Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Ungovernable Podcast Ep. 13 – Kassandra
Recorded: April 10, 2025 Duration: 1:07:58Space Recording
Full Transcription
Thank you. Music Thank you. Thank you. Music Thank you. Thank you. Join the ungovernables.
No filters, no apologies.
This season we're cracking open Uniswap governance.
Welcome to a new episode of the Ungovernable Podcast.
Hey, hey, hey.
Welcome, everyone, to a new episode of the Ungovernable Podcast.
Oh, I think Joe cut out.
Uh-oh.
Can you hear me? You're back. You're back. Oh, perfect timing for Joe cut out. Arrakis Finance. Uh-oh. Can you hear me?
You're back.
You're back.
Perfect timing for a cutout.
That's great.
Love it.
I don't know where I lost everybody, but Cassandra is with us today, founder and CTO of Arrakis
Finance and active delegate of the Uniswap DAO, cryptographer, software engineer, founder,
always learning, previously an engineer at Gelato,
researcher of multi-party computation and zero-knowledge proofs and other emerging
cryptographic tech. We are Joe and Austin of Alpha Growth, your premier DeFi ops and growth firm.
Cassandra, thank you for joining us. What else should we know about you?
Hey, thanks for having me. Best intro I think I've ever gotten. Very thorough. I don't know where
you found all that. Maybe I sent you a blurb. I don't know. But anyway, that was great. Oh,
nice. AI knows. So yeah, no, that's me. Nothing else to report. That's perfect intro.
nothing else to report. That's perfect intro. What about outside of the realm of the really
exciting, riveting cryptographic technology? Yeah, well, you know, I certainly got really
nerd sniped by that. Maybe an interesting factoid is before I went down this crypto rabbit hole,
I was actually a totally different domain theater director. So very,
very large pivot to get into computers, software, crypto. It's very mathy. So yeah,
got a wide background, I guess.
Truly. And how'd you make that jump? That's a tough one.
Yes. Super weird one. Everyone always asks. It's pretty funny. I was working on some like sort of experimental theater
that was actually dancers.
They were like computing something
like a really small logic gate based computer.
I didn't know much about computers at the time,
but I thought it was cool that like the dancers knew the rules,
but they would either be a zero or a one
and then they could go through the logic gates
and the dance would look a little different.
And then I said, oh, okay, i'm kind of like a technical person and this
is really interesting theater work maybe i should learn more about how computers work so i can like
deepen this and then i just the moment i started wrote my first line of python i don't know i was
hooked and i just never looked back so i'm on this very long tangent into computers because of that that is that is
pretty impressive that you didn't you know that you didn't have like the bill gates moment where
you were you know first thing when you were in fifth grade or eighth grade or whenever that was
where he was like oh going to the computer lab and learning about this really young that's pretty
wild that you got drawn to computers not that long ago. And then now you're like deep into the zeitgeist of cryptographic computing.
That's so nuts.
Yeah, I'm just like a very research.
I would say I always was a very research oriented person.
So always like to just go super deep on niche topics.
And so, yeah, this was the new one.
And I mean, yeah, I was really, I guess maybe to bring it it towards crypto i was like really taken by how everything happens in the open so like i was starting to learn about bitcoin
and made a little tool to encode transactions myself by hand and just thought it was
so neat that i was interacting with this system that nobody controlled and nobody knows knows
who started it and like you could go so deep as some random person in your underwear in
the basement um uh with this like yeah this special new public thing so yeah i never look
back from there would you say that that right now is the easiest time in the history of of humans to
become an expert in a really niche thing very quickly. Yeah, for sure. I think for sure. Of course. Yeah. It's like, now there's so much,
uh, now the question is like what to focus on is like the most important thing actually.
Uh, because once you know what you want to focus on, yeah, there are so many things that can help
you, you know, AI, et cetera, ways to easily be self-motivated and go deep. But I still think
nothing compares to the world of computers.
Like, yeah, I was just, I was like a junkie with computers at the beginning where the fact that it was something where if you ran into an error or you got stuck, you could just go on Stack Overflow,
just like take whatever you saw and Google it and just like brute force yourself into going deeper.
I feel like there are a lot of other things. It's still
like knowledge is siloed a bit and you have to know the right people or something to go deeper.
With computers, it was always like, if you have the desire, there's infinite things to read in
public forums and ways to tech up by yourself. So yeah, I still think computer science is special
in this regard. Out of all the different categories that you've researched and rabbit holes that you've gone down,
what to this day has still blown your mind repeatedly?
Yeah, that's a great question.
I mean, there's this quote from Hal Finney in the Bitcoin talk forums where he says, like, I don't want to exactly, I'm probably going to misquote it, but something like, I've always been taken by the mystery and the paradox of cryptography.
Something about going all the way deep into the math of how these cryptographic primitives work and then building a POC and then actually seeing, you know, that zero knowledge proof checks out or the signature checks out was always like some crazy high for me.
So, yeah, cryptography is really cool.
I don't know why, but I just something about the fact that we can actually hide things with mathematical precision
and that this is like a new thing that only emerged in the public in like the 1970s.
I find really, really fascinating.
But yeah, I'm also, true, this is also like a deeply nerds and I buy other things too.
I don't know, like philosophy, there are some niche philosophers that I just can't stop.
I can't believe what I'm saying now, for i just can't stop i'm just i can't believe
i'm saying it now for instance can't stop talking about um so yeah i'm like a um an uber fan of
michelle foucault for instance really random thing so anyway i don't know okay wait let's i want to
i want to click back on the zero knowledge stuff because when i first heard about zero knowledge it was
it was fully homomorphic encryption via uh zomo no uh zama you remember the company's ah
that was yeah there's this like uh they're like in some sort of patent war this thing yes yeah yeah
so i i heard about them like three years ago this whole idea of zero
knowledge proofs was introduced to the world and especially with the defy uh the defy use case
and then you have things like zk evm um all these different these different chains that have come
out and do you think that it was do you think that it's here to stay do you think that it was, do you think that it's here to stay? Do you think it's going to come back with a vengeance? Because it doesn't really seem to be sticking. And then
correct me if I'm wrong. Oh, yeah, yeah, I do think you're wrong. I mean, I think
like the people who are really deeply invested in zero knowledge proof, like at the,
at the research and math level, which is all super new stuff I think
people sometimes again like even just basic asymmetric cryptography like digital signatures
based on elliptic curves that you know bitcoin is absolutely based on in ethereum 2
are only as old as 1970 something and zero knowledge proofs are like really really new at least ones
that are actually useful and at scale and like people who are um who are into the research side
of that i think are very convinced that maybe even more than blockchain itself zero knowledge proofs
are here to stay and are going to transform the world and transform the way we we like collaboratively
uh trust and and do things in the world and with computers.
For instance, imagine going across a border and just, you know, getting a green check mark
without having to show your name or go and have the mask to see your phone or whatever they do
in order to vet that you're actually, you know, not on some sanctions list, right?
Now they can do that, just get a green check and they don't know anything
else. But they're absolutely certain that you're not on that list. This can be applied in so many
different places, in my opinion, and could really transform the social fabric.
That was absolutely the same narrative that I subscribed to when I first heard about it. And I
still do subscribe to the use case a lot, pretty far outside, but more so with what I was bringing up was the DeFi use case, because it doesn't really seem like any of that has stuck.
Yeah, I think that's interesting.
I think with DeFi.
So, yeah, also, it's really important that in the context of blockchains, when we say ZK, it's actually like two parts, right?
There's just the succinctness.
There's the part that's just about compressing computation and then being able to verify that it was done correctly so that it's like you can do a very fast computation
instead of doing the really long one and then there's the zk that actually hides the inputs
and which is actually about privacy, which is like learning something
without revealing some of these inputs and learning something in zero knowledge. So a lot of times when
we say ZK, it's actually not ZK. It's like we say ZK snarks, but we're really only using the snark part.
I think for DeFi, actually, the privacy part is really important. And I have a feeling eventually,
I think privacy has taken a long time in crypto to grow, certainly in the Ethereum space.
But I do think maybe then you'll get this unlock, where if you have these actually functioning dark pools,
if you have new institutional actors coming online on-chain because of the fact that they can actually keep what they're doing private,
then you might see this proliferation of zk use in d5 but like uh
yeah in terms of getting the succinctness i mean you know you hear about it i think slightly um
uh yeah maybe it's outside of my exact knowledge but it's true that maybe it's not being used too
directly in d5 things just yet that would be maybe my guess is when when
we start having like their real zk like with privacy as well maybe there will be some unlocks
but basically on the app layer and on the d5 layer like yeah users you need to work backwards from
what the users actually want and what they want to do and um uh until we have that sort of unlock that actually helps someone achieve something
we couldn't before with apps.
I don't think just because it sounds really cool
or like it's a really interesting math problem,
you're gonna get a lot of traction
and maybe this is kind of where we're at
in the DeFi, ZK space, I don't know.
It might be a little too late
because we just talked about this for five minutes,
but will you explain zero knowledge to someone who has never heard of it yeah i can try so um uh there are some great i
can say also there are great tools like on the internet i think the zcash founder has like a
really good um like really uh approachable examples that make you understand zero knowledge in like a simple way
or a non-technical way. I think zero knowledge for privacy, I'll try to give a thought experiment
maybe. It's like, let's say me and you, we play this game where I take a coin and I put it behind my
back and I put it in one of my, my right hand or my left hand. And then you pick if it's in my right
hand or my left hand. And what you tell me is that, you know, a tell of mine, I have some special
physical tell. Um, so, you know, whether it's in my left or my right hand,
but you don't want to reveal the tell that you know to me, but you want to prove to me
that you do know this tell. And so we can play a very simple iterative game where I can just
put a coin behind my back, right or left, and then you pick left and, oh, it's actually my left.
And we do that 10, 100, 1,000 times in a row. And if 1,000 times you're right, after a while,
you'll sort of have probabilistically convinced me that for sure you know my tell without revealing
to me what the tell is. So this is sort of like the maybe a little bit of the intuition of what a zero knowledge proof is all about.
And original zero knowledge proofs were actually interactive like that.
You'd have to play something over and over in order to convince me.
And then there are just some cool math tricks to basically make it so we can do this in one round instead of having to play an iterative game over and over.
make it so we can do this in one round
instead of having to play an iterative game over and over.
So what do you think the big use cases are
going to be for zero knowledge proofs?
I love this because I just have to say,
this isn't directly my subject, so I'm going for it.
But yeah, I think there are tons.
Today, there's this huge unlock, right? I mean, ZK rollups and like, there are some things that are being used in production today, pretty meaningfully, like rollups with applications on them.
as well, which is being used a lot all over blockchain to just speed things up and deal
with like some of the bottlenecks that blockchains have.
But with the actual privacy zero knowledge proofs, I think, like I said, I think there
will be private DeFi, private financial applications on blockchains that will really explode where
I'm very excited for things like the launch of Aztec to shill a random thing that I have
no affiliation to, but i think is really cool um but then i think outside crypto there's also a lot um and uh yeah like i said i
would think about um ways where we're we're so used to in order to verify things in order to prove
that we are really authenticated that we're really allowed to do
something that, uh, right. To prove that you have enough money or cashflow to rent a, an apartment,
to prove that you're, uh, that you're a member of a certain set, all of these things we're like
totally used to in order to do them, just giving away a ton of information and just saying, I trust
that you won't like abuse this extra information I'm giving you in order to do this. So I think all
of these things, if we if if those kinds of things become way more scalable, like maybe you can
just like rent an apartment for a year and get approved for that just with an app, like in a
in a way that doesn't involve any third parties and is just really seamless and allows for like new kinds of experiences, that'll be powered by this kind
of technology, in my opinion. And so, yeah, it's hard to even say because I think it's really
radically transformative and we're like stuck almost normal people in this mental model of
like the only way to prove stuff would be to actually reveal information. See it. Yeah. But like, yeah, maybe this gives some color. I don't know.
No, I love it. And then I feel like there's huge unlocks too for, you know, different hacks where
data gets passed around like social security numbers, for example, most social security
numbers are on the dark web and like to fight off against identity theft.
It becomes really difficult for different banks or different financial institutions.
But I know we could probably do a deep dive on that for like a whole hour in itself.
So I'll hit on something that you are a specialist in and would love to maybe hear a little bit
from you.
What is Arrakis?
And maybe simply kind of explain like an. What is Arrakis? And maybe simply kind of explain
like an overview for what Arrakis does.
Yeah, totally.
So Arrakis is a liquidity layer, in my opinion,
and we help all kinds of token holders,
let's say, especially we look at like token issuers and DAOs are sort of the tighter set of like the main customer,
but really anyone with token inventory who wants to supply liquidity to markets on decentralized exchanges.
We help them do this in an efficient way, in an automated way, so they can express their liquidity-providing preferences on-chain in a more fine-grained way. complex over time, allowing would be liquidity providers to harness that complexity and to
provide liquidity to markets on chain in the kind of best and most optimal way possible for their
goals. So at the end of the day, we're really working with especially people like token issuers
and DAOs that have a token to help them figure out how to deploy this
liquidity on and across chains so that, you know, swappers can easily swap their token. And, you
know, that this liquidity is accessible across many venues and traders have a good experience.
So it's like all about the LP side of decentralized exchanges.
Okay. And how would you like differ compared to other liquidity solutions?
Yeah, I think we started similar to a lot of others. So another thing like in the idea of concentrated liquidity, providing liquidity to a decentralized exchange now became much more sophisticated.
You had the ability to set specific ranges to set much more fine grained preferences.
And so we weren't the only ones who, as a response to this, said, OK, providing liquidity to to DEX was always awesome because anyone could do it.
We want to preserve that, but now it's getting more complex.
And so maybe there's this room for this sort of layer in between the unsophisticated LP and the DEX who can help that unsophisticated LP be more optimized.
LP be more optimized. I think where we really differ is we eventually found our niche
specifically in these private offerings for the liquidity provider who is somehow
deeply incentivized for that specific token market to make sure the liquidity is good.
So rather than focusing on retail Lps who might want to try to farm
maximally the most incentives or you know just earn a yield through trading fees on like blue
chip pairs we're really focused on young tokens that um need to structure the liquidity and
working with like the token issuer or as close to that party as possible to help them figure out the
like base liquidity for their new token market um so we we specialize in things like token
generation events how to have how to deal with dex liquidity at like the moment the token starts
trading as well as yeah working specifically with the dow token issuer or uh you know a project
that has launched a token to figure out how to structure their market.
So I think a number of other similar liquidity solutions, I think, are more focused on like specifically yield capture for like the retail participant.
We're much more focused on like the setting up the initial market structure for young tokens.
Yeah, that makes a lot of sense.
Do you feel like, because I feel like these services,
especially with Uniswap, for example,
V2 was super easy, right?
Put two tokens together, you're able to understand your LP.
There is some impermanent loss, but it's quite simple.
V3 became a little bit more complex,
and then I feel like now you're
starting to add hooks to it, which can change the game in an infinite number of directions.
With that, do you feel like liquidity provisioning starts to become an institutional game,
moving away from retail? And if so, why? I do think so.
I mean, I think the beauty of Decentral...
I'm a real believer in automated market makers
and truly decentralized exchanges.
I think what they enable is more variety.
So I think you'll always have, at least on the fringes,
retail actors and retail money
piped into these decentralized exchanges.
And this is a benefit
this is what's so cool and like disrupts the market structure of like just the way things
work in trad fi um but i do think that now there is a will be a lot more competition with um like
much more sophisticated financial entities and like sophisticated actors and traditional financial institutions who will
potentially dominate the LP side. And I think this could also be okay. This is not necessarily
like a horrible thing. This should help. I think most retail participants are more interested in
being like on the taking side of being a trader. And this should help make those markets vastly more efficient. As you said, the reason why it's sort of you already said it,
right? The complexity is just massive. And even we saw the leap from UniB2 to UniB3 was a really
big jump for just like your normal person who is, let's very defi literate really interested in all this stuff
but gonna try like to naively uh handle their liquidity strategy and changing their ranges
and like all of a sudden i think people really realize this is very complex we realized it as
well as like we were there from day one on v3 and we were learning as we go and we we uh i think in the very first days
before v3 even launched like i thought it'd be easy like oh there's gonna we're all gonna within
a few months someone's gonna figure out like the perfect kind of range rebalancing solution and
we'll just do that and be more about like who will have like the best automation and optimization of
this stuff it turns out it's a super complicated open problem that I'm sure people who knew more
about market making than I did when I was just starting there already could have identified.
But yeah, we saw just like tons of friction just in this jump in terms of usability on the LP side.
And yeah, like also friction with now this liquidity wasn't as composable.
You used to get back a fungible token.
Now you get back a non-fungible token.
And so it wasn't as composable in all of DeFi if you wanted to incentivize liquidity on a certain pair.
This was very complicated for people at the beginning.
So we really saw how much friction was introduced just from moving V2 to V3.
saw how much friction was introduced just from moving v2 to v3 and i think the leap between v2 to
v3 and then v3 to v4 is actually v3 before is much greater because with v4 and hooks you literally
have an infinite design space for people to design sort of whatever decks they would like to on top
of the uniswap v3 primitives or even kind of fork them out completely and just use the
singleton for like how you handle token balances, but do pretty much whatever you want in terms of
market making on top of the unit before decks. So I think that now this complexity, like if it was
hard in terms of usability, just going and trying to set your ranges manually and check in on your positions.
I think now when you're also going to choose a specific hook
for a specific pair, you have to know exactly the features of that hook and then how you might adjust your strategy for those features.
Maybe even you want your own set of features.
You're going to be like a developer developing your own hooks
in order to deploy these things.
And I just think this really leans towards
sort of well capitalized, sophisticated entities figuring out how to maximally
squeeze efficiency here. And again, everybody benefits because while market makers
definitely have, you know, reputation of potentially doing shady things or like
people are getting mad at them all the time they also provided this like integral service to everyone else which is make it possible for you to swap
on just go to uniswap or go to your favorite front end and swap and uh you can't do this without
people injecting liquidity into the market and those who are doing it who are going to survive
in the long term have to be sophisticated managing their risk and so forth. So yeah, I think it's totally inevitable.
Yeah, I'll stop there. No, there's a lot of wisdom in that. Do you have, does Iraqis have
any sort of retail facing product that more of the middle market, even small whales, big sharks could be interfacing with?
We have some we have some old ones. But right now, it's really
not our focus. I think we're looking we want to figure out we
kind of only want to really push that when we know it's something
that is like long term sustainable for retail. So it
does crop up occasionally. And you can see we do have like currently some
public vaults. If you go to our app.aracus.finance.com, you can see some offerings,
some different public offerings for like anybody who might want to pledge some capital to a pool
and earn some certain different kinds of rewards. But I would say, yeah. And over the
last year, we've really pivoted our focus to raise our focus on these DAOs and token issuers.
And it's not like we were abandoning retail. It's just, we believe that soon it might be possible
with more structured products and hedge things to give them an offering that's like long
term stable for them. For now, the only things that are actually interesting to retail that like
are in expectation, profitable and yield bearing for you are really only when there are these
third party incentives. And the thing is, those usually only last for a certain amount of time,
or those yields are only lucrative up to a certain amount of capital and for a certain amount of
time. So it's sort of like we've done it before and we may do it again, but it's often been just
kind of like short term boom and bust. And I'd only be excited about really focusing on retail
if we can find something that's like more long term and sustainable. Fair enough. What is the
relationship between Arrakis and MEV? How do those two play with each other?
Yeah. So, I mean, I think like one of the big topics in decentralized exchanges and especially
on the LP side, though really on both sides of the market is MEV, right? And MEV is this,
is this it's this design and structural problem that's really not um doesn't necessarily need to
be surface to users in the end it's like i i believe that just by designing better apps and
better primitives mev will be like a much much smaller topic eventually it's just that you know
in crypto it's still so early and so we're designing these totally new primitives like AMMs.
And there are some side effects that simply can't be, couldn't be known in advance.
Would you mind breaking down MEV for the layman?
Yeah, sure.
MEV that's known as maximal extractable value now came from minor extractable value.
And it comes from the value you can extract by being the one to order
transactions in a block. So all transactions and blockchains first go into a mempool,
and then they end up in a block. And that block is ordered usually by one privileged entity,
whoever has the right to propose that block. An Ethereum, one validator per slot. And you know
a little bit in advance that, okay, that's my slot. I have total rights over ordering that block and Ethereum, one validator per slot. And you know a little bit in advance that, okay, that's my slot.
I have total rights over ordering that block.
And it turns out if people are doing financial activity,
like you can see that everyone is about to, I don't know,
let's say buy ETH in the next block because those are the pending transactions.
Because you have total control, total rights,
like a feudal right over this block space as yours, you can decide how transactions go
into that block. So if everyone else is buying ETH, maybe you can insert your own transaction
to buy ETH ahead of all of them. And maybe you can even do more sophisticated things,
which aren't just like betting on a certain token where you can buy ETH before everyone, allow everyone else to buy, and then sell on the way afterwards and make a profit from the price at the beginning of the block and the price at the end of the block.
And this would be known as like a sandwich attack.
But essentially MEV is just all around the value you can extract by ordering the way pending transactions actually end up in a block.
You explained that really well, by the way. That was perfect. Very easy to digest that
explanation of MEV. So cool. But then, so like, as you can already see, it's like only like when
there's trading and so forth happening, MEV usually proliferates even more on blockchains.
With AMMs, we've learned there's a lot of painful MEV.
There's like this sandwich attack that I mentioned
because prices on AMMs just like directly move
with how much volume goes through the AMM.
You know that some trade is going to move the price.
And so if they have some slippage conditions on their trade, you can abuse
these conditions to just like extract value risk free. And the other one that really affects
liquidity providers is the fact that prices on AMMs, we don't know anything like they're totally
neutral to the real world and how prices are moving on other venues outside of the blockchain.
And so because blocks on Ethereum come only every 14, 12 to 14 seconds or so, every 12 seconds,
you have this risk-free trade you can take at the top of the block by basically,
if the price moves enough on other venues, you can move the price on the AMM and basically take this stale price liquidity on the AMM at the top of the block.
And this, if you're making money from this trade, now you have to be sort of sophisticated because you'd have to have money on another venue as well to sort of do the full ARB and make this profit. But if you have the sophistication, you're making risk-free money when you do this,
or like low risk. If you make a profit when you do this, where is that profit coming from? It's
not coming out of nowhere, right? It's coming from someone's pockets. It's coming from the LP's
pockets on these two markets that you're equilibrating. And so, yeah, there has been tons of ink spilled. People are still
working on and thinking about it and tons of brilliant people figuring out solutions to this.
And I think we will have great problem, great solutions to this problem.
I'm going to push back a little bit on that. And I'm going to play devil's advocate and say that
I think it's a little bit loaded to assume that this actually is a problem.
I don't think everybody would agree that MEV is actually a problem. Well, the people profiting
from it, of course, are not going to think it's a problem. And the more that you have extractable
value, the more likely people are to participate in the blockchain and decentralization and push
these other things forward. So there are some
arguments that MEV is not entirely a bad thing. What would you say to that?
Certainly. And there are also like non less toxic forms of MEV as well. I mean, I think in this case,
I think the thing that people don't like, which I think I personally can understand pretty well.
Now, maybe there are some interesting arguments I haven't thought about in the vein you mentioned.
But on an AMM, because of this arbitrageur actor, the AMM only converges to the real price because of this arbitrage.
arbitrage, it's kind of like we're paying a random third party, anyone, whoever wants to do it,
to help make these AMMs function, to help them price different tokens at the right price as
information in the world evolves and people go oh no this thing has been hacked and
now everybody thinks it's worth a lot less or i don't know they fired the ceo whatever it is um
these things the blockchain doesn't know about them and doesn't have a way and this this information
is delivered by these arbitrageurs which sounds cool but if you're paying like the goal of a market would be to pay as little as possible to these like why pay a
random person uh I I believe it was like Max Resnick of once ethereum fame who said like AMMs
are kind of like a leaky bucket so this is just value that's flowing out of the system it's not
going to the traders it's not going to the traders. It's not going to the LPs. It's just going to these
random participants who are sophisticated enough to spin up these bots. Why pass the value there?
If you could avoid it, you would. I believe, sure, there can be value generated from these
things. LPs take a fee on every trade, and that's great. But it's all about like, where does the value actually flow? And today 99% of this MEV value flows to this validator who has, or maybe not 99%, but a lot of value
flows to this validator and these participants in the MEV pipeline, builders, validators, searchers.
And that's not to LPs, not to swappappers so not to like the real retail participants but to
just these third parties yeah sure it's cool that they're playing but they would play anyway they'll
they'll play if they can make one penny of risk-free profit so if we're paying them tens of millions
of dollars a year on the eth usdc mainnet uh pool maybe we could just change some of
the uh the logic in these far contracts so we pay them less and um uh they would still come anyway
right because again if they're making risk-free profits even if it's like you you will just want
to make sure you make a really competitive environment for them so that as much value as
possible flows back to like normal users which should increase
blockchain usage much more than if you get lots of value extracted to just like uh sophisticated
participants so it's a little bit of like kind of a necessary evil but you want to minimize
that necessary evil if possible um yeah like in market design you often just want to
create as much efficiency as you can as much competition among different parties to to uh to
like maximize your efficiency in my opinion what we have with the current mev landscape on dex is
it's like a bunch of deadweight loss and kind of like a market that's that's designed it's like you have a bucket that carries water and uh it has
a hole in the bottom like one thing you can do is just try to patch the hole yeah plug the hole so
i guess when you're working with these different protocols and different dows like when it comes
to liquidity management and you know plugging the holes in the buckets.
Do you have any like top tips you would recommend?
So, yeah, that's the thing.
I really feel like it's not, I'm hoping that it's not a problem that's like has to be
swallowed by like the outer parties, but it's really a problem for people who are designing
UNIV4 hooks or designing
things like Uniswap itself. Hopefully we can kind of solve these things at that layer,
because sure, there are definitely things you can do to avoid major snafus. I think in the real
world, in practice, what we help DAOs with a lot in terms of MEV is like foot guns, right? Because
actually more than you might
imagine, there are like major mistakes that end up having huge MEV, like supplying liquidity to
a new market at like completely the wrong price without noticing it. And then just exposing
yourself to a massive arbitrage and like creating really weird effects in the market and losing a
lot of money like instantly. These things happen all the time.
So like these really basic things about understanding the adversarial environment
and like having proper checks and so forth in place to be doing like relatively sane things
is something that I think concretely we end up helping a ton with.
On top of that, yeah, there's like, that's the problem is from the LP side with this, like,
this kind of arbitrage that I'm talking about, there's not much that you can really do today.
Because it's literally sort of in this market structure that you have to deal with this.
I think still there are some, yeah,
there are like secondary things you can do,
but the core problem can actually be fixed by having cooler hooks,
cooler functionalities in blockchains themselves.
Maybe things like what's being done on Unichain
and Optimism with Flashblocks,
plus using MEV taxes inside of a Uniswap before hook would just kind of
magically make a lot of these problems go away um and we could click into what exactly that is but
i mean maybe it's not worth it i don't know yeah i hope maybe like tldr what's an mev tax because
i haven't heard of that before yeah this is a really cool idea and i love it because it's so
simple from uh dan robinson from uniswap was i believe the one first really thinking about this I haven't heard of that before. Yeah, this is a really cool idea, and I love it because it's so simple.
Dan Robinson from Uniswap was, I believe, the one first really thinking about this. The cool idea that they realized is that L2s today have centralized sequencers,
and they basically have sequencing rules that are way different than Ethereum L1.
What I was talking about MEV was in the Ethereum L1. What I was talking about MEV was like in the Ethereum L1 context where you don't know
who the validator is and a random validator selected every block and they can do whatever
they want with ordering the sequence. And what they do today is actually they broker a deal with
these specialized builder actors with MEV Boost and like that's how blocks get ordered on L1.
It's basically very anarchic. You can do whatever you want with your block space. with MEV boosts, and that's how blocks get ordered on L1.
It's basically very anarchic.
You can do whatever you want with your block space.
But because L2s have centralized sequencers, some of them,
and hopefully maybe we'll see L2s with different designs,
but for now the big ones do, you can leverage the fact that the sequencer can now have specific rules
about how they sequence blocks that they'll just commit to.
Now, you are trusting the sequencer to do this, but you already trust centralized sequencers on l2s anyway for instance not
censor you and like include your transactions and so if these centralized sequencers commit
to saying hey we will definitively order our blocks based on priority fee which is like how l1
validators used to do it vanilla.
Like if you just use the normal vanilla code, you would do this.
But the thing is you can actually run this sidecar and maybe boost and do whatever you want
and sequence things actually not based on this priority fee ordering.
But on the L2, this sequencer could commit,
I will absolutely use.
So if you put a higher priority fee than someone else,
you will go ahead of them in that block.
And then the cool idea, if you make this commitment,
which Optimism and Unichain and all the Optimism chains,
I think, or most of them are taking, the OP stack chains,
you now have this amazing property, which is inside a smart contract, you can actually check the priority fee. So like there's an op code you can use to check the priority fee. And Dan Robinson's extremely simple idea was, what if you take whatever priority fee the user used, the user is using to express,
hey, I really want to get into this block because I really want to trade before someone else, or I really want to be the first trade in this liquidity pool so I can arbit. Well,
now you can multiply the amount they're willing to pay in priority fees by 99 and add that as a fee that they must pay to the liquidity providers in order to take the trade.
So whatever you put as your priority fee is also going to affect the fee on your swap.
the fee on your swap. That's the idea. So if you put a high priority fee, your swap fee goes up,
That's the idea.
which means now you don't want to put to, like, if you start putting a higher priority fee,
and now what I said is it's 99 times larger than whatever you added as priority fee
is what you will now pay to the LPs. So this means that every time you add one unit of priority fee,
add one unit of priority fee, you add 99 units of more fee to the LPs, which means 99% of all this
MEV should get recaptured by the LPs. So the issue is like this MEV isn't that bad. It's just like
it's flowing to the wrong parties. It's what a lot of people say. And now with this simple mechanism,
if you have sequencers that work like this, you can be pretty sure that the MEV will flow back to the application.
And what's really cool about this MEV tax idea is it works for like any, it's super generalistic.
It could work for liquidations on lending markets.
Any DeFi primitive could now slap this simple thing on the top of their smart contract functions that checks the priority fee and says, if you're paying priority
fees, you also owe a fee to the application. And the application can thus claw back the MEV they
generate. So yeah, that's the idea of ME2B taxes. Yeah. It sounds like it'll plug some of the holes
in the bucket. That is the goal, right? That is their idea. And it's actually pretty
interesting that it hasn't really been, I think it's still like super early days. So we're not
even really seeing this much, that much of this being used yet in production, but I think we
definitely will. I think it's a very exciting line of research. So I guess maybe one more question
about kind of AMMs and DEXs. How do you see the competitive landscape growing with DEXs?
And what do you think the focus is going to be?
I know with Uniswap, it's kind of been more infrastructure.
But how do you see the DEX landscape growing?
Yeah, it's so interesting.
It's really hard to know. It's a really exciting time because I'm a believer that Univ4 is a really interesting and strong design for the future, nice principles.
I've always been drawn to Uniswap because I feel like they come from a market design first approach over everything else.
approach over everything else, just like what's the right market structure and deeply thinking
Just like what's the right market structure?
about that rather than thinking more from kind of a business angle of like, how do we capture
more users? Of course, they want to do that as well, but they lead with this really kind of
research first, like what's the right design rather than like focusing a lot on the kinds of
incentives they can use with their token and so forth um so i think that's
really cool and i think uni before has a really uh interesting design and a shot to be potentially
like a major hub not only for spot market spot decks spot market swamps like a dex but even
other kinds of d5 primitives which i could see kind of forcing themselves into Univ4 hooks just to access these network effects of this singleton, right?
Where all the tokens are in one single Univ4 smart contract.
And this means that tokens don't have to hop and you don't pay like extra gas fees for all these transfers.
There's just a transfer at the end of the big operation you want to do. And so like, if you get a lot of network effects in Univ4, you could see this crazy
compounding where every other DeFi primitive sort of has an incentive to actually build it somehow
on top of Univ4 or connected to Univ4 in order to access some of the liquidity network effects and
like some of the gas optimization you get out of the the singleton so like yeah i'm really excited about this architecture but at the same
time i think in the real world right things are very path dependent and you have to also win the
market share and it's totally not clear that this is just gonna happen by default right just because
uniswap is this big name. So I think you see a lot of
competition as well. A lot of people innovating sort of in other ways. It kind of reminds me of
like Ethereum versus Alt L1s as well, where people, the sort of smaller competitors are using
different approaches, maybe cutting some corners, but also finding a way to be innovative, right?
And sort of, you know, question whether the big incumbent is really going to be innovative right and sort of uh uh you know question whether the
big incumbent is really going to be the future thing uh so i think you're seeing this today
with spot markets i feel like one thing that really has to be sorted out is interop across
chains in the at least in the ethereum world evm world without that right now things are
at least in the ethereum world evm world without that right now things are really fragmented and
so you you also see things where it's really hard to judge whether something is really going to be
big or or whether it's just kind of playing a short-term game with uh you know you see some
dexes that's you know target a very specific chain and ecosystem and community. This can work really well, but is that really helping us solve the big problem of figuring out how to make blockchains awesome and make our markets better than TradFi markets?
I'm not sure, but maybe.
And so, yeah, there's tons of path dependence.
And you see a huge success to think projects like Aerodrome, which have taken a very different approach than Uniswap in a lot of ways but have absolutely found ways to kind of out compete them in the market
um and uh so the arrow gonna be really arrow velo the aerodrome velodrome model is i think
really interesting because it's making they're taking like a short-term incentive a short a short-term model and stretching it out into the
long term somehow they've managed to really thread that needle and it's it's the most sustainable
it's the most sustainable ponzi that i've definitely ever seen i don't i no i don't
don't give me i mean those it's great team We love working with those guys, but it's pretty wild.
That was a pretty revolutionary way to –
but when you really peel back the layers, it's like, hmm, that's interesting.
You're playing short-term games, but somehow it does work
as long as everybody continues to play the short-term games into the future.
Yeah, exactly.
I mean, to me, and I totally agree.
I mean, I don't want to be, like, making any claims about other people and so forth. But I do think like when you focus on, and there is really cool innovation since Curve onwards, right, of which I feel like even Aerodrome is like a bit somehow related to like, you know, they use the VE model, maybe more like solidly is more than what they were directly based on, but still that comes from this curve world.
And I feel like these, working on the incentive layer isn't meaningless.
It's really cool.
And there are ways to superpower your product and sort of get traction by focusing on actually the kind of incentive part of, oh, we have this token and we can distribute it
in certain ways to incentivize capturing more LP
and maybe, yeah, optimize the usage.
And like this has proven to be very interesting.
I just think, as you say,
one issue is the sustainability in the total long term. And nobody really knows, because it can be really hard. It can, as you said, maybe it's working, it's working well, but is it only working well, because we're not in sort of the, the decline part? And if the decline happens, does it unwind like a Ponzi. I feel like you never know what's a Ponzi until the unwinding happens,
right? And maybe in 50 years, we're going to say the US dollar was a Ponzi. You hear it all the
time, right? But nobody will really feel that. Mom and pop, your parents won't feel that until
they actually feel the pain of the unwinding and go, oh crap, I thought this was real, but now
you guys are, now a few insiders are exiting and we're left with nothing
you know um so i definitely yeah i take back my comment ve tokenomics is not the most sustainable
ponzi the united states dollar is the most sustainable ponzi that has ever been released
yes well yeah so yeah we'll see but i think um uh i think the issue is if you get too caught up with the incentives part, it can really, really work. But like you're kind of putting a bandaid on top of these other structural issues that you can just fix of like the core market design itself, like how the swapping works. Sometimes it feels like you're putting a hat on something that,
like to sort of hide that snuggly face or something. I don't know. I don't know.
And we, yeah, we've, we've definitely, we've definitely seen a lot of that all over the place,
different chains, different protocols. It's, that's the oldest trick in the DeFi book for sure.
Even though there's not that many pages in that book.
Let's double click into the incentives though, because that is a topic that
we've all decided is going to be the main growth lever that the foundation uses to grow some of these big projects, right? The V4 Hooks hooks and unichain and they've got a massive
incentive budget to really kick start that party and the key there is that that is going to be used
to kick start the party right so we've got like we've got kegs we've got pizza at the party it's
great people show up people drink the beer people eat the pizza but what's gonna keep everybody what's
gonna keep everybody and their bags committed to v4 and committed to unichain it's a great question
i mean i think the first step that i really like about the approach that uniswap is taking and
like with their partnership with like gauntlet and so forth. I remember two years ago at ECC, I saw one of the Gauntlet people describing how they were working
with Uniswap to think about how to do liquidity mining incentives in a more principled way.
So like, I feel like the first step is acknowledging that it's not just about
making those tokens go around and kind of selling the idea of we're all going to start earning and
locking and
so forth and thinking more about like, okay, the game is you want to put incentives to bootstrap
and then you eventually want to remove the incentives and see if you actually bootstrap
some activity that is now sustainable. And so like the fact that they were doing that experiment in
that way was already a great signal because it seems like it's focusing not on incentives, right?
Like these other things are literally,
they're structured to be like incentives forever, right?
Curve and so forth.
I mean, maybe curve ends, I don't know.
But I feel like Aerodrome, I think,
is like gonna be always incentivizing.
Like it's embedded inside the structure of the market.
So this is like a very different idea.
It's like, okay, we put incentives
truly as just a
bootstrapping mechanism, which was always like kind of the original idea of this token, using a
new token that you issue to bring liquidity to your DeFi idea was really always supposed to be
this like, first phase acquisition phase. But a lot of times, you know, people kind of psyops themselves into saying, like, let's never turn this off. Basically, afraid to see what happens if you start reducing those incentives. Can you actually bootstrap something that now, like, the idea would be in a DEX that you incentivize LPs so that they come in in the first place. And then that actually stimulates volume
because you have a chicken and egg problem. There's no volume if there's no liquidity yet,
but maybe once there's a good enough liquidity, all of a sudden you really stimulate volume that's
willing to pay fees to these LPs. And then even if you remove the rewards now, the LPs are actually
making a profit just from the fees. This would be like, as long as you're pointed in that direction,
I don't know how easy that's going to be, But like, that seems like the right experiment to run. And so I'm pretty encouraged by that. I think it's still really hard, right, for a DAO to coordinate exactly how to tune all these variables. And, you know, it's $100 million budget or something that you try to allocate in the right way through a bunch of forum posts. Like, I mean, there's still definitely going to be some,
it's not going to go perfectly. But yeah, I think having that approach where you're actually trying
to stimulate something that will become sustainable without incentives and sticking to that and then
really measuring where it's working and where it's not. I think that's a nice start. something that will become sustainable without incentives and sticking to that and then really
measuring where it's working and where it's not. I think that's a nice start. Do you think there's
any room for leveraging like automated liquidity managers to improve incentive programs?
I think so. I mean, it's just really interesting. It all depends. There's a lot of
like minutiae, how you do it. Like, I think they're going to use like the Merkle way of
doing things, which is nice because it's great because it makes it so that both anybody who
wants to participate and like liquidity managers who, you know, on behalf of retail, optimize the positions can all try to gain from these incentives, like play the game of these incentives and provide liquidity.
It's interesting about whether they, we've thought a lot about the different ways that this can be done and what it creates.
ways that this can be done and what it creates. I think what's interesting is,
yeah, right? Once you put incentives, you create a game that's just a pure financial game to just
maximize your share of these incentives. And that's not always the same goal as making this
liquidity really sustainable. So we've seen in the past things like,
you know, liquidity providers that put super concentrated liquidity to get tons
of the incentives, but over a very short timeframe. And it's like, does this actually
help bootstrap that market or not really? Because sure, I mean, it is really good liquidity for
those moments, but it's sort of very erratic. So it's not like over the long term, they're really
helping the sustainable liquidity there. They're kind of being more mercenary and just trying to
be more effective at farming these rewards than others. So yeah, the reason I mentioned that is
because I mean, liquidity managers sometimes like maybe because of like social relations or other things like that, they can be more principled about how they do this thing.
But on the other hand, they could also be the ones to like figure out how to super optimize the like extraction of these rewards in a way that's maybe just like great for you as a yield earner, but not necessarily as great for the market structure.
And those automated managers can do that too.
And if they do, maybe that could create some weird effects as well.
So it could help.
But I think really it's like a liquidity manager is basically just,
you know, a conglomerated liquidity provider, right?
So it's just another actor
and i think they have the same incentives as normal lps is arrakis planning on directly
leveraging v4 for sure before we're already so we're about to roll out a uh before hook
most of the liquidity this like uh liquidity with DAOs and token issuers and institutional liquidity that we have, that's almost all of it's in V3, will migrate to UNIV4 pretty soon.
And then, yeah, use our hook, which allows us to do like dynamic fee setting.
So yeah, we're definitely going to use V4.
In terms of getting these incentives, that's an interesting interesting question too is like what pairs will it be on um if it's mostly like blue chip pairs potentially we
will do something but as i said this isn't our main priority because it's always this thing where
yeah we'll do that for the one for the three months while those are really lucrative but you
kind of have to keep chasing the different pairs and things and we don't see that as our main forte is like being the ones eyeing these
different third party rewards. And so we'll see, depends sort of how lucrative they are,
what pairs do we see an offering that's good for our users. But maybe the other thing is maybe
there'll be incentives on actually the pairs that we use with some of these token issuers,
in which case for sure, we'd be very excited about these incentives. What is something that you wish more people
challenged about DeFi? Yeah, so I mean, I'm like a real like crunchy person, I would say,
or like I got into a real, really principle. Like I got into into crypto really from the idealistic vision of it just like
i said like some guy in my basement completely nerds i use theater director nerds night by the
bitcoin white paper saying like i'm gonna learn every single thing about this um and so yeah to
me it's really about way bigger than these companies, their governance tokens.
We're trying to build something totally new here
that we don't even know what it is
that hopefully can stand the test of time
like things haven't before,
like more than governments have before kind of thing.
And I think in DeFi we sort of it was such an amazing explosion in the first DeFi summer
that everybody kind of uh maybe too easily sort of settled that like this is how DeFi works there
are these VC backed companies and they launch tokens and create these protocols and because they create these protocols and people use them their tokens go up and I think it's not that that doesn't exist or can't be cool
but I think we sort of just we're like oh okay there's this new Silicon Valley of DeFi projects
that totally makes sense it's like this is completely nascent it may not that may not be how development in crypto
actually like stands the test of time even in d5 um and uh so yeah like my biggest the biggest
questioning i will have people want to i would have people question is like are these vc backed
are these VC-backed startup-style projects,
of which Arrakis is among them,
is that what DeFi is and what DeFi needs?
Or are there other approaches as well to DeFi that are really important
and that could really work?
I almost feel like sometimes you can leapfrog some of the things
that these sorts of go-to-market strategies have
if you have a totally different approach.
Who knows?
In my crunchy utopian world where you just build stuff as a developer
and retroactive public goods funding works perfectly, and if if everybody uses it you're actually really well incentivized to just
make it and give it away and not have to figure out why it has like where the business model is
for you um uh like i think in this major unwinding period that we're experiencing right now right um it's a moment where we can also
like chart different paths through d5 um and yeah i mean that's the that's the big challenge i would
give to everybody is like um how close is crypto and d5 to silicon valley and how is it similar maybe in some ways but how is
it different too there are some ways that is very different and maybe we were sort of skeuomorphic
right now and it'll it won't be like that in the future so yeah it's my take that's awesome
you're uh you're a brilliant guy we really appreciate you coming on this was uh the
hardest my brain has worked in a long time.
Yeah, I was talking a lot of random stuff there. No, you're very articulate. You're very articulate and you speak about these very complicated concepts in a way that even I can understand. So I appreciate you.
Great meeting you guys. This was really fun. Thanks for having me.
All right, guys, that's a wrap of another
episode of the ungovernable podcast cassandra thanks again that was awesome for our listeners
please don't forget to subscribe on spotify youtube drop in new episodes every week and
follow us on x at grow uniswap to stay up to date that's another episode. See everybody next week.
This was the Ungovernable podcast by Alpha Growth.
