Steven Goldfeder(Arbitrum) on Why the World's Biggest Institutions Are Coming On-Chain

Speaker A: The question was how much money would a nation state, say, have to pay in order to destroy Bitcoin?

Speaker B: Do you think scaling is solved?

Speaker A: I don't think scaling is ever really solved because, uh, scaling is always like a bit of a cat and mouse game.

Speaker B: Having the ability to stop it and having that report going to become not only an institutional need, but also probably a compliance requirement.

Speaker A: Core systems should be as final and permissionless and trustless as we can get them to. Can you freeze these funds that are suspected to be from North Korean hackers?

Speaker B: Arbitrum was one of the very first L2 chains that ah, targeted the blockchain trilemma of security, decentralization and scalability. Stephen Goldfeder, uh, along with his mates at uh, Princeton, his PhD mates and even his professor went deep into it. They wrote a paper and then launched uh, the Off Chain Labs company which went on to launch Arbitrum. Arbitrum has come a long way since then from solving and fixing scalability to now positioning itself as one of the preferred chains for financial market infrastructure as more and more financial markets go on chain. Let's look into this with Stephen on Unhashed. Hello Stephen and welcome to Unhashed.

Speaker A: Hey, it's great to be here.

Speaker B: Stephen, you recently, uh, posted about a rebranding of Off Chain Labs to just off chain. So what's. So there's obviously some, you know, thinking around um, how the future of um, Arbitrum and Off Chain is and m moving beyond um, uh, deeper R and D thinking around scaling Ethereum more into infrastructure for uh, financial markets and stuff like that. So what exactly is this rebranding all about? What is different about off chain versus off chain Labs?

Speaker A: Yeah, so on its face it's a relatively simple rebrand. We dropped the word labs and dropped uh, some color. But uh, actually I think there's something a lot deeper there which you've hinted at in your question, which I'll elaborate on. And the idea is that actually if you look at the evolution of off chain labs and actually the evolution of Arbitrum as well, we started off literally in a lab at Princeton University. I was a PhD student there, uh, together with Harry, my co founder, another PhD student, Ed Felton, um, my third co founder who was a professor at Princeton. So we started off in a research lab and uh, we founded the company in 2018 and Arbitrum was still very much a project in the lab, if you will. It took us three years, but not until 2021, three years into the company till we actually launched the product on Net and of course we are a full enterprise uh, services company. What that means is if you are an enterprise, so say you're Robinhood and you want to build on the blockchain, you want to build on Arbitram, whether that's your own private chain like Robinhood's doing or that's you want to build on the arbitrum. On Arbitrum 1, the public blockchain like so many others, you know, in the tokenization space. Blackrock, uh, Franklin, Templeton, Wizntry, Invesco. Right. So whether that's your own chain or the public chain, we at off chain can provide you the full stack of infrastructure to do that. And that's basically chain stack if you need that is uh, wallet level infrastructure. If you look at how the products we have at off chain you basically operate at three different levels. So of course we're most well known for Arbitrum which is layer two, so the middle layer. But actually we also build Prism which is the core Ethereum client that powers the layer one and around every big Ethereum upgrade we've been instrumental from the merge to every upgrade you've seen since then the Prism team has been one of the most instrumental teams in helping those uh, big Ethereum upgrades come about. And then actually that's so that's the moving down from Arbitrum and if we move up from Arbitrum we also have the zero dev platform that we build in house and this is the leader in smart wallets and smart accounts. So if you want to do things like gas abstraction or build wallet experiences with policies that, that would take a lot of the difficulty of using crypto away from users and have um, smart wallets with smart contract base as well as TE based policies. We build those as well in house. And the idea is if you're an enterprise that wants to launch on Arbitrum today or on Ethereum today, uh we have the full speed of services and vertically our research team and our engineering can go across that. We can give you all the resources you need and customizations as well. Whether it's privacy you want or maybe it's compliance, whatever you need. We are the one stop shop to uh, bring the biggest institutions of the world on chain. We're doing it with Robinhood and we can do it with, we uh, will do it many, many, many others as well.

Speaker B: Okay. I think there's so much to unpack in everything that you sort of said in that opening uh, uh, response and uh, I think that that will probably Take up the role of uh, an ASH today. But yeah, it's very interesting to sort of touch upon all of those um, surfaces. I'd like to ask you, uh, most people in the space know about the Princeton story and how you guys got together. Started off, uh, with very deep thinking around scaling, uh, solutions. Take a step back uh, there and share with me what is it that inspired you? What was it about cryptography or what exactly was this sort of solution? What did you want to fix that really inspired you, uh, to think through this and all of you to come together?

Speaker A: Sure. So I'll go back a little bit before. Where I usually start, um, actually as an undergrad is when I got interested in cryptography as a subfield of computer science. I was studying math and computer science and got pretty interested in cryptography and actually um, in a summer research internship, stumbled my way, uh, on uh, MPC or multi party computation. Um, and I was working with a, ah, really top uh, professor in the field named Yehuda Lindell, um, who is one of the leaders in this space. And I was just an undergrad, very interested in this and trying to essentially uh, learn uh, more and uh, something that fascinated me. And at that time, this is probably 2011, I had never heard of uh, bitcoin or cryptocurrencies. Really uh, knew nothing about that. And then uh, over the course of the next year or so, uh, I became more interested in cryptography. Uh, and when I applied to PhD programs, I decided to basically apply for crypto PhD programs. At that point crypto meant like uh, cryptography exclusively, at least to me. Um, so it was uh, 2013, actually it was a visit. When I visited Princeton, uh, as an admitted student before, uh, I actually went there so early 2013, that was when I actually was first introduced to this idea of Bitcoin. And this was they had these lightning faculty talks. So they had all the admitted students together and different faculty members gave these five minute talks. And this guy that I never heard of before, or maybe I heard of, I definitely met him before, named Ed Felton, gave this five minute talk on bitcoin. And actually I remember the talk he gave was about a paper he wrote on what's called the Goldfinger attack. And the question was, how much money would a nation state say have to pay in order to destroy bitcoin? Remember this was like 2013 and the number was like in the hundreds of millions of dollars. Actually, uh, kind of shocking if you think about that today. And then that night actually at the uh, the Princeton Security students played, uh, there was a poker game for them where um, the, it was played in Bitcoin. And I think the department, uh, gave each person a half a bitcoin to uh, uh, you know, uh, as everyone, to sort of to uh, enter the game. So uh, that was uh, really my first notion and introduction to cryptocurrencies. But I was still very interested in crypto. And ultimately I decided to come to Princeton. And it was there during my first uh, year there that my advisor, uh, another professor named Arvind Narayanan, was beginning, uh, to explore cryptocurrencies himself. And we went on that exploration path together. So uh, we did some interesting work, uh, at Princeton on cryptocurrencies. It started off with, um, I was fascinated by this question of securing digital assets. Right. The idea was that there were so many hacks, even at the time, uh, people didn't know how to secure these digital assets. It was single key. You lose your key, someone else accesses your key, you either lose all your money or um, uh, someone else gets access to all your money and can steal all your funds. And this is happening all over the place. And even institutions had basically no idea how to uh, secure their digital assets. So I, you know, putting back together my mind from my, for my internship a couple of years back, this idea of multi party computation. I became very interested in this idea of can we do threshold signatures or signatures that distributed the signing authority to multiple parties for the signature algorithm used by Bitcoin, uh, which is ecdsa. And it turned out that there were really no protocols at the time that could do this. Um, there had been like some work in the literature a decade earlier, um, that touched it and did like preliminary protocols that worked in other settings but not really in the cryptocurrency setting.

Speaker B: And this is basically multisig. It's.

Speaker A: Yeah, basically a multi party computation multi sig. So not a multi sig in the sense that the blockchain has multiple signatures, but basically you produce a single signature. Um, but multiple people have to combine their information and their secrets in order to do that. Exactly. Um, and interestingly enough, just how, you know, funny the world works. That paper from 10 years earlier was written by this, uh, one, uh, of the authors was this person named Rosario Gennaro, who I had actually met the prior year when I was uh, taking classes at Columbia. Um, so I got back in touch with Rosario and said, hey, let's like dig up that research from a decade earlier and try uh, to reapply it to this Space. And that led to many years of fruitful collaboration. Uh, Rosario and I published a lot of papers on multiparty uh, computation and threshold ecdsa. And actually fast forward to today Rosario, that research team I mentioned, which is Corda Off Channel Labs. Rosario was a member of that research team today. So we still have the very fortunate opportunity to work together. But that was my sort of first uh, interest in crypto. Over the course of my years in Princeton we also went on to write uh, the leading textbook in cryptocurrencies, Computer science textbook and cryptocurrencies are potentially the only textbook. Uh, and cryptocurrencies, ah computer science textbook and cryptocurrencies that's used in hundreds of university courses, uh, in several languages today. And uh, got involved in many other different uh, interesting areas. But one of them was this Arbitrum project. And Arbitrum actually started off with my co founder, now my now co founder but then a professor in the group, Ed Felton had this idea and said hey, um, before smart contracts were even a thing, right before any smart contract platform existed but people were writing about them. So there were papers that discussed them discussing the limitations and understood the limitations and basically said like we need to build this two layered uh, basically fraud proof uh system in order to uh, make smart contracts actually scalable and work. And again this was very theoretical at the time. There was no scaling issue because there was no actually deployed system of smart contracts. And actually if you go back and you look the very first, uh, this wasn't the very first time Ed talked about it but uh, early on there was a class seminar where Ed led a project with some undergraduates. And this was I believe the fall 2014 semester in Princeton. There's actually a YouTube video that has like the class, uh, if you look for the first mention of Arbitrum, that has the undergraduate presentation from them from back, I think it was early 2015 and that date is interesting because that was six months before Ethereum went live, which went live in July of 2015. So the idea was that Ed was thinking about this problem very very early on before there was any system like this. Ed went off to the White House for a few years and Arbitrum basically just uh, lie fallow. Ed was the uh, cto, deputy CTO of the United States under President Obama. And when he came back to Princeton in 2017 that's when Harry and I knocked on Ed's door and said well now there are deployed systems, smart contracts. We're seeing some of these scaling issues, or at least we thought we were that you, that you, that you hypothesize uh, would exist. Let's actually take arbitrum, build it in uh, an academic context. You know, with rigor, uh, uh, at a graduate level you look to publish a paper. And that's exactly what we did. We published it in UserNext Security in the summer of 2018, in August of 2018. And actually in August of 2018 is when we launched off chain Labs and moved our work from the Princeton Research Lab to the off chain Research Lab. Research lab.

Speaker B: That's a fascinating um, story and sequence of uh, uh events and it's very interesting that uh, Ed Felton actually thought about scaling problems um, even before actually the infra existed and uh, there was an actual live situation of uh, uh, L1 uh scaling problems.

Speaker A: Yes, it was definitely uh, very very early. And that's one of the things uh, as an academic that makes you a successful academic is being able to see some of these problems really early on. Um, it's been uh, eight years in the company now working with Ed and Harry and we still work together very closely every day and still feel like it's early on and we uh, solved many problems but uh, we have a lot more problems that we need to uh, continue to solve in order to make the industry ready for the next wave of mainstream adoption.

Speaker B: So clearly from there to now and sort of dropping uh, the labs and getting more application focused in many ways. Um, do you think scaling is solved? Is the trilemma solved?

Speaker A: Um, you know I don't think scaling is ever really solved in some sense because uh, scaling is always like a bit of a cat and mouse game where you solve it and then more demand comes and then you solve it again. So I don't think we're ever at a point where we're scaling a solve. That's number one. Number two is any blockchain system is always going to have some limitations. So enough spam on the network will always uh, get it to reach its limit. That's fundamental because um, you have to have ah, a metering resource, a metering tool for what's essentially a limited resource. And even if you, you know, make it really, really, really large, there's going to always be some limits. So I don't think scaling is ever solved. But I think we're at a very, very good spot now. We're continuing it off chain to invest in, you know, what is the next generation of scaling. How do we uh, increase the capacity of our chains from you know, 10 uh, gigas to 100 gigas to 10 mega gas to 100 mega gas to a gig of gas etc and this is work. We made a ton of progress in uh, both things. We've already launched and also internally built systems that will continue to increase the capacity of these chains. So no, I don't think scaling is ever done but I don't think it's the biggest problem today. Today I actually think the biggest problem is uh, what we're seeing today is institutions are coming on chain and they're very, very uh, excited to come and bring their users on chain. But they need to have a user experience that matches uh, essentially the experience that their users expect from Web2 and some of the basic requirements, uh, that just their users, whether they're institutional users, business users or just retail users are going to want to and expect. And you need often um, very custom and clever solutions in order to get there. Right. So we have the core blockchain infrastructure, we have a lot of block space. But the question is if I'm a Web2 institution or um, I'm a fintech institution, I have these users and I have these assets. The question is how do I bridge that, how do I bring those users on chain? Because a lot of the institutions, they understand the benefits of coming on chain. We call this the programmable economy. Right. Today in the Digital but still Web2 like system, a lot of the systems are manual, they're not programmable. You can um, you know things are digitized but there's no like uh, the markets are very, very manual. And you can't just program uh, and enter the system and automatically connect to the liquidity of others and just build on top of the work that others have done. And the question is how do we facilitate institutions who want to come on chain, who to wants to be able to do that and bring their users on chain. So we have to understand what are the expectations and need of institutions. That can be privacy, those can be compliance, those can be just user experience. And how do we actually bridge those things together. And that's one of the nice things that we're building at off chain because like I said we have expertise across the stack from the layer 2 down to the layer 1 up to the wallet smart infrastructure account layer. And we can build custom solutions for enterprise across all these categories and basically help them join the programmable economy. Programmable economy no uh, matter what their needs are.

Speaker B: No, um, exactly. As institutional adoption happens with programmability in financial markets, uh, clearly user experience that Abstraction is important. Privacy is important. We've spoken about all of these things. Uh, you know, these are like upfront right now. Perhaps one of the other things which is less discussed is, um, um. How shall I phrase it, let me call it, um, recourse. Uh, uh, so when um, mistakes happen, for example, uh, or breaches happen, having the ability, in spite of being a decentralized system, uh, having the ability to stop it and having that recourse, that's also pretty much going to become not only an institutional need, but also probably a compliance requirement going ahead. Uh, I have two questions, uh, for you here. Um, I think one is the obvious one. You know, how difficult was the decision more recently and just to give context when this whole DAO issue happened, uh, and uh, our different foundations stepped in, froze the assets, um, we are disturbing assets. So walk us through that whole decision, uh, making system and where, you know, and how, how, where you stand on this whole requirement, uh, for recourse. Uh, and secondly, you know, taking a step back, uh, as somebody who comes with sort of the old foundation of decentralized systems, distributed architectures, how do you balance this in, In. In uh, the, the very core architecture of uh, uh, uh, you know, blockchains.

Speaker A: Uh, yeah, so it's a great question. Um, and actually, you know, this was back to my early work on mpc, uh, threshold signatures. One of the basic, if you read our early papers like this is like one of the, the biggest, um, you know, motivations that we use for that work and said if you make a mistake here, there's often no recourse. Right? You can't. In your bank, if you send the money to the wrong person or so. And so you can usually figure it out, right? You know, especially, particularly for large numbers, they track the money down and often it works out well. But obviously in crypto, um, many of the systems are just fining. But of course some of the systems are not actually, um, that final. So there's a few questions and the question is like, number one, you could break this down, like, what is the ultimate goal for finality? Uh, number two is like, um, what should we do in situations where there uh, is something that we can do to sort of help a user or help a class of users or you know, stop a nefarious, uh, you know, nation, uh, state from gaining, uh, access to funds that will use to say, support terrorism? Right. A lot of different, uh, um, uh, issues that are sort of mungled together and they might not necessarily have the same uh, answer to them, but Fundamentally, I guess I'll answer the second part first. Where um, this is like core to the programmable uh, economy. And the idea is that you can come onto the blockchain and you can actually program your rules and your policies in ways that uh, actually have the protection that you want. Right? So for example you can spin up a zero dev wallet and have some spending policy for users that requires maybe multiple sign off, that requires maybe spending uh, limitations per day, that has limitations of where they can send those funds to. So actually you can put guardrails in place. And this is one of the things that I was referring to which is, you know, the blockchain is not, I don't view it as like here is a solution that fits everyone. I uh, view it as programmable. So what is an institution needs to come and say is hey, these are our requirements and I'm not going to judge. Any institution tells me hey, we have compliance requirements. We need to have the ability to do A, B, C, D, E. As long as those are transparent, they're implemented in code and anyone can opt in or out an institution to do whatever they want. Right? I'm not coming here with judgment saying no, no, no, everything needs to be absolutely fine or you can' Any controls, no. Institutions if they want to come on chain, will absolutely need to build in controls around compliance, around privacy, um, around user access, um, a lot of different uh, levers. And our goal is to give the institutions the full suite to be able to program these and actually in doing so not be, not say okay, well my requirements, the cool part of the programmable economy, it's not programmable economies, right? If you have a different set of rules for me, it doesn't mean you're on your own island. No, you can still tap into Arbitram liquidity, you can still tap into Ethereum settlement, but actually your users will be in a zone where they have maybe other wallet level or even sometimes chain level restrictions of what they can do. So fundamentally I don't think that there's a one size fits all solution here. I think that we need to make blockchain systems amenable for any set of institutional requirements and recourse is one of those interesting things that uh, some institutions will have deep requirements about. Sometimes that's implemented as uh, front loaded policies of what you can and can't do and sometimes it might be implemented as levels of controls or sort of time delays for things to happen to allow, you know, either humans or machines to be in the loop, to uh, to analyze Things before they become very final. So while the blockchain infrastructure generally is very final, I think that we need to understand that we have to actually build systems on top of that that are, you know, that can be less final and can basically meet whatever regulatory, uh, institutional compliance requirements that uh, the actual users of these systems need to get into. Your actual, your, your other question about uh, what happened recently. So um, referring to the, the Kelpdale hack. So number one it's, it's important to, to mention that you know, Arbitrum was really, what I like to say is, you know, an innocent bystander here if you will. Right. Arbitrum wasn't hacked uh, in any way. Arbitrum wasn't attacked. In fact, uh, the, you know, the, what happened was uh, particularly for these funds, these funds were bridged onto Arbitrum because Arbitrum has deep liquidity and the attacker realized that they could um, access the liquidity in Arbitrum to swap, to swap the RS uh4 ETH. So um, they basically accessed uh, arbitrary liquidity and then they left the funds on Arbitrum. And then this very big question came. Um, okay, well what do we do? Uh, what is we as an ecosystem? We uh, as a, you know, an arbitram dao, uh, that control the, the ecosystem, uh, both technical and um, non technical decisions. Like what is the best thing to do in this case? Well, what are the range of options? Let's see, uh, the dao by the way has the full power to always vote and say hey, we as DAO vote to say move these funds or lock these funds. But the problem is you can't do that because uh, and uh, the problem is DAO votes by design take a couple of weeks to uh, actually go into effect. And this is important because if you're a member of Arbitrum, you have funds in Arbitrum and the DAO does something you don't like, you want to make sure that they have a delay there. So you can always get your funds off. Right? So if the DOW decides to change the rules on Arbitrum, there's always going to be a multi week delay. So no user is ever locked into those due rules. If they don't like them, that's critical. But the problem is, you know, if you go and put on the DOW forum and say hey, you know, we're going to move these funds off, what will happen? Well, the attacker will just move them off right away and you'll actually never have that ability to do that. So you need to Make a very, very quick decision. And who has the authority to do this? So not me. There is this, uh, and not anyone in Off Chain and not anyone at the Arbitrary Foundation. But there's this independent, uh, um, council called the Security Council, which is made up of. And it's elected. Um, elections happen. Actually just had an election. They happen twice a year. Um, there are two cohorts of six members, so 12 members total. And they basically, uh, as you know, can do these emergency actions on the Arbitrum Network. Um, and that's actually, you know, what happened here today. So the question came to the Security Council and said, hey, Security Council, um, can you freeze these funds? Essentially freeze these, these funds that are suspected to be from North Korean hackers. Um, can you freeze those and make sure that they can't get access to their money? And the nice thing is, and an Off Chain Labs, by the way, full transparency. I'm not on this council. I've never been on this council. Uh, we have one member, an engineer at Off Chain Labs who ran for the council and is on the council. But basically, uh, it's a nine out of 12 also. So you need like nine people to sign off. So we basically had a very limited, uh, role in actually, uh, deciding this. What we did have a larger role in, uh, back early on when we were a partner in creating the Arbitrum Dao and creating these rules, was setting up the system that actually enabled this to take place, right? So that, you know, I and my co founders and my company had a very large role in putting the Security Council into place, putting the Constitution into place, together with our launch partners that actually define the rules of what's supposed to happen here. And what I can say is I think that the system that we put in place worked very, very well. Right? There was an independent Security Council. They had access to the best information from law, um, enforcement, from lawyers, from ourselves, and technical information information, and ultimately had to make, um, what I call a relatively difficult decision. I think that they made the right decision, but ultimately I didn't make the decision. And have they made a different decision? Like, it wasn't like a very, very obvious. There were, there were, there were loud voices on both sides, but ultimately I think that they, um, made the right decision. And I'm very proud that we put a system in place that allowed this decision to happen and gave it, um, you know, the space to happen. And also, uh, we gave them the technical support once they decided, hey, we want to do this decision. Someone needed to help with the Code to write the code, uh, to make this happen. And that's where we at off chain also supported as well supporting them once they decided they wanted to do this and take this boat. Um, they needed uh, technical support in order to actually um, pull this off, which we were happy to facilitate. But ultimately the decision didn't lie in our hands. But I'm very proud of the process we put in place that allowed this decision to happen.

Speaker B: Right. No, um, you know, purists have been crying foul about uh, this, but I also so firmly believe that this was the right decision in the context of everything uh, uh, that transpired. Can uh, pretty well imagine how difficult it would have been for the Security Council to actually evaluate all aspects of it.

Speaker A: Absolutely. And I should mention that I do look forward to a future in which there is no Security Council. So, so why does a Security Council exist? Um, not particularly for this issue really for critical bugs in, you know, in the Arbitrum, uh, code base. Right. So let's say someone finds a critical bug, you can't again go to the DAO forum and said hey guys, I found a critical bug, let's patch this. The second that boat hits the forum, a nefarious actor goes ahead and drains the funds. That's obviously not feasible. So you need some emergency upgrade path. Why don't layer ones? Why doesn't Ethereum have a Security Council? Because Ethereum can hard fork and they can basically say hey, um, all nodes move over to this software we're hard forking away. Layer two is because Arbitrum's rules are defined not by a direct consensus of its own, but in an Ethereum smart contract. The only way that Arbitrum can hard fork is either by an action of the Security Council that can change the rules in those contracts or actually by a hard fork of Ethereum itself. Where Ethereum would say hey, we're going to hard fork Ethereum in a way that actually changes the Arbitrum contract and changes those rules. I actually have been advocating for some time and I think that um, we should look forward to a path where Ethereum adopts its largest and most successful roll ups by some opt, you know, some uh, um, fear and uh, objective measure. Right. So anyone can qualify for this. And the idea is that Ethereum can now take a head, take, take control uh, of these contracts. Um, who is Ethereum? How do you agree to this? That's a very good question. Right. And like, you know, how do you actually get consensus on this? That's a very fair question. And, and you know a, ah, point of difficulty but I think in principle that would you know be nice. The same way Ethereum, you know will hard fork for a critical bond in Ethereum software to get to a future where that happened for Arbitrum as well. Um, I don't think, we're obviously not there today and today therefore the Security Council is a necessity. So you have. The question is well if Security Council exists and it has this power, what should it do in this situation? That's the context of this. But there is a world which I look forward to in which uh actually uh the Security Council doesn't exist and we have other paths to mitigate say critical bugs in the arbitrary software. And then the question would be basically uh, the same as if for Ethereum, right? Does Ethereum hard fork to save um, funds? Uh well if it doesn't Ethereum maintenance it could do it in arbitrum if not not. But ultimately uh, I think you'd be in the same position. Uh, I think that that's you know a um, a very plausible future we can get to. I don't think it's necessarily um, you know the most popular path that's going to happen right now, but I will continue fighting for that.

Speaker B: Right. But in some sense that's sort of pushing that decision further down onto the uh, onto Ethereum. And it's not so much a question of um, I used to word finality, um unless I'm understanding it correctly it's not so much a question of the finality of the transaction. You could still have for institutions who have always guardrails and uh, uh everything programmed into uh how their walls or smart contracts are supposed to uh operate. But in spite of that, if there's uh, you know a straight out uh security hack or there is a social engineered um, uh compromise it'll ultimately um, need um, a freezing of assets uh essentially now, now uh, that decision ultimately again go close to whoever is sort of controlling the sequencer, the uh, or the asset itself for example if it's stablecoins sir Tree could have acted on that as well. So someone has to take that decision. Uh right. So is it more a question of pushing this up to or down to Ethereum or is it more about uh as a sort of self governing um mechanism um coming up with a consensus on how players, at least all large players need to operate whenever something like this happens.

Speaker A: So a few things there, I mean like there already a lot of the decision does lie on Ethereum if you will. Right. 70 million of the funds were on Arbitrum 2 uh, hundred plus were on Ethereum. Uh, but no one really suggested uh, and I don't disagree with this, no one really suggested that Ethereum should hard fork uh, again to capture its funds. Um, so yes, it's putting that decision down but it's also moving it to a much more decentralized sets. And the expectations uh, that come with that set are often you know, just different, uh, um, it's different levels of coordination, ah, etc. So there's just a different set of uh, considerations when, when you're asking that question.

Speaker B: Yeah, I mean it's. Something has to emerge and yet something um, uh, it has to still remain um, you know, sovereign control free. So how do you balance that? And uh, I think that's the argument of purists on the other side that if you can stop this tomorrow, governments can ask uh, uh, to their own uh, um advantage certain transactions to be frozen as well, which uh, benefit one versus the other. So I think those are all sort of uh, not hypothetical questions. Those are like real life situations that everyone in the system has to face.

Speaker A: So, so this is the way again I think about it, which is there's you know, back to what I was saying before, there's sort of two levels of thought here, right? There's what properties should the core systems have? And again my thoughts are that the core systems should be as final and permissionless and trustless as we can get them to. Um, then you have to ask yourself the question, okay, but if you're, you know, if you're not 100% there today, what do you do in the interim? But like ultimately to me that is the goal, to get um, fully uh trustless and get as little control as you can. And then though the question is, okay, if I'm an institution, uh, or I'm building an application on top of this system, how do I architect my system that it taps into those controls but gives me whatever levels of, you know, levers or levels of control that I'll need. You mentioned stable coins, right? Stablecoins are a great example of this, right? You have stable coins that sit in Ethereum, but they still have central kill, uh, switches or free switches where the operator can do whatever they want. Is that good? Is that bad? It is what it is and whoever's using the system should understand that. Hey, you should know that if you're using stablecoin X on Ethereum, um, uh, you're actually not getting the raw guarantees of Ethereum finality. You're Actually getting something else which is Ethereum finality plus the decision of whoever's operating the stablecoin that they can uh, seize your funds. And you just have to understand that it's not good or bad but you understand that holding you know, USDC on Ethereum is not the same as holding Ethan Ethereum.

Speaker B: Right.

Speaker A: These are, these have, these have different profiles to them. So ultimately I think um, the core, core infrastructure should be as trustless, as permissionless as uh, trust minimized as possible. And uh, then we need to build systems on top of that that can uh, scale and meet the needs of individual institutions. And again I don't judge institutions at all that you know the fact that an institution might need to have deep compliance requirements or they might need to know you know, KYC their users. Right. A lot of the uh, at least funds that are launched and uh, on, on Arbitrum today and on all blockchains today, they're operated in a pretty locked down way where they have a kyc, uh, set of users or set of users that the operator uh, knows and has uh, you know, identified. And I don't view that as good or bad. Is that like permissionless defi? No. Can the blockchain rails provide something to a wide class of people? Yes. Some people want permissionless defi and privacy, full privacy. Others want assets that are KYC but still get the benefits of being able to trade on chain, being able to trade 24, seven still get all the benefits of tokenization. And I think we'll see a spectrum where some of those will cross and move over the chasm at some point. Right. Some of the funds that have started today as lockdown and permission and KYC that think ultimately either in direct form or in wrap forms, we're seeing this already will move over to more permissionless. Some of them won't. But again that's like application level decisions. So for me the question is we need to make the base layer as secure and decentralized and that's why we're building on Ethereum because we believe that the most secure and decentralized way to build Arbitrum is by building it and selling on Ethereum. And that was true in 2014. Well 2015 I should say that was true in 2018 when we uh, published a paper and that's true uh, in 2021 when we launched Arbitrum and that's true today in 2026. That being said, I think we also need to build infrastructure uh on top of that and we're doing a Lot of this at off chain so that an institution, when they want to join the programmable economy, they can program it. They can actually build uh, whatever rules in code to tap into the system, but also encode their complex set of rules which will by the way be totally different than those rules of any other institution. And that's why the system is so powerful. You have all these different institutions, all that have different requirements and rules, but they can still. You use shared infrastructure, access, shared liquidity and ultimately benefit from each other. And users can go back and forth and use basically all of these at the same time.

Speaker B: Right, uh, we should segue into uh, you know, Robin Hood from, from here. And that's a phenomenal step that happened in the whole ecosystem when a large uh, fintech like Robinhood not only coming on chain but also going with uh, Arbitrum, um, where are you on the whole tokenization, um, uh, vector. How do you see Arbitrum being a um, major infrastructure for tokenization? Tokenization, I mean just to put things in perspective, has been growing significantly. I mean the audience who are familiar with this, it's uh, grown almost 10x in the last uh, two years. And um, in the more recent past adds a billion dollars of TVL almost every week. Uh, that I look back, I think now we're over 30 billion uh already and projections are significantly robust. Uh, and one of the smaller pieces of this whole tokenization vector right now is stocks, uh, and equities uh, being traded um, on chain. And you know, tell us about that, Stephen. What are, what's your alignment with uh, Robinhood? What's your thinking around um, tokenized uh, equities? And these are still right now more like uh, wrappers or uh, uh, sort of mirrored representations. When can we see for example the first IPO happening on um, you know, Arbitrum infrastructure, uh, or of primary issuance happening on um, Arbitrum infra.

Speaker A: Yeah, I am a very, very big fan of tokenization and I think that um, tokenization and blockchain rails are just the next form of evolution for equities and other tokenized assets. Um, and we're seeing this progression really happen in the ecosystem. It started off with the programmable money, right? Stablecoins. We just discussed the idea that you can move money around in a programmable way. And today we're actually seeing this being extended not only to money, but to all sorts of sorts of different assets and asset classes. Um, and again this idea of the programmable economy you mentioned, Robinhood, which of course is bringing equities on chain, right? If you just think about for a second, um, the ultimate goal of having, you know, defi and perps and derivative markets with access to, you know, the full, the full, um, you know, gamut, uh, of, you know, us and eventually global equities on chain, it's extremely, extremely powerful that you can do. So we're seeing this basically, uh, this idea that you can join, you know, have programmable markets, you can bring your assets, uh, on chain, you can solve liquidity problems, uh, that you have elsewhere by, you know, by sourcing liquidity on chain, by, uh, coordinating liquidity coordination on chain. So coordinating some of these, you know, lending markets like usdai, using blockchain infrastructure, solving issues like, um, availability. Right. Uh, of markets, solving issues like, uh, you know, back to uh, stablecoins like settlement today. If you take some of these funds that, you know, take the, you know, some of these treasury funds off chain, they're like five days, you know, T five days to settle and get your money back. For an institution that's trading in size, that's a ton of lost interest just waiting to get your m. Your funds back. And there's really no reason for it, right? We have blockchain rails where you can get your funds back and like literally the next block. So, so, uh, stablecoin settlement and you know, instantly have access to your funds and not have that, that large gap that exists. So, um, we're seeing a large transformation across basically every asset class out there where, um, we are taking assets that, you know, I hate the term real world assets, but real world assets, whether that's equities, real estate, uh, tokenized funds, bringing those on chain, giving users access to these. And again, I think we're in a very early, early part of this transformation in the sense that a lot of these are somewhat locked down today, right? In the sense that you, like I said you have to have KYC or they only work in particular products. But the vast power of what you're able to do with this is like, extremely, extremely encouraging for what will come in the future. And again, that's up to these different issuers. If you want to keep your funds locked down, that's fine. Like, we're not judging, we're providing the infrastructure to do what you need. But ultimately it's very clear that the way that the ecosystem will go is more and more assets will come cha will come on chain. They will freely trade, uh, in between each other and each other's systems. You'll be able to build, you know, A perp dex that has access as the best equities in the world and precious metals and things that literally markets didn't exist for. There's one really ah, interesting one. Uh, there's a company called steelcoin that's actually tokenizing steel. Today steel is the second most traded commodity in the world but there is no like liquid market for it, for it at all. Um, and they're using blockchain rails to actually go ahead and tokenize this asset. So I, the sky's the limit. We're going to see assets that were literally never tokenized before, um, come on chain. And I'll end with this one other point here which I think is really interesting, which is you know, back to Robinhood. What is Robinhood doing today? So if you go ahead and Robinhood app in Europe, you want to buy a share of Apple, you're going to get a blockchain um, asset essentially. So you're going to get a share of Apple on Arbitrum, a tokenized asset, same flow. It looks basically exactly the same as the US app, but you're going to get a tokenized asset instead. And if you're a user, say in Europe, you might not even know that you're getting a tokenized asset. And that's the power here, that you're going to basically use the blockchain rails to provide a real experience for users. They're benefiting from this technology. They don't even know that they're using it. That to me is a zero to one moment of is this technology useful beyond, you know, the core set of um, enthusiasts? I think the answer is resoundingly yes.

Speaker B: Um, no, no Stephen, that's, that's obviously sort of exactly how things are heading. But when will sort of native issuance happen on chain? I asked you about when do you think for example the first primary issuance will happen? And of course we've been talking about opening up the entire defi stack composability to uh, financial market uh, securities or instruments that will be much more seamless when these instruments are natively issued as opposed to wrapped structures. Right. So what are your thoughts on that?

Speaker A: I think so. Two things. Number one is we're seeing um, already some of the uh, nyse, for example, uh, nasdaq, we're seeing a lot of movement on behalf of um, native uh, exchanges, um, but also some native asset issuers that are increasingly moving towards doing more things on chain. But that being said, I think, actually I don't think that native issuance is going to be. First of all, I think there will be some time until we have um, native issuance of most assets on chain for sure. But I actually don't think that that's going to be uh, a blocker. Um, where are we today and what are we missing? And it might not be like, oh, let's just rip up this whole system. We might say, hey, we have something that works really, really well. And um, this is just maybe the end state.

Speaker B: Right. So let me ask um, you this, Stephen. So obviously tokenization is uh, something that um, many of the top blockchain infrastructure, blockchains, uh, lwom, lbuz are kind of competing for in many ways. Uh, and uh, how, if you set aside the US market, for example, if there is a platform which is uh, taking uh, any other, uh, large financial markets, uh, structured products, credit markets, capital markets on chain, what would be Arbitrum's compelling uh, sort of uh, story for a platform like that? Which is taking for example, let's take assets, financial market assets from India or Brazil or one of the other large Asian markets, for example, what would be Arbitrum's compelling story there?

Speaker A: Yeah, a few things. Number one is um, there's a network effect that exists here, particularly when it comes to liquidity. Arbitrum has very, very deep liquidity and also has a very strong uh, DEFI ecosystem. A lot of defi innovation happens in Arbitrum. Perps as a product took off first on Arbitrum, uh, before they were cool and everyone else was talking about perps. Right. We had gmx, uh, of course, uh, really leading the way with perps. And Arbitrum today still has a large hand and most of a large uh, perp dex is that exist on chain today. Um, that's uh. So number one, it's the access to the innovation, access to the DEFI ecosystem. Right. A lot of the uh, asset issuers that I talk to, um, that are bringing their funds on chain, they understand that what they're doing today is the baby step, but the real goal, and sometimes it requires them to work with say the SEC or get regulatory approval and there might be hurdles in place, but they understand that the real goal is actually say, how do we actually make this one composable ecosystem? How do we not have, you know, Tradfi and Defi? How do we just have PHI1 financial ecosystem where uh, blockchain assets and blockchain rails are just a core part of it. And that is something that is happening um, already today, even if it's, you know, relatively um, uh, small steps, um, you know, to, to get there. But we're seeing essentially emergent, the merging of these two, of these two different systems, uh, and then coming on chain. So that's number one, Arbitrum has a very strong ecosystem, particularly in Defi. And that's the goal for a lot of these funds to be. Uh, you know, they see where this is going, they want to be part of that ecosystem. And even if that's not step one, even if it's like step 15, that's ultimately the prize for many of even the largest asset issuers out there. Uh, number two, um, you know, the arbitrary technology stack is uh, by far the most customizable and flexible. What I mean by this, you know, say Robinhood. Robinhood. They announced, they're launching Robinhood Chain, but actually they first launched and today are launched. They have over 2,000 equities on Arbitrum 1. The idea that Arbitrum has something very unique in the ecosystem. It has a top five blockchain by economic activity in Arbitrum 1, but also the number one blockchain stack for building your own blockchain. And you can seamlessly go across those. And that's something in one word is optionality. And number three is Arbitrum's technology is just still far and away the best technology for institutions. It has the fastest block times in the industry, native block times, um, as low as 100 milliseconds for public chains, for custom chains, 250 milliseconds is what it runs with on the public chain. Um, we have uh, infrastructure like Time Boost for um, monetizing your ordering as well. And uh, we have something called Stylus, which we haven't talked about yet. But Stylus gives uh, developers the ability to write smart contracts in Rust and C and C. And you know, uh, last June when Robinhood announced that they chose Arbitram, one of the things that they mentioned that day at their event was Stylus. And the reason why Stylus is so powerful is because you have institutions that have so much legacy code and so much legacy knowledge in these traditional programming languages. And they now want to build, build, build, uh, on Chain, you can basically use the same technology bridge that gap and Stylus makes it a lot easier for you. So ultimately, uh, a lot of different things. But we also, like I said, wrap it all together in offchain and are able to offer this to people, to enterprises, and they know that they're in good hands across the entirety of the stack, across the entirety of their needs. And that will grow with them. Their needs today might not be, uh, need their needs tomorrow. They might not even know what their needs are tomorrow. But both the Arbitram stack and the offchain enterprise platform have enough optionality built into it that we know no matter what comes, we'll be able to support you.

Speaker B: So with that in a few lines, what's your vision of the sort of financial market infrastructure and how tokenized markets will look like in the next five years?

Speaker A: I think, um, in the next five years, you know, um, we will see again this mergence, merging of these two ecosystems. There's going to be less tokenization of real world assets, right? Real world assets will just become assets and the blockchain uh, infrastructure will just become one of the core Rails that you'll need in order to run uh, financial infrastructure, uh, in 2030. Let's say we work with so many institutions, but 90% of institutions, they have their core business and then they have their like blockchain project, you know, that's like this little thing on the side and the little playground. It's cute. Uh, maybe someone gave them a grant for it, it makes them some money, it's nice, it's an experiment and ultimately I think that's okay because I think over time those experiments will win out and they'll understand the power to them. One of the really amazing things about Robinhood is they're forward thinking and understand, you know, up from the CEO, the CEO level from Vlad, understand that tokenization, blockchain infrastructure is the next phase of this evolution and that's what I think we're going to see. We'll continue to see institutions come on board uh, at record pace, uh, a lot faster than you know, I personally uh, expected, um, you know, just a couple years ago. But ultimately I think we're going to see the lines blur. And crypto is not the side project. It's critical infrastructure for the future of finance.

Speaker B: Stephen, fascinating conversation, lovely having you on an ash.

Speaker A: Thank you so much for having me. Likewise.