liquidity

providers need to coalesce around projects and and sort of think about not just current profits but you know future profits and i think that there's a really interesting angle where multiple sort of fragmentation across multiple decks creates um exponentially many arbitrage opportunities which you know furthers mev and i think that you know ultimately there there have to be both incentives to build and have stable ground upon which to to feel that you can invest and and write you know integrations for and so i think that the token is tied into all of these sort of areas and very intricate delicate interesting ways and so i think that we see it as this like meta incentivization governance layer for a core protocol which ultimately is agnostic too to this to this layer got it

let's go into the price curve ticks the ranges there's a few kind of words that are used in the new documentation that i actually don't know for example like ranged liquidity i actually don't know what that means so maybe you can start to break this down a little bit for me what are these doing maybe on the on the user level what would i see if i'm using this now uh so in unit uniswap v1 and v2 uh when you provide liquidity your tokens can be bought and sold at any price so let's say it's ether and us dollars you you're providing liquidity and anyone can buy or sell ether and us dollars in the pool whether the price of ether is one dollar or you know a million dollars

so you don't have any control over where your liquidity is actually used whereas in unispot v3 you can restrict the range of prices where your liquidity is actually used and not worry about providing any capital for liquidity outside that price range um so let's say you only think that ether is going to be between the prices of two thousand and three dollars you can provide liquidity in that price range and get a greater capital efficiency get a greater portion of the fees that are earned in that price range it starts to sound a bit more like an order book model then like where you'd actually say you know i'm i'm ready to sell at this price and you're ready to buy at that price and

then you match it but you're still an amm right so what what is this how would you describe this this combination yeah i i would say one of the core differentiators is that once someone like takes your order it doesn't go away so they can always just take the other the other side of the order so if they buy something they can sell it right back to you as long as you don't like remove your liquidity um so you're still participating in market making you're you're still buying and selling this like spread but you can limit the range within which you do it and and if you if you do limit it enough like it is kind of like an order book where like if you just limit it within you know point zero zero zero zero one percent

which is smaller than you can do in v3 but like theoretically then it is kind of like a limit order taken to the extreme but but there's still that that case where like if somebody takes that limit order they can just like go in the other direction and give it right back to you as long as you don't remove your liquidity are there some like benefits like so i think you can kind of imagine why this would be useful to people that like you have a bit more control i'm kind of curious what does that look like does it sort of say also that you'd receive a different percentage if you did a different price range like do you does that affect actually how you earn the fees it does so just just very quickly i mean in unison it was very easy to think

about lping because you just provided 50 50 value right so you you just provide equal amounts of both assets essentially in v3 that's that's no longer the case and you can actually take fully asymmetric positions you know in the limit actually as moody mentioned you know that becomes a limit order um and so you can come to the table with you know differing amounts of assets and and in addition to that sort of location parameter which is you know where on you know the price curve whether you're providing you know from prices that are between the current price and higher prices or the current price in lower prices there's also sort of a scale parameter right and so you can choose your location and your scale and the scale is what determines the sort of risk and concentration of liquidity so the tighter your range

right the smaller price movement is required for your range to be exited and right that corresponds to a higher risk of you know what some people call them permanent loss or um you know inventory risk or or what have you and so there's there's two sort of elements at play here location and scale and so i think that they both can affect fees differently but but essentially you're only earning liquidity provider fees you know for as long as your liquidity is actually you know supporting trades and so that only happens when you're when the price is actually inside of your your specified range was the goal of introducing this partly to like mitigate impermanent loss or as we defined on previous episodes like opportunity cost was that something you were trying to do in creating these different ranges where like you could

either expose yourself to more or if you go i guess wider less i would say it allows liquidity providers to like specify their risk preference like their their preference for impermanent loss so like if you have a higher tolerance for it then you might choose a more concentrated range um because you'll be earning more fees with the same amount of capital and if you have a lower tolerance for it then you might do something like you swap v1 and v2 where you you don't ever want to sell one asset at too lower price so you you provide liquidity across a much wider range and actually yeah can you still use it as as you did with l1 like can you still set this setting so that you're

like if you're just really used to the way it was can you still do that yeah so if you provide liquidity across the entire price range then v3 is more or less equivalent to v1 and v2 with a caveat that the gas characteristics are slightly different and maybe what the lp tokens look like are slightly different too right yeah that's correct uh v3 also uses nfts and then there's one more caveat that v3 doesn't automatically reinvest your fees which you know some people might think is a is a bad thing uh i think it's kind of a preference thing because i actually like that v3 doesn't reinvest my fees because because then my fees aren't subject to

impermanent loss hmm before it was an optional to do that i guess yeah before the lp tokens had this automatically reinvesting property where you're actually accruing pro-rata shares of liquidity and not not underlying tokens yeah well i guess one one other thing is like how do you view kind of the landscape of protocols that build on top that will be kind of rebalancing for people right it seems like based on at least sort of what it seems like sentiment is is people kind of want these sort of passive investing vehicles that will like go and reallocate to pools do you envision a world where the original uniswap team is is maintaining things like that or do you envision a world where there's like

500 different kind of strategies each focused on a different subset of the market and kind of how do you think about that because right at the end of the day i think there's still going to be demand for passive investing vehicles and you know there's only a certain subset of the market that can actually be active in managing all this liquidity i i think some of us have different different opinions on how much uniform should be involved in in that particular domain my opinion is that it's like a very opinionated subject like how you how you deal with reinvesting these fees and how you how often you should adjust your liquidity ranges and how aggressive they should be in terms of tightness or concentration and because it's so opinionated there's

going to be many different implementations and people will have different preferences and sometimes one will perform better than the other and then other times others will perform better i think unit swap should take a role and this is just my personal opinion i think uniform should take a role of enabling all these different strategies to exist these automated passive strategies and like building all the right tooling for them so that it's easy to make those um but like building an actual automated strategy i think we should leave up to the integrators would you expect there to be additional products that live on top that actually articulate one of these not just the individual liquidity providers who they themselves are

playing a special strategy but like actually products that play that strategy for you or does that already exist actually yeah i mean i think people are thinking about this exact problem and you know the the really great news from my perspective is that we don't have to worry about it i mean not to say that unisop doesn't want oh you go right i mean not to say that doesn't want to productize this or think about it we do and i think you know we want to support people and and build tooling that that enables this but like ultimately you know it's not we don't have to be a monolith right and you know we can be we could sort of pursue that traditional sort of growth path and sort of try to try to offer um you know a whole range of products that appeal to

you know investors and traders and basically the whole gamut of people use uniform but the really beautiful thing about open source and defy is that we we can sort of put things out there and then um work on a very high level with other teams whose relative expertise might be you know a bit more suited towards you know these these automated strategies and and we can give them the tools they need to make decisions on behalf of their users right and i think that's what is extremely interesting and powerful about building in crypto in general and i think that you know that's not to say again that unisof isn't interested in this or that we won't ever you know come out with some kind of product that does this right but i think what we can say definitively is that our like razandetra is to let people who do have opinions

and who do want to see very specific strategies implemented and give people that opportunity to invest in those or um or hedge hedge their positions or or what have you we want to let them do that and so that that's that's powerful as you describe it and you describe these different strategies for me it sounds more and more like a game that like people have been playing a game that's quite easy and then some people often those with like a lot of funds can play it well but they don't have to put very much energy and now there's like all these different new ways to play this thing and even if you don't necessarily come from one of the more legacy groups you could potentially also play this game so well that you maximize it is that is that actually the case or is because that's what it sounds

like or do you think it's still sort of like really hard for the little guy to like truly participate in this no no i i still think that unisop can be a medium for passive and you know passive quote-unquote investment that's the primary sort of motivation um i don't think that the existence of the ability to potentially you know earn outsize you know apy by being active is at all a fault or a failing in the universal core model i think that you know those positions are almost inherently taking on more economic risk and so you know being compensated for that isn't bad i think and and you know it potentially might even net out in in the long run right as we see these strategies you know rise and fall and win and lose and i think that

the important thing to note here too is that you know traders really win in these scenarios like when people are sort of playing games and i wouldn't i don't think games is an inaccurate characterization right like when you decide to sort of make a big bet on what you think is gonna happen in in a pool um and that doesn't go too well for you right like you know that permanent loss is sort of translated into great prices for traders too so it's like you know we it's you know we're not like tried to distribute you know loss from you know lps to users or anything like that but i think that like the net outcome of this we're hoping obviously is positive and i think that so far the early results seem promising and i think that it's a very interesting very multifaceted complex

you know system and i think that we're you know we set up the rules of the game in a very abstract sort of generic way having built v3 and i think that we're going to start to see very diverse strategies um on top of that and i think that passive lping either meaning just like single sort of range just set it and forget it style or participating in sort of an index fund that you know that has some sort of automated rebalancing associated with it it is still going to be a valid use case one kind of other big change uh was the licensing in in v3 and as as someone forked it and put it on bsc at this weekend or friday i guess ooh how do you view kind of the future of of licensing you know i think the interesting thing of course from a

governance standpoint has been uh this kind of snapshot vote on allowing arbitrom to kind of have usage within the license at least relative to the team's deployment so maybe from the development standpoint how do you guys think about the license and how does it either impact features you want to make or things you want to do and and what was sort of the thought process behind it sure yeah so so just you know on a high level the v3 code base is um was issued under a business source license which was something that was developed by by maria db um and basically what this license is is it's a and it's a time locked gpl right so uh it devolves to gpl after about you know after two years which is specified in the original contract and

before then um it's it's more or less restricted to people who both who either unisoft labs you know the company or unicef governance specify right and so so it has this exemption clause by which um governance is able to make exceptions to the license but you know by default any sort of commercial use of the code is prohibited within that time frame so that's that's the statement of what the license is but you just sort of said that recently there was like a governance decision to actually open that up is that how people can use it so what happened with arby trim is there was some community interest in deploying to arbitram there was a thread on the governance form and then they created a snapshot proposal and because it reached the threshold of

40 million votes which is what's required to get an on-chain proposal passed we figured we'd just make it easier and and deploy to arbitram for the community um since the community signaled strong support and so say say that unit swap unit swap labs the company decided that they didn't want to deploy to arbitram what the community could do is update a reference which the license points to to include a carve out for off chain labs or an arbitrary deployment and then anyone could just go ahead and deploy the unisop code to arbitrim so they could kind of go around the will of the

company if like through this governance mechanism this is due to the license the exact way it's described yep the technicalities of it is that the license actually points to an ens name which can resolve to a file containing additional clauses to be appended to the license there's one more thing which isn't relevant it also has another uns name reference which points to the date at which the code becomes gpl so you can override the date so that all the code becomes gpl on a faster timeline but then what happened with binance with the b with the bsc finance smart chain uh basically someone just forked it and put it put it up yeah what does that mean though like if somebody just forks

it and puts it on pse what happens lawsuits if you find them uh let's see okay okay so like there is basically now the community does have something to protect themselves maybe but it is open sort like yeah so like where are you guys sitting on that well i mean the community has this license to the code so the community can enforce it if it wishes to uniswaplabs also has copyright and i think as far as like what we'll actually do i think that's not it's not really up to the two of us it's sort of a team conversation but but yeah i think from my perspective the license isn't necessarily about enforcement it's

about legitimacy and it's hard to rally a community around sort of an illegitimate legal project which is not respecting the copyright or the attribution of like the work that it's using so i'm not as worried about like the community rallying about like you know like hiring lawyers or something but i think it'll work out just just because you know they're not playing a fair game let's touch base on the nfts as or lps as nfts because you mentioned that before how does that actually what does that look like what are these things and maybe what were they before what were lp tokens originally and what are they now yeah so lp tokens originally were erc20s which meant they were fungible

so if you have one lp token i have one lp token we can swap them and as long as it's for the same pair like they're they're completely equal in value but with v3 because fees aren't automatically reinvested one one user's liquidity position isn't necessarily fungible with someone else's even if it's the same tick range even if it's the same starting capital because you also are entitled to these fees as much as we like wanted to make liquidity fungible for all for all the benefits it has like being able to deposit into something like compound and borrow against it it was just very difficult from a technical perspective to to make that happen um so like the the

easiest thing to do was to make them non-fungible so that means when you create when you provide liquidity to universal v3 which you get back is this non-fungible token which represents your position and the parameters of your position so what is the price range in which you provided liquidity how much liquidity did you provide yeah but worth like given that there was a lot of projects that use the old lp model like what does the new nft model do with those and i have an example here like maker using uni pool lps as collateral yeah maker maker was a little sad when we told them because they had just you know because their their governance system is you know it's a bit more conservative it's a

bit slower right and they had just before we announced v3 they had just gotten around to like you know starting to to get v2 pool tokens into their system but um it's okay we're we're still super excited about the long term of that general phenomenon it basically just makes it a bit harder right yeah it's sort of a speed bump you know ultimately there's nothing preventing a team maker you know someone else from you know choosing a method of making lp tokens fungible um basically the core problem is right more or less um dealing with fees which are collected asymmetrically as moody mentioned and so you know if you're willing to sort of commit to a strategy for um reinvesting these fees you you can actually make certain positions fungible and i think that this is particularly interesting and relevant in the context of psm and not to not to like get too

sidetracked but yeah it's basically this like uh program that maker has um dumped a bunch of collateral into that lets you swap between dye and usdc and you know it's a pretty powerful sort of backstop for um for these trades and i think that there's a really interesting you know reparameterization of that system that could happen on v3 um and i think that a tightly concentrated v3 position um in die usdc around the price of one could i think serve a lot of the same purposes as the as the psm does right and so i think that there's you know there's that and then there's the the whole again universe of possible ways of fungiblizing ultimately non-fungible positions by by sort of being opinionated about a february investment strategy so i think that as people start to see needs for um you

know borrowing against lp tokens supporting things like the psm we're gonna see strategies evolve and develop um it's a bit early to say you know who's gonna sort of which of these are gonna come out on top or which is maybe most important but you know they're gonna come around sooner or later do you think something like fractional like what we were talking about with andy would sort of be one of those types of things where you take something non-fungible make it fungible again yeah i mean in the context of a single sort of shared canonical position that multiple users are pulling into still that's you know just represented by a single nft i think yeah you could you could apply fractional to that nft and then sort of use that as the like plug and play mechanic by which you're allocating fractional ownership and

that's again just the beauty of d5 like it's so cool that even is a possibility of working so yeah that might be that might be awesome those nfts as they are are those also like getting mixed up in the regular nft markets like is it starting to be like there's these new nfts let's put them in with the other nfts and trade them around like what's happening like are they yeah are they falling into that category too they are and they actually it's unfortunately a little confusing i don't think everyone realizes that the nft represents your liquidity in uniswap and so if you sell it you no longer have the rights to that liquidity and so somebody ended up selling it selling

one of their liquidity provider nfts on openc uh for much less than the capital was actually worse worth not realizing thinking it was just this like cool new thing to sell yeah but but they have real value and actually like the image and the description that's all like secondary to the core purpose of them which is to like to tokenize your ownership of this liquidity and so that's so you can send it around and uh see it in your wallets and and all that good stuff that comes with that do you want to see like a world where people are like pooling together these nfts selling them in batches for the perceived value in the future something like this i mean that in some ways that's what owning a share in the

index fund is yeah so there's an index fund that's allocating to the pools and it also has a lock up then it's exactly that's what it sounds like this is old finance right didn't that exist in like original finance the the difference the difference is at least you anyone can use this right like i would say like it's not like anyone can go buy a vix etf right like you can't really right you still need a third party who has a broker's license right like a lot of that stuff is removed so you know i think the the game aspect you should view as a positive because it kind of means more people can play the game than normal finance but you know i think it inevitably will converge to the same thing there's not really

there's no way around it i know there was kind of the dream sold of not coming into that but what do you guys think about it are you where are you at on that yeah i mean i was we were super worried for a while that you know v3 was going to be like too powerful for um professional market makers right like we really were it was a very serious concern that we had i would say an extended internal debate about um moody's smiling for those of you who aren't on the video call um you know because we all feel really passionately about this and i think that especially given flashbot's like totally meteoric rise to sort of prominence and and all that i think there's a really very real concern about you know the fundamental stability of blockchains in you know in a setting where

massive amounts of value can be sort of extracted somewhat for free from you know quote unquote retail by quote-unquote you know professional in in this case not market makers necessarily but you know just professional sort of miners or um arbitrage viewers or what have you and so you know the jury is still out on on all of this and i think that they're i'm pretty optimistic about cryptography and like open source ethos and sort of like adversarial thinking like getting us past the sort of threat that um mev and professionalization in market making poses to like passive investment or to just blockchain users as a whole but it is definitely like a scary problem and you know in theory like in the limit you know unisop could just you know v3 could have just like devolved into this

like weird sort of like in a flashbacks bundle like everyone adds liquidity before a single trade which is observed in the mempool and then like withdraws it immediately after to like harvest all the feeds like you know that hasn't happened at least not yet um and nor do we think it will but like in theory that's a possibility right and so like there are some scary like edge cases where where things really do go down there off the rails pretty quickly and um i'm not sure it would converge to traditional finance but it would certainly converge to like powerful players having extremely strong control over the system and that i would argue that's like an aspect of traditional finance as well so i think in my opinion like the most important aspect of d5 is that anyone who has skill and drive can come in

and compete with those large players and i think that's not going to go away even if we give these players like more tooling more control over the liquidity positions and how how they trade that bar might be a little bit higher for for what is a competitive market maker competitive trader you kind of spoke about these edge cases or like you know extreme cases and one i wondered about was you know you said you launched may 5th and there was a pretty intense drop obviously maybe not as intense as you know march 2020 but that was i guess felt what would that look like internally was that something would that have been a possible like breaking point did you feel like the

the systems in place were stressed during that drop in price does that even have an impact yeah you know for my personal sanity i really don't even look at the prices um uh maybe this sounds bad but it sort of just doesn't really register i think for us in a lot of ways like and actually there is this sort of counter cyclical aspect to uniswap by which like when the prices are falling people typically want to be in stables right or um they want to sort of hedge out their positions so yeah i guess the question here is did it impact anything it sounds like it didn't like i'm thinking sort of from like the maker like things getting there's nothing in uniswap that automatically calls things i guess i mean there's no leverage right okay so one of the advantages of a

concentrated fee model versus like borrowing against an lp share to increase your share to get fees is that there is no kind of liquidation yeah however there is this kind of potentially worse loss due to lack of fee reinvestment and oh and or the price moving really far out right so that's the thing you might be more concerned about which i think there are people who have been complaining a little bit about that but that's because they didn't read the code and they weren't able to ever think about it very much it's basically like people wanting to get their lp tokens or get their funds out well they just felt like they missed out on a bunch of fees on like a mean reversion right like if we for instance like take a coin

like comp which like i think took like a 70 nuke and then went back up like 40 right like people want the fees on the retrace but if they didn't adjust their liquidity they lost out on those fees and so that's usually where i think you've seen some but that's just because people were just aped into things and didn't read the code that's that would be my but me you know do your own research it's supposed to be a buyer beware marketing crypto i mean i i think there's some things you could do from a ux perspective to to address those things like uh v3 is a very different animal from b2 for for the user especially as a liquidity provider so yeah there's a ton of user education to be done there it's it's a rabbit hole

you could go down it does definitely sound like you've just unleashed a much more powerful beast i mean to be honest i have not tried it i did try the the v2 and was a liquidity provider for a short amount of time so i do know what that looked like i'm actually after this interview i'm super excited to go try it does the dad like does everything don't sell your nft on open c don't sell this don't sell the lp token got it okay rule number one yeah it's you know there's a couple more levers i'd say in the interface one thing that's interesting to think about and maybe a good starting place is like stable stable asset your liquidity provision because you know that's the real like home run uh use case for v3 right is um you know if die

diet doesn't break its peg like diet usdc is literally free money right and so you know if you want to put those assets in a very tight narrow range around the price of one um you're almost certainly going to see some volume there because you know there is this persistent interest in swapping between those two assets so you know they're they're these differentiated use cases in v3 like stable stable like you know general like sort of just defy token eth there's like socks bitcoin sacks like provision along a curve you know from a starting price up like they're you you can now sort of like nv3 sort of expand your definition of what it means to lp and you can sort of pick subsets that you are more comfortable with um if you don't feel like you are capable of like picking prices in um in optimal manner and making sure

that your ranges are appropriately updated when prices move so there's options out there cool cool yeah sounds like i'll try something here as i understand it recently there was a proposal or like some encouragement from vitalik to have like the oracle world live more in the uni-swap world and i'm wondering like given that there's constant trades happening like how do you actually do pricing in the past you weren't using external oracles were you no so there's sort of two aspects here i mean the the use of the term oracle is has been a bit a bit muddied but so for for pricing v3 and you know v2 and v1 are fully algorithmic the prices are determined by the underlying liquidity that's being

provided by end users and v2 and v1 that's quite simple it's just you know the asset value a and asset value b and v3 of course it's this aggregation across potentially varying tech ranges that end up providing an aggregate quote for for people looking to trade and then there's this second layer where you know because of the incentives um in in uniswap uh the current market price on unisoft is typically tracking the price off chain or you know the true price of assets quite closely and so this is a valuable um thing that we're getting literally for free and so if we can figure out a way to expose this price to other applications that's potentially valuable and so there is this like internal mechanism by which unisoft e3 pools

offer historical prices to you know other people looking to use those prices to inform you know leverage liquidation things like that and then there's this third layer which is i think what you're referencing where vitalik and a post on the governance forum maybe a month or two ago um proposed that yeah unisop become a i think a dollar price uh oracle was the proposal and i don't have too much of an opinion about that but it would certainly stray from the fully algorithmic ethos of uniswap uh the protocol and it would it would instead rely more on sort of economic assumptions about the the market cap of the token and the behavior of token holders in a in an adversarial sort of voting or oracle updating system and so i think

the core protocol is like it's its own little bubble and then this oracle proposal i think would would build on top of sort of the quote-unquote core competencies of unisop namely like thinking having thought very deeply about um you know pricing and mev and oracles and things like that and then offer this service that would basically i think expose the the dollar price oracle um using uni as sort of a backstop for this system and i don't know you know i don't know to what extent that's actually going to happen or you know if that's going to be us the bills that are again it's a fully on permission like anyone could do this right like we can't stop them which is cool but you know there's um i think there's a conversation to be had there because i mean over the last two three

months i actually did a series of i think six episodes all about l2s and i know that two of them have publicly announced with you that uniswap is going to be deploying on them and that's optimism and arbitram and maybe some others are you're in talks with them now my question here is how do you deploy on main chain mainnet l2 and then another l2 like do you end up with liquidity pools on each one of these do they all have their own like yeah i'm quite confused about what this actually means when you say it's like deploying on this l2 yeah so it's it's actually not that different from deploying to a test net i

mean it's an evm compatible chain so everything is completely separate that means the liquidity pools are separate and so the liquidity is fragmented across l2 optimism l2 arbitrary and layer 1 ethereum and so for in the case of arbitrim trim like we have this deploy script and we literally just pointed at the arbor trim node and fired it off and it pretty much just worked and optimism requires a few more changes just because they made some different architecture decisions but otherwise it's it's a lot like deploying to a test net and then the big difference is that you can bridge these like mainnet assets over to these layer twos whereas you know with the layer one test nut you can't do

that in a secure way at least i wonder from the user's perspective what do they look like like does it do they choose that's a that's a great question that's like something we're sort of figuring out in in terms of like ux um i think early on it will be mostly user driven the the choice between like which l2 they they get um and even like the experience of you know connecting to a layer two may not be quite as polished as as it is if you just like go to uniswap interface to connect to l1 so definitely a work in progress and uh i i think it will get a lot more difficult to like create a really nice polished ux as as we deploy to more l2s it may

become a little confusing and that's like a problem we're sort of working through today yeah i think one one further question maybe both from a mechanical perspective and also the user perspective is in the multi-l2 world right there's kind of a natural triangular arbitrage of like i have wes wbtc on arbitrom i have west wbtc on optimism i have wet wbtc on mainnet and so in order for me to actually kind of synchronize prices i'll either have to go through mainnet or you know theoretically like a state channel relay like connects or something but like for what it's worth i think the connects thing seems to be at least starting to make l2 relaying possible but

i guess my question is from the end user's perspective there's a couple problems here one is well there's a lot of mev that will be lost to if you have to do the leg through the main chain right because because like front running the sequencer is now just way more valuable than any front running any individual transaction in a bundle and then the second thing is kind of when you show a user a price there's a sense in which the ux could actually go do the arbitrage and kind of provide the median price as the execution price a little bit like what robinhood and siddell's arrangement is in some ways how do you view kind of this revenue sharing model that will now have to inevitably have to kind of come up given that the end user now has to like manage this one and it's a little bit

kind of an extension of your previous question there's a lot of a lot of things there so feel free to take whichever ones you feel most comfortable with yeah i think the mev thing is is is very interesting and you know each additional altitude introduces quadratically more layers of mev and arbitrage opportunities here so i think it's it's something pretty serious to think about um you know it sort of comes down to whether we think uniform is like a piece of code that can just be deployed anywhere or whether it's an experience an app you know something that users can sort of bite their teeth into um and i think that it's easy for us to think about it as a piece of code because you know we wrote this and we specifically set up the license so that like these exemptions could be made

we're like we were very thoughtful and careful you know to the extent possible about you know about making sure that this thing was going to be portable the license was was you know you could disagree with the sort of motivation but you you can't you know it was it was quite thoughtful it was quite intentional like we did this for a reason and i think that it's going to come down to a question of like how much our company wants to be involved or how much you know the unisof community wants to be involved in making unisop a cohesive experience across l2s and across networks you know to be frank like we haven't deployed to any other chain other than ethereum for a reason and i think that's because you know there was there were security concerns there were community concerns there was you know there was just a very

real risk that you know unisop wouldn't be living up to its ethos to what we want united swap to be if we were deploying on these other chains and um i think that's no longer the case with l2 because these these you know systems are inheriting a lot of these security guarantees that ethereum provides you and so i think that this is a really positive step forward and ultimately it's going to be a challenge and a test for these protocols to you know see if they can manage cross chain with like little like you know stapled together glued on like connect style chains um or whether something like compound chain right which is which is another alternative right like if you just bring compound into its own little universe and sort of talk to different um people as or different chains as outlets you know

that's another option as well so i think really i don't have an answer for you but i think my you know that's where our heads at really is like whether unisoft just you know is its essence um you know just a piece of code that can live anywhere and and um you know fragmentation is like this external concern that we don't have to worry about or whether whether we can think about products that span um l2s and and sort of encompass the whole of ethereum and you know by ethereum i mean like not just l1 but you know these these l2 systems that are hopefully just as secure if not more uh and so yeah i don't know i don't have an answer but that's sort of like that's just what comes to mind and we'll we'll see what happens um as time goes on i'll just add that

my sort of early thinking on this is that it's hard to say that you know l1 ethereum eth is fungible with like optimism eth especially because they have like different security guarantees and then even with like these fast withdrawals like fast exits there's going to be different amounts of liquidity you know based on the the confidence that these market makers have and the all twos and the different parameters that go into security and the architectures which makes it feel to me like it's kind of unavoidable uh this choice that you give to the user like they have to you know pick arbitrary optimism but um i know that that's product people would probably hate me

for saying that but like i mean i don't know if that's true because like look at the stake d thief pool right it's proven that people are willing to actually hold an lp share of um the multiple types of ether and use as ease right like the curve st east pool is like through all the crashes it was the most low variance pool in all of all of ethereum on unislop or curve or balancer sushi swap whatever so i think that's a testament to the fact that people are willing to actually hold the lp share of a bunch of different types of wrapped assets as the real as the real asset and so then you get then you have a settlement asset that's effectively the lp charon on there one so you i'm just saying there are some weird

things that might occur like that that you know lido i think has kind of shown that the thing about listening to all of your answers is and i i say this given that i you know i think i maybe a year ago was quite quite negative on this ecosystem but i kind of now i'm very positive is that the introduction of mev auctions and kind of the fragmentation and doing having to do all this stuff is very bullish for polka dot because polka dot already built that in in their consensus like it actually looks very similar like everything we're talking about looks like what dots look like with crowd lens i've never been more positive on on on polka dot until this conversation not what i was expecting for you to ruin i know i

the way this evolved it was like it's like whoa like the compound app chain model mixed with kind of like this settlement and multiple eats is like kind of what polka dots spent a lot of time trying to solve and their auctions are very similar in in some spirits to the mev auctions for the sequencers so yeah i i think it's actually kind of interesting to see these design spaces converge and in the case of ether it's like you know unisop sort of enabled that uniso plus mev sort of kind of forced the hand towards that but i feel like this this is going to get retweeted by some polka dot shills out there i know i'm sorry i'm sorry i wasn't i wasn't trying to like give a shill message i'm just trying to point out that the polka dot auction plus many chains thing is kind of

similar to mev deter auctions determining which ordering of l2s gets executed yeah okay but bringing it back to uniswap and the project like people who want to get involved or want to start building maybe even some of these like strategy bots or i don't know like how can they actually start participating in the ecosystem what do you recommend they do yeah i mean that's a great question and i think that it's important to understand um that the things you know we have all been saying this whole podcast about you know being involved and community growth and all this stuff like it's not just sort of abstract it's actually quite it's quite concrete and so you know specifically the unisoft grants program was created it was the first governance proposal that passed um about six months or so ago now and it

just funded um a multi-sig um you know comprised of some some great team members some community members that is and it's funding you know all kinds of programs that are contributing back to uswop in the unisop ecosystem and so you know i don't know the numbers um off the top of my head but you can actually just go like it's unisoft grants on twitter i think um there's a whole like notion doc of all the projects that have been funded um you know what they've you know the deliverables like what they've produced so far i mean we've seen so many cool community projects come out of that in just just the early sort of few months that that it's existed so far and i think that you know we we really want to make this treasury you know valuable to you know to ethereum to uniswap

and i think that putting it to good uses is actually quite quite hard and i think that this this grants program is a great first step toward just like really being inclusive and engaging to developers and you know as moody said like maybe just you know the bar is just going to be higher to getting involved in d5 and you know we do want to be inclusive and stuff but it but it's also like it's it's really pretty novel that um if you're interested and engaged you can get access to like you know funding and and you know start working on these ideas and like be very very close to the metal and so um yeah i think that that's something that to look out for if you want to get involved and you know really it's just about enthusiasm and time too like if

you are willing to like put in the hours like we're not we're not at the stage yet we're like you need to be a rocket scientist to get involved you can you can hop in and make an impact very nice cool want to say thank you to both of you for coming on the show sharing with us the journey through v1 v2 v3 of uniswap and uh definitely like giving me a chance to dig in ask you some questions i've had for a while uh yeah thanks so much for being on the show thanks anna this was great and thanks durin thanks for having us cool thank you to the podcast producer andre podcast editor henrik and to our listeners thanks for listening let me say that again thanks for listening and actually i have to repeat something because i said i did

eight episodes on eltooth but i only did six i like went and counted right after i said it and i'm like wrong so i'm just gonna i'm gonna record this you so one question i have is about this idea of uniswap on l2s because i mean over the last two three months i actually did a series of i think six episodes all about l2 systems l2s and i know that there's two of those l2s arbitrarium and optimism who've publicly publicly who've publicly announced that um uniswap will actually be deployed on them

now what i don't really understand is like what does that mean to have uniswap on them is it one sort of joint uniswap that runs on them or is it like separated yeah what what does that look like