- yesterday
What Will 2025 Unravel for Rollups & Appchain Stacks? 🚀
2025 is set to be a pivotal year for rollups and appchains, with breakthroughs in scalability, interoperability, and user experience. Industry experts weigh in on key trends:
🔥 Hyper-Scalable Rollups – Next-gen ZK and Optimistic rollups push TPS limits while slashing costs.
🔗 Seamless Interop – Cross-rollup and appchain communication becomes frictionless with advanced bridging.
🛠️ Modular Stacks Rise – Customizable appchains gain traction, powered by frameworks like Cosmos SDK, Polygon CDK, and Arbitrum Orbit.
💡 User-Centric Design – Simplified onboarding and account abstraction drive mass adoption.
Will 2025 be the year of the modular, multi-chain future? Tune in as top builders break it down!
🔗 Join the discussion at Cogitus.io
2025 is set to be a pivotal year for rollups and appchains, with breakthroughs in scalability, interoperability, and user experience. Industry experts weigh in on key trends:
🔥 Hyper-Scalable Rollups – Next-gen ZK and Optimistic rollups push TPS limits while slashing costs.
🔗 Seamless Interop – Cross-rollup and appchain communication becomes frictionless with advanced bridging.
🛠️ Modular Stacks Rise – Customizable appchains gain traction, powered by frameworks like Cosmos SDK, Polygon CDK, and Arbitrum Orbit.
💡 User-Centric Design – Simplified onboarding and account abstraction drive mass adoption.
Will 2025 be the year of the modular, multi-chain future? Tune in as top builders break it down!
🔗 Join the discussion at Cogitus.io
Category
🤖
TechTranscript
00:00So with that, let's move on to the next panel which is really around the role of an AppChain
00:15Stacks. We've seen a lot of activity with role of an AppChain Stacks in 2024, lots of innovation
00:27and we believe that in 2025 it's really going to the next level. So I'd like to invite the following,
00:38Cooper Medroni, Senior Product Manager at Off-Chain Labs. Hi Cooper. David Pavek, who is the Director
00:48of Ecosystem Development at Polkadot. Yeah, David, you're here? Oh yeah, great. Nikki Chalabi,
00:57who is Head of DeFi, Tyco. And my dear friend, Matt. Matt, are you there? Oh great. Matt's
01:04made from Avalabs and this is going to be moderated by Dhan Varshish, who is the Co-Founder and CTO
01:11at Zee. Yeah. Over to you folks.
01:32I say this is one of the hardest panel in Denver, regardless of how much snow it was there last
01:37week and how much it was going to be the next. And I'm thankful that, you know, my panelists
01:42are not wearing boxing gloves, but that's a good thing as well. We'll have all the vocal
01:47chords on. We have, you know, a powerhouse for a discussion to talk about what is going
01:53to happen in the Web3 growth. And most of the growth we all know is happening through scaling
01:58solutions like layer 2 chains, layer 3 chains. We roll-ups or different options. We have
02:04Harbitrum, Tyco, Polkadot, Avalanche here. What else do you need? I think I'll start with the
02:11question and I think I'll throw it to, you know, pretty much all of you. And the question is, what
02:17are we looking at in 2025 and beyond? You know, people say, there was a time when they say,
02:22OP roll-ups, that is a thing you do. And they say, ZK is the end game. But then we have
02:28base roll-ups as well. And now we have Avalanche L1s, you know, a different way of
02:32permissioning, permissionless approach towards app chains. So, Cooper, I think we'll start with you
02:38in terms of, you know, what is your take? What is going to happen?
02:43I think right now we're in an era where there's an explosion of stack options. And I think that it's
02:48great that we have all this product diversity because builders should have as many available
02:53options to find the exact right combination of, you know, centralization, security, you know,
02:59block speed, all these parameters that they want to use to tailor the experience of the
03:04applications. So, I would expect that this diversity increases. I would expect that we
03:09see only more stacks enter the game. I think that there are, you know, network effects
03:13presence when things like, you know, shared liquidity or, you know, in Ethereum, we see
03:18a lot of this aligning around interoperability solutions between common stacks. So, I think
03:23there will also be kind of a unification of the experience, which is mainly coming from
03:28interoperability, hiding some of the differences. You know, we hear like chain abstraction as a
03:33term that is thrown around a lot. Basically, the idea that we hide the experience of being
03:36on the chain. You don't need to really remember where your assets were left. You can kind of control
03:41them and have this really unified and clean app experience. So, I would expect both the stacks
03:46to increase underneath, but I'd also expect that users actually get to be hidden from a
03:51lot of that complexity and it really just becomes more options for builders and developers.
03:56Super. David, I think I'll pass on this question next to you because we've heard
04:00Polkot are doing a lot of new things. It is not the old Polkot we heard about, you know,
04:04two years ago. There's Jam which is picking up. So, how do you think Jam is going to fit into this new,
04:09you know, era of actions and scaling? Yeah. Basically, I think there are, what you also
04:16mentioned is there are so many different use cases where you need different solutions and
04:23the great benefit is you can adapt, you can use whatever you need at the moment. So, if you're
04:29just a startup, a Solidity smart contract might be all that you need. If you actually have users
04:35and you need the huge, huge throughput, then a Rollup or Parachain as we call it in our ecosystem might be
04:44might be a better fit. Yeah.
04:46It's funny. I think it was two years ago, or a year and a half ago, I was at DevConnect and I was the only L1 on a panel of all L2s. And I think we've gone kind of full circle where we're seeing the downfall of general purpose L1s and L2s. We have
05:02way too many of those. I think it's an oversaturated market. And we're really seeing the rise of like application specific rollups or L1s or Parachains for instance. And that's actually why I got in the space, not why I got in the space, but before I joined Ava Labs four or five years ago, I fully subscribed to the thesis of application specific blockchains. And so it led me down the path of Cosmos,
05:05Polkadot, or Avalanche. And I think it's, I think it's, I think it's way too many of those. I think it's an oversaturated market. And we're really seeing the rise of like application specific rollups or
05:14L1s or Parachains for instance. And that's actually why I got in the space, not why I got in the space, but before I joined Ava Labs four or five years ago, I fully subscribed to the thesis of application specific blockchains. And so it led me down the path of Cosmos,
05:32and I think it's, I think it's nice to see things like Unichain or Hyperliquid or DYDX kind of get product market fit on the permissionless monolithic environments and then migrate over to their own application specific environment. And so I think we're going to continue to see that in the future. We'll probably only have five or 10 general purpose L1s and L2s in the next two years in my opinion.
06:02But we're going to see an abundance of application specific rollups and later on blockchains.
06:08Well, these are super exciting. So Nikki, I'll come to you now because we talked about, you know, different type of solutions. And you, you're building a based rollup solution. So how do you see, you know, the pigment for projects?
06:19When do you think these projects will choose a base rollup? And what kind of project these can be?
06:24I mean, from my perspective, the future is based, and there's no discussion around that. So, and why I would like actually, I mean, generally speaking, right, it really depends on your use case, as many people mentioned here, like, you can go for a ZK rollup, you can go for an optimistic rollup. If you want to go, you can go for volition.
06:45For volition, it's up to you to decide. And maybe different use cases in the end of the day, like, might require different solutions. So, like, it's not like one size fits all.
06:55But when we're talking about, like, base trollups, what we really need to measure, like, basically, we're inheriting everything pretty much from Ethereum.
07:05So when it comes to settlement, consensus, layers, and data availability layer as well. And what's really important is that, like, the base sequencing inherits the centralization of the layer one.
07:19So that's our sequencer is centralized. And hence, we do see that the narrative is shifting towards base rollups. And essentially, like, the centralized sequencer is becoming a topic.
07:34And I do hope that one day, like, end users will do requests or will do demand, like, censorship resistant L2s and centralized sequencers will become obsolete.
07:48So somehow, you know, hearing this, we are seeing that, you know, there was a discussion long time ago that was about scalability and decentralization.
07:55And, you know, with new L1s and L2s coming around, we are getting to the same discussion again that, you know, how do we achieve more decentralization?
08:02Because, you know, when we start solving the scaling problem, decentralization started getting compromised again.
08:07So what is your, of course, your take would be on, you know, decentralization through decentralized sequencing.
08:14But I'll pass this question on to, you know, David as well and Cooper, you too.
08:18Like, what is your solution to decentralization? Because, of course, in the end, the projects would want decentralization.
08:23They would want trust security, which would not come out of decentralized ordering sequencing or other type of solutions.
08:29So basically, with Polkadot, we learned the hard lessons that the end users don't care so much.
08:37Basically, we really focused on decentralization, security and so on.
08:42And users don't necessarily care about it.
08:45But what we realize now, with also the shift in the regulatory environment, that companies care.
08:54So basically, maybe the end users, the consumers actually, they, for them, it doesn't matter.
09:02But whatever product they are using on-chain, that needs to be secured.
09:07As a company, you don't want to have something that can easily be exploited, that is centralized.
09:13You don't want to depend, in the end, on one single entity that runs something and so on.
09:19So, yeah, that's my take.
09:22Yeah, Cooper?
09:23Yeah, I agree completely.
09:24I think, you know, from seeing just how people use the product and how people, you know, interact with, you know, having their own roll-up stack.
09:31Definitely, the most popular configuration is people building an L3 with, you know, our DA layer.
09:38They're basically trading off for lowest possible cost, fastest possible speed, best user experience.
09:44And I think they're trying to win by capturing users and, you know, basically building successful app ecosystems.
09:50And so that really sometimes has not a lot to do with decentralization.
09:54But the groups that do care are, you know, institutions.
09:57I could definitely see something like base sequencing mattering a lot to a traditional finance institution that's sometimes even too scared to move their funds from L1 to L2.
10:06And I also think that, you know, for us, we recently launched Bold, which is our permissionless fraud proof protocol.
10:12So it's kind of a step change from an institution being able to, you know, monitor the chain and, you know, run a validator knowing if something is wrong to now actually being able to submit the fraud proof themselves, which gives them just that much more security.
10:25So obviously validation is just kind of one axis of decentralization.
10:29Sequencing is another.
10:31But definitely we think that the users that we are currently have are pretty much choosing to just optimize the user experience and minimize cost.
10:39So I guess optimized user experience is the priority, is the important factor.
10:44And then comes the decentralization, maybe as an obvious evolution of the solutions themselves.
10:50Now, Matt, what is your take on, because Avalanche has been, I would say, it's permissionless from day one.
10:57L1s are permissionless.
10:58So what can you add?
11:01So although I would love there, I would love for them to be permissionless since day one, we actually just pushed the POS upgrade probably six months ago.
11:11Before that, when we first launched Avalanche, Sumnets, which are now rebranded Avalanche L1s.
11:17They were all POS, meaning like it wasn't permissionless in terms of validation.
11:22And if you acquire enough of the tokens to validate, you had to still be permissioned or whitelisted in terms of validation.
11:27It's funny.
11:28I have a question actually for you over at Arvitron.
11:30I think you guys were the first L2 to have permissionless fraud proofs, if I'm not mistaken.
11:39Is that correct?
11:40Technically, we have complete permissionless fraud proofs first.
11:46There was a push from OP where there were permissionless fraud proofs, but there was kind of a bottleneck at a multi-stick at one point.
11:53So it depends on your definition.
11:55And, you know, I think that we'll leave it at that.
11:58Gotcha.
11:59Do you think outside of like arguments, say on crypto Twitter, like, it's kind of funny.
12:06I feel like users just didn't care.
12:08Like, we're all at the point where we're all arguing amongst ourselves, when in reality the people that we're trying to appeal to, to your point, just don't care about decentralization.
12:16Like, decentralization only matters when it actually needs to be decentralized, whether it be for regulatory reasons or, I don't know, if there's a war in the world, you need to decentralize consensus so a network doesn't go down, something like that.
12:30It's a little bit interesting because I don't know if you guys saw Base actually announced the app chains.
12:35I think it was like two days ago or yesterday even.
12:37And they're using AWS for data availability.
12:41And so it's a little bit sad that a centralized company running at L2 is now running on a centralized data availability layer.
12:51And so I don't think users actually care.
12:54And so it'll be interesting seeing like when decentralization actually matters again.
12:59I know, I mean in the case of like North Korea for instance, the Bybit hack like decentralized bridges that are probably being approached by three letter U.S. agencies.
13:13Like, hey, can you censor this transaction?
13:15And most of the time they're not decentralized enough where they could actually censor the transaction because they have, they're running like most of the nodes for consensus.
13:24They're fully decentralized interoperability layer or something like that.
13:27And so even in that case, like they're still censoring transactions.
13:31I mean, they should be like North Korea shouldn't be allowed to just like fully wash or clean those funds.
13:37But I don't know. It's been interesting because I think the ethos obviously that we're all here for in the first place is decentralization.
13:45But we've gone so much in the opposite direction of decentralization in the past two years.
13:51Yeah, I would say that my main slant is towards, you know, satisfying what these customers and users want.
13:59And then we have to build technology that doesn't compromise on these values that they have, you know, cost, user experience.
14:05And that's going to take time. And it's also going to take a certain amount of scaling of different layers of the ecosystem to get there.
14:11But you can always, you know, hopefully make more compelling products later to shift them onto decentralization, which is a better model than not getting the users or apps at all.
14:20So I think that I totally agree with your take. I think we just need to honestly lean in to the users, you know, what the users are telling us while also, you know, remaining somewhat principled.
14:32You know, there's always an opportunity to get them onto that new platform later because they just want to build apps.
14:38You know, builders want to have their apps succeed. They want to have, you know, millions of users.
14:42And sometimes to do that, they need to use these systems that, you know, look like shortcuts in the short term.
14:48But I would also say like, you know, we're talking about decentralization from a technical technological perspective, but decentralization is not only like technological matter.
14:57It's also a cultural shift, and I think we haven't still achieved that.
15:01So it's also like the same, you can relate to the privacy, right? Like we do believe that people care about privacy, but in reality, they don't.
15:10Like everyone accepts T's and C's, right? So doesn't read the messages. They're like, okay, accept, accept next.
15:17And I think this is the message is very clear. Build the apps, go out there, worry about the decentralization later.
15:25And maybe let's talk about decentralization two years later on this panel.
15:28But maybe quickly to add, I think actually enterprises do care, at least in my conversations, because they look at regulatory compliance,
15:39then it's definitely necessary. You cannot simply tell them, hey, we are running a database here.
15:45And that's how we are allowed to share the data with everyone and things like this.
15:50And also like, you don't want to depend on a single entity to shut down your business and things like this. So, yeah.
15:59I think that's all right as well. Institutions are definitely exploring more reliable solutions.
16:03And one of the aspects of reliability comes out of decentralization.
16:07But we'll know when we'll know. Now I'll come to the burning question.
16:11None of us, you know, I haven't used that keyword yet.
16:14And I've heard a lot out there that they say ZK is the end game. ZK, you know, that keyword.
16:20So I would like to know your take on it, you know, maybe one by one, maybe you can start this time.
16:26What is this thing about ZK? Is it the end game?
16:30Yeah, I mean, regarding ZK technology, like ZK circuits more specifically, I've always believed like they're kind of the gold standard in terms of privacy.
16:40But in terms of actually helping scale blockchains, you know, roll ups, they've always been a little bit too costly in terms of compute.
16:49So I don't actually think it's the best case if you're using a roll up to scale blockchains.
16:55I think optimistic roll ups, to be honest, have kind of won out longer term.
17:00At least long term, if you say over a year, year and a half, which is like five years in our industry.
17:06But yeah, I think ZK is still the gold standard in terms of privacy, especially if you're trying to prove certain private information that might be stored off.
17:18Chain and bringing on to blockchains and only allow like one one way informational flow just in terms of like checks, meaning like in the US, you have to be 21 years old to drink.
17:28I can just upload my identification and be a ZK circuit.
17:33It just says yes or no. And you know nothing else about my identity in terms of like being over 21 years old or something like that.
17:40Yeah, I fully agree. So with the next version of Polkadot Jam and so on, we specifically for the scaling part haven't integrated zero knowledge because we think it's too costly and so on.
17:54And you can create as long as you can avoid it and integrate sharding solutions and so on that are ultimately as secure as a ZK based solution.
18:05And you save all the computation costs. That makes a lot of sense.
18:09But obviously there are a ton of use cases for ZK technology.
18:13Yeah. Yeah, Nikki. Yeah, sure. I mean, from my standpoint, it's like this. I mean, I guess you guys heard about Tycho Strico, which is the multi-prower system and essentially like we're working with the likes of SP1, SGX and risk zero.
18:32And we do believe that in the end of the day, ZK is the end game to a certain extent. But of course, it also depends again on the use case, right?
18:43I will not say like, it doesn't mean that everyone should go for it. Certain use cases might require different solutions. You know, there are optimistic relapse, there are ZK relapse, there are volition, as I mentioned before, there are volition.
18:55So it's up to your use case to decide which one fits you the best.
19:00So a true believer. But Cooper, I think your take would be the most interesting one.
19:05Yes, I felt that tension as we went down the line. I think, I'm going to say I think ZK is the end game.
19:13I think that, you know, it's true that the optimistic model of doing things is always going to be cheaper because you're not computing everything.
19:21You only compute in the worst case scenario when you need to submit a proof.
19:25And that's just a certain security profile that some people will accept. And it makes sense in certain applications to have that low cost.
19:33I also see the value in everything being proven. But the issue with that is that proving times are quite long.
19:40And even when they get down, it's unlikely that you're going to be having proving times as fast as we have blocked speeds today.
19:46So there's always going to be some degree of trust that you have until that proof gets delivered and posted.
19:51But I do think that that's where, you know, other solutions come in to, you know, make a balance where we hopefully can bring all the finality times down on a lot of chains using ZK.
20:01Like it would be great if every rollout could have, you know, sub hour finality because the effect of a long challenge period is that funds basically have to get carried as a carrying cost for any liquidity provider, a solver, a bridge.
20:15So when that goes down, everything trickles up and becomes a lot cheaper.
20:18And then on top of that, there's kind of the built-in, you know, interrupt benefits of being able to kind of use other proofs, you know, to kind of work with data or assets across different chains.
20:29But it will always be, I think, a kind of heterogeneous solution underneath where people are making different trade-offs just based on their security model and preferences.
20:38Is Arbitrum doing any research in the ZK space right now?
20:42Yeah, we have a really, really great research team.
20:45You know, we were the team that actually invented optimistic rollups.
20:49And so we're very on top of ZK solutions.
20:52I think it already just comes down to when do the customers ask for it?
20:55And, you know, we have, you know, quite a deep venture of ZK researchers on the team.
21:00That's great to hear.
21:01But is Avalanche going to be doing ZK ever, Matt?
21:08So I don't know, like, what's public, but we actually acquired a team that does ZK for privacy.
21:13I think it was like six months ago.
21:15And they're building basically a privacy-based payment solution, but then they can reuse that tech stack in our Avalanche L1 architecture for enterprises or institutional use cases.
21:26And so we are experimenting quite a bit in terms of ZK for privacy, but not actually for scaling.
21:32I don't worry about the clock.
21:34You know, I know this discussion can go on and on and on for hours.
21:37And in fact, it might continue after this panel outside.
21:40But one last question to all of you.
21:42You know, people who are building projects, they're building DeFi, they're building gaming, they're building, you know, Tradeify, they're building PayFi.
21:50Some of the enterprises are coming in, like, you know, some telecom company, energy gas companies and all.
21:54So what is this thing you want to tell them that you would be doing in 2025?
21:59And that makes sense to them to join Web3.
22:02I don't care which stack, any stack, but we want them on Web3.
22:05So what is your message to them so that, you know, they start seriously thinking about that Web3 is a playground.
22:11They can actually come on and, you know, build.
22:16Yeah, I can kick off.
22:17I think that, you know, my message would be that the people that are playing with this technology, they're not doing it because they, you know, want to take part in a speculative ecosystem.
22:28They're doing it because this technology actually reduces costs and creates more efficiencies for them internally.
22:33And I think that that is kind of the validation we've been waiting for as an industry to know that this technology will be adopted.
22:40And it's really just a matter of time, penetration, you know, the business conversations to proceed.
22:46But there's really real value here.
22:48And, you know, I think that if you're on the sidelines, it might be time to take a look.
22:52Yeah, just to add, I think something very important is also that this technology will really start to shine and outgrow once we have these network effects.
23:03It's like if different corporations are able to work together on chain that are currently don't trust each other necessary.
23:11And then basically you are either a part of this and you grow with the network or you aren't.
23:18So yeah, I mean, it takes actually time for the innovation to get adopted.
23:23Right. So like it doesn't happen in one day.
23:26So essentially it's important for people to come and explore it.
23:30Generally speaking, the technology to try it out for builders to build.
23:33And for us actually as an industry to educate people around us like and tell them about the benefits of it.
23:40Right. So that's how important, but also like as mentioned was had a bit of a discussion around decentralization.
23:46So also it's our job to teach them why decentralization is important.
23:51Right. So and for and for them to understand that they actually should go for censorship resistant solutions.
23:59Yeah, it's funny. I think I've come like full circle within this space where I think the way forward is like payments, remittances, efficiencies,
24:09for settlement of like RWAs, tokenized assets, et cetera, which like everyone has talked about since 2017.
24:16But in the 2021 cycle, we kind of made our way elsewhere in terms of like Ponzi clicker games or PFP NFTs or kind of other use cases,
24:27which might be fun in the short term, but not have any sort of longevity.
24:31And so I think we're going to basically come full circle here in the next couple of years and see blockchains just being used for the use of these.
24:38cases that they're designed for. And I mean, I think stable points are still the number one source of adoption within our industry.
24:45Yeah. So I think we've got several many insights in several directions, but all pointing towards a better future in Web3.
24:52That's what I believe. And I think that's what we all believe. So thank you for coming here, joining us and sharing these valuable insights.
24:59We can continue outside in case you know you have. So yeah, thank you very much.
25:04And we'll host you again probably and maybe hear about what is happening more and more next year.
25:09Next year.
25:10Yeah. Thank you so much folks.
25:12Thank you so much folks.
Recommended
14:50
|
Up next