Optimism 联合创始人：L2 发生了什么？

Nothing said on Empire is a recommendation to buy or sell any investments or products. This podcast is for informational purposes only and the views expressed by anyone on the show are solely their opinions, not financial advice or necessarily the views of Blockworks. Our hosts, guests, and the Blockworks team may hold positions in the companies, funds, or projects discussed.

All right, gents. Excited about this. We have Karl, Ben, two of the three co-founders of Optimism and OP Labs—what am I supposed to say there?

A little bit of both.

A bit of both.

Yeah. How are we doing?

Great. We're doing great. Very excited to be here.

Your vibes are surprisingly high considering—

We're optimistic about our company for a good reason, you know what I'm saying?

Always the optimist. I actually had someone message me on Telegram the other day. They said, “Do you really not understand what's going on in the market? Why are your tweets so optimistic? It's very annoying.”

That's how we feel.

Sorry. Unapologetic.

Calm it down, guys. We're at a 10; we have to bring it way down.

I actually thought we could start on a more depressing note, and then we can bring it up. There's been—I don't know if this is the right word, and I mean this in the most respectful way—a reckoning of the L2s, I guess I would call it. I'd love to hear you guys walk through what happened here. If you go back to maybe 2021 or 2022, L2s were the hottest thing in all of crypto: $5 billion, $10 billion valuations. Then things changed. Solana got hot, then Hyperliquid, and now maybe it's back to the Ethereum L1. I'd love to hear how you think about what happened to the L2s over the last couple of years. Then we'll take the audience out of it and do the rebirth, obviously.

Yeah, I think, in short, it's hype meets reality. Crypto goes from being speculative and blue-sky—“We're going to conquer it all”—to actually trying to deliver on its value to end users. It turns out that delivering value to end users requires real product development, actually knowing what value you're going to give and generate for your users, and providing a good and cohesive product.

Let me tell you, if you've been in the crypto space, you know that that's a serious shortcoming of many of those chains or hype projects that really have not weathered the storm because they don't have those fundamentals. When you don't have fundamentals, the floor can fall out from under you.

Yeah. What do you think, man?

Yeah, I super agree. One way to frame this is that crypto has lost the ability not to hold itself accountable for the crazy that it does. There's one framing of this that's very, “What the heck were you doing?” On the other hand, we finally reached a point where people outside of our little bubble were taking us seriously.

It's not surprising that you would have a little bit of deflating as you entered the real world—not just meme-coin farms, circular dependency, and virality, but actually having real businesses that are considering using real products. At the end of the day, we came a long way as an industry with some of those vibes, and it got us to where we were. But to take it to the next level, you need a little bit of a reset.

Yeah. Professional. Button up that last button. Look at this guy. Yeah, yeah, yeah—zip it up. Zip it up, Karl.

1. What L2s Got Right vs Wrong

What do you think? We'll talk generally about L2s, and then we'll take it into Optimism. What do you think L2s got right, and what do you think L2s got wrong?

Yeah, great question. I think, fundamentally, L2s got the delivery vehicle for providing a good product correct. When you're building an L2, you're extending the capabilities of the L1 and being able to differentiate your blockspace to actually deliver a good product to your users. You're focusing on the correct thing, because the correct thing is differentiated products that your users want.

And not reinventing the wheel with yet another validator set and hype cycle—“Buy my token”—because that is ultimately unsustainable, even if it is wildly successful in the short term. What I think L2s got wrong is that it turns out you need to differentiate your blockspace. You can't just launch a chain, launch a token, and then throw up your hands when nothing happens, because it turns out that to deliver a good chain, you need to deliver a good product. Ultimately, a lot of the hype cycle wasn't doing that.

Do you guys think of your customer as the user? You mentioned user and product. Who's the user? Who's your customer? Your customer is not the end user, right? You're in the B2B2C game.

Yeah, of course we have OP Mainnet, which is growing super well, and of course there are B2B relationships there. Ultimately, you have to take a more holistic view of these products and try to understand them more deeply, because the answer is not one-size-fits-all in any way.

There are many relationships, and the power of crypto is the DeFi LEGO world. You're stacking up all these LEGOs. Should you call that a B2B, B2B, B2B, B2B2C product? That framing is a little bit reductive, because effectively what you're building is a network of participants. You need to apply the product framing and understand the relationships and the value being provided from one user to another. Ultimately, we are sitting at the infrastructure layer that's serving a lot of levels up. That's true, but it's also not one-size-fits-all where you just say, “B2B, B2C—these are the 2 options in crypto.” It's a little bit more complex.

And is your product blockspace, or do you think about it differently now?

I think that is probably the closest thing to our product. It is the infrastructure to host blockspace, at minimum, and the ease of creating net-new blockspace that is differentiated and enables your product to do whatever your product needs to do. Ultimately, blockspace is only as good as what the users are using it for.

If you have to put it in a word, blockspace is a reasonable word. It also begs the question: What the fuck is blockspace? In some cases, that is an asset-issuance substrate on which you're providing assets. In other cases, it's a substrate for building financial products.

So, yeah, it is blockspace, but you have to dig a little bit deeper if you want to more deeply understand how people are using the blockspace. That's where you need to make the product decisions in the first place.

Yeah. When you guys were trying to figure out how chains will look and what chains will be used for—how has your thinking evolved? You can go anywhere you want with that question.

You can think of it as a natural evolution of most technology stacks: bespoke to standardized to specialized. This is a very counterintuitive thing, but in some ways it's a reverse cycle from what you would imagine, which is standardized to specialized, or even specialized to standardized.

We started out building really bespoke technology stacks that were hyper-focused on a single use case and only enabled that one thing. For example, payments—that was the first use case for L2s. It was, “How are we going to scale up a payments network to global scale?” We knew we needed a Layer 2 for that. That required a huge amount of technical development to build a single-use-case solution.

Then we went into the specialized camp. That is a technology stack that can technically be used for anything, and this is where the trap happened. You ended up with 100 different clones of the exact same product: “We are going to be the standardized blockspace for all things to all people.” That is not differentiated, nor does anyone want that.

What we're seeing is a transition from standardized into specialized again, but it's different from bespoke. Standardized to specialized means that I can take a general-purpose platform and tune it easily to my particular use case. I don't need to spend a 5- to 10-year R&D cycle to do it. I can extend it and make it different in a product cycle that's a few months, as opposed to multiple years. That's the transition that we have seen and that we are really trying to accelerate at this point: enabling everyone to build their differentiated blockspace across the board.

When you guys were trying to figure this out, I'm assuming Blockworks has had these times when things were really hot, and then times when we're like, “Oh my God, it's all over,” and then it's really hot, and then it's all over. I'm assuming at some point in the last couple of years you went through the “Oh my God, it's all over” phase. You guys seem good right now.

But how did you guys figure out what to do? Even the bespoke, specialized, standardized stuff? I know you have great investors, like Andreessen, for example. Are you sitting down with the Andreessen folks and they're like, “No, no, no, no, no, Karl, this is actually a very normal thing”? Was it the internal team? Are you talking to the whole company? Is it just the 3 founders? Are you talking to customers?

Can you just walk me through how that works? Maybe it's a really nerdy, founder-only question that no one will care about the answer to, but personally, I'm very curious how you guys went through that journey.

I mean, it's a mix, right? One of the things that we realized relatively early on was that we needed to actually build out a pipeline for talking to customers at this kind of scale, at the institutional level. I know, right? What do you mean? We don't just get in random Telegram group chats and try to find other anons with, you know, like—

We really need a CRM. How do you—

I know, yeah. It's like, whoa, we're growing up, right? That's back to the bubble going down and then you playing it back up.

So, you know, it's obviously a mix. The other tricky thing about it is because you have this more complex network of participants that I was talking about, where it's—

Right.

It's not just one layer of B2B to C. There are many routes through the network that a blockchain is building. You definitely have to talk to a lot of people. You want to understand what traders are doing at the same time as you want to understand what asset issuers are doing, at the same time as exchanges.

Yeah.

One part of it was just building the reps of being able to talk to the serious players that are getting into crypto now and being able to understand what they need. But then you also have to go at different layers of the stack. It's not just the chain, the next chain purchaser, but who they're going to talk to.

Yeah, it might sound like very silly advice: “Oh, yes, we talk to customers. We need to talk to customers.” We've known this from day 1. But something that I learned recently, which I thought was a great little insight, is that the difference between a professional, world-class athlete and your average Joe is how incredibly perfected their basics are.

So, we ultimately need to get in those reps, and that's how you can actually build something that people want, because this is really hard. Talking to customers regularly—I struggle with that more than most—is really hard.

Yeah, it's tough. It's also unlearning the crypto-bubble language for things, like starting to understand—Karl knows this really well. We've spent days trying to understand the traditional financial system, only to realize it's actually quite obvious, other than the fact that it does things exactly the opposite of how crypto does it in the first place.

We were just assuming that the way crypto does it is the way that everyone does it, you know what I mean? It's crazy. Just being able to translate between those 2 worlds is super challenging.

Not just finance. Traditional SaaS go-to-market motions, customer success, BDRs, and SDRs. There's a lot there, which I actually want to get into, because I met your guys' chief revenue officer, Kyle, who's fantastic. I want to get into the transition to this enterprise-scale business in a second.

If someone's listening to this thinking, “This all sounds great, but the L2 meta is over. L2s are dead,” how would you respond to that?

2. peaq Ad

It's like the server meta is dead. Okay, what are you building? I think it is true that selling a server in the abstract to a hype machine is dead. But you need to sell a product, and to build a product, you need a server. So, we're here for you.

Robots and machines will outnumber humans on chain, and Peak is how they get there. Peak OS enables any machine to do business on any chain. It turns machines into autonomous actors and liquid assets with ease, giving robots access to capital and the ability to compound their value. That means funds, allocators, and institutions can underwrite, finance, trade, and route capital to machines the same way they handle any other asset. Millions of new potential consumers for any Web3 service from compute to storage to anything a machine may need without a human in the loop. Head to peak.xyz or click the link in the show notes to find out more.

3. Optimism’s Strategy in 2026

Okay, let's talk about this evolution of Optimism now. How do you think about the business model of Optimism evolving? Did you guys even think about the business model at the beginning, or were you just thinking, “Mass proliferation of this block space, and eventually the money will come”?

I mean, ultimately, we are still in the early innings of crypto adoption, right? When we're talking about being taken seriously as an industry in the TradFi world, or just anywhere outside of our own little bubble, we are still—even with the proliferation of stablecoins and everything else—in a fraction of a percent of building the markets for the world.

In some sense, you have to keep an open mind for things evolving over time. One of the simplest understandings that I think we came to more recently was that we went, first of all, from having 1 OP Mainnet chain and just building that out—and that's still around, and that's doing really well.

Eventually, we hit a point in time at which it became time to scale that up. Now we have built 1 server, and that server runs well. Now we need multiple people to be able to use this server technology, and we're ready for that.

So, now we're in the world of OP Enterprise, where we have a managed offering. You can basically say, “Hey, I want a chain,” and we say, “Don't worry, we got you. We'll handle all the technical details for you. Just focus on what actually matters, which is—”

Your product, your differentiation.

Specialization, baby.

Mhm.

Was Coinbase the breakout deal where you guys realized, “Oh, this is a business model: providing specialized block space, a customized deal to an enterprise that needs block space, which we can deliver to them”?

It was absolutely. It was absolutely a clarification of how these technology stacks are going to be used in the future.

I will say we definitely had a bit of an unfair head start in that process of actually professionalizing, because it turns out that to sell to enterprises, you need things like SLAs and SLOs. You need all of the actually useful things that the traditional stack provides to its customers. We really didn't have a way to replicate that, and so it's been about building that muscle.

It's been exactly that. It is forced professionalization. Now, of course, it means you have to talk to customers to get there, but it's really being guided in this direction, I think, by the demand and by market signals.

How has the way that you guys spend your time changed?

Wow. The role of a founder changes literally every 6 months. You have a new job title. It's like, “Okay, this is the most pressing thing for the business. Go do that. Go learn how to do product,” et cetera.

I wasn't the one who learned how to do product. I'm living in that research mode a little bit too much, but the useful thing about having a bunch of us is that you find the right hat at the right time. We evolve, and the business evolves, because the business demands fully different skill sets at every stage of the cycle.

Yeah, yeah. I've seen that with Mike and me. We've completely shifted our roles every 9 months or so. What are the roles for the 3 of you right now?

I've recently been deep in the engineering weeds. It is a real change in the way that we've been building. With the advent of these incredible tools, the way that we develop software, the way that we think about securing our software, and all of these processes and pipelines have massive opportunities to optimize.

We're shipping some really ambitious features that require some deep, in-the-weeds attention. That's been my focus in the past 6 months. But before that, it was totally different.

So, I mean, I'm also in a little bit of a research journey right now. I think the—

So, where do the customers come in now?

No, we—

That's a very good team.

Yes, very good team. We also have a recent habit, which has been the happiest thing ever, where we have a competition over who can bring in customers every single day to the office.

We have a conversation on a different topic. This person works in finance, this other person works in another sector, and we're just gaining all of these insights and piecing it together.

And to be clear, the important research problem that I think crypto has to solve right now is understanding—or coming to terms with—the different market structures that TradFi has, where they’re actually massively beneficial. I don’t mean just, “How can we get more TPS out of the L2?” It’s, “What can we learn from the market structure of broker-dealers being separate from exchanges?” And settlement. We in crypto have tightly coupled all of these things together, and this gives us this very cool, permissionless, decentralized substrate.

But there are also economic inefficiencies that come out of that, and it’s very important to learn how the rest of the real world has solved those inefficiencies in practice and try to figure out how to incorporate those.

Yeah, and then how to build a product that can serve them.

Exactly.

And there are some things we do better, some things that they do better, and we need to do those things better.

And those are opportunities. Ultimately, the pace of innovation that is possible in crypto is something that other sectors are not going to be able to match. So, yes, we have these inefficiencies. Yes, we have these crazy inefficiencies in a lot of cases. But we’re also patching those holes every day, and the people who patch the holes first get rewarded, and we’re trying to enable those people.

Yeah, 100%. Can we talk about the business model of Optimism today? The way I understand it is, it’s almost an enterprise SaaS model. You have this great block space that you spent years building, but there’s all this other stuff that comes around building on a chain. So, you guys have built this full-stack solution where, if someone wants to go build a blockchain, they can come to you and you offer everything.

I don’t know how much you offer, but maybe someone needs a wallet—you can probably offer that. You offer all this stuff. So, it’s a blockchain in a box. You charge a lot of money for that, which I don’t know how much, but I’d guess five to six figures a month or 7 figures a year. I would guess your target customer list is actually not 10,000 people. There are probably 100 to 200 names on a spreadsheet. Is that roughly the business today? Is it an enterprise SaaS model with 7-figure annual deals?

Yeah, that’s exactly right, and that is something that we have a foundation for because we have experience building out our own chain and building out other chains. It turns out that chains are a great opportunity. They allow you to launch products that otherwise would take an enormous amount of time and effort to launch. However, if you’re doing it from scratch, they can be a massive hurdle.

You don’t realize how many integration partners you need to work with, how many things you need to stand up, how to balance the various liquidity incentives, how to make the lending pool work just right with your spot exchange. Token availability is an incredibly important situation, and fragmentation is massively inefficient. All of those things are learnings that we’ve accumulated and therefore can leverage in all of these enterprise deals as we bring these enterprises on-chain.

Because guess what? They have enormous potential to reach a large number of users with incredibly useful financial products and other products that otherwise they would not be able to offer, and to offer new tokens and new experiences that are just not super feasible unless you’re on-chain. And so, we’re trying to accelerate that transition. We’re not kidding ourselves. We’re not the biggest exchange. We are instead the people who know how to take that distribution and leverage it for a new product line.

Yeah, and I think for the customers listening in on this, the only thing we might add is that there’s still a little bit of solutions engineering in there as well. Especially now, with the regulatory landscape coming into focus across different jurisdictions, there is specialized work that we can do to extend the chain technology to meet people where they are.

Yeah, for example, if you have a token and you need to apply some kind of compliance controls on your token, the best place to do it is in the sequencer itself, so that you don’t have any risk of that token spreading and getting into the wrong hands. You can impose trading locks where you lock down trading for all of the different pairs for your particular block space. There’s so much more customization that you get when you actually have control over your block space, where you can customize your block space and access the on-chain distribution at the same time.

What else goes into this thing that you’re selling? A financial institution comes to you and says, “We want a chain, and we want the full stack. We want everything.” What’s in there? A block explorer? Maybe bullet points: a wallet? What do they actually get? What are the deliverables?

Block explorer, running the actual sequencer. There are a few tiers, by the way, depending on the throughput scales and things like this that you want. Block explorer, compliance modules, the sequencer itself. The other biggest one, quite frankly, is just RPC providers. You have to keep in mind that whatever app you’re building on top needs a connection to the chain state to read it all down, and that’s a huge industry.

DeFi protocols—there’s a whole laundry list.

You would give them lending and borrowing. They might say, “I want native lending and borrowing. I want a native DEX.”

Oh, yes, this is another thing that we are actively exploring. That is a level of customization that is absolutely the frontier of block space. How do we build native lending into the chain? How do we build in your particular use case?

We’ve seen what Hyperliquid has done.

Exactly. And so, really, that is part of this bespoke engineering solutioning. Part of this is, “Okay, you want a chain, but why do you want a chain?” If we had the one-size-fits-all solution for you, then we’re recreating that standardized, nondifferentiated block space that has plagued L2s and L1s, candidly, since crypto’s genesis.

So now it’s really about, “Okay, what is the particular use case that you are targeting, and what are those experiences?” Then we just happen to know the laundry list of things that you need to actually land that particular beachhead product.

And are you guys white-labeling other people’s products? For an RPC, are you going to Alchemy, Infura, or QuickNode or something and saying, “Hey, white-label this thing”? And for the wallet provider—I’m not sure who provides wallets right now, but—

We’re the one throat to choke when you’re trying to work with a bunch of providers. We have these agreements with a bunch of providers, and then you can come to us. We know how that process goes down and can make a change.

A provider here, and then it goes—okay.

Exactly. And so, it’s really about collapsing the chain-onboarding process from 9 months down to a couple of weeks.

There are 2 business models you could maybe go down. One would be fees: “We will give you a chain. We’re not even going to charge you, and as you scale, we’re going to take a percentage.” The other model is enterprise SaaS: You can scale, you can make hundreds of millions, and we’re not going to take any fees. We’re just going to use an enterprise SaaS model. It sounds like you went enterprise SaaS.

Yeah.

Is that right? And why?

I think one of the things that’s really important, and that crypto has really undervalued, is legibility to existing businesses. If we’re going to say, “Hey, we’re going to take a percentage cut of all of your MEV,” people are going to be looking at us like we are crazy.

If we instead say, “Hey, we have this very predictable pricing model. This is what you’re signing up for. You can get control over your infrastructure at any point, but in the meantime, we’ll charge you this much per month based on usage, et cetera, et cetera,” that is something that someone can read and understand versus a kind of sneaky, jargon-filled contract that doesn’t make a whole lot of sense.

You add the fees and the data fees, then you subtract it by the amount of L1 data that you spend, and then you divide by another number, and there’s a—what are you talking about?

Yeah, exactly. And to be clear, when you’re working with an institution, they want to pay you. They want to understand why you’re not going to go out of business in a couple of months. So, part of having a legible fee model is actually so that you can have a trusting relationship with your customers, so you know where the lines are.

What’s been the toughest thing from a culture perspective? Blockworks has seen multiple iterations of the company.

They're like, “Media and events,” and then we got into data, and then we cut the news business. There are big cultural shifts that have to happen there. We've felt these big cultural shifts, and they're externally maybe not as seen, but internally they're felt, right? What does that feel like internally?

I think it is a new experience, a new challenge, all of the time. I would say that this recent cultural shift away from the kind of hype-cycle version of crypto into the kind of enterprise, rubber-hits-the-road version of crypto has been a particularly challenging transition. Ultimately, it is challenging for everyone, but it is also a really fun thing to do: you come into the game saying, “I want to deliver impact to tons of users. I want to have a real change in the world. I want to make something positive happen.”

It starts out as a dream, and then when it becomes real and you start seeing people using your tech stack to do incredible things across the globe, en masse, that is a very satisfying experience. It's a different kind of satisfying experience from dreaming about the car and then having the car, but it is an incredibly valuable journey. I feel like the folks who have been along for that ride are very grateful for the experience. At least I am. I don't know if I should speak for others.

The transition is a little challenging, but then you start to see the payoffs, and everyone can really get aligned behind that and say, “Oh, we are having an impact here. This deal is closing because we are providing a valuable service to these people.” That's where it's positive.

Now, I saw 2 announcements go live recently. One was Bitpanda's Vision Chain.

Yeah.

Upbit, is that another one?

Yep, exactly. Upbit.

Did you guys do OKX's chain?

Yeah, OKX.

I want to get to the customer stuff in a second. I was talking to the founder of an OG DeFi app, and he was saying, “It's interesting. We have almost 2 sets of employees now. One was the dreamer set.”

Yeah.

“Which is the initial group, and they're here for the dreams and the vision. Then we have this new class of employees, and they're actually almost more productive today. They're almost better for the stage we're at now. There's not a clash—we're rolling off the first set and onboarding the second.”

Yeah, and I think the dreamers can become practical implementers. In some ways, that's the point of being a dreamer: to see the dream manifest. You can't get stuck in the dream cycle.

Yeah, true.

I think we're lucky to be on the infrastructure side as well, right? These shifts aren't as dramatic as a complete 180-degree pivot away from the fact that block space exists. Yes, “block space” is a reductive word to use, and what we're talking about is actually getting to the deeper value that we're providing, but you still have chain infrastructure. There's a nice bridge there.

All right, we'll get to the customers in a second, because I want to hear what they need and what they want. What they need in an exchange is something different from what they need in a fintech. What does the revenue mix look like today? How do you guys break it out? Is it fees from OP Mainnet, enterprise contracts, ARR, and transactional revenue? When you show the P&L on a quarterly basis, how do you look at it?

It's a majority from the enterprise sales motion, and we're actively growing that business right now. Historically, it has been a lot of on-chain fees, but ultimately that has served as a bridge to what I think is our more sustainable business model long term.

Yeah.

The only other thing I would add is that we haven't actually talked about this yet, but one of the other key product and go-to-market realizations people need to have is that not everybody needs a chain, at least right now. I know, I know—that's a time-frame caveat right now, right? But we also have OP Mainnet as a really successful, growing chain.

Folks like Ether.fi, for example, just moved their card business over and migrated 200 million, so they're—

I saw that—not without some drama.

Oh, my God. Yeah, so anyway—

Moment of silence.

Moving on. The reality is that we also have this OP Mainnet product that is block space, which we view as a launching pad for people. Today, Ether.fi doesn't need its own chain. In the future, it could be the case that they need one, so you have that transitionary motion to lock in as well.

4. What’s Optimism’s Moat?

Interesting.

Why should someone come to Optimism and work with Optimism? I would think that, in these enterprise motions, you probably compete with 3 or 4 other chains. Sorry to the folks I'm going to forget here, but it's probably Arbitrum, Canton, Solana, and ZKsync. I'm sure someone's going to send me an angry message because I'm forgetting people, but those are the folks that come to mind. Why would someone come to you? Why do you win these deals?

It is the production experience. Being able to talk about the ways that we have transitioned people over to their own chains, with more than 50% of all L2 transactions—it really comes down to experience, time, and reps working with all of these enterprises. It's about making sure that we're able to differentiate their block space for what they need and do that in an efficient process, with a long track record of folks we've worked with.

Yeah. I would also say that it does become the case that human trust comes back into the picture, right? That's an important thing to provide and actually act on. One of my favorite meetings that we have, which is in a really good place now over the past 6–12 months, is our monthly engineering prioritization meeting.

Basically, there's all the work that we could do, and we are ruthlessly going through that list and saying, “What is the value of this work? Who is the customer that it's providing value to? How urgent is it?” Then we prioritize exactly according to that list.

Yeah. Being able to act and build trust with what we build is, I think, super important.

Did you have an investor who told you to do that meeting?

We've been blessed with some incredible executive talent, some of whom you have met, who have taught us, as very inexperienced folks, how to actually run businesses.

Yeah, what a concept.

It's harder than it looks, to be clear, and I'm not saying I'm doing it.

Yeah. How big is the team right now?

It's about 90.

90 folks, okay. What's the revenue, roughly? It's funny—it used to be all on-chain, right? Now you have this off-chain revenue, which is a weird thing. The industry is going—there's more and more revenue happening off-chain, which we're trying to figure out as a data business. How do you figure that out? I know you can't tell me exactly what the revenue is, but what's the range, roughly?

Your estimate is probably about as good as mine. Your description of roughly what we're charging—you can probably work it out. But, yeah.

Who are your favorite customers? You can't pick favorite children here.

It seems like you're having the most success with the centralized exchanges. Is that right?

Exactly.

Yeah. And why is that?

We have a huge amount of experience working with exchanges, building their own chains, bringing their distribution on-chain, and enabling the products that they're trying to build.

I mean, you could list them: Coinbase with Base,

Kraken with Ink.

OKX.

OKX and Upbit and Upbit.

Upbit.

Upbit with GIWA.

Yep. These are the biggest chains.

We've been doing this for a long time. In fact, we have a target list of all of these different customers, and we are building relationships with each one, understanding their product needs, and enabling them to deploy their products on-chain.

Outside of exchanges, who do you guys think you'll have the most success with? Let's fast-forward Optimism 1 or 2 years. Who's the group that you'll have the most success with outside of exchanges? You have fintechs, payment providers, financial institutions, crypto companies—let's not forget an Ether.fi.

Oh, sure.

Who do you think will be the second-biggest customer? Who will be the fastest-growing customer segment in a year?

I think it's natural to go to folks like fintechs, based on their desire to onboard on-chain. However, I will also say that we are seeing a huge amount of interest from traditional financial institutions that are coming on-chain with a whole range of products. Whether it be the NYSE, Nasdaq, or ICE, all of these folks are absolutely making crypto investments.

And so it’s really a matter of when. At that time, depending on how it plays out, that could be a massive, massive opportunity for all of Web3.

Yeah. And then there’s a little bit of agility factoring into that equation as well. Some of the bigger players—and the more TradFi you are—the slower the sales cycle ends up being. You know the game.

It’s fun.

How do you think about incentivizing people to work with you? For what it’s worth, most of the big deals that people see in public—“This chain won this huge deal”—are paid. There’s some payment happening. That gets a really bad rap, but it’s not actually the worst thing. A lot of that is actually how AWS landed a bunch of its early customers: by paying. It’s not actually the worst thing to do, but it gets a bad rap in crypto. What’s your philosophy on that?

There is absolutely a time and a place. However, that time and place is on the way out, I say. One of the things that candidly loses us deals—deals that we would make in an ideal world—is that some of those deals are won and lost just by how much you want to pay. At the end of the day, if you lose a deal like that, then you should feel good, because that means you are not incurring this massive debt on your balance sheet.

We are transitioning, really importantly, to look for, believe it or not, paying customers. That has been a—

Crazy concept.

Crazy concept, but candidly, even if it’s an extremely exciting opportunity, sometimes you really want to ask, “Do you want to pay for this?” Because that is a signal for whether or not you want the thing in the first place. A lot of folks will be tempted to launch a product just to get the payout.

Yeah.

The other thing to look out for is the blank-check deal. There’s a big difference between putting some payments between milestones and having incentive-aligned success, where both parties are doing really well. That’s much more synergistic than just having a number on the line to get the signature.

5. Who Needs To Build a Chain?

I got a call from somebody the other day. They’re being offered $125 million by a chain. I’m sure you’re talking to these guys, but they’re being offered $125 million, and they’re like, “Should we do it?” I was like, “If they’re offering you $125 million, you probably have to take that.” What do you think? Why does someone need a chain?

At the end of the day, they need a chain to build a product that is going to be used by end users. Having a chain is incredibly important relative to just deploying on someone else’s block space if you want the distribution to the on-chain user base and the ability to deploy new applications. So, there are 3 ways.

Make sure I understand that. Because if you deploy on Ethereum L1—

Sure.

—you get the distribution of Ethereum L1. It’s actually almost hurting your distribution if you take everyone into your own chain, no?

Definitely. That is the case in the status quo today. That is certainly a bug, not a feature, of the tech stack. The vision, and the reason why layer 2s were even successful in the first place, is that you were able to leverage the Ethereum distribution when you deployed your layer 2 in the first place.

Is that the same level of interoperability as deploying a net-new smart contract on Ethereum L1 today? No, it’s not. However, is that eventually going to absolutely be the case? Yes, it absolutely is.

So, someone should be able to opt into the interoperability or keep it private on their own—

Exactly. And regardless, still be able to connect to the liquidity and tap into all of these other chains. Ultimately, that is a technical problem that will eventually be solved, both on the financial and the infrastructure sides.

Once we solve that, asking why you want your own chain is almost like asking why you want your own website. It’s because I am a business, and I care about owning my own space where my users go. I can control it, I can customize it, and I can distribute it just as well.

Yeah, I agree. I also think implicit in this question—this is the golden question of what block space means. What are we caveating away or silently ignoring when we describe it? That’s something that’s really relevant to this conversation as well.

A chain for an asset issuer versus a chain for a lending protocol is going to end up, in this end game that Karl is talking about, looking like very different things. They’ll both be block space, because block space is just a server. But look at all the different types of servers and ways that servers are used. There’s a big difference between a big data back-end server and a Twitch server that streams out videos and signals to all the people watching some viral stream or something. These kinds of differences matter here, so it can’t be one-size-fits-all.

I was talking to one of your customers—maybe it was Bitpanda—and they were telling me that they wanted to have a European stablecoin. Did they end up doing this?

Yeah, yeah. They’re going to be the first L2 with a euro stablecoin as the gas-paying token.

Okay. Small customization probably matters a lot to them.

Yes.

A massive impact. So, what’s the actual—so they want to have a European stablecoin. Is it Circle’s euro stablecoin or something else?

I don’t remember off the top of my head. The whole thing for them is about compliance as well. The gas mechanism is part of that.

The gas mechanism for that?

Yes, exactly. And they’re doing the full MiCA compliance suite, so they’re going all out for the EU regulation. With the advent of MiCA clarity and all of these different regulations, we are seeing a huge number of folks who would otherwise maybe be a little bit unsure about coming on-chain. They’re coming on-chain, but they need their own chain. They need to be able to customize it so that they can comply with the various regulations that are coming out today.

If you had to rewrite the Coinbase deal—I know that was the first big enterprise deal you guys did—what did you learn from that deal, other than the generic answer that you learned how to work with an exchange and how to tell them how to do things? Tell me what you guys would have done differently with that deal.

Empirically, it is this: create a deal structure that is legible to traditional financial institutions, other centralized exchanges, and generally well-run businesses. That would be ideal, rather than a deal coded in the language of crypto that we grew up in and obviously know like the back of our hands, but that ultimately is not a scalable way to talk about these enterprise deals.

We were like, “Okay, now we’re learning. This is how we actually speak to enterprises. This is actually what enterprises want to use our tech stack for.”

Yeah, and you are—I’m sorry. I should have caught up on this before, but are they rolling off, or are you still working together?

Yeah, we still work together. The general way that they’ve taken their chain to the maximum degree of customization is by making changes to the Base stack and to the protocol. They’ve rolled out this really cool Azul hard fork recently.

They’re kind of at the end of the customer life cycle that we’ve started to see. A lot of customers want to come in, want a chain as quickly as possible, don’t want to host it, and don’t want to make a huge investment, but they want that product payoff. So, we enable that and bring them in, give them their chain, get them set up, and grow their chain.

Then, like Coinbase, they have a big team of engineers that want to make massive customizations and differentiate their block space even more. We’re an open-source stack. That is why we are open source: so that you can make these customizations. They’re kind of at that tail end of, “Wow, their chain is so customized that they just want to do whatever they want with it.”

That is ultimately part of this whole cycle that we’re starting to see as we transition to enterprise sales.

Step 1: an Ether.fi comes and deploys on OP Mainnet. Step 2: they use the full stack. Step 3: if they get to this stage—which probably most won’t, but they could if they wanted to—

It’s amazing that they can. That is a real case study in what it looks like to be a successful chain.

What do you think of some of these? Anything else on the Optimism business model or customers that you think we should talk about?

Wow, we did go through a lot. My gosh. Open-source standards are one thing that I would say, of all the things that we changed, has paid off the most dividends. Even when you look at Base making these customizations, we’re still able to collaborate with them on new protocol features that are going to be useful to others and things like this.

And so, I don't know. Shout-out to open source, because we didn't do it yet.

Yeah. True, true. Speaking of open source, decentralization, and some of these other topics, what do you think of some of the new chains that have gone in different directions?

Different strategies make total sense. I think it is great to see, again, differentiated block space. Sorry to be a broken record here, but we need some closed-source chains to have a chain ecosystem that's actually highly differentiated. Is the closed-source chain going to be the standard, open standard for printing your own chain or creating your own chain? Probably not, but is it an awesome thing to see? Absolutely.

I'm really excited, candidly, to see all the experimentation and people trying to find different use cases for crypto that will hopefully expand the market and the market opportunity. Because once you do make that expansion, once you show that opportunity, that is an invitation for another person or another company to come in and try to run a competitive product. We want to make sure that we're there and talking with all of them.

Yeah, that experimenting is super key. When we were all in a little crypto bubble together many years ago, I feel like the idea of a permissioned chain was this—

Oh, yeah. It's because enterprise blockchains had one wave of complete failures in 2017. Yeah, blockchain, not Bitcoin.

Yeah, and the reality is these are two points that are very far apart, on extreme ends of the spectrum. Going back to what I was saying earlier, there are things that TradFi gets right that are strictly more economically efficient than what we do in crypto. These are the middle grounds that these experiments are trying to tease out, which is super relevant.

It's super useful to think about having a chain that is more public and is used for settlement, with a matching engine that is a little bit more private. That kind of nuanced take is not where we were in the failed 2016, 2017, and 2018 enterprise permissioned chains, where you would write your validator set into a database. So, it's super important to find those middle grounds. They become useful products.

Is there anything that you've learned from Solana, Hyperliquid, or Canton where you thought, “I actually really like the way that they do this. We should think about doing our version of that”?

Absolutely. First off, I think Hyperliquid, for example—one of the darling children of crypto today—is really showing that you can build a competitive exchange that is decentralized—well, is Web3, on-chain—and really can provide a good user experience. That is opening up a new design space. It's related to what we talked about earlier with all of the enshrined DeFi, with all of the enshrinement and customization talks.

It's really created a new surface area and de-risked that surface area to provide even more sophisticated on-chain products that maybe people wouldn't have thought were possible previously. Canton is obviously similar in a sense. We're going to have many chains, they're all going to connect, and that's something in the water.

I think it's very obvious that this is the future. There will be many chains, they will be interoperable, and they will be a massive network of differentiated block space. So, really, it's just about enabling that future across the board.

Yeah.

My brain went to Solana for the same reason of enshrinement, but it's the same reason, right? Enshrine it. It was a very interesting decision for Solana to take some of the core primitives, like the idea of a token, and embed that into the native definition of a chain, as opposed to just saying, “It's all smart contracts. Your tokens, your everything, are smart contracts.”

It's probably an interesting trend to imagine the world in which Ethereum did that. Ethereum probably wasn't ready for that, but when we're talking about specializing—now that we've gone from bespoke to general and now we're back to specializing—that is a very interesting use case.

6. The Current State of Etheruem

I'll talk about ETH for a little bit, maybe. Karl, I first heard of you when you were, I think, pre-Optimism, doing research at the Ethereum Foundation. I'd love to hear how both of you guys think about Ethereum the network, ETH the asset, Vitalik, and the whole Ethereum Foundation. Just unpack your thoughts on Ethereum right now.

I'll give some hot takes. Generally, I think Ethereum is wildly misunderstood by the ecosystem in a lot of ways. The ETH asset is also very misunderstood by the ecosystem in a lot of ways, and I think that's very natural, but it's really worth stating.

Ethereum, in my mind—hot take—is competitive with Bitcoin. That is its primary competitor. It is a decentralized store of value: maximally distributed, maximally credibly neutral, maximally crops, for example. Its competitor is Bitcoin, this mysterious Satoshi. That is actually a very compelling narrative and a very compelling product in a lot of ways. That is where the networks go toe-to-toe, in my view.

There has been a rise, with the advent of all of this differentiated block space, where it's like, “No, Ethereum needs to compete not just with Bitcoin. It needs to compete with Solana. It needs to compete with Hyperliquid. It needs to compete with”—the list goes on. At the end of the day, that kind of competition is, first, not Ethereum's specialty, and second, a losing game.

This is also why Ethereum is going in the direction that it is. It's going toward decentralization, credible neutrality, robustness, and security. The Layer 2s are going toward differentiated block space, fitting in chains across the spectrum, competing with the Solanas, competing with the Hyperliquids, and all of that stuff.

You can't have it all in one product. You really can't compete with everyone, because then you lose to everyone. Ultimately, that is the misunderstanding, I think, of Ethereum among many folks in the Ethereum ecosystem.

It's normal. Even companies get really confused about their product positioning. There are internal questions: Do we compete with them? Do we do this? Think about how hard that dynamic is, and then multiply that toward a global, decentralized ecosystem with a whole bunch of opinions. It is a little bit chaotic.

That's why when you go on Twitter, you're not going to get that clear-cut story, but that's at least what I think is the consistent through-line for the Ethereum and Layer 2 space.

So, if you were Ethereum or Vitalik—whoever you want to be—would you stop trying to bring people onto the chain and build on the chain? Would you more try to be a store of value?

The idea of bringing people onto the chain is absolutely not in conflict with being a store of value. It's just: Who are you bringing on-chain, and why are they using that chain?

Ultimately, Layer 1 is useful for a certain class of things. That class of things is absolutely really important, widespread, and going to have a massively positive impact on the world. Is that class of things literally everything to everyone? Absolutely not. It doesn't make sense.

These are not really in conflict. They're only in conflict when you think about it from a zero-sum, reductionist viewpoint. Then you look at it more holistically and with a little bit more of a growth mindset, and you're like, “Oh, wait, the pie is way bigger than a single type of block space.” That's literally the mantra that I've been repeating, honestly. I apologize for repeating it throughout this entire podcast.

Yeah. Are you in direct competition? Is Optimism in direct competition with Ethereum?

Ethereum?

Yeah.

Yeah, that's the most insane thing. I think it's a Solana psyop. I actually think originally it was some crazy community members who were like, “Yeah, we know how to take out ETH. We're going to convince them that they're competing with the Layer 2s. Divide and conquer.” It's the classic strategy. It goes way back.

So, anyway, that's what I think happened. From Twitter, you get some crazy stuff. The relationships—that's what matters.

What do you think the role of Vitalik should be?

That's a good question. His role is a leader. His role is to set the direction, and his role is to make unpopular decisions. That is ultimately a very challenging role, but an incredibly important role.

Saying no is the most important thing, right? Being able to say, “This is what we are doing. That's where we are. That's our differentiation. That's what we're investing in,” and sticking to it. I think that is what Vitalik has been doing and continues to do.

Yeah. What do you think?

I mean, we've had a lot of late nights together on this one. We're pretty aligned on the Solana psyop, really.

Yep. Yep.

Yeah, I think the only thing I might add is, to your question of bringing people on chain, does that mean Ethereum shouldn't be growing its ecosystem if it's focused on being a store of value? Of course, the answer is no, and these 2 things are not in conflict. They could be in conflict if Ethereum tried to do that so hard that it sacrificed some of the valuable decentralization and sound-money store properties, right?

I don't see Ethereum doing that, but if it did start doing that, that would be where I started raising red flags. I'd be like, "Whoa, we've got to shift." But for now, we're getting what we need. It's good.

5 years from now, who's more successful, Hyperliquid or Solana?

Ooh, wow.

Spicy one. Should I say shout-out to Hyperliquid? I feel like that's kind of the consensus—maybe semi-consensus—but, yeah, I'm bullish on Hyperliquid. I'm also bullish on Solana.

Do you feel more competitive with Solana than Hyperliquid? You kind of grew up at the same time: more competition, more animosity. Whereas Hyperliquid's, you know, 5 years—it almost feels like a different era of crypto.

Definitely a different era of crypto. I would say that's probably true to an extent, but I don't feel really competitive with any chain. I think it's really about enabling people to deploy their chains. The kind of competition that I feel with Hyperliquid is more an admiration, where I'm like, "Ah, yeah, that's a good use case. Let's enable that one," right?

And that goes for really all of the chains. I want to make sure that the delta between having a massive distribution and then launching your own Hyperliquid killer is minimal, right?

I mean, we've gone through so many cycles now of asking that question about 2 of the hot contenders of the month—the up-and-coming ones. Yeah, exactly. So, after you watch this story enough, maybe you want to ask the deeper question of where the equilibrium rises, or whether it will just always be, you know—

We don't ask deep questions on this show. Just superficial this-versus-that, man.

Well, it's like, will there be 1 winner? It's like, well, markets are always changing, bro. Okay, I have to pick 1. Yeah, yeah, but I have to pick Optimism.

There we go.

We do it for the short clip on X.

Yeah. Actually, Optimism competes with Bitcoin. You heard it here first.

He's a store of value.

7. Launching a Token

Oh, yeah. Exactly.

All right, that's my transition into talking about the token, which I'm guessing you guys don't want to talk about. But I have to ask—less from an investor's point of view and more from a founder's point of view.

In 1 case, you have this amazing benefit, which is that you launched a token at this peak, hot time. That's awesome because you get a lot of money, a big treasury, and a big balance sheet, which lets you do things. The downside is that you're essentially a public company. It's tough to raise and tough to capitalize the company now. Maybe I might be wrong about that, and it can be demoralizing when something goes down a lot for internal employees and for you guys.

How do you think about just having a token? Was it a good idea to have a token? Do you wish you didn't have a token? I'd love to hear your honest thoughts here.

Definitely, I think it's been a great thing overall. Capitalization—we're doing great. In terms of morale, watching a price chart and having that in front of you is absolutely brutal, for sure. I know that Jing went years without checking the price, and in general, that's been our sanctuary: let's focus on fundamentals and not watch, candidly, the craziness of the market.

Ultimately, so much of it is just macro. It's just, where's the new hotness? Right now, it's investing in AI. That is totally fine, but it doesn't change what we came out here to do. So, we're moving forward. We have a very clear path ahead in terms of how we actually achieve the goals that we set out 10 years ago to achieve: bring the world on chain, build the world computer, all this stuff.

Ultimately, it's a positive thing overall. It can be a distraction and it can be emotionally charged, but you just have to keep walking through and focusing.

I think if we were—if there was a crunch spot where it felt like there was a big runway problem, that would be a lot more painful. But we're not in that world. It's nice to have the recurring revenue and be figuring out the business model.

Yeah.

I also think I'm very thankful, and my answer might change. I agree with Karl, but my answer might have changed if we had just become laser-focused on "number go up, number go up, number go up." Sticking to the fundamentals—that we're building some decentralized governance and there's some value behind this—is a dramatically different story from the course that you see some other projects take, where they're just like, "Now the token is out, and it is the bottom line. It is everything that we are doing."

Everything is about a discussion of what's going to happen. That's not a sustainable way to build a sustainable business. It's good in the short term; it's bad in the long term.

One more on the token: do you guys think about this at all? I know you're not focused on the token price, but if you were a publicly traded company, it's really important for the CEO and the employees not to focus on the price. But you do have a set of people who own your token and might actually want to buy your token, and traditionally that's called an investor. You have investor-relations teams and quarterly calls. Do you guys do that? Do you think about it? Is it not the right time?

Not the right time. We've got to stay focused. We've got to make the actual business—the underlying promise of all of crypto—real, because we're not there.

8. What Defines Success in Crypto?

Okay, maybe getting out of token land and getting away from that. One more question on the customer base: it seems like the product-market fit for a blockchain is capital markets, right? Even as I ask you, "Who's your customer?" I'm still saying, "Is it a fintech? Is it a brokerage? Is it an exchange?" These are all really capital-markets players.

Do you guys think that blockchains are just good for capital markets and finance? Do you think that at some point we should talk about AI and decentralized AI? There's a whole gaming thing that completely died. Do you think that comes back? I'd love to hear your use case for blockchains and where you see this going.

Okay, I think crypto has a very particular use case that everyone uses it for. Why? It's because it's the only thing on the market that can provide that particular use case: decentralization.

Now, we have paid a price. Whenever you're building software, there's a complexity-and-capability trade-off. The more complexity, the less capability. The less complexity, the more capabilities you can build. I genuinely think that we in crypto have not actually achieved the final form or the realization of the vision, and that's what's missing. We have to reduce that complexity ceiling so that we can increase the capability ceiling.

I genuinely believe that we have a future and a path to no-compromises crypto. That means if I'm looking at deploying my application or putting my IP somewhere, instead of using a centralized Web2 hosting platform, a standard run-of-the-mill database solution, or whatever, I would actually lean toward bringing my new IP on chain, deploying there, and composing with the broader ecosystem.

It might be hard to convince the existing builders, but I do think there is a new class of builders—folks who are leveraging AI—who will gravitate toward that if there are no compromises. There are compromises today. We are not there today.

To ask why you don't just deploy every crypto social-media platform—again, I'm not proposing that we do a crypto social-media platform—but all of them have died. I believe that's because it is ultimately just not a better platform to build your social-media platform on.

I genuinely still believe it is a counter-opinion these days that composable block space will be a better web platform for all sorts of usage, if we can break that complexity and make the capabilities match and exceed Web2.

So, you think decentralized social will exist if you guys are successful? We should have decentralized social. We should have on-chain gaming.

We should have—name all of your read-write-own best-case scenarios. Everything. We should have it.

Will it look like your decentralized social platform from the ground up? Maybe not. Maybe it composes in some fancy way with Twitter, and your AI bot generates a magical front end customized to you. But ultimately, we need to be a back end. Crypto needs to be a back end for the best web platforms.

I think we have the foundation to be the best platform because we have something that no one else can provide, but we don't have the tech.

Soon. Maybe as we wrap up, I'd love to hear your guys' thoughts about AI in general.

You guys are both engineers, I think, and researchers at heart. I would love to hear your guys’ thoughts on AI.

Yeah, you should talk.

I don’t know. If you want practical advice, we’re shooting for it. Beware the slop and beware the hallucination.

Yes. The power of agentic engineering—vibe coding, right—is there, and you can create massive amounts of code. Oftentimes now, the bottleneck is your understanding and confidence in it. I think this is the failure mode that everybody gets into: “Oh my God, the zero-to-one is so dang fast. It’s so damn easy. I’m going to just do that 100 more times, and now I’m at 100,” right?

Ultimately, what you need to do is not that. You need to be very thoughtful. You need to be very methodical. It’s really easy to go from zero to one, but you should probably, like in traditional engineering, go from zero to one many times before you actually consider yourself at one.

This idea that you rewrite—it’s very standard practice in traditional software engineering to end up rewriting code multiple times before you have something that you’re really excited about. And yet we go into the world of AI, where you can produce 1,000x the amount of code. People are like, “Oh my gosh, there’s so much more to go do here.”

Well, okay, you’ve got to move slow to move fast. That is still true of AI. It’s just that moving slow is about your own comprehension and confidence in what is being built in the first place.

Yeah. Ben, Karl.

Thanks, gents.

Thank you.

Cheers. Thanks, gents.