We continue our Q&A interview series, in which our co-founder Adam Goldberg speaks to founders about the most important inflection points in their careers, projects, and protocols.
Today we speak with Steven Goldfeder, the CEO of Offchain Labs, the team behind Arbitrum, the leading Ethereum Layer 2 blockchain, and a company at the forefront of Ethereum scaling.
Startups are a series of unknowable choices. All great startups have inflection points where builders are faced with difficult decisions that will compound for many, many years, for better or for worse. In this series, we will explore a range of inflections a project has encountered, some in the past with the benefit of hindsight, some in the present, and others with an eye towards the future.
This transcript has been condensed and edited for brevity and clarity.
Arbitrum's Early Choices
Adam: Hi, I'm Adam Goldberg, one of the co-founders of Standard Crypto. Today, we're continuing Behind Crypto Builders with Steven Goldfeder from Offchain Labs, the company behind Arbitrum.
All great startups have inflection points where builders are faced with difficult decisions that will compound for many years, for better or worse. Thanks for joining us, Steven. Let's talk about some of those inflections for Arbitrum.
One that I wanted to kick-off by talking about is that when you look at L2Beat deposits today, which are admittedly only one way to measure usage, Arbitrum One is the number one overall. In fact, it is even larger than numbers two through five combined. I bring this up because I think Arbitrum had a unique go-to-market challenge. When you rewind back several years, you can say Optimism had all the blue-chip project brands. They had the blessing of the Ethereum Foundation and so much more.
How did you attack a category that already had a broadly anointed winner with such a big lead on mindshare?
Steven: First of all, it's great to be here. Thank you for having me. For us, it was very much about building in a very principled way – in a way that was very values-driven.
I always like to say that if you’re building a business, it’s like selling diapers on Amazon. You care about how many diapers you sell. That's like the end product, and that's what matters; that's all of it. So if you have more revenue and more diapers sold, wonderful. If not, then that's bad. If you are building, though, in crypto, it's very different, right? Our goal isn't just growth. We want to grow, of course. We want more people to use the technology, but there’s a set of values that are very deeply intertwined there. That is, the values of decentralization or the result of decentralization, the value of censorship resistance, and some of the benefits of decentralization when it comes to being able to use the blockchain in a way that can't be censored by some centralized third-party intermediary or the like. And for us building out the core security and the core technology were the most important things.
That's why when we built, we didn't take any shortcuts there and we'd been designing the system for years. It came out of Princeton; it is an academic paper. We said, okay, we need to build this. We just kept our eye on that prize, and sometimes, it felt frustrating because, to your points, definitely it felt like early on, inexplicably, some things were stacked against us in the way that if anyone is examining this, Arbitrum clearly had and was, and still today is the better technology, the best technology out there.
It was hard to understand why not everyone was seeing it that way. Although I think if you zoom out and look at any successful story in any industry, there's always that. So, it's very easy to get into it in the moment but for us, it was more about keeping building. I always like to say we had the benefit of being able to launch Arbitrum as the first open general-purpose rollup and being able to do that with security proofs built out and not compromised.
I like to think if we were in a different position and we had to choose between those two things that I still would have chosen security priorities, even if it meant that, hey, we're not going to be first to market. Even if it meant we're going to lose some of that land grab because that's why we're here. Again, we're here for a set of values. We know how to build centralized systems—we have centralized systems—but our goal is to build decentralized systems. And I'm not here to say anything negative about any other team. Other teams have different approaches, different values, and different things that they do.
But if you ask me what was the driving force behind us, it was basically staying true to those values. Even the way we launched back in 2021 with the fair launch strategy that didn't prioritize anyone. This was very different than the way some others launched, where they would roll the red carpet out for some and not others.
We took a very egalitarian approach, a fair launch on the same day that SushiSwap went live, that Uniswap went live, and another exchange called Swapper went live. Our job, we believe, the infrastructure of that layer, was to roll out the red carpet for everyone. Not to say, hey, that because you succeeded on Layer-1, you’re going to have a priority position on Layer-2. That's not our job. It’s probably because, hey, you were succeeding on Layer-1, you built an incredible business, you’re probably going to do well on Layer-2 as well. As we've seen, fast-forward, Uniswap is one of the largest protocols on Arbitrum and a lot of the big Ethereum blue-chip OG protocols are some of the biggest Arbitrum protocols, but there are also new ones, right? So, had we taken that strategy, we would have missed projects like GMX and Treasure and said, no, you guys aren't important.
We didn't take that approach. We took the approach of everyone’s on the same footing. Again, the theme between that and building security is a set of values—openness, permissionlessness, and decentralization— that drove everything we did. Growth was only secondary to that.
Adam: That's interesting. Did you ever tell the story as such to these partners early on? Or was it that once the vision that you had articulated was live, it was such an easy choice for them? Did you take them along on that journey as you did with me?
Steven: Not really. I don't think we understood it, to be honest, at the time. It wasn't that we decided and said, hey, we're going to do this radically crazy thing by prioritizing security. We're going to do the thing that was natural, and we just thought that was the default and thought that was obvious. We're going to prioritize security and make sure that the security is a priority and paramount from day one. We didn't go in and say, hey, we’re not do this fair launch thing. This is radically crazy. We thought that was the correct and default approach. And, again, obviously, there's no correct way here. Different teams have different approaches and might have different ways they view things. I'm not here to say one is right and one is wrong, but I am here to say we did what was clear to us but only if you look back in history that was actually really important, right? That was actually somewhat controversial, but I don’t think at the time we realized this. We never sat down and said, hey, should we make certain compromises either in the system or in the ecosystem openness? It wasn't like that.
We just went on the path we felt was right. It turned out that it resonated with a lot of the community and helped build up the community, but hindsight is 20/20. It was only looking back and you'd see, oh, that was an interesting decision and it's not the same as everyone else made. But at the time, I think that was a lot less obvious.
Starting-Up From Academia
Adam: So, I want to pull on another thread that you brought up with Arbitrum’s origins being from Princeton. I think you would say that Arbitrum is a "professor coin" that worked when most did not.
What are your lessons learned in spinning out a university research project into a successful commercial endeavor?
Steven: Yeah, I remember first hearing this term, maybe in 2018 or so, a friend of mine actually that told me he was in the Chinese language message boards, which he understood. He said, “Everyone's calling you the professor coin”. I got the sense they didn't mean it in the nicest way—looking back, they weren't wrong. If you look back at myself, Ed, and Harry. We came from academia, didn't have extensive business experience, and there were reasons to, I think, be skeptical and doubtful. I think it ultimately turned out to be very helpful because, again, in a space that’s values-driven and those values translate directly into technology and building technology that enables censorship, resistance, decentralization, et cetera. Our academic experiences proved to be very valuable. We had to learn a lot on the job for those other pieces that we might have been missing. But I don't have any resentment towards that early characterization because, to be honest, it was a fair characterization.
If I were an investor early on, those were very reasonable things to raise, but again, we built a technical product. The way we built the company, we didn't have a partnership team until we had a technical product that people actually wanted to build, and it was too much. You can't have technical people taking every partnership call. So, we needed to build a partnership team. Similarly, the marketing team only came later when there was a technical product, partnership teams, and marketing needs, both from us and partners, and that were like, hey, we want to get the word out, we want to co-market this, and we need a marketing team. It was very much led by technical readiness, and at every step, the tech is driving the need for partnerships because the technical people don't want to be in these calls anymore and same for marketing. That's basically how it evolved, and that's how we overcame it.
So, we kept building the technology. It was early. It was right. It made sense to a lot of people. The need for it built out, I know that need demonstrated itself over time, and we built out those functions and overcame that initial perception of being a professor coin. Not that we aren't a professor academic project, but we were able to backfill those pieces, but technology was always driving the train in the case of Arbitrum.
Adam: How do you think about building the nucleus of the team in the sense that I think you talked a lot about building partnerships and scaling marketing over time, but is it important to have a combo of professors and, say, a grad student protagonist on the founding team? We've found that outside of crypto that has been a helpful starting template, but I'm curious to explore your experience.
Steven Goldfeder: Yeah, for us, particularly in this space, that's where my experience is limited. So I can’t speak for others, but everything is so technical, and the technical details matter, right? It's not just the bottom line. We made this much money, sold this many widgets or diapers or whatever you have. We built a product that has certain guarantees of security that hopefully will never get invoked.
It was a very different endeavor, and I think the academic expertise and focus we had with Ed and Harry, working on blockchains and deep in their security models for years before we even dreamed of building something commercially, proved extremely valuable. Definitely, I think that recipe works a lot for us. As we said before, it had its risks, and they had doubters early on, and those doubts were often very fair. The team having this sort of mix was absolutely fantastic and wonderful.
For me personally, having two co-founders who I had known for many years, was close with, and have only gotten closer to over time. People like to say, “Co-founders are like a marriage”. You spend a lot of time with them; I've been very fortunate definitely. If it wasn't for Fred and Harry, I wouldn't have been able to do this and I think we really work together as a team. If I remember our initial roles, people sometimes ask me, “How did you become CEO, Ed Chief Scientist, and Harry CTO?”. I remember vividly there was this conversation in Princeton on the lawn right outside Sherrerd Hall, which is the building that we worked in. It went something like this where Ed said, “Hey, Steven, you talk the most on the phone calls; you should be the CEO, and Harry is a student who writes the most code”, and Ed is, of course, a professor, so he's the Chief Scientist. It all just made sense, but it was a very delayed reaction to that decision because, at the time, there was maybe one or two phone calls a week, and we'd all be on the phone calls.
Everyone was contributing to the technology, writing the paper, and core design. Sure, we took on these titles, but then they didn't really mean much, right? Nothing changed. Then as time developed, as the team grew, as we started having more phone calls, started raising capital and building up a team, it was like, “Hey, this is your responsibility. You're the CEO, hey, you're the CTO, you've got to do this. No, You’re the Chief Scientist, this is on you.” And we grew into those titles that we took onto ourselves earlier with zero immediate ramifications. I would say of the three of us, I had the biggest shift from what we had done before, but speaking for myself, I would also say that for Ed and Harry, it was a wonderful experience.
We've all grown and learned a lot, and again, even today when we're more specialized, we're always there to support each other, which is extremely valuable.
Adam Goldberg: Yeah. Maybe the secret is that grad students are professional learners and are really good at learning new things. I'm curious. On a more personal note, what was it like for your first job out of your PhD to be CEO of a tech company? What lessons did you personally learn through that journey?
Steven Goldfeder: I was fortunate to learn a lot of lessons. Also fortunate to have really good mentors and investors.
Actually, from Standard Crypto, Alok. He's one of those people who I like to call when I need some sage advice. I was very fortunate to have really good, helpful people showing us the ropes, teaching us, and saying, hey, we made some mistakes here; you might not want to do those. So that was very helpful.
Also, it's ironic, but I had done a few internships as a graduate student, and I wasn't intentionally learning at the time, but I did three internships at Google and an internship at Microsoft Research, and so much of what I knew was just from seeing these companies. I ended up applying learnings there in ways that I didn't realize I was learning those summers. I was there for my coding or research internships and ended up applying some broader topics of management and the like. I was also fortunate, both in graduate school and in a postdoc at Cornell with Ari Juels to have good mentorship and see professors run effective teams.
So, a long time ago, I thought I would be an academic. That was my personal goal if you had asked me eight years ago, “what are you going to do?” Obviously, I’m going to be in academia. “Do you want to launch or run a company?” Absolutely not, I would never do that. People say running an academic lab is like running a small startup. I don't know if that's true now that I've seen both. I’m not sure I didn't necessarily agree with that, but there's a good parallel there. My point is I was fortunate to have very good mentors in the university setting who I had seen do that as well. That was very helpful for me. As well as my parents growing up. My dad manages a software team, which is also where a lot of the learning came through that.
I had a lot of times when I learned without realizing I was learning. It was definitely a big journey and a big change but again, it was very gradual, like I was saying before. It wasn't, okay, you're the CEO of a large company. It was you have this CEO title, and in the early days, I was almost embarrassed to use the title. It was like, oh, it's only three people. This seems a lofty title for this. We’re all just doing the same thing. Then people would say what are you? And I would just say I'm the co-founder.
It wasn't until the company really started to grow. Then I actually started to act in that role and feel it's time that I can start using this title. It just felt a loftier title than the rolling up my sleeves position I was in, but over time, gradually, it’s definitely evolved to that over time. It's been an immense change if I look at it day to day. What I'm doing now, what I was doing five years ago, it's vastly different but I don't look back. I'm super happy, and I love the challenge of every day and look forward to continuing to do this for, hopefully, a very long time.
Fundraising
Adam: It's really interesting to hear your perspective on these attributes of a strong founder that are naturally homed in an academic environment. I'm curious to also talk about one in particular that I think is a really big phase shift for people – how you think about fundraising.
When you're in an academic lab, fundraising is something that is relatively repetitive, right? The goal is to do the work so that you get more money, and you spend all the money you get, and then that's that. When raising a seed or an A for a project with ambitions like, of course, you've had for Arbitrum. It's a different pattern, right? If you were to just say, hey, we raised $5 million to spend all $5 million this year, and once we've spent that and hired a bunch of people, we can raise another $100 million, which might put you in a sticky situation.
So, what advice would you have to someone coming straight out of academia and stepping into a founder role about how to think about financing?
Steven Goldfeder: Two things. Number one is that I think the big difference between academia and startup fundraising is the separation of roles. In an academic environment, the professor who's doing all the research, the PI or the advisor, whatever you call them in your group, is also the one filling out the grant applications, and often, it's the most dreaded part of their role. They hate that part often, and then it's nice to get a good win. You get a nice NSF grant, and you feel good, and then there's press.
But often, you don't see the loss. You just see the ones that they get. There are a lot of ones that they don’t get, and it's the most dreaded part of their role.
So, it’s not your question, but just aside for a second, we have a research team. We have ex-professors on our research team, and one of the draws is that you don't have to worry about that. You just come to work and do all the fun stuff. The company's business team and I are going to raise funds for you. That's alluring for academics because they don't enjoy that part of the job.
Getting back to what you actually asked. There's a big focus on sustainability and growth and, eventually, turning a revenue; although, as you know better than I do it, it's not monetization too quickly. It’s really a path to monetization and the path to growing value to not necessarily cash in on the value you've created, but really growing that value, growing the brand value, and growing the value of what you've built so that there are revenue opportunities there.
I think part of it is also conserving, like you said. You can't just go ahead and spend all your money and then just try to get more money. You actually have to understand investing your money in people and projects, understanding the ROI, and also making sure that you have sufficient runway.
So, my advice to founders here is really the question is, how much do you want to raise? I've seen people err in both ways. You don't want to raise too little; ironically, you don't want to raise too much too often. I think that's obvious, but I'll be more specific.
If you're going ahead and doing a fundraise, it's a lot of work. You want to make sure you take in enough capital so that you can comfortably execute your vision. It doesn't mean that you'll have a decade of runway, but it means that you don't want to be razor-tight and make one wrong decision. One technical project has been delayed by six months. Hey, those things happen every day. In fact, in your seat, I guess that you probably just add six months whenever a founder tells you a project deadline because you know these things slip.
My point is you want to ensure you have comfort within reason to execute, and often, it can be a learning experience of I don't need more money but I'd rather take less dilution. It's looking at the big size of the pie in the big picture. If you're success aim is to be successful. Do you want to take a chance of losing everything because you're just trying to conserve the most possible value, right? Optimality versus practically making sure things work is one big thing. So, raise enough.
Number two is also don't raise too much because when you raise too much, what happens is particularly when the market goes up. You can probably do another raise and raise like a gazillion dollars or a 10 X valuation; the problem is the market's always going to go up and down and up and down, and you want to be at a point where you feel comfortable with what you've raised, you can give a good return on that, you can build, and you have what you need to build.
You don't really want all these investors that have given you this big valuation, and the market turns, and then it's “Hey guys, where are you?” You always want to ensure you're on the right part of that curve where you're delivering value to investors, and they came in at a good time and got a good deal. I think not overraising is a big part of that. Just not over-diluting yourself if you don't need to be is also the other part of that. My advice would be to figure out what I really need to get this done. Add some comfortable cushion to that, and that's the number that you should try to raise, and then make sure you budget appropriately. Have two to three years' runway budgeted so you can comfortably deliver, get to your next raise, and not run a fire sale saying, “Oh no, we will have to lay people off or raise in the next six weeks.” That's a big problem.
Adam: I wholly agree, and I think that one major shift for many folks is when an investor sees something that has, say, nine to 12 months of runway left.
I think we're thinking that thing is on a path to dying. That feels like a really small amount. But I also agree with you on the flip side if you're raising, at the early stages, and in a position where you have eight years of runway, in many cases that works against you because you don't feel enough pressure.
So there's a happy medium in the middle.
Steven Goldfeder: Yeah, absolutely. Feeling a little bit of that pressure is definitely very healthy and very helpful in motivating you to actually take care of the right bets and put everything in. We've been in that, early on, we were definitely, there have been times before we did our series A, where in that position looking at the numbers and were going to have to raise at some point relatively soon because otherwise, the numbers aren't going to work.
So yeah, that's not a fun position to be in either, but it also makes you really work and put in those crazy hours and crazy days, taking those bets, which hopefully pay off.
Placing Bets on Technologies
Adam: I guess we touched on this at a very high level in terms of how you thought about deploying and spending money.
But if you take a step back more broadly entrepreneurship is about encountering inflections and making bets, and how you know when you're right and when you're wrong? How do you know when to keep going versus when to retreat on something? I'm really curious to talk through an example at any point in the Arbitrum journey.
Steven: The answer is you won't always get it right and the ability to constantly re-ask yourself that question and not be stuck in the decision that you once made is, I think, critical to succeeding. I can give you a few examples.
One I'm still playing out people often ask "ZK-rollup or Optimistic rollup" and I will give a 10-minute argument easily of why I believe today sitting here in 2024 Optimistic rollups are the best technology and the Arbitrum Nitro Stack is the best possible way that we have that's currently scaling Ethereum and looks to be the most successful way to scale Ethereum in the long run.
I think one of the things people don't understand is that they sometimes assume that I made that decision in 2018 and am following through with it today. The answer is absolutely not. If I believed that ZK was the right technology today, I'd be advocating for moving Arbitrum’s technology stack to it.
I don't believe that today, but it’s also not something I've written off. We have a research team. We have Rosario Janeiro on our research team, a notable cryptographer and one of the initial developers of the SNARK, the first modern zero-knowledge proof, and Victor Shubin, a cryptographer.
We have very good cryptographers with deep expertise on our research team who can evaluate the space, reevaluate, and say, hey, are there places where we can use new technology? Be it zero-knowledge proofs, something else in our tech, or our technology stack.
So that's one that I would say is playing out but the critical thing I like to say: if you look at the name of our company, we're Offchain Labs. If you look at the name of some of our competitors, often they'll have the name either Optimistic or ZK or SNARK or STARK built into their names.
My only point is that from our perspective, we haven't dug our heels into a specific bet. As I said, this is the future of the company. Our company's mission is to scale Ethereum with the best technology available, not with Optimistic rollups.
We believe today that the best way to carry out our mission is with Arbitrum's Optimistic roll-up technology. Crucial to that is the flexibility that if something else comes along, we can change it. So that's an open-ended example. It's one that's playing out. I can't tell you what will happen, but I can tell you that I'm looking at it today.
To give you two examples of past events and things that we've re-evaluated and changed. If you look back to the Arbitrum paper from 2018, this is the USENIX security paper, the initial Arbitrum academic paper. It had something in there called the AVM or the Arbitrum Virtual Machine. An incredibly cool, from a virtual machine design perspective, incredibly cool design, an innovative design, and I would probably argue it's the design that got the Arbitrum paper accepted into USENIX. It was novel. It was interesting from an academic perspective. It was very neat. And it worked. In fact, when Arbitrum went live in 2021. It shipped with the AVM, the Arbitrum Virtual Machine, and it worked and was in production for about a year.
One year later, this is in August 2022, just about to the day we launched Arbitrum Nitro. Arbitrum Nitro took the AVM and retired it, replacing it with WebAssembly Design. And three years earlier, I didn't know it was possible to build optimistic roll up with WebAssembly design. It's something we figured out over time. We realized, hey, this is better because it allows us to use WebAssembly to get all language support from WebAssembly to use Geth, Ethereum's node. It just had a lot of better things. Now, you might have thought we would say, hey, AVM, this is our baby. We spent years designing this like Offchain Labs is the AVM, right?
You could have imagined us digging in and saying it's impossible for us to change. We're too far into this, right? We've spent years, literally going back to Princeton in 2016. The answer was no; we were happy to retire the AVM because the goal wasn't to launch the AVM into production. It was to scale Ethereum.
When the AVM was our best path to get there, we were very happy to push that, and when it wasn't, we were very happy to retire it, even though we had built it. So, I think that's critical. The ability to throw out your own work and evolve your own work and say, that was great.We wouldn't be here today without the AVM, but that doesn't mean the AVM will keep us here forever. The staircase is the metaphor that I like to use. It works for ecosystem growth as well, right? So, you climb stairs and look back; those were some silly stairs. But we wouldn't have gotten this high if there were no stairs in the beginning.
Another place I looked at in the ecosystem is when we grew it. We did this thing called Arbitrum Rollout. Internally, we called it DAPPS Extravaganza. This was on our testnet. We did a full DeFi ecosystem; no one wanted to build on Arbitrum back then. This was 2020, late 2020. So, we took all the apps; we took Aave, which we called Arbe. We took Uniswap, which we called Arbswap. We took MakerDAO, which we called FakerDAO.
We just took all the DeFi apps and put them onto this Arbitrum chain, and we said, okay, people will build on top of that. I remember we had one big win then, so I will never forget it. We had a big win, which was Burgerswap. I don't know if you remember them. They said, hey, we're going to endorse this. We're going to be part of this. We were so excited that Burgerswap endorsed that. It was a big, very big moment for us, and you look at it today. Arbitrum has every protocol on it, and you say that was a pretty silly milestone that was so important to us. We got one. The answer is no, not really. Yes, we're on step 100 today, but had there been a blank space between steps two and four, we would have never gotten to four. So that's critically important. So, I think understanding the roles over time and the evolution over time is important, and the ability to question our own assumptions and evolve them and not just say, okay, we decided this in 2018, and therefore, that is our decision, and we're going into autopilot. That's a recipe for a disaster. So it's okay to make mistakes, but you constantly have to look at the data and ask if that was a mistake or if that was correct. If it was a mistake, you have to say, okay, let's do something about that.
Arbitrum's DAO Governance Approach
Adam: That's super cool. Those are instructive examples. I also want to talk about one of the largest challenges that any network faces, which is the governance handoff from the company that built the network to the decentralized community. To talk about a concrete example first, I remember that AIP1 was met with a fair share of controversy, and some folks weren't sure if it was a ratification of an already preordained decision or an actual vote to determine an outcome. We can talk about that specifically, but when you look at the governance handoff broadly, what, if anything, would you have done differently in retrospect now that you have some of these learnings under your belt?
Steven: To talk about AIP1 one for a second, I think it was more of a comms issue than anything else. As we mentioned, ratification, which is certainly presenting it that way and doing that process, definitely was a mistake. But actually, if you look at what happened, Arbitrum today, this was in the proposal as well, has onchain controlled in the onchain wallet. It got three and a half billion ARB tokens which is, I don't know, six or six or so billion dollars today. Not only that, Arbitrum also gets all the fees from the chain, Arbitrum DAO. Just this past year, right? Arbitrum has been alive and DAO controlled, launched March 16th of last year. Since then, over $50 million dollars of ETH have also gone into that wallet, and this is all controlled onchain by the Arbitrum DAO. This is an experiment in decentralization that no other layer-2 is. There is no other layer-2 that I'm aware of that actually has not just foundation control but also on-chain governance-controlled wallets and governance-controlled parameters. So Arbitrum has self-executing on-chain governance.
What does that mean? It means that if you go ahead and say, Hey, Arbitrum Stylus is a big update that's coming to be available on the network, ultimately, we at Offchain Labs don't have the keys to launch Arbitrum Stylus and Arbitrum One. It will be up to each individual network, and now, with Arbitrum Orbit, there are dozens of them whether they want to take Stylus.
So if you take Arbitrum One, it's the DAO that's going to have to vote, and the way that vote's going to work is there's going to be an actual update. That code is going to be pushed on the chain, and there's going to be a vote about that. Do you want to launch this? And if the DAO votes yes, that becomes executable and actually executes itself.
That's a big difference between how some other DAOs work, where there is a sort of vote, and then there's someone with a key, and the person with the key is expected to turn the key in the direction that the voters say. There's no glue there between the processes, as opposed to the self-executing governance, which is the vote turns the key, right?
That's, I think, a big difference. So, the Arbitrum DAO is an experiment in decentralization of one that I don't believe we've ever seen on this scale, (particularly with the amount of control it has) anything like it. It's a big experiment of decentralization. So, I think there were some communication issues, and those were regretful. Certainly, we wouldn't repeat those, but we think the product was and is more decentralized than any DAO that exists at this scale today.
That being said, there are some pros and cons, right? So, a lot of times, people like decentralization in theory, but there are some nice benefits of centralization, right? It's efficient. One person tells you what to do, and it works. They don't have arguments when it comes to grants, for example. So, the Arbitrum Foundation gives grants, but the Arbitrum DAO gives bigger grants. So just last week, the Arbitrum DAO proposed a 200 million ARB $300 plus million dollar incentive program for gaming on the Arbitrum ecosystem, and they have to set up structures with the foundation to give this money out. If it was all just controlled by the foundation, which is how some other DAOs do it, it's easy. There's just a team that goes into a back room, writes some checks, and doesn't tell anyone about it.
As opposed to here, the community has that control. So there are pros and cons. It's definitely less efficient, but it's also decentralized. And we believe again, going back to one of the first things I said, the values and the importance of decentralization, that's why we're here. And that's important.
Decentralizing, giving over control to the community, I think is critically important not just from a values perspective but also from a decentralization perspective, right? I don't have keys to the Arbitrum One blockchain, and I think that's a very good thing.
That's both from a technical perspective but also from a funding decision perspective. The community has a ton of power here; to some people, that looks like chaos. So that would be the con, but I would say that that's good.
If you look at a democracy, it often looks like chaos, and the reason is because people have voices. There is value to raising their voices, and people don't always agree. They may often be very aligned in their values and goal, which is to do the best thing for this country, this blockchain, or whatever the system is, but they may have very different views of how to get there. In a dictatorship, you might have a lot of silence because those who disagree don't need to raise their voices because they know there's no point to it. So I think there's balance, of course, but some sense of chaos is a very direct indication that those users are engaged. They have opinions, and they understand that being loud and vocal matters. They can affect change with their opinions. So I think the right way to balance this is that it’s all the parallels come from state and country level governance, but delegated democracy there are delegates and rights. So users can back a delegate, that’s going to speak to the issues they care about. If you care about gaming, you might back Treasure. If you care about DeFi you might back GMX or Camelot or one of the others; there are obviously different choices. May back, ?sobe? or Xai, if you care about gaming. There are a lot of different choices that you have, but you can also say I don't want to go to Capitol Hill myself and read every proposal, but I want to align with someone that I believe will.
I think that's probably the best way, but these things take time to figure out, right? So sure. If you have a centralized DAO, that sort of just has everything happening based on a team of one or two people; they make all the decisions. It's going to look more efficient. The Arbitrum DAO is taking a radically different approach, and I think it takes time, but the structures are emerging, right?
The fact that they got out this proposal and it's a very detailed proposal, with different buckets, et cetera, different management structures, sustainability, to offer grants to gaming, they've done this already for DeFi, as well. These short-term and long-term incentive programs.
I think it's wonderful to see it taking off, and it will become more efficient over time. Personally, while some will probably argue and say, Hey, it would have been more efficient if the foundation just controlled everything in a centralized manner. There's probably some truth to that in the short term, but I think the few heading into the long term. It’s going to take a little more time to find this footing to set up these processes but the checks and balances and the fact that the community is empowered will set it off in my opinion on a much better trajectory.
All that being said definitely some regrets in the way things were communicated early on but in the actual structures. They’re radically different and I think radically important and will lead to real decentralized ecosystems in ways we haven't seen before.
It's going to be chaotic it’s going to be a bumpy ride in the beginning but that's because it's real and needs to find its own footing not someone else dictating what it is.
Adam: That makes a lot of sense, and I also agree with the notion that Arbitrum's governance structure is more powerful and thus more complicated.
I'm curious: What advice would you give to, say, a founder building an app on top of Arbitrum where it's much more vanilla governance and they're thinking through how they make that transition from being a company to actually going through governance handoff or are there specific phases that they should keep in mind?
Steven: Yeah, so number one is, again, I think, really understanding where you want to be in this trade-off. So what I just described is, I think is great, but there are definitely other DAOs that want to be on a different side of the spectrum. Not to say that just because you're on Arbitrum, which it’s DAO does things one way that you can't have autonomy there. Decide how to set up your governance. There are pros and cons and trade-offs obviously in each way, and you need to work that out with your team.
The idea is what I described as a radical vision and many land on different points of the spectrum, and that might be just fine. I think the critical thing to realize if you’re a founder setting this up is that you can’t have your cake and eat it too, right? It's either decentralized or centralized. Obviously, there are different points on this spectrum, but if it's really decentralized, a lot of these more "chaotic elements" or elements of people rising and having opinions were part of the process. It's inevitable. It's actually to be celebrated. So I would say not to get too discouraged by that and that's important. The other thing I would say to someone building on Arbitrum is to learn how to make sure you’re represented as part of the larger Arbitrum governance system. As I mentioned, it's controlled by the community, and in Arbitrum, there are often sub-communities, and that's, again, one of the things that shows it's a robust ecosystem. There's not one Arbitrum community; there's the Treasure community, the GMX community, and, I can go on, the Xai community. All of these are aligned in some ways but not necessarily fully aligned. They have different interests and things that they'd like to see out of the DAO. Figuring out where to make sure you have a voice, you’re represented in the DAO and that can come from a few things. Maybe you want the DAO to fund initiatives related to your projects, right? So if the DAO is doing gaming initiatives maybe you have some other vertical you want funded, or maybe when the DAO considers technical parameters in the chain, like the DAO just launched ArbOS 20 "Atlas", which launched EIP 4044 support and reduced fees in certain ways. You want to make sure that, hey, my app is being considered in these considerations and we're looking at the full picture and asking how these changes affected this vertical and that vertical. So, having a voice in the DAO is important and that comes from both the developers and their community. Engaging in your community, encouraging your communities, maybe setting up a delegate for your app saying, Hey, community members, if you hold Arbitrum tokens, can you delegate them here to make sure that your interests are represented?
I think that's very important, too, to make sure that you have a big say in how the ecosystem underneath you is operating, will evolve over time, and make sure that you're at the table.
Adam: Yeah, the DAO politics is evolving pretty quickly.
Steven: Yes, absolutely.
"Community-Source Software"
Adam: I guess to talk about something more forward-looking or even explore the roadmap piece a little bit more.
How have you been thinking about the recent proliferation of app-specific rollups and rollup as a service platforms? Some folks would say that you could view this as a threat to Arbitrum as execution tech becomes commoditized but how, if at all, does that change your thinking about building a general-purpose layer?
Steven: I think you'll need both. That to me has become relatively clear. Arbitrum doing Arbitrum One and Arbitrum Nova are sort of the flagship chains of the Arbitrum DAO, but now there are 50 plus Arbitrum chains of development that I’m aware of. A few dozen that have already been announced. Arbitrum Orbit chains that are doing app-specific things and innovating in very, different ways. I'll get to that in a second, but the idea is that I don't think it's one or the other. I think some applications are better off on the public blockchains sharing liquidity and being resident with others, say, DeFi apps. A lot of the magic of certain verticals, particularly DeFi, is the synchronous interactions, the composability. That’s something which you really get on sort of one chain.
That being said, I think there will also be other verticals and other opportunities for chains to be dominant in their verticals. So Arbitrum Nova, for example, is doing really well in the gaming sector. It has many high-profile launches there, but there are some games; Proof of Play is a great example. They're a really strong gaming team, and they recently graduated, if you will, from Arbitrum Nova to their own Arbitrum Orbit chain because they were doing a lot of transaction volume. It made sense to have their own sort of chain their own environment that they control. They're not competing for fees with others, and the nice thing is it still uses the core Arbitrum technology to secure it. I think that there's a nice balance there is what I'll say. Also it gives you the ability to innovate in many different ways. So you have some chains like Rarible, RARI Chain, or Frame that are both building on the Arbitrum Orbit stack that are, for example, building in creator royalties and directing into the protocol.
You have a chain called Kinto building on the Arbitrum stack that is targeting institutional investors in a KYC environment. You have a chain called Fhenix, F-H-E-N-I-X, and F-H-E is also a cryptographic term for fully homomorphic encryption. They're building a privacy-enabled chain with deep cryptography baked into the chain. Innovation and customizability that's core to the Arbitrum technology and that's something which the Arbitrum stack gives you.
The ability to use a really solid base but also change it and customize it as you will. There are certain customizations that are sort of built-in. You could change the fee token, the governance token, the block time, etc., or the confirmation time. That being said, there are certain customizations like Fhenix, which is not built in, but you can just take the code and use it, rip out parts, add parts, and do whatever you want with it. You'll have a really solid base to do something truly innovative and not have to start from square zero. So, I think the Arbitrum stack's ability to give builders customizability and allows them to really do what they want and build things that are different and differentiated is incredible. I think it's a wonderful thing for the Arbitrum ecosystem.
To your question of is this a good thing? Is this competitive? That's the Arbitrum ecosystem was set up to address this problem. The term I like to use here for what Arbitrum has done is "community-sourced." So, typically, in software, there's open-source software and closed-source software, right? It's a very binary world. Other people can use your software for free. They have to pay you nothing, and it's just open. Or it's closed-source, and maybe the source is not even published. Maybe it's published, but it's all rights reserved. No one else can use the software you wrote. It's all yours. That's the binary in the sort of historic world.
What web3 and crypto enables is an idea of a community. An aligned community that shares resources in a very real way and has a shared ownership, and you can have what I like to call the community-source license, which basically says; if it is not closed-source cause you can use it, it's not open-source because you can't use it, and give nothing back to the community. It's community-sourced, which means you can use it but you have to join the community and actually contribute back. Cause when you see an open-source development, often you have a tragedy of the commons.
If you look back at the bugs and SSL libraries you see there’s trillions of dollars from the biggest tech company in the world being secured by this, and no one's actually contributing to it, right? They're having trouble raising basic funds for that, and that’s a problem. The way that the Arbitrum licensing is set up is if you want to build a Layer-3 on top of Arbitrum chains, Arbitrum One, or Arbitrum Nova, you can do that. You don't have to contribute anything directly because indirectly, you're already aligned. You’re already contributing to the ecosystem. You're building an ecosystem, bridging via the ecosystem, et cetera. If you want to go ahead and, to your point, build a competing Layer-2, you're not naturally aligned with the Arbitrum ecosystem, you can actually hand Arbitrum technology to its competitors, and Arbitrum technology is much better than its competitors.
It's not something that necessarily be in the best interest of the Arbitrum ecosystem to say, okay, to its competitors that have not built out all of the advanced features from security and just the UX perspective that Arbitrum has. You can use this technology for free. You have to contribute nothing back. So, there's something called the Arbitrum Expansion Pack, which says, okay, you can use it to build a Layer-2. You can even use it to build Layer-3 on a competing blockchain like Base. You can go ahead and do that, too, but you have to contribute something back as part of the license. In this case, it’s I believe about 10% of the sequencer revenue, 8% goes to the DAO, and 2% goes to the developer guild to build a fund where teams can contribute and get paid to build the core technology. It's a very aligned structure. You do what you want keep 90% of the profits but 10% you're asked to contribute back not to Offchain Labs, not to myself, not to any particular individual developer, but to the community of developers and dapp community that's building the software.
I think that's a very aligned model because, again, it's not holding anyone back. It's free to use. It's open; it's permissionless. You don't even ask permission. You just go ahead and use it, and the licensure says, but contribute something back or be aligned. That could be your alignment can come implicitly by just building on the Arbitrum ecosystems. You’re naturally flowing through contributions to the DAO, or it could be explicit by saying, okay, I’m not building on the Arbitrum ecosystem, but I'm going to contribute something back directly, and I think that's a fantastic model and builds that alignment because you don't have that risk of, okay, someone's to take this technology and make it obsolete. No, Everyone's aligned and contributing to building the core technology and raising everyone up together as part of that.
Adam: I love the term community-source software. In many ways, I think. We at Standard Crypto think a lot about founders needing to navigate community market fit instead of product market fit here. I think you could argue that is the output of community market fit.
I know we only have time for one more question, and I think we talked about this a little bit at a high level earlier, but I feel like I often hear you say Offchain Labs isn't a rollup company; it's an Ethereum scaling company. And there's a grander vision that ties into Ethereum's long-term scaling roadmap. Can we dig a little bit more into that?
Steven: Yeah. Offchain Labs is core contributor to the Arbitrum technology, right? We build Optimistic Rollups. Will also build internally Prysm. Prysm is the leading consensus client for Ethereum. We just had Dencun go live EIP-4844, and for us, it was a great experience and really a celebration of two layers. We contributed to the core Layer-1 software that made the update happen, and we also contributed to the Arbitrum changes that utilized those Layer-1 changes to make the rollups cheaper.
Basically, we had our hands in both of those and contributed to both layers of the stack. We're also, for example, contributing to the Espresso protocol for shared sequencing. So Offchain Labs’ mission is not to use an Optimistic rollup or a ZK-rollup specifically but to scale Ethereum, scale the security and decentralization of Ethereum, and also give a great user experience at the same time.
So, there’s core pillars of security, decentralization, scalability, and of course, UX is key as well. We're contributing everywhere we can. We're also building interoperability technology to help that vision come to light. So, that's where I view Offchain Labs. We contribute directly to two products in-house but also to others like Espresso. You'll see us do more of that over time and contribute to more. We're very aligned. We're not going to start contributing to something that is not aligned with our core mission. The idea is that as time goes on, as you solve one problem, often other problems arise.
I remember when we were raising funds for the company in 2018, people would say I don't think that scalability is our number one problem. It's something else. I remember this very well, and then 2019 and 2020 came, it was clear to people that scalability was our number one problem.
You can argue today that from a core computational scalability perspective, we basically solved that. Particularly, costs are down now among the Ethereum rollups generally, and that's relatively a solved problem. Then, there are other problems that are very much aligned like the user experience, for example. And just state growth; generally, things that affect every EVM chain, we have to solve as well. Our goal is not to stop at one very narrow vertical but to be broad in contributing to Ethereum. The easiest way to see that is via Prysm and Arbitrum, which are just two radically different points of the stack, but like I said, you'll also see more of that.
Adam: Awesome. Thank you so much for taking us through the Arbitrum journey. This was a lot of fun, Steven, and I appreciate you coming on.
Steven: Of course. Thank you for having me.
Disclaimer: Offchain Labs is a Standard Crypto portfolio company.