Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Oasis x Liquefaction: The Future of Ownership
Recorded: Sept. 16, 2025 Duration: 1:02:58Space Recording
Short Summary
In a recent discussion, experts from Cornell Tech and Oasis unveiled 'liquefaction,' a revolutionary wallet system that redefines blockchain ownership and enables private delegation of account control. This project, supported by IC3 grants, aims to launch a decentralized exchange, promising to reshape asset management and trading in the crypto space.
Full Transcription
Thank you. Thank you. Hey everybody, thank you for joining our space on liquefaction with some great speakers,
Donnie and James from IC3 in Cornell.
We're going to give it another minute or two for the space maybe to fill out a bit, and
then we'll start with some introductions.
So yeah, just please be patient for another minute or two and we'll get right into it.
Awesome.
Can you guys hear me?
Yes, we can hear you, James.
Perfect.
Good morning, Will.
Good morning, Donny.
How's it going?
Pretty good here at Cornell. Yeah what's is the what's the city like
today great weather lately not too hot not too cold yeah i don't know i think you bet you had
the right choice going uh to europe in the summer the city can be pretty rough during that time of year. Yeah, I've heard that. What about you, James? Are you located
on Roosevelt? Is that what it's called? Roosevelt Island? Yeah, yeah, I'm at Cornell Tech. Okay,
and Ari is not located there, or he's doing like here, he's doing a mentorship remotely right now?
or he's doing like here he's doing a mentorship remotely right now yeah for this semester um
we're hoping he'll be back in the spring got it okay well um people will fill in as the space
continues uh we'll also you know post it for later and probably make some snippets um but i guess um
right now let's start with some introductions.
For everyone who doesn't know me, I'm Will.
I'm the ecosystem growth manager here at Oasis.
I work on pretty much everything I can to help grow the Oasis ecosystem and product space.
We're joined here today with Danny and James from Cornell Tech, also representing IC3,
to share a little bit more information on liquefaction.
Liquefaction is something we've talked about a bit in the past,
but I think it never hurts to revisit such a groundbreaking technology.
But maybe before we get into all the liquefaction stuff,
maybe James, Danny, if you can do a brief introduction to
yourselves as well as what your guys are doing at Cornell Tech at a high level, what your PhDs are in,
and then what IC3 is as well. Maybe start with James. Yeah, sure. So I'm a PhD candidate at Cornell Tech. I'm advised by Ari Jules, and our lab does blockchain research.
So my latest project has been liquefaction,
which is point of discussion today.
But I have a lot of interest in smart contracts,
and especially the Ethereum virtual machine.
So there's a lot of cool interactions that we study there.
And then to give a brief background of IC3, IC3 is an initiative for cryptocurrency research
headed by a bunch of faculty members at universities across the country and even
across the world. So that includes Cornell, Yale, Princeton, UC Berkeley, University of Illinois, EPFL, ETH Zurich, University College London.
And I think it was created because at the time, like there wasn't actually a whole lot of rigor in these white papers that were coming out from blockchain projects.
And so it was founded to kind of advance the scientific foundation of blockchains. So Flashbots, Avalanche, Arbitrum, the term MEV, and I think even the Ekaden paper that Oasis was kind of formed around all came out of IC3.
Yes, the Ekaden paper. I think that had some heavy hitters i think andrew miller ari don all contributed to
that and it was uh i mean it's not one-to-one exactly like what the oasis consensus is but
our consensus was very much modeled around uh the ekadim paper and i don't know if i'm pronouncing
that right but uh yeah if anyone wants to find it uh maybe we'll share it on Twitter later, or you can look up EIKEDN white paper and you'll find it.
Danny, chance for you to introduce yourself and a bit about what your PhD is in.
Thank you, Will and Oasis for having us here.
It's always good to kind of be able to spread our research.
And I'm also a PhD student here at Ari's lab, Ari Jules here at Cornell Tech.
I work with James mostly in blockchain research, but I'm also interested in.
Other things like cryptography, general privacy, and governance. I also am part of IC3, and
that's what basically has been funding the research here at Cornell in a great part.
Yeah, maybe before we get into liquefaction just like a maybe a thought question um
and i'm sorry i didn't prepare you guys for it so you're on the spot a bit
but um i feel like and maybe i'm wrong but crypto is one of the industries where
academia bleeds very much into you know the. You know, you have Emin, who is at one point co-director of IC3.
Now he's at, you know, the founder of Avalanche.
Ari's working as chief scientist at Chainlink Labs.
Dawn was the founder of Oasis.
Why do you, I think Philip Guyan was in your guys' particular research group and now is the head of Flashbots.
So why do you think it is that in crypto specifically, it seems like academics bleeds into founder or bleeds into the product and space,
whereas in other industries, you see it more like they, you know, maybe you're an advisor or something like that, but you don't normally found a protocol or they're not normally tied so closely together.
Maybe you guys have some research working on the inside, maybe insights working on the inside of this like sort of industry.
Is there any like thoughts you could share about that?
Yeah, that's a really good question.
I've actually thought quite a lot about it because I was surprised when I got into the field how much impact you can have by doing research.
I think it's the fact that security is a very delicate topic.
If done incorrectly, it can bring a lot of harm and many bad things can happen.
We have all seen in
the past and this makes for this topic being researched like the impact of research on the
industry bigger because you can kind of generally trust high academic papers to be safe and reliable more than you can
usually do in many industry companies because an industry company being
successful does not mean it's going to be safe in the long run.
While many papers include security proofs and many conference have very long
proceedings that require them to at least have some level of assurance that what
you're actually building
will be safe in the long run? I have a slightly different perspective, actually. I think
blockchain and crypto as a whole has always been very open. All the way back to the early days of Bitcoin, Bitcoin was like this giant open source project and a way of
opening up finance. And I think everything we do as academics is completely open. So I think it
kind of lends itself to that sort of category. And it's cool that as PhD students, we actually get a lot of independence. So you can
kind of think of it a little bit as training for being able to lead our own projects in the future.
So in a sense, I'm not too surprised that there have been these large crypto companies that have come out of academia.
Yeah, yeah. I mean, it's it's something that I think has only been the benefit of the space
is having the people who are the actual experts be the ones who are sort of leading the charge.
And I think it's something unique about our industry.
And I sort of hate to think of where we would be and how many exploits would exist and what the security would look like had that not been the case.
Maybe one last question before we dive into more of the topics, but more of a surface level question.
What do you guys think? Do you think that, you know, at the end of your PhD, like that might be something that you'd be interested in is taking either liquefaction or any of the other subjects of your research into a product that you're a founder? Is this sort of founder, product manager type role seem interesting to you? Or do you really want to focus on the academia as you move forward?
on the academia as you move forward?
So for me, like there's definitely a tension
between these blockchain companies that need to make money.
And so figuring out like a healthy way of doing that
would probably be a priority of mine
before I try to found a business.
Like in some ways, blockchains have eliminated middlemen,
but right now we have blockchain companies that are basically creating them in order to survive. So if I do go that such an open protocol environment, I would be pretty interested in that for sure.
Yeah.
Finding product market fit, finding the business model is not so clear right now in a lot of industries you know people have ideas but yeah it's still not clear like where
if for example like intense service providers and all where like a lot of the cutting edge right now
in crypto um where the where the way to make money is um other than like you said like acting as a
middleman which is what you actually were initially aiming to to solve um i think that's a friction
but i do actually think the industry will solve it.
It just might take some time to mature.
And then yeah, Danny, any thoughts from you?
Any desires to be the next Emin, the next Philip Dayan,
or yeah, are you gonna stick to the more academic side
or haven't decided yet?
Yeah, I actually have no idea.
I mean, never say never,
but I guess i'm still a
second year phd student so got a lot of years to go and i'll decide when i get closer to the end
what my future path will be okay um well yeah that's i think that's a good intro to set the stage for where blockchain is and what your research is and what IC3 is.
So without further ado, I mean, maybe you guys can give us an update on what your current research is, liquefaction, how it came to be and what problems you're solving.
Yeah, sure.
So what is liquefaction? Liquefaction is a wallet system that challenges existing notions of blockchain ownership. And it also challenges an assumption that most people have about blockchain ownership itself, namely that addresses are controlled by individual entities.
And let me unpack that a little bit.
So when you use a Block Explorer and you look at an address and you see all the transactions that were sent by that address, you assume that that's the same person who sent all of
them, right?
So generally, we don't even trust.
And it allows you to delegate portions of your account, like not everything, but maybe
you want to let someone else sign a particular set of transactions or a particular set of
messages from your account.
And you can even do this in a way that's completely exclusive, like you block yourself out from having control of that anymore.
And when we get into this a little bit further, we'll further explain what that means.
But essentially, you can create new types of blockchain assets at the wallet level using liquefaction that you couldn't do before.
that you couldn't do before. And the other interesting thing about liquefaction is you
can do this basically private delegation of your account or delegation of your account
completely privately. So you can hide to the rest of the world that you're even entering these
agreements for your account. And maybe, Will, you can help me because I know I get straight
into all the technical
terms. So you could probably help me unpack that a little bit. But that's kind of the broad look at
it. Yeah. I mean, I think we'll get into use cases later. But I guess the key is this challenging
the assumption of the single wallet owner right this like you
said you know if you look you would make that assumption and because I think up
until now it has been a safe assumption to make but yeah this challenges
challenges that and with challenging it causes problems but also opens up a ton
of new use cases before we get into maybe some of the use cases,
Danny, maybe do you have any other thoughts that you want to share on like what the research on
liquefaction is? Maybe if you can give a history to the research and like what caused you to start
looking into liquefaction? Yeah, so well, first of all, it's surprising how this really simple assumption that makes so much sense to have.
Like just if I have some coins, then I must have the keys for those coins.
How easily it gets broken and how many kind of repercussions this can have in the ecosystem and different existing use cases that we'll dig into later.
But this actually came from back when Phil Dian was a researcher here at Cornell Tech with Ari Jules.
along with Ari posted a blog post talking about something called the dark DAO.
That was basically a way to privately sell or rent your DAO tokens for specific proposals.
And this is clearly an attack on the integrity of many voting systems
that maybe don't want people renting the tokens or selling them
because that could greatly increase
the ease for bribery.
And the fact that this is done privately can also make it so that the DAO coordinators
cannot even kind of stop it from happening.
That's back in 2018, if I'm not mistaken. So fast forward a couple of years ago, our lab worked on a mitigation for that kind of
attack that we call the encumbrance attack or liquefaction-like attack.
And that mitigation, we'll go back into it later, but it's called complete knowledge.
Basically ensures that the funds are not kind of, you're not able to rent those funds privately.
But while doing that research, the group found that this kind of property or kind of framework that you can use to sell DAO tokens can be used in many, many
other instances. So that's where liquefaction actually comes from. It's kind of a generalization
of this renting DAO tokens concept that has been extended into many other use cases and
formalized into a nice theoretical framework that can be kind of expanded into anything that almost anything that you can
think of now as we'll talk later today. Yeah, maybe we're going to go a little bit out of order,
but I think probably the best way, maybe each of you can pick, you know, two of your favorite
use cases, and then we'll explain on the backend how those use cases work using Liquifaction.
So James, yeah, what are two use cases
that you found powerful or just interesting
using Liquifaction?
Oh yeah, that's a good question.
So yeah, the one that started it all
was the DarkDAO application,
basically private vote selling.
So that one's pretty cool.
I like that one.
And then, so yeah, basically this means you can privately sell a DAO vote without actually making it public that you're letting someone else vote on your behalf.
you're letting someone else vote on your behalf. And then the way this is different from something
like LobbyFi is that you actually, when you sell your vote, you have no ability to override
the vote that you sell. Whoever you sell it to has ultimate control over that vote,
and there's nothing you can do about it afterward. And then the other application,
I'll have to look at the list. There's a lot of cool ones. I know Danny likes the overlay smart contracts. I'll let him talk about that. But let's see. There's a lot of privacy ones we've created. I'll just go to airdrop rights. Basically, with liquefaction wallets, you can sell your entire account
exclusively. And something that we'll get into is with a liquefaction wallet, you actually don't
ever know the private key to your account unless you liberate the account from liquefaction.
So you can completely sell the entire account to someone else and the counterparty
can get the assurance that you won't be able to sign transactions that would steal stuff that's
in it in the future. So what you can do is you can do these things with airdrops where you actually
know that your account is going to be eligible for an airdrop. So you sell the ability to claim that airdrop to someone else before the
airdrop even happens. And I think that's pretty cool.
Yeah. There's so many cool use cases of liquefaction.
There's, we actually propose both constructive use cases and adversarial
use case. One constructive use case I really like as James mentioned is the Overlay Smart Contracts.
It's basically a kind of cross-chain interoperability framework that allows all the blockchains
to meet in the same place and talk the same language.
It allows for blockchains that don't have same place and talk the same language it it allows for blockchains
that don't have smart contracts to have smart contracts and use them in the same way you would
use them in any other smart contract based chain and it allows all the assets to kind of
meet in the same place and you would be able to safely trade bitcoin with ethereum and do
some pretty cool stuff that uh that you could not do before unless like a specific
company had implemented it for that use case
yeah and so so i mean breaking that more, when you say overlay smart contracts, basically what you're saying is you can give Bitcoin all the same capabilities as you have on Ethereum, or you could do it for XRP or Stellar.
Like you can basically create smart contract capabilities for any network and then even cross network. So the smart contracts from one network can interact with ones from other ones.
Is that correct? Or is there anything you added to that?
It's exactly that from an iVert perspective.
Right. I mean, I agree. I think that one is extremely interesting.
I think that on a technical level, you know, we still have a lot to work on in terms of how do we actually implement that, that it's useful and not overbearing from the developer experience or from a user experience.
can tell it's like the first let's say cryptographic primitive um that really makes sense for building
this you know right now there's a lot of things like side chains that exist and people are trying
to do like zero knowledge on bitcoin and stuff like that um but like this is uh at least from
a high level seems like to be a much simpler much more more ready-to-day solution. So yeah, excited to dive into that more.
But so maybe for these three use cases, we can then unpack what an encumbered wallet is and
what is liquefaction doing on the back end. So dark DAO was the first thing that we talked about.
What's actually happening on the back end that allows the dark DAO to have that
capability yeah so um so just to as a reminder like with dark DAO you have some DAO tokens that
are that you um and some DAO proposal that's coming up and you want to sell those DAO tokens
to someone else because maybe you don't even care about this proposal, but someone else really does.
And the person you want to sell it to wants the assurance that you are going to vote the way that they want to vote.
And the way this would work with liquefaction is liquefaction is a wallet system. So, and as you remember with Liquefaction, you don't actually know the private key to your own wallet.
And this is kind of what enables you to enter into these side agreements is the wallet will itself will enforce whatever access control policy is operating on it.
So what you can do is you can create a new policy that says that this person that wants
to vote on your account is now the-
James, maybe before we go further, can you explain how it's possible for you not to know
your private key?
Like what is happening?
What is an encumbered wallet and how is it generated so that you don't know the private key? Oh yeah, great question.
So there are multiple ways to do it, but the way that's easiest to do today with good enough
security assumptions is with something called a trusted execution environment.
And trusted execution environment
is like a secure area of a computer processor
that gives us certain confidentiality
and integrity guarantees about what operates inside.
And this is what helps power Oasis Sapphire, for example.
And we built our liquefaction system on top of Oasis. But
basically what happens is the TE can, it has special hardware protections to make it pretty
difficult for anyone, even the physical owner of the hardware that's running like the software,
like the software to know what's happening inside this TE. And so the TE can generate a private key
and it can encrypt it so that only it can decrypt it. And we can use this so that we know that
this account, for example, is controlled by a private key that's only known to this TE.
And what we do is when we create this liquefaction wallet, we do it in a way that the TE agrees
that it will only allow people to use that account as long as it follows this policy that we give it.
as long as it follows this policy that we give it.
And so that's how you can create a wallet
without knowing the private key
is you now, instead of going to MetaMask
and MetaMask being able to sign transactions itself,
you instead go to the TE
and ask it to sign a transaction or a message
and it will output the signature
without giving you the private key.
So like maybe at a high level
or a different way of looking at it,
would it be accurate to say that it's essentially
a smart contract that has the ability
to have encrypted state?
And so within that smart contract,
you can actually put the private key to a MetaMask
wallet or a Bitcoin wallet or any sort of wallet. So basically the smart contract itself becomes
that wallet. Yeah, that's right. But you need everything to be, you need the privacy part
because the key needs to be kept secret. So you wouldn't be able
to do this using account abstraction stuff on Ethereum or a smart contract wallet on Ethereum,
for example. You need, because we're dealing with private keys and signatures. So we need that to
be stored somewhere such that only the TE can decrypt it. Right. But yeah, it's completely right.
only the TE can decrypt it. But yeah, it's completely right.
So then from there, what's happening within this smart contract wallet or encrypted smart
contract wallet when you give your rights away to vote in a DAO?
Yeah. So basically, that means we're telling the Liquefaction wallet contract to accept this policy over signatures that are made regarding DAO messages.
So basically, previously, your DAO message, maybe you had access to sign it, for example, for previous proposals, DAO proposals that were coming out.
But in future proposals, your wallet will agree that you don't have access to sign those DAO proposals that were coming out. But in future proposals, your wallet will agree that
you don't have access to sign those DAO messages anymore and will block any of your own attempts
to do that. And instead, you're giving it to someone else, maybe for the next proposal,
for example. And so the wallet won't let you sign voting messages, but it will let that other person sign those
voting messages.
Very cool.
On behalf of your own account.
Yeah, it's very cool.
I still am a little bit nervous about the use case.
You know, I think it could be considered potentially a beneficial one but also definitely controversial one if people are
selling their votes and it's happening entirely confidentially and no one knows whether this is
actually you know uh i don't know i always use stani at ave as an example or you know just someone
who bought his votes for the day now i don't think stani would ever actually sell his votes but it's
a little concerning to know that he could and no one would be the wiser.
Yeah. Yeah. There's like there's definitely like existing vote selling things like Lobby Fi.
But with those, you can you can measure the activity that's happening and and stuff.
So with liquefaction, you would be able to do it in a way that's invisible.
But I was going to say something else. What was I going to say? I don't remember. It was something
to do with dark DAOs. But oh, yeah, we actually don't think dark DAOs are a huge threat just yet.
And the reason is that we don't think a whole lot of DAOs are actually very decentralized.
So you can still have the whales just kind of coordinate in a telegram channel or something.
And so we think once DAOs become more decentralized, then dark DAOs will become more of a threat.
That makes sense.
It's like people don't need to buy votes yet because it's already too easy to control
the DAO.
Yeah, pretty much.
But maybe a less adversarial, more beneficial use case.
We'll go to Dani's.
So on this overlay smart contract, it's going to be a lot pretty repetitive, but maybe it'll
hammer the point home.
What's happening on the back end, Dani, when you're building an overlay contract,
let's say on Bitcoin?
Yeah, so the idea is pretty much the same.
We just take it a step harder.
So let's go back to what was happening with voting.
You would create a new address within a TE.
These private keys kept hidden from the user and everyone.
And then you add some kind of policy or logic
on top of the address that restricts the access
to the signatures.
So the logic could be only a person that has sent some funds
to a specific address can claim the signature for the next proposal,
or you could have it be chosen randomly between different set of addresses,
those who can vote on the proposal and those who cannot.
But after all, what you're doing, you're basically adding some kind of logic
that decides who get to access the
signatures or not that interact with the assets that lay within that address. If you think about
it, it's pretty much the same as smart contracts do. Smart contracts basically add logic on top of addresses and decide how the funds or assets that are
stored within that address gets treated or get sent around. So we could actually, instead of
restricting ourselves to just having DAO tokens, we could have any kind of token within that address
and restrict who can send
it, who can receive it, what's the condition for this to happen.
And after all, we could have an actual smart contract on top of a general address.
The cool fact here is that this address does not have to be on top of a smart contract
based blockchain.
Because what we have been talking
about is DAO tokens. Those are in Ethereum, but all the TE logic actually happens in a TE that's
outside of Ethereum. And if we want to have some liveness guarantees, we could even do it in a
TE blockchain such as Oasis. So what if we, instead of doing on top of Ethereum DAO tokens, we did this kind of encumbrance
on top of Bitcoin tokens.
Now we have an address that's generated within the TE, a Bitcoin address, and any fund that's
stored within that address can be controlled by logic that's specified within the T in terms of a smart contract.
That's actually why we call it overlay smart contracts, because we are basically adding smart contracts to non-smart contract blockchains.
Yeah, so maybe what is an example of like an overlay smart contract um i guess you know we've
talked about dexes um in the past we've also talked about you know cdp stable coins um maybe
give an example of like what you could build on bitcoin and how that would work. Yeah, I'll give the most simple example I can think of is imagine Alice and Bob want to trade an Ethereum token for a Bitcoin token.
So that you would need really complicated bridging infrastructure to be able to do that. But if you have this liquefaction framework set up, you could just encumber the two tokens.
Now, their logic is defined by smart contracts in the T.
And now you could have both addresses commit to sending the transaction to the other address.
So how that would work is basically the logic that's implemented for each token is send the token to Alice once Alice has committed to sending the token to Bob.
committed to sending the token to Bob.
Right.
So, I mean, what I think is really cool about that is what you're essentially describing
is what happens on centralized exchanges.
So what you're selling is you can build a decentralized exchange that is as functional
or offers the same security guarantees as a centralized exchange and potentially is far more performant as well
because you can get sort of instant finality maybe not like you know a 10 000 transaction
per second instant finality but you can get the same instant finality you get on a centralized
exchange using this decentralized exchange which I think potentially has the power to
transform, you know, the A and M space that exists today.
Yeah, there's definitely a lot of things that can be studied on top of decentralized
exchanges. I also really like the flexibility task is on a centralized exchange, the kind of
ask is on a centralized exchange, the kind of logic you can add on top of your funds or what you can any developer can implement their own
product on top of this exchange and be able to interact with different funds and be able to
exchange them securely and do whatever they want without the need to actually implement
all the bridging infrastructure right I mean I, yeah, you can build, for example, functionality that's associated with KYC. You can build lending, you can build trading, you can build all of that into like one sort of platform while remaining decentralized.
And yeah, not having to trust like right now, you know, the comparison is like the Bitcoin to Ethereum transaction that you just talked about would require like a rapid coin, which requires BitGo, which requires which means you have a third party custodian that you have to trust.
And then some bridging solution that you have to trust in order to get wrapped BTC on Ethereum.
But then how do you get the BTC back to Bitcoin?
You'd probably sell that on a decentralized exchange
or a centralized exchange,
and then take your stablecoin in the middle
and then buy Bitcoin on the Bitcoin network
or on a centralized exchange.
So it's a lot of steps,
and there's many third-party trust assumptions
that have to be made.
But you're saying that with basically
this sort of overlay smart contracts,
this can be entirely eliminated.
Yeah.
I mean, there's always a benefit to having a company,
generally liability assurance, but decentralization has great benefits too.
So, yeah. Yeah. So, yeah.
Yeah.
Okay, cool.
I mean, I guess changing it up, we talked about some use cases.
I'd like to talk a little bit more now about why Sapphire,
and maybe going a little bit further than why Sapphire,
why these use cases need what we call a persistent state.
Why can't you just do this on Google Confidential VMs or AWS Nitro?
So why did you guys choose one Oasis Sapphire for your TE provider?
And then going further, why you need persistent state?
Maybe James?
Yeah, so, well, I guess we'll actually start
with the persistent state because that's kind of the reason we need a blockchain to begin with.
So when you create a liquefaction wallet, it needs to follow policies, right? But it has to
follow the policies consistently. If, for example, you were able to enter into a policy
like that lets you sell the Dow tokens, but then somehow go back to the state where you didn't have
that policy anymore, then you could almost double spend your liquefaction agreements, which would
not be good. So we need consistency there. And that's why we
need persistent state. We need to protect against any sort of policy rollbacks, because that could
be devastating if, for example, you sold some assets that were in this liquefaction wallet
to someone else, and then could just reclaim them like double spend. So we don't want that.
So we don't want that.
And so there are a couple TE-based blockchains which we could use.
One of them was like Secret Network.
One of them was Oasis Sapphire.
And the reason we chose Oasis was just that it was easy to use and EVM-based.
That was basically the two reasons.
I was experimenting with different TE-based blockchains
for an IC3 summer camp
hackathon project. And that's how I got exposure to Oasis. And that's how we started working on
Sapphire. Well, I hope that since then, maybe there's more reasons to continue with Oasis
Sapphire. I know, yeah, we work a lot on our security and things like that, but I'm glad that you chose Oasis.
We were happy to work with you guys on this research, and we're excited for all the things to come.
Maybe, yeah, going further into that, you know, so we talked about why persistent state, but then at the same time, there are interesting use cases that maybe you would want to do off-chain
computation.
So we have our ROFL framework, what stands for runtime off-chain logic, that also uses
TEs and confidential compute.
And you've said in the past that like this is also interesting for liquefaction.
Why would that be interesting?
And why would that not run into the same issues
that you said with AWS Nitro or Google Confidential VMs
or one of these other TE cloud providers?
Why is one, what are the interesting use cases
that can be done with ROFL?
And then why does ROFL not run into the need for persistent state?
Not that it doesn't need it, but how does it flexibly handle that in ways that other confidential compute providers cannot?
That's a good question.
So yeah, ROFL and then DSTAC was the other one I looked at. Like these are really exciting, like off chain compute things,
because basically in my opinion,
they're fulfilling the original vision for smart contracts.
Like think back to when smart contracts were just coming out and people had
all these like crazy ideas that they wanted to implement on smart contracts,
but like you couldn't actually do all of that today because it would just be so expensive. And because you need
to replicate the same compute on every single Ethereum node or blockchain node that's running
that blockchain. And that's really expensive. Like you don't want to have to recompute like,
I don't know, something like a game server on everyone's computer, that's just not necessary.
And so Raffle is going to be really important for things that only need to be computed once,
but which we can create proofs for that we can put on the main chain like Sapphire.
For example, in our liquefaction project, we rely on this trusted block hash oracle to tell us what the recent
Ethereum block hashes are. And this is how we can kind of establish this root of trust for
establishing on Oasis that Ethereum transactions actually were processed on chain, on Ethereum.
And currently we don't have like a really neat setup for this. We just have like a trusted Oracle that's just kind of a node that uses a couple RPCs to determine which block hashes relate to actual block numbers.
But something we can do with Raffle is we can reduce our reliance on those trusted RPC operators and create our own lightweight block hash Oracleacle using raffle and a light client for
ethereum so this is definitely going to be necessary for these overlay smart contracts when
you want people to be able to add their own chains and you need a root of trust that is not something
so centralized as some rPCs, for example.
Yeah.
Oh, sorry.
You go, Danny.
No, I just wanted to add that I'm really excited for Ruffle.
And also, it's integration with many liquefaction applications that can come up with some of its use.
Like a simple one so people can get an idea of the cool things you can do with it is you could decide who can vote on the next proposal based on who wins a minecraft uh pvp fight uh so that you
can run the minecraft server within a ruffle and then this will kind of generate a proof on the
winner of the fight and will give access to the signature to the winner.
You can make all kinds of goofy and useful and fun things that you could not do by just
having Sapphire alone.
Yeah, we actually, like I actually implemented this with Andrew Miller over the summer.
And to clarify, Mojang slash Microsoft, like they don't like
blockchain. So we actually used open source Luwanti instead. But it's basically a Minecraft,
open source Minecraft game. And so yeah, like gaming and a whole lot of other applications
are now possible in blockchains thanks to things like, like Raffle and Dstack
being able to give us these, these trusted off-chain compute modules that, that let us run
these game servers. And then you can trust that the outcome, like the server was operated fairly,
like previously you'd have to trust that whoever's running the game server was, was fair and wasn't
like rigging the game or something. But now, now you don't have to do that whoever's running the game server was fair and wasn't rigging the game
or something.
But now you don't have to do that anymore.
You can just create this kind of easy to set up and verifiable way of doing things now.
Yeah, that's really cool.
I mean, I guess it's taking, I think last year, Flashbots, I think using D-Stack, but maybe not.
And Andrew Miller, they basically allowed you to take control over a Twitter account for one tweet or for one hour or something.
If I remember correctly, I think the Azuki NFTs did a demonstration on this.
the demonstration on this.
What you're saying is you can take this a step forward
and actually allow using Raffle or D-Stack
the ability to take over assets
or not even just assets,
but you can take over specific subcategories of an asset.
So I don't have the ability to send tokens,
but I have the ability to use a board ape
or something like I can win a raffle
or win a Minecraft tournament and then have access to a board ape or something like, like I can win a raffle or win a Minecraft tournament and then
have access to a board ape for, you know, a week so I can get into their private server or something
like that. Yeah, exactly. So liquefaction lets you like create these really fine-grained policies
that even the assets themselves don't let you get so deep on. Like, for example, currently, if you wanted to do some sort of delegation for your NFT,
the NFT has to support it somehow.
But with liquefaction, you don't even have to worry about what the protocol
or the smart contract that you're using can do
because you can create your own policies that work at the wallet level.
So you can create these policies that are really specific that the existing contracts would only
authenticate via address. And now you can create policies for specific capabilities that your
address can do and then sell that or rent that to someone else securely. Right. So, I mean, the way I look at that, I always try to think about it,
like where's the breakthrough in that, is that you're actually,
even if the, like,
it's almost like you can create an Ethereum overlay contract on Ethereum,
like add functionality to smart contracts that already exist and do things in
smart contracts on Ethereum or any network
that like previously were sort of impossible or like too intertwined to separate can now be
separated into like all the individual aspects of like a token or a smart contract.
Exactly. Like the DAO token example is a really good one. DAO tokens let you transfer the DAO token,
basically like sell it.
Or like in addition to that,
you can vote with your DAO token.
But with liquefaction,
you can split those two into two different people.
Like maybe one person has the ability
to sell or transfer the token.
The other person can vote on the next proposal.
And the way these votes are set up, those two actions don't actually conflict. Whereas in the existing system, you have to have
the same address and you can do either one of those, but you can't split the two.
Right. And all of that happens confidentially. And so we did have a question earlier that was
talking about Vitalik's recent post post i don't know if you
guys seen it i know you guys focus mostly on the academic side so maybe you're not like in crypto
twitter as much as the rest of us but vitalik recently you know shared some thoughts about um
you know the need for private for privacy on ethereum and basically from everything that i'm
hearing with look affection this is a way that you can sort of
without upgrading Ethereum or without, you know, requiring like Ethereum to enable their own TEs or
build, you know, massive ZK layers, which even the ZK layers would have problems because you
can't really share a state. But yeah, without like a massive overhaul of how Ethereum currently functions,
you can basically achieve, you know, any level of privacy on Ethereum
with these sort of encumbered accounts and liquefaction
and then sort of the separation of sub policies.
Is that something maybe you guys want to go into
or explain the nuances there a little bit more?
Yeah, I guess the broad picture, in my opinion, is that liquefaction enables private off-chain coordination with people you don't trust, pretty much.
So kind of that was previously only something that like these centralized exchanges could do, although you had to trust them. Like you can kind of think of that, how like you deposit to a centralized exchange,
it's basically going into their account and then you can withdraw it later. But everything you do
on the centralized exchange is kind of private to the centralized exchange. Of course, the privacy guarantees aren't as clear
as you can get with Liquip Action.
Yeah.
Yeah, I guess also that privacy is also a delicate topic.
We have seen that with TornadoCash,
making legally compliant privacy when it comes to financial things, it can get
difficult. And I think liquefaction provides with a really flexible environment where you can
implement CYK and integrate that with your application really easily, and even use Ruffle for off-chain CYK assurances and stuff like that.
So I think liquefaction is going to be pretty much the way to go
unless Ethereum does a big change on their whole protocol.
Yeah, and I know Vitalik sometimes is involved with IC3 so I'm counting
on you guys to to petition him I'm just joking but yeah I mean it's exciting to see just in
general the privacy is becoming you know a bigger trend in web3 and it's coinciding with this sort
of breakthrough research that you guys and others are doing.
I guess, you know, we're getting to only having like eight minutes left.
And I want to give some time to discuss what comes next.
So we've sort of built the we didn't get into Board API Club, but I think people have, you know, in our community have already heard about it.
You can sort of rent an ape for two minutes for the price of like a couple pennies you can enter the server um you can go to exclusive events
and prove that you technically own that ape for that you know period of time that you've rented
it for and it all it comes on the back of liquefaction that was sort of the proof of concept
um we've gone over some use cases that was published in the paper what's the next step for your guys
research how do we start building a lot of these cool ideas that we talked about like
overlay smart contracts or you know like the minecraft using raffle where are you guys going
from here and how do you see that working? Yeah, so we're working on our next
liquefaction projects and it really takes the overlay smart contracts idea and runs with it.
Kind of as we've discussed before, you can use liquefaction to basically create a decentralized
coinbase and we're trying to do that. We're working through some of the technical
issues and the architecture that would be best, but kind of there's this other component,
security, and anyone who's done research with trust execution environments knows that there's
been a bit of a checkered history there where you've had side channel attacks on TEs. And I think Oasis's design
has already helped eliminate some of the security issues, but we want even greater assurance
for when people are now having these high value assets on liquefaction wallets. So we first have
to get through the security, like how to upgrade the security and get better guarantees there. But then also like think more about how assets from
different blockchains can interact and what sort of policies would enable that. But yeah,
that's our next project. Basically a decentralized centralized exchange on top of liquefaction.
centralized exchange on top of liquefaction.
And any insights right now as to like, well, not like insights, but like any gut feelings as to
like what the first step is in improving security or improving upgrade and policies?
So we've come up with a couple different security techniques and we're working on
synthesizing them and figuring out what's
actually compatible or not. There are a couple of things you can do to basically eliminate the T
holding onto the key. But if you don't have a T holding onto the key, then other people
won't be able to sign directly through liquefaction that way. So there's a couple
things you can do. You can have pre-signed messages
that you coordinate with the TE, but the TE only knows the signature for it. It couldn't
reconstruct the private key itself. So using key sharing, for example. And then there's a couple
other ones. We have on-chain logic, like an Ethereum smart contract that helps enforce some things
even further in case, for example, the TE system gets compromised, then maybe you can fall back
on some on-chain smart contract to protect against attacks from the TE system that was compromised.
to protect against attacks from the TE system that was compromised.
And then we've also come up with some other defenses like, for example,
basically an early warning to the TE system being compromised and using that as a hedge against
total failures. Like if you can compromise this
small part of the T, that's probably easier to do than the rest of the system.
And you get a bounty for that, then that might protect or at least warn early that the rest of
it's in danger. So there's kind of a lot of different components that we can compose that we're trying to figure out which ones actually can work together for this cross-chain decentralized exchange specifically.
One follow-up question there is like TEs, side channels, and the fact that it is a trusted hardware, that's sort of like the one point that everyone attacks whenever they want to oppose TEs or any of the technologies built on TEs.
Even outside of liquefaction and like the AI compute space, people are still trying to fight this friction of like TEs are ready ready today, but they're still maybe not as ideal
as people want them to be.
And sort of the consensus seems to be that like,
one day everything will be ZK or FHE.
If we do find,
like if in this liquefaction research,
you do find, let's say a solution for TE security,
at least for liquefaction,
do you see like there being any downside to TEs going forward then?
And why you would prefer, let's say, an FHE solution?
Or probably do you think that liquefaction will always run on TEs at that point?
I guess sort of they have to, but yeah, maybe just share some thoughts about that.
Yeah, that's a good question.
I think we have to work with what we
have these currently like are secure in many ways and they could get broken eventually it's there's
no like uh mathematical proofs uh proving their security because it's based on top of hardware security. And while I think that we can, or people can find use case specific ways
to reduce these security threats,
I don't think there's going to be a general way
to securities in the long run.
It seems like there's always going to be
an underlying hardware assumption.
And you can have different use cases and each use case can have different ways to kind of
make it safer to the threads on that specific use case. But having a general way to secure
hardware systems, I don't think that's going to happen personally.
I actually think, like, I think there are advances in TE design.
I know there's a group that's been working on, like, trustless TEs, and I'm thinking about ways of constructing them that are, like,
especially difficult to break, like using certain hardware techniques,
like Puffs, for example example but there are limitations to
all of these like they they don't have all the properties that we need for liquefaction for
example so like I'm optimistic that there will be more developments in the TE space
certainly TEs are the most efficient way of doing what we need to do for liquefaction. And so for the time
being, like we're going to be using TEs because they're just easier to use and they give us the
speed and cost that we are looking for. Yeah, it makes a lot of sense. I mean, it's a hard
question to answer to always to predict the future, right? But yeah, I appreciate both your insights and yeah, you know, sort of explaining the
nuance of cryptographic proofs versus hardware-based proofs.
So we're at the hour now.
I will, I guess for now, this is the end of our conversation, but we'll probably have
a lot more in the future, hopefully soon as your next phase of research comes out.
Maybe just a chance to sign off and plug, you know, any of the upcoming projects, upcoming
events or anything that people can look forward to or where they can follow you guys for more
information.
Yeah, sure.
So yeah, we're working on this next liquefaction project. I definitely want to
remind everyone about Take My Ape that basically lets you become the owner of a board ape on
Ethereum and like it literally gets transferred to your liquefaction wallet. And you can go into
their exclusive discord and other things for fun, which is pretty cool.
And it's pretty cheap.
It only costs like basically an Ethereum transaction
plus a couple of Oasis transactions to do.
So that's at takemyape.com.
You can follow initc3.org,
which is the account I'm on right now, IC3,
for more updates about our research.
And we'll definitely post any new projects we have
on the Twitter page.
Or X page, I keep forgetting.
Yeah, absolutely.
Definitely check out TakeMyApe,
whether you're just someone who's always wanted
to see what TakeMyApe, what a board ape is who's always wanted to see what, like, TakeMyApe, like, what a board ape is, and, you know, you wanted to enter their member space, or maybe see if you could enter an event using one of these apes.
That's one use case.
Or maybe you're an early adopter of technology, and you want to just get a concrete, hands-on experience with some of these more abstract ideas and see, like, what's, you you know what does it feel like to use a liquid faction app and maybe that'll inform you when when when this uh these more uh
let's say developed concrete versions of these other use cases we've talked about come out
um i definitely recommend you do that and definitely recommend you follow ic3 on the
oasis side you know we're going to be um
thinking korean blockchain week in the coming weeks uh and then definitely we'll be in buenos
aires for devcon so if you're going to either one of those please reach out to us maybe we can
you know see you at an event or host an event together um and yeah thank you guys for joining
this space uh see you at the next one have a great rest of your day thank you guys for joining the space. See you at the next one. Have a great rest of your day.
Thank you for having us.
Yeah, thank you very much.
Bye James.
Host
Oasis
from
Oasis Network
