Latest / Crypto Coin Show / Secret votes, public proof: Gnosis Guild's Auryn Macmillan on The Interfold
Transcript
- 0:00Speaker 1: Every worthwhile democracy on the planet uses a secret ballot
- 0:05Speaker 1: for a really good reason.
- 0:07Speaker 2: Even if I want to, I cannot prove to you
- 0:09Speaker 2: how I voted.
- 0:10Speaker 1: We have these competitive, mutually distrustful institutions. This type of
- 0:14Speaker 1: infrastructure is kind of the only viable option. He's shown
- 0:19Speaker 1: a real desire for this type of thing to exist
- 0:21Speaker 1: for almost a decade.
- 0:24Speaker 2: I trust me, but why should you trust me? And
- 0:26Speaker 2: why should anyone else trust me?
- 0:36Speaker 3: I'm Ashton Addison from the Crypto Coin Show. And today
- 0:38Speaker 3: on Blockchain Interviews, we have Oren McMillan, co-founder of Gnosis Guild,
- 0:43Speaker 3: the team behind Interfold, a network built by Gnosis Guild
- 0:47Speaker 3: that lets people and organizations work with secret information like
- 0:51Speaker 3: votes or bids and still verify that the results are
- 0:56Speaker 3: legitimate and real. This sounds like an extension of ZK Proofs,
- 1:01Speaker 3: but with getting people together and ensuring that votes count.
- 1:06Speaker 3: And I feel like the government should probably pick up
- 1:08Speaker 3: on this eventually, which maybe we can dive into. The
- 1:11Speaker 3: network alpha for Interfold went live in August alongside of
- 1:15Speaker 3: the Fold token. And now they are rapidly expanding. Oren,
- 1:20Speaker 3: welcome to the show. And thanks for taking the time.
- 1:23Speaker 2: Thanks for having me, Ashton.
- 1:25Speaker 3: Yeah, I would love to start out with a high
- 1:28Speaker 3: level on what exactly are the use cases for this
- 1:32Speaker 3: in working with secret information and And why do we
- 1:35Speaker 3: need cryptography for that?
- 1:37Speaker 2: Yeah, absolutely.
- 1:39Speaker 1: So I guess, yeah, at a really high level, there's
- 1:42Speaker 1: this broad class of use cases where you want to
- 1:46Speaker 1: take many private inputs from many different parties, arbitrarily compute
- 1:51Speaker 1: over them, and then produce a verifiable outcome.
- 1:54Speaker 2: So you hinted at two of these already, voting and auctions.
- 1:59Speaker 1: I think of these two really canonical examples, right, where Voting,
- 2:02Speaker 1: every worthwhile democracy on the planet uses a secret ballot
- 2:07Speaker 1: for a really good reason. A secret ballot is collusion resistant,
- 2:12Speaker 1: coercion resistant, bribery resistant.
- 2:15Speaker 2: By divorcing your vote.
- 2:19Speaker 1: From your ability to essentially prove how you vote, making
- 2:21Speaker 1: your vote receipt free, you disarm this whole class of
- 2:25Speaker 1: attacks that can fundamentally break the integrity of these voting systems.
- 2:30Speaker 1: The same is true for auction systems. Typically, a sealed
- 2:37Speaker 1: bid second-price auction is considered the most economically efficient auction type. Traditionally,
- 2:44Speaker 1: to run these, you've got to have some trusted third party,
- 2:48Speaker 1: some auctioneer, either as a person or as a hardware
- 2:53Speaker 1: device that takes all of the bids and computes the outcome,
- 2:58Speaker 1: the winner and the price they pay, and then reveals that.
- 3:01Speaker 1: But if that auctioneer is compromised, then it can kind
- 3:08Speaker 1: of leak strategy and break the economic efficiency of that auction.
- 3:11Speaker 1: And so, yeah, in both cases here, the kind of
- 3:13Speaker 1: integrity of the system, the secret ballot or the sealed
- 3:16Speaker 1: bid auction is determined by its capacity to keep inputs
- 3:22Speaker 1: and intermediate states private and ideally to make those inputs receipt-free,
- 3:27Speaker 1: such that as someone participating, even if I want to,
- 3:31Speaker 1: I cannot prove to you how I participated, how I voted,
- 3:35Speaker 1: or what my bid was. And it turns out that
- 3:39Speaker 1: this is a really, really hard problem. You mentioned that
- 3:41Speaker 1: it sounds somewhat like ZK, and I mean, rightfully so.
- 3:49Speaker 1: ZK is a component to it, but it turns out
- 3:51Speaker 1: that ZK alone can't achieve that set of qualities. In
- 3:57Speaker 1: order to prove something with a zero-knowledge proof, you must
- 4:02Speaker 1: know the data that you're trying to prove. And so
- 4:04Speaker 1: it was actually this dilemma that originally inspired the Interfold
- 4:08Speaker 1: to exist at all. We were playing with another technology
- 4:12Speaker 1: called MACI, the Minimal Anti-Collusion Infrastructure, and that is voting
- 4:19Speaker 1: inside of a zero-knowledge proof. and had this fundamental trust
- 4:23Speaker 1: assumption where an honest coordinator needed to receive all of
- 4:29Speaker 1: the encrypted bids, decrypt them locally, run the circuit over
- 4:34Speaker 1: those inputs, produce a tally and the corresponding proof that
- 4:38Speaker 1: the tally is correct. But it meant that that coordinator
- 4:40Speaker 1: could see all of the inputs. And so for a
- 4:43Speaker 1: handful of quadratic funding rounds, I ran that coordinator software
- 4:47Speaker 1: and kind of felt this pressure very personally, right? I
- 4:51Speaker 1: trust me, but why should you trust me? And why
- 4:53Speaker 1: should anyone else trust me? Not to leak that data
- 4:56Speaker 1: and kind of compromise the integrity of the quadratic funding round.
- 5:01Speaker 1: And so this was kind of the inspiration then for
- 5:04Speaker 1: us going out and saying, how do we improve on
- 5:05Speaker 1: those trust assumptions? How do we distribute trust in that
- 5:09Speaker 1: trusted entity, the tallying entity and voting, the auctioneer in
- 5:13Speaker 1: an auction? How do we distribute trust in that such
- 5:16Speaker 1: that no single party has the capacity to distribute, sorry,
- 5:20Speaker 1: to decrypt and then leverage that information, you know, ultimately
- 5:24Speaker 1: become corrupted. And so it was the addition then of
- 5:28Speaker 1: fully homomorphic encryption and threshold cryptography that allows us to
- 5:32Speaker 1: get this combination of guarantees in terms of verifiable execution,
- 5:36Speaker 1: encryption over, computation over encrypted data, and then threshold decryption
- 5:41Speaker 1: of the output.
- 5:43Speaker 2: So that was a huge monologue.
- 5:46Speaker 4: No, it can get technical, definitely.
- 5:48Speaker 3: And I've heard that fully homomorphic encryption, FHE, and some
- 5:53Speaker 3: of these more new age 2026 projects, that's, I don't
- 5:58Speaker 3: know if it's a step up from ZK, but it
- 6:00Speaker 3: sounds like it's a more holistic approach and it's the
- 6:03Speaker 3: next level of technology for encryption for private data.
- 6:07Speaker 1: Yeah, I mean, I wouldn't say that it's a step
- 6:08Speaker 1: up necessarily, or like a different level. More so, I
- 6:12Speaker 1: would say it's a different set of affordances. So the
- 6:16Speaker 1: two are very complementary. In order to do verifiable FHE,
- 6:21Speaker 1: you need some proving system. And so you end up
- 6:24Speaker 1: using some flavor of ZKPs in order to prove things
- 6:27Speaker 1: about private data individually. And so a good example of
- 6:30Speaker 1: this is the distributed key generation ceremony that our nodes run.
- 6:37Speaker 1: they are all passing information back and forth to one another,
- 6:40Speaker 1: and every piece of information that they pass back and
- 6:42Speaker 1: forth carries with it as well a proof that they've
- 6:45Speaker 1: done their part of the job correctly. And so, yeah,
- 6:50Speaker 1: zero-knowledge proofs are kind of baked in at a very
- 6:52Speaker 1: core level to each stage in the Interfold's kind of
- 6:56Speaker 1: end-to-end process.
- 6:57Speaker 2: The same is true for an end user who's.
- 7:00Speaker 1: Wanting to participate in a program, so say a voter
- 7:04Speaker 1: or a bidder, they end up generating a bunch of
- 7:08Speaker 1: proofs client-side to essentially do things like input validation, to
- 7:14Speaker 1: prove that they're a valid bidder, a valid voter, that
- 7:19Speaker 1: their bid or their vote or whatever it is is
- 7:21Speaker 1: correctly formatted, that they're voting with equal to or less
- 7:27Speaker 1: than their amount of vote weight and all this kind
- 7:29Speaker 1: of stuff. So you do all of this input validation
- 7:31Speaker 1: at the edges And you prove that your ciphertext is
- 7:34Speaker 1: correct rather than having to do input validation in the
- 7:38Speaker 1: kind of core FHE program.
- 7:42Speaker 3: Now, Gnosis Guild has been around for many years. And
- 7:46Speaker 3: I think many people I know have heard of Gnosis,
- 7:50Speaker 3: but they may not know exactly what the team has
- 7:53Speaker 3: been working on. And it's evolved over the years. It's
- 7:56Speaker 3: been many crypto cycles. Can you talk about some of
- 7:59Speaker 3: the main initiatives of Gnosis Guild? and how that led
- 8:03Speaker 3: to finding this important issue that the Interfold is working
- 8:08Speaker 3: on with the private data?
- 8:10Speaker 2: Yeah, absolutely.
- 8:11Speaker 1: So Gnosis Guild, we spun out from Gnosis in 2023.
- 8:15Speaker 1: This is around the same time that projects like Calswap, Safe,
- 8:21Speaker 1: Karpacki spun out from Gnosis as well.
- 8:23Speaker 2: So there was a bit of.
- 8:26Speaker 1: Just an organizational shift in how Gnosis was operating at
- 8:29Speaker 1: the time and kind of made the decision to spin
- 8:31Speaker 1: a handful of these teams out. We, internally at Gnosis,
- 8:37Speaker 1: prior to spinning out, had been focused on designing and
- 8:41Speaker 1: implementing the Gnosis DAO, the kind of infrastructure for the
- 8:43Speaker 1: Gnosis DAO. And this led to us building a suite
- 8:48Speaker 1: of tooling called Zodiac. It's a smart contract layer for
- 8:53Speaker 1: essentially nuanced access control on top of the safe. And then, yeah,
- 8:59Speaker 1: spun out from Gnosis in 2023, brought Zodiac along with us.
- 9:04Speaker 2: And so that was a kind of primary focus.
- 9:07Speaker 1: And actually they just, the Zodiac team now just spun
- 9:09Speaker 1: out from Gnosis Guild now. So three years later, after
- 9:13Speaker 1: kind of coming along with Gnosis Guild for this journey,
- 9:17Speaker 1: they've now spun out and just announced their Angel round,
- 9:20Speaker 1: which is really exciting. The Zodiac's kind of growing up
- 9:25Speaker 1: further now, um, which is exciting. And I think we'll
- 9:28Speaker 1: see some, some really cool stuff from, uh, from the
- 9:30Speaker 1: VODAC team now. Um, the, we've had a handful of
- 9:35Speaker 1: other projects that we've worked on in the interim. Most
- 9:38Speaker 1: of them have, uh, we've canned it at various times,
- 9:41Speaker 1: you know, there's just haven't, haven't, uh, panned out the
- 9:44Speaker 1: way we wanted. I haven't got the kind of traction
- 9:45Speaker 1: we wanted, or we've just, uh, eventually kind of shifted
- 9:48Speaker 1: focus elsewhere. Uh, and the interfold is, is really Gnosis
- 9:51Speaker 1: Skills kind of primary focus, uh, at the moment. Um,
- 9:54Speaker 1: And it really came around, I think, as an extension
- 9:57Speaker 1: of the same kind of obsessions that we've had collectively
- 10:01Speaker 1: this whole time, which is just this kind of access
- 10:04Speaker 1: control obsession. And in particular, access control and system integrity
- 10:09Speaker 1: from the perspective of large distributed organizations or large distributed
- 10:14Speaker 1: groups of people. And so, yeah, it's kind of an
- 10:18Speaker 1: extension of and reflection of that, that ongoing obsession.
- 10:22Speaker 4: Yeah, that sounds really exciting. I would love to speak with,
- 10:26Speaker 4: with Zodiac as well.
- 10:27Speaker 3: And it's great to have a lot of spin outs
- 10:29Speaker 3: because there's smart people always coming up with good ideas
- 10:32Speaker 3: and building quickly, especially in the age of AI, build it,
- 10:36Speaker 3: figure out the product market fit. If it's, if there's
- 10:39Speaker 3: no issue, if there's an issue or no traction, move
- 10:43Speaker 3: on to the next idea and keep building. So sounds
- 10:45Speaker 3: like fast paced moving team.
- 10:48Speaker 2: Yeah, yeah, exactly.
- 10:49Speaker 1: It's, I mean, validating ideas and then, yeah, either, either
- 10:54Speaker 1: kind of letting them run or kind of killing them
- 10:56Speaker 1: off is just part of the, the, the, the cycle
- 11:00Speaker 1: and the process.
- 11:00Speaker 2: Definitely.
- 11:02Speaker 3: And privacy was one of the earliest narratives in, in
- 11:08Speaker 3: cryptocurrency 15 years ago. And other narratives came up in,
- 11:13Speaker 3: as the industry grew and as though in the last
- 11:18Speaker 3: two years, compliant privacy has become a huge narrative as
- 11:22Speaker 3: well and regrowing now that institutions, enterprises are getting involved.
- 11:27Speaker 3: They want to not show all their cards while they're
- 11:30Speaker 3: trading or doing their business and all of these things.
- 11:35Speaker 3: How does what Interfold is working on fit into that
- 11:38Speaker 3: compliant privacy narrative for large institutions?
- 11:41Speaker 2: Yeah.
- 11:43Speaker 1: So I think The two types of systems that I
- 11:46Speaker 1: described earlier, so voting and auctions, fit into this kind
- 11:50Speaker 1: of broad class of systems where the integrity guarantees of
- 11:55Speaker 1: that system are really, to a certain extent, dependent on
- 11:58Speaker 1: its capacity to keep secrets, right? I think there's a
- 12:02Speaker 1: whole other class of use cases for the Interfold. where
- 12:05Speaker 1: the value proposition is more to do with the access
- 12:10Speaker 1: to or liability from data that you want to produce
- 12:14Speaker 1: some insight from. And so a concrete example of this
- 12:17Speaker 1: would be something like, say, a researcher wanting to do
- 12:22Speaker 1: some research over data held by a handful of different
- 12:25Speaker 1: medical institutions. in order to gain access to that data.
- 12:29Speaker 1: So first and foremost, the researcher cares about the insight
- 12:33Speaker 1: that they get from that data. They care about the
- 12:36Speaker 1: result of the regression that they run over this data
- 12:38Speaker 1: or something like that.
- 12:39Speaker 2: They don't care about the data itself.
- 12:42Speaker 1: And in order to produce that output, historically, they've had
- 12:45Speaker 1: to go and take custody of all of this data,
- 12:47Speaker 1: which means a whole bunch of HIPAA compliance type hoops
- 12:52Speaker 1: to jump through in order to take custody of this
- 12:55Speaker 1: medical data. And then once they've got custody of it,
- 12:57Speaker 1: they've got this huge liability.
- 12:58Speaker 2: On their hands.
- 12:59Speaker 1: If they mishandle it and it leaks, then they are.
- 13:03Speaker 2: Liable for it.
- 13:04Speaker 1: And so this both access and liability is kind of
- 13:08Speaker 1: burdened to ultimately produce the insight that they care about,
- 13:13Speaker 1: the kind of results of their research. And so I
- 13:16Speaker 1: think what I hope here is that this type of
- 13:18Speaker 1: technology will ultimately reduce both of those things, right? By
- 13:23Speaker 1: never having to take custody of the plain text data,
- 13:26Speaker 1: hopefully reduce the compliance burden that you have in order
- 13:31Speaker 1: to run your computation and get your insight. And then
- 13:35Speaker 1: also by never having to take custody of the plain text,
- 13:38Speaker 1: you hopefully reduce the liability that you have for that
- 13:44Speaker 1: data potentially leaking. And so that same pattern, I think,
- 13:50Speaker 1: applies across a whole bunch of different fields where it's this, again,
- 13:54Speaker 1: access and liability burden that hopefully we can reduce and
- 13:58Speaker 1: reduce make a really compelling way of producing insights without
- 14:05Speaker 1: those reduced access and liability issues.
- 14:09Speaker 2: Definitely.
- 14:10Speaker 3: The HIPAA-compliant and medical information, that's something that, although more
- 14:16Speaker 3: niche for researchers, is important, of course. And I feel
- 14:20Speaker 3: like that would also apply to personal information, passports, documents.
- 14:25Speaker 3: We're seeing leaks inside of the crypto industry, at least.
- 14:29Speaker 2: Of addresses, names. Yeah, this is a huge issue. And
- 14:33Speaker 2: I think.
- 14:35Speaker 1: The KYC AML complex is a massive potential opportunity here
- 14:41Speaker 1: as well. And similarly, the broader fraud prevention complex, there's
- 14:48Speaker 1: a really interesting opportunity there for better identity and credit scoring,
- 14:54Speaker 1: risk scoring, fraud prevention scoring standards to yeah, improve the
- 15:00Speaker 1: liability issues here while still giving the, again, AML, KYC
- 15:05Speaker 1: fraud prevention assurances that institutions need for their compliance.
- 15:12Speaker 3: It would be great to get these major institutions on board,
- 15:15Speaker 3: credit scores and all, but I feel like we have
- 15:17Speaker 3: a little bit of work to go to get to
- 15:19Speaker 3: that level. We're still in the early adopter, early innovator phase.
- 15:24Speaker 3: The network alpha went live in August for the Interfolds.
- 15:28Speaker 3: along with the fold token. Now, what does that mean
- 15:32Speaker 3: in terms of what stage it's at and what does
- 15:34Speaker 3: live mean operationally?
- 15:36Speaker 1: Yeah, so the fold token is a critical component to
- 15:41Speaker 1: the network. It's what node operators stake in order to
- 15:45Speaker 1: spin up and run a node. and it becomes then,
- 15:47Speaker 1: or I should say they bond to spin up and
- 15:50Speaker 1: run a node, and that is what is essentially at
- 15:53Speaker 1: stake for misbehavior. So there's a system of economic rewards
- 15:59Speaker 1: and penalties for good and bad behavior, respectively. The network
- 16:04Speaker 1: being live means that the contracts have been deployed, they're
- 16:08Speaker 1: switched on, and E3 requests can be made, and E3
- 16:11Speaker 1: is an encrypted execution environment. It's the kind of whether
- 16:14Speaker 1: we bundle in the various different cryptographic and economic components
- 16:19Speaker 1: into something that is reasonably cohesive. And yeah, where we're
- 16:26Speaker 1: at right now is essentially in a dogfooding phase. So
- 16:29Speaker 1: we have the Interfold's own governance built on top of
- 16:34Speaker 1: the Interfold. And so, so far, the Interfold is the
- 16:37Speaker 1: only entity that has made E3 requests. And this is
- 16:41Speaker 1: partially because For the first couple, we unpaused the protocol,
- 16:45Speaker 1: made the request, and then paused the protocol immediately after
- 16:48Speaker 1: so that we could have a very constrained set of
- 16:51Speaker 1: first attempts. It's now unpaused more unilaterally. And so what
- 16:57Speaker 1: we'll hope to see here over the next couple of
- 16:59Speaker 1: weeks and months is more protocols, more teams coming in
- 17:03Speaker 1: and starting to experiment with it live, starting to integrate
- 17:06Speaker 1: it into their own governance workflows. And this is both
- 17:10Speaker 1: within the Web3 ecosystem via the Aragon plugin, Aragon extension
- 17:14Speaker 1: that we have, and then also beyond the Web3 ecosystem.
- 17:18Speaker 1: We're really looking to push into some more... more traditional
- 17:24Speaker 1: use cases really cross the chasm into the rest of
- 17:27Speaker 1: the world. And so looking at pilots and things like
- 17:30Speaker 1: local government and state-level government, political polling, and then, yeah,
- 17:35Speaker 1: we'd love to push into things like corporate governance as well.
- 17:39Speaker 1: For the voting use case, the auction use case is
- 17:43Speaker 1: coming up very soon.
- 17:44Speaker 2: We are.
- 17:46Speaker 1: Finalizing the last handful of cryptographic components here. So there's
- 17:50Speaker 1: an extra layer of proving that needs to happen and
- 17:53Speaker 1: an extra layer of FHE components that need to be
- 17:57Speaker 1: generated during DKG to enable auctions.
- 18:01Speaker 2: That circuitry is all done now.
- 18:04Speaker 1: And so it's a matter of getting it into the
- 18:06Speaker 1: kind of production code base and getting the necessary parts
- 18:09Speaker 1: through audit so that we can switch.
- 18:11Speaker 2: On auction and make that another use case as well.
- 18:15Speaker 1: So yeah, there's a lot of, I guess, exciting progress
- 18:18Speaker 1: coming up here over the next few months.
- 18:21Speaker 3: And as this progresses into larger corporations, companies outside of
- 18:27Speaker 3: the early adopters of crypto, what is the value for
- 18:34Speaker 3: the companies that are using it to use a shared
- 18:37Speaker 3: infrastructure or a decentralized network versus running something like this
- 18:41Speaker 3: on their own private servers?
- 18:43Speaker 1: So I think where it's most valuable is where there
- 18:48Speaker 1: are very strong needs or kind of desires for, where
- 18:54Speaker 1: there's a benefit to collaborating between competitive or mutually distrustful entities.
- 19:01Speaker 1: And so you can imagine things like, say, a handful
- 19:04Speaker 1: of competing banks, right? They have a bunch of proprietary data, say,
- 19:09Speaker 1: on their risk scoring for various different customers. They also
- 19:13Speaker 1: probably have a huge overlap in their customer base. And
- 19:17Speaker 1: so they could collectively benefit from aggregating those risk scores,
- 19:21Speaker 1: but traditionally they would not because none of them would
- 19:24Speaker 1: trust the other's infrastructure in order to share that secret data,
- 19:28Speaker 1: produce the aggregation. And so in this kind of scenario
- 19:31Speaker 1: where you have these competitive, mutually distrustful institutions that can
- 19:35Speaker 1: collectively benefit from an aggregation, this type of infrastructure is
- 19:39Speaker 1: kind of the only viable option.
- 19:40Speaker 2: Okay. Okay.
- 19:46Speaker 3: You mentioned with the fold token that's used to run
- 19:50Speaker 3: the nodes and lock up and have your stake at
- 19:54Speaker 3: risk for misbehavior. What, you know, what companies, if they
- 20:02Speaker 3: don't want to go through those processes and what if
- 20:05Speaker 3: they want to pay in ETH or stable coins, is
- 20:08Speaker 3: that Is that a possibility in the future or is
- 20:10Speaker 3: this going to be stuck to using the fold token?
- 20:13Speaker 3: And does that increase barriers to entry for larger enterprises
- 20:17Speaker 3: to use the protocol?
- 20:19Speaker 1: Yeah, so the fold token is not a payment token
- 20:22Speaker 1: for the protocol.
- 20:23Speaker 2: Payments are made in a stablecoin.
- 20:25Speaker 1: So when you make a request to the protocol, you
- 20:27Speaker 1: pay the protocol fees in a stablecoin. It's USDS, Sky's
- 20:31Speaker 1: stablecoin at the moment. And node operators earn a portion
- 20:37Speaker 1: of the protocol revenue.
- 20:38Speaker 2: So they get paid in.
- 20:42Speaker 1: Proportionally to the number of node operators and the complexity
- 20:44Speaker 1: of the job that they're performing. They get a portion
- 20:48Speaker 1: of the protocol fees. And so, yeah, from a barrier
- 20:51Speaker 1: to entry point of view, I think it's as low
- 20:54Speaker 1: as it can be while being a kind of Web3
- 20:55Speaker 1: native protocol. the fold token itself operates kind of at
- 20:59Speaker 1: a layer below that as part of the economic security guarantees.
- 21:03Speaker 1: And actually, kind of diving into that further, node operators
- 21:06Speaker 1: have a kind of dual staking model where they stake
- 21:09Speaker 1: both the fold token as their core component, that's the
- 21:15Speaker 1: kind of core bond for them running a node, and
- 21:17Speaker 1: then they also stake a stablecoin bond, they put down
- 21:21Speaker 1: a stablecoin bond to... augment essentially their likelihood of being
- 21:26Speaker 1: selected for a committee. And so essentially for each thousand
- 21:33Speaker 1: SUSDS that they stake, they get an extra ticket into
- 21:36Speaker 1: each round of sortition. So each time a committee is selected.
- 21:41Speaker 1: And this is essentially a soft civil resistance. It makes
- 21:45Speaker 1: it more economically efficient to stack your bond on top
- 21:50Speaker 1: of one node rather than spreading it.
- 21:52Speaker 2: Across many in an.
- 21:56Speaker 1: Attempt to kind of get more likelihood of being selected
- 21:59Speaker 1: for a given committee. So yeah, this was a very
- 22:03Speaker 1: explicitly designed mechanism to make it more economically efficient to
- 22:08Speaker 1: kind of keep all of your bond in one place.
- 22:13Speaker 2: And then on top of this, it means that.
- 22:14Speaker 1: There's less volatility in the economic security of the protocol
- 22:20Speaker 1: because the stable coins as well as the fold tokens
- 22:23Speaker 1: are kind of at risk. So there's always fluctuations in
- 22:28Speaker 1: the fold token, which obviously has a kind of floating price,
- 22:32Speaker 1: don't dramatically impact the economic security of a given E3
- 22:36Speaker 1: because there is this associated stable coin bond as well.
- 22:40Speaker 1: And then I guess the other component to this from
- 22:42Speaker 1: a node offerer's perspective is that bond is in the yield-bearing.
- 22:47Speaker 2: Version of the stable coin.
- 22:48Speaker 1: So state such that they are at the very least
- 22:54Speaker 1: earning the stable coin kind of savings rate on that token.
- 23:00Speaker 2: And so not kind of eating that opportunity cost.
- 23:04Speaker 3: Okay. Well, I think stable coins, that's going to be
- 23:08Speaker 3: a major improvement for companies that are, as the industry
- 23:15Speaker 3: progresses in the next few years, Stablecoins, I think many
- 23:19Speaker 3: companies will be very comfortable with stablecoins.
- 23:21Speaker 4: At least I hope they will, because we're.
- 23:23Speaker 3: Getting to that point where the dollar and stablecoins should
- 23:27Speaker 3: become synonymous.
- 23:29Speaker 2: Right, right. I agree with this. I think that that
- 23:32Speaker 2: as a barrier to entry will become less and less
- 23:34Speaker 2: of a thing over the coming months and years.
- 23:38Speaker 3: Now, I think it was about a year ago, Vitalik said,
- 23:43Speaker 3: His latest paper was talking about privacy and integrating privacy
- 23:48Speaker 3: into the core of Ethereum, which I think is important,
- 23:52Speaker 3: not just for enterprises, but for everybody. I understand that
- 23:56Speaker 3: Vitalik participated in the auction and in some way in Interfold.
- 24:01Speaker 4: Can you talk about that? And does he have a
- 24:03Speaker 4: belief in this kind of protocol?
- 24:05Speaker 2: Yeah, I mean, I think he's shown a.
- 24:10Speaker 1: real desire for this type of thing to exist for
- 24:14Speaker 1: almost a decade at this point. Um, so he, Vitalik, uh,
- 24:20Speaker 1: Barry Whitehead and WJ Coe were the, the original authors
- 24:23Speaker 1: of Macy that I spoke about earlier, the minimal anti-collusion infrastructure, uh,
- 24:27Speaker 1: which was an explicit attempt to solve for this class
- 24:30Speaker 1: of problems. Um, And then, you know, I mean, in
- 24:34Speaker 1: Vitalik's writings, ever since then, periodically, he's brought up this
- 24:39Speaker 1: as a problem that doesn't yet have a good set
- 24:41Speaker 1: of solutions for, but, you know, something that's very, very
- 24:45Speaker 1: clearly needed to just improve the state of the world.
- 24:48Speaker 1: So it seems to be something that is pretty well
- 24:50Speaker 1: aligned on. In terms of participation in the Interfold here,
- 24:55Speaker 1: gave some really great feedback early on, has periodically chimed
- 24:58Speaker 1: in with kind of feedback on the protocol with encouragement
- 25:02Speaker 1: and just generally has been really supportive. And then, yeah,
- 25:07Speaker 1: I believe participated in one round of the auction, which
- 25:10Speaker 1: was great to see that support as well.
- 25:14Speaker 2: But genuinely, the.
- 25:17Speaker 1: Protocol feedback and the willingness to kind of engage and
- 25:22Speaker 1: be supportive and be critical as well on things actually
- 25:26Speaker 1: we've designed and rolled it out.
- 25:28Speaker 2: That's been, I think, the most valuable contribution.
- 25:33Speaker 1: And then, yeah, touching on the kind of privacy at
- 25:36Speaker 1: the base layer, this is something I'm super excited about
- 25:41Speaker 1: and is very, very complementary to the interfold, or I
- 25:44Speaker 1: should say, or the other way around as well, kind
- 25:46Speaker 1: of vice versa.
- 25:46Speaker 2: They are very additive.
- 25:51Speaker 1: The more privacy that we have at the base layer,
- 25:54Speaker 1: the less, I guess, additional hoops and Rube Goldberg machines
- 26:00Speaker 1: we have to set up at the Interfold level to
- 26:03Speaker 1: mask where interactions come from and ultimately make the information
- 26:10Speaker 1: that's being aggregated in the Interfold.
- 26:14Speaker 2: Less traceable, more receipt-free. A good example of this would
- 26:19Speaker 2: be for an auction.
- 26:25Speaker 1: The Interfold can make the bids themselves receipt-free, can make
- 26:31Speaker 1: the auction verifiable, can do all of this. But if
- 26:36Speaker 1: you can see that an account on Ethereum has deposited $ 1, 000
- 26:42Speaker 1: into the auction contract, you can make a very educated
- 26:46Speaker 1: guess that that's the upper bound of their bid in
- 26:49Speaker 1: the auction because you see.
- 26:51Speaker 2: This transparent deposit flow in.
- 26:55Speaker 1: Or if you've architected it such that it's based on approvals,
- 26:58Speaker 1: then you can see what's the max approval that they've got,
- 27:01Speaker 1: what's the max balance that they've got in their account.
- 27:03Speaker 1: And so you can make some very educated guesses about
- 27:05Speaker 1: the maximum possible bid that they could have made, and
- 27:08Speaker 1: so you can reintroduce some of the inefficiencies. And so
- 27:12Speaker 1: one way that we've talked about solving for this is
- 27:15Speaker 1: through integration with Railgun, having bids come out of something
- 27:20Speaker 1: like Railgun so that it's much more opaque who's participating,
- 27:25Speaker 1: what kind of balances are they participating with, and all
- 27:27Speaker 1: this kind of thing. And I think this, doing it
- 27:32Speaker 1: with Railgun would be awesome. I don't think Railgun's going anywhere,
- 27:34Speaker 1: and I think in terms of anonymity sets, it seems
- 27:36Speaker 1: like it's got the largest, most active anonymity set of
- 27:41Speaker 1: kind of modern privacy protocols. But yeah, more optionality on
- 27:46Speaker 1: this type of thing, and particularly if it starts to
- 27:48Speaker 1: become more viable at the base layer, is beneficial as well.
- 27:52Speaker 1: And again, it's purely additive. If we've got multiple anonymity
- 27:55Speaker 1: sets that people can draw from in order to participate
- 27:58Speaker 1: in financial tools that are built on top of it,
- 28:02Speaker 1: then yeah, this is a huge win. And then from
- 28:06Speaker 1: the other perspective, the privacy at the base layer will
- 28:11Speaker 1: carry the same set of limitations that we talked about
- 28:14Speaker 1: earlier in terms of it would deal well with individual
- 28:21Speaker 1: and kind of party to party privacy. but will not
- 28:25Speaker 1: be able to handle the kind of aggregation of state, again,
- 28:31Speaker 1: for the same fundamental reasons, right? In order to produce
- 28:34Speaker 1: a proof of something, you have to know the data
- 28:35Speaker 1: you're proving. And so without some kind of NPC layer there,
- 28:41Speaker 1: that type of information aggregation, maintaining the kind of receipt-freeness
- 28:49Speaker 1: property is just not possible.
- 28:51Speaker 2: There's a kind of fundamental trade-off there.
- 28:55Speaker 3: There's a lot to work on, Oren, and the team
- 28:58Speaker 3: sounds like they're working pretty hard. But I'm curious on
- 29:03Speaker 3: the next stages for the Interfold and moving out of
- 29:06Speaker 3: the dogfooding stage. What are the next early adopters and
- 29:10Speaker 3: how will the team grow this beyond just the own
- 29:14Speaker 3: use case?
- 29:16Speaker 1: Yeah, so as I said before, probably the next handful
- 29:21Speaker 1: of adopters Organizations that will end up using this will
- 29:24Speaker 1: come through the Aragon integration. So we have a plugin
- 29:30Speaker 1: already set up for Aragon, and so it should be
- 29:34Speaker 1: relatively easy for organizations to come in and essentially, if
- 29:37Speaker 1: they're already using the Aragon stack, then plugin. in their
- 29:40Speaker 1: private voting extension and fire that up offer for new
- 29:46Speaker 1: organizations to kind of make it part of their deployment flow.
- 29:51Speaker 1: We're also looking to push into a handful of, again,
- 29:55Speaker 1: use cases outside of the Web3 space.
- 29:57Speaker 2: I have a couple of.
- 29:58Speaker 1: Pilots that we're exploring there that we can't talk about
- 30:01Speaker 1: just yet, but I'm really, really excited about.
- 30:05Speaker 2: And then, yeah, the other side of things is auctions.
- 30:07Speaker 2: So as I said before, the.
- 30:11Speaker 1: Circuitry that there's a kind of components for making auctions
- 30:14Speaker 1: viable is all done, actually.
- 30:17Speaker 2: It was RR and our team sent me a message
- 30:18Speaker 2: last night that was really surprising.
- 30:20Speaker 1: It's much more done than I thought it was, which
- 30:23Speaker 1: is great. I was expecting the DKG timing to be
- 30:28Speaker 1: in the order of five or six hours or something
- 30:30Speaker 1: at the moment, and it turns out it's maybe down
- 30:33Speaker 1: to more like an hour and a half, which is
- 30:36Speaker 1: great news. And so, yeah, just a matter of kind
- 30:40Speaker 1: of polish and integration into the protocol there. and getting
- 30:44Speaker 1: through whatever audit pipeline that we need to to get
- 30:47Speaker 1: that into production readiness. And then I think the other
- 30:52Speaker 1: challenge there is just finding who is the first integration partner.
- 30:56Speaker 2: So auctions are.
- 30:59Speaker 1: different beast of a product to governance because we don't
- 31:03Speaker 1: have an immediate opportunity to dog food it ourselves.
- 31:07Speaker 2: We need something to be selling for auctions to work.
- 31:11Speaker 2: And so we'll probably look to find an integration partner.
- 31:13Speaker 1: There to do the dog fooding on our behalf, to
- 31:18Speaker 1: do the first handful of auctions.
- 31:20Speaker 2: And so Yeah, figuring out who that is.
- 31:23Speaker 1: We have a handful of leads that we've been working
- 31:25Speaker 1: with for some time already and just need to do
- 31:28Speaker 1: the prioritization and figure out who is the best option there.
- 31:32Speaker 3: Yeah, I would love to follow along with that and
- 31:35Speaker 3: potentially participate privately in the auction. What's the best way
- 31:40Speaker 3: to learn more about the Interfold overall and to follow
- 31:43Speaker 3: along with the progress?
- 31:45Speaker 2: Yeah, absolutely.
- 31:46Speaker 1: I'd say jump over to theinterfold.com. or follow us on
- 31:50Speaker 1: Twitter at The Interval. We're pretty active on both. And
- 31:55Speaker 1: I think, yeah, if you're interested in the protocol, the
- 31:57Speaker 1: docs are incredible. Dive in there, just docs.theinterval.com.
- 32:01Speaker 2: They're really, really thorough.
- 32:03Speaker 1: Shout out to Hamza and Ali and Marv on our team.
- 32:07Speaker 2: They've been just polishing these relentlessly. And so.
- 32:13Speaker 1: Yeah, it's a really good spot for anyone who wants
- 32:15Speaker 1: to dive in and learn more. There's also a link
- 32:17Speaker 1: to a Telegram community with a handful of operator side
- 32:23Speaker 1: and just kind of general question and feedback and all
- 32:25Speaker 1: that kind of stuff. So plenty of opportunities to jump
- 32:28Speaker 1: in and engage for anyone that's interested.
- 32:31Speaker 2: Sounds great, Oren.
- 32:32Speaker 3: I can leave those links in the description as well.
- 32:34Speaker 3: I appreciate you taking the time. All of the best
- 32:37Speaker 3: to you and your team and building out these privacy
- 32:41Speaker 3: protocols to More than just the early innovators of crypto,
- 32:46Speaker 3: this is something that the world can benefit from, governments
- 32:49Speaker 3: can benefit from, everybody. It's just a matter of breaking
- 32:54Speaker 3: down those barriers to entry, making it easy, extrapolating out
- 32:58Speaker 3: the crypto tech side and getting it adopted. So thank
- 33:03Speaker 3: you so much for taking the time and we'd love
- 33:05Speaker 3: to follow up again in the near future.
- 33:07Speaker 2: Yeah, absolutely. Look forward to talking again and thanks so
- 33:09Speaker 2: much for having us.
- 33:15Speaker 1: Thank you.