The Rise of Modularity: Scaling Beyond Monolithic Constraints | Nick White, Celestia

Transcript

monolithic chains have this negative Network effect where the more users there are the more applications there are the the slower everything is or basically the more expensive everything is it becomes like crowded and congested and in a modular setup we actually have these positive Network effects where the more users there are more applications the more people running nodes the more block size we can have all right everyone welcome back to another episode of zero X research today is April 10th and we have an awesome conversation lined up with Nick white the the former founder of Harmony and the current co-founder of Celestia where he spends his time building out a network of modular chains with Celestia focusing on data availability uh but

before we get into that as always we're joined by two members of the blockworks research team Xerox pibbles and effort Capital to join us for a little segment of Hot Seat cool Throne uh effort I'll pass it over to you to start us out well I have my hot seat this week is flashbots and the Emmy Mev boost relay uh so about a week ago today uh from today four to five Searchers were actually front run by a block proposer for approximately 20 to 25 million um so just a brief background on on I mean V boost and PBS proposer Builder separation so one of the core ideas of PBS is that the proposer the boss proposer is not allowed to see the contents of a block they're signing until based on the Block reason being is

if they see the content of a block then they have the opportunity to actually front run whoever put the transactions in that block and ultimately can profit from the opportunity um theoretically the way PBS is set up today it's extremely hard for a malicious block proposer to deconstruct a bundle see what's inside the block content and um and actually front run uh that block Builder and Searcher um for them to do this they have to either double sign which is a slashable event or they have to submit a block before the Mev boost relay I was able to submit it to the network which is extremely unlikely um what ended up happening was this block proposer was able to reorder the transactions in a block and included new

transactions that was not inside the block uh when the Builder gave it to them as part of the Mev value chain in ethereum today um pretty much it was a bug in the system for me Mev boost relay um there's a ongoing patch that the community has been working on I'm not sure if it's live today like I said this happened about a week ago I'm pretty sure it's live today but as of today we found out that tether actually created a blacklist and has blacklisted the address that performed this exploit on me boost and that address had approximately three million dollars of tether on it and it's pretty interesting for tether to actually create this Blacklist because they're very careful with doing so um I think one of their arguments against usdc was that they were very

much more careful with blacklisting than usdc historically so for them to do that I think it's like a a pretty big concern and on top of that flash Bots if you actually go on their GitHub they just created their own Blacklist uh for anyone to use flashbots relay moving forward if there's any Bad actors in the space so I think as a whole uh PBS and Mev boost is like used by over 85 of the validators in the network today everyone thought it was relatively battle tested at this point but we're already starting to see some uh some issues in the system and obviously down later in in the roadmap for ethereum we want to see PBS enshrined in the protocol uh following the footsteps for all the work that flashbots has done but

maybe it's time for the community to realize that maybe it's just not yet ready to and Shrine it into the protocol um because there's a lot of actors in the space that are all trying to figure out ways to exploit the system so there's a lot of profit to be had yeah this is a pretty crazy instance of the uh the attacker being the attack e or a little bit uh what's the saying a taste of your own medicine right so methbots notoriously tear apart users and extract value from their transactions um and put in most cases they put up a ton of capital in doing so right you can put up a million dollars just to Come Away with like a five to ten dollar profit and that's kind of what happened here this Mev bot put up a significant sum of value to extract just couple

dollars of value and the validator reordered the transactions in such a way that they could basically drain all of the funds from the Mev bot uh so I don't know it's just like this little constant Battle of how like the just a value chain of ethereum and the players that are always looking to extract it one thing I want to just make sure we point out is The flashbots Blacklist but specifically for the Relay so it wasn't like excluding uh that validator from the Mev boost protocol as a whole which is important distinction right because the relay is just one piece that any like anybody can pick any relay of their choosing uh but I just wanted to point that out that there is a little bit of a distinguish to make here right and and I did see some of that going around

Twitter that you know like oh they blacklisted ethereum's not decentralized and like that's not really the takeaway to have here but I agree with you effort as well as you you mentioned that the The usdt Blacklist so they basically froze the funds of this this um validator that had taken These funds and that's interesting thing right and that almost I don't want to say this like for sure implies that there was a court order or some legal ruling that forced them to do this but that's exactly where my head goes is tether probably had a reason that they had to do this I don't think they're trying to make a statement with this um but I don't know so that's kind of my takeaway is like we I expect more news to come around the legal side of this which sends a really interesting precedent right if validators kind of

you know misbehaving within a design system is considered an illegal offense then I don't know if the takeaway of that is a good thing or a bad thing uh but it's it's certainly something to chew on it's interesting because ethereum is trying to create credibly neutral Mev Solutions and you could argue that they didn't do to your point Dad they didn't do anything wrong they did exactly code his law right they did exactly what the code allowed them to do and now they're getting penalized for it which I would argue is the exact opposite of credibly neutral it's an opinionated approach to Mev saying you're only allowed to do what we think social consensus allows you to do not necessarily what the code says you can do so I think this is a really interesting uh you know event that that happened and again I think it's good

though it's good that you know block creation is a purely purely adversarial environment people are going to try to poke holes in the system to find ways of exploiting it um it just shows that PBS and I think the entire me value chain for ethereum just needs a lot more work until it's ready to be enshrined in the protocol yeah also positive too that it wasn't in China the protocol in the current state and there's kind of a time period where it can be battle tested until it actually is so I guess we can count our stripes and and feel lucky about that one and I also agree with Dan in respects too I'd be really hard for us to believe this didn't come from an order like I don't think tether is in the business of you know monitoring ethereum and and making sure everything

is working exactly as intended so definitely agree with you there Dan yeah and one more interesting takeaway uh that came out of this was flashbots actually introduced just a little bit of latency into the process of mbb Mev boost um it's communicating with the relays to prevent this from happening again it was kind of like a quick stop Gap uh while we think of like a more long-term solution and that actually just so barely increased the block time of ethereum from about 12 seconds to 12.2 which like if you actually look at a chart of blocks graded per day it is like a noticeable downtick in the amount of blocks produced per day so just as something to flag if you see a crazy chart floating around with it like you know all of a sudden there's less blocks being produced on ethereum it was like

oh that's a red flag but there it was a uh it was intentionally introduced because of this incident yeah that was a good find there Dan I guess I can take it over for my hot seat I've got Jared gray and and Sushi swap in the hot seat this is pretty easy one for for anyone who hopped online Saturday night saw that there was an exploit in a recently deployed routing contract um not not exactly everyone's favorite experience if you'd been using Sushi swap over the last you know three or four days since it was deployed but yeah everyone basically needed to jump on and revoke their approvals to avoid you know being uh falling victim to the exploit and I think it was in total about 1800 eth that was taken away from from Sushi swap as a result of this this also comes

shortly after Jared gray the the head chef at Sushi swap got served as subpoena so things aren't really looking great for sushi swap I know in our newsletter today Westy highlighted everything Sushi's been through over the last year and it is an exhaustive list it's pretty crazy um but a good fun a good portion of the funds were recovered and they're working with the Lido Dow to actually recover some more of the funds because some of them were sent to the lighter Dow as Mev payment so it'll be interesting to see if they do return some of those funds to Sushi kind of sets an interesting precedent I don't really know where I fall on you know how I think this should play out it seems like a a bad precedence in some ways but in other ways you want to make the sushi users

hold it's like an OG D5 protocol um so I guess I'm curious do you guys think they should do it and then also just on the topic of sushi if you had to take a guess what would you say in terms of price do you think it's higher or lower in a year's time you just got to throw the whole Sushi protocol away I'm sorry I just uh start over um I think uh light Island should not return uh those proceeds you know that's just that's business kiddo yeah I actually agree with you there people's like it I don't know it's like this responsibility taking back the airdropped responsibility meme that's kind of what just happened to this Lotto validator right they're just doing their

job validating blocks hit a hit the one the Mev Lottery and now they have this problem on their hands like what do we do with these funds like they were exported funds um and now they're mine but I did nothing wrong and I kind of agree with you it's business as usual this is part of the process like would it be the right thing to do to give it back yeah there might be some like moral or ethical argument here but I don't think they should have any legal responsibility it's kind of the way I view it what about you effort yeah I mean I second Pebbles uh notion about throwing out sushi personal opinion I think it's going to zero nobody uses it it doesn't have a it's not like winning product Market fit like in any category unfortunately it's an OG D5 protocol I would not mean any like

personal disrespect it just I don't know anyone that's using Fuji it's called what it is uh when it comes back to Lido paying back Sushi holders I think again it's it's nice if they do that um but I'm really curious like what happens today like the real world if I go if I steal money and let's say I go buy a house um is there some type of callback mechanism like in the real world for like a legal entity like an old house buyer that sold their house uh did they have to give back like the funds that were used in that transaction payment I'm not sure but again it comes back to the rules of like code is law and or is it social consensus if ethereum try to be incredibly neutral uh platform then it's just business get up

just like little said like you can't say it any better that's literally what it is and I don't think Lido should have to pay back we're bringing the FDIC to the blockchain yeah I strong agree with everything everyone said I don't think lighter should have to return those funds it's just kind of the way the road um and also just uh for the listeners who aren't familiar Sushi's on a ton of different chains so like that was just a mess for anyone who'd use she swap for the last four days like it's on like 10 to 12 different chains so everyone was kind of worried if they had used Sushi swap over the last few days but uh pibbles I think I can kick it over to you now for your hot Cedar cool Throne yeah hopping off of ethereum we have um xnft backpack on Solana is set to launch

in like the next week or two they got a big uh Saga launch for their their crypto phone um this is a pretty big step for Solana it's it's kind of the Do or Die moment I think like they they put all their eggs in this basket uh let's see what happens uh I think xnft backpack is super neat there's gonna be like the first nft mint on the wallet uh Happening by the end of this month and it's really neat because you can just throw different apps in there like they have like stuff like a Rory which is like a Solano play during game and even like dexes they have all the dabs living natively in the wallet and then that wallet can also live on the Saga phone so it's a totally new

like mobile experience and we do everything on our phones nowadays so really excited to see how everything pans out but this can also turn to a hot seat real fast if everything is just a catastrophe so she's super cool The Saga phone specifically is something I'm super excited about like I don't it's certainly not something that I purchased solely for the reason of I'm I'm of the belief like I even took this with like Tesla cars right like that when you're making this like crazy revolutionary product the first gen iteration is probably not something you want to own but is like very cool and Tesla proved to be that like the cars fell apart they're kind of shitty but like they worked and they operated and like that was the major invention and then you

fast forward to today and like the car is you know in a much better place and that's kind of how I view this phone is like yeah it's gonna be a phone and it's gonna work and it's gonna have really cool features related to crypto but like it's still going to be clunky ux and like not that smooth so that's kind of like where my head's at and thinking about how it's gonna work but in the if the off chance that's not the case and it just like a fully functioning awesome to use phone uh that has these crypto native features I mean that will very quickly get lead me to purchasing one I'm pretty excited about it yeah stronger I'm excited as well like for smart contract platforms like ethereum like I look at them as really focusing on trying to lower the validator

requirements to actually participate in verifying the network but Solana for me is the one chain that's actually trying to build products built towards consumers that people actually want to use and like if we be completely real with ourselves we have pretty much zero real world application for crypto except stable coins I think is like maybe the one you could argue that has some really good product Market fit outside of you know Bitcoin I think there's something there too um but yeah I like the approach Salon is taking and I'm excited to see how the soccer phone ships and I also want one of those mad lab nfts I think the art is actually pretty sweet it gives me some uh peaky Blinder Vibes for anyone who watched that show on Netflix yeah I can uh I can jump things over to to my cool throne here and I'm I'm also going out

on a bit of a limb here and doing a preemptive cool throne and I'm gonna put anybody who held through the chappella upgrade uh in the in the cool Throne because in my personal of you it's not it's going to be just another nothing Burger event you know there's all these major hyped up things uh that we'd love to do that in cryptos put all this attention around the next major thing and then it happens and just like this release like not a huge bearish event not a huge bullish event it's just like operations as normal and that is exactly what I have to expect to happen when the chappella upgrade goes live and so right now just to give some stats to give some context there's about 18.1 million eth staked which equates to about 34 billion dollars and 1.1 or 2 billion dollars of

additional eth rewards so the actual staking rewards that validators have accrued since being live on the beacon chain so the in my view that with rewards are very likely to be withdrawn in like a high capacity right you know if you staked let's say six months ago seven months ago now you're like actually trying to realize some of this value um and so like the the rewards yeah two billion of cell pressure okay that's like how many minutes of proof of work consensus uh security like okay not a huge deal to see two billion dollars of cell pressure in my view um and I really just don't I'm not buying Into The Narrative of like everybody's super eager to unstake and sell that just that like the logically that just doesn't add up and I'll give

you a bit to the color as to why right if this upgrade was happening exactly 12 months ago and let's go eight months ago and we're two months after the terror collapse I'm terrified because the sentiment in that time was like everything's going to zero it's the end of the world like buckle up and that was pretty much the case right then we fat like three months after that FDX blew up things just kept getting worse on a consistent basis but if you made it through all of that you're at 12 like we're not what like 12 months 13 months into the bear market now like the light can be seen the light at the end of the tunnel can be seen by everybody and it doesn't I don't I just really struggle to make the logical Gap that people are

going to withdraw their entire stake position in mass and proceed to just sell it um and Mark Zeller uh who works for Ave Chan over in the Ave protocol had a really good point on this and you can actually see uh through some of the youth clients uh the software you're running to operate a node you can actually see if people have already opted into getting into the line to withdraw so you can't withdraw yet until the upgrade goes line live but you can actually get in line um and by getting in line you do sacrifice your future rewards and so the upgrade is slated to go live Wednesday uh April 12th at about 5 or so a.m Eastern Time and this episode is going to go live somewhere around seven hours after that

so we'll see how horribly this uh this cool Throne panned out but so far there's been a very very small like an immaterial fraction of validators have already opted into that line and at this point all you're doing is um basically electing to not receive one day and 12 hours worth of rewards and if you wanted to be first in line to dump your whole position and sell it that is very much a I need to be first game and nobody's taking up that opportunity so personal view it's going to be a nothing burger that doesn't mean that's gonna be some massive massively bullish event unless you're under the idea that this massive bearish event is already priced in and then just doesn't happen but in my view we're just gonna keep trading sideways and that's what's going to

happen is again just a big nothing Burger recommend no uh no preemptive cool Thrones like these for the sake of our industry and certainly not Financial advice either yeah we're actually gonna just brick the entire ethereum network that's what effort said earlier and um this is going to be the biggest Black Swan in the history of ethereum let's hope not yeah like effort said for the sake of our for the sake of our industry though Dan like overwhelming majority of Institutions that are not necessarily crypto native but all in Bitcoin and eth they're not staking today they're not taking the

risk they don't understand the technical like uh they don't understand like the technical hurdle I guess related to chappella they don't trust the ethereum community they don't trust crypto natives they're just like I'm gonna wait and see if it goes well I'm gonna stake if it doesn't I'm just obviously going to probably sell my eth but assuming everything goes well I I have to imagine steak to eat like goes up to over 20 maybe not by year end I mean as of right now it's on Pace to do that uh even before shapella but heading into Chappelle you're still seeing more eth gang State uh it's not slowing down so that leads me to believe that it's more of like a confidence event where people are going to realize that everything's good let's take our eth becomes even

more money it becomes more of a yield bearing asset in the ecosystem um I I see it as a bullish event for for eth and crypto as a whole um and obviously today we'll see how things pan out there on Thursday but as of right now the market was moving up higher on the news like leading into chappella uh whereas typically you would think that it would kind of be like a risk-on event where the mark would kind of sell off heading into it so I think it's just an overall interesting uh Market climber right now yeah I think like the main withdrawals is honestly going to come from people who locked in their Eve like a year ago like into like really shitty Structured Products like so for example I've got 150 of each staked on coinbase that's just been stuck there and

actually now that I think about it I haven't tried to actually convert that to CBE but I think generally speaking mainly people are going to be selling that that entity is going to have to withdraw based on you know customer deposits and then that customer is ultimately going to go to a decentralized derivative because it's a lot more useful and can be used in different applications um I guess that is another interesting thing to think about though um with CBE that I just thought of is like if there is no way to convert that staked eth into CBE if coinbase doesn't automatically convert that to CBE for you and then people do sell to get into decentralized Alternatives then maybe there'll be a nice little discount on CBE like I

expect the the real plays to actually come from like little outlier situations like that more so than I do like if there's actually anything here with with the unlocks actually opening up so agreed Dan nothing Burger yeah and over this weekend I actually saw cbeath was still trading at about like a one one and a half percent discount which is pretty interesting because if you look at Lido steak teeth that's at a I think the peg the last checked was 0.9966 and fraxi for example 9991 I just wrote something about this earlier today which is the only reason I just know those numbers off the top of my head but then you see a whole one and a half percent discount on on CBE so maybe there is something there I I never dug in it was over the weekend so just didn't really uh open that chapter but

that is an interesting thing to think about and I want to say but not 100 sure that there was like a terms of service agreement that you had to like opt in if you had that OG version of coinbase stake teeth to kind of move that over into CB liquid state D so that could be something to at play as well but I think that's a great place to to wrap the intro guys this was a fun one and uh super excited for the listeners to head over to this interview with Nick white it was a super exciting conversation and you can really hear how motivated he is to just like talk about Celestia so super exciting and without further Ado Nick white welcome everyone well we are joined by Nick white the co-founder of Celestia Nick thank you so much for joining us today we're super excited to

dive into the Weeds on everything Celestia is doing but we'd love a quick little intro from you uh kind of your background and what led you to founding Celestia oh thanks for having me guys I was listening to some of the previous episodes and uh overall like the block Works podcast has really taken off as one of my top listen to podcasts recently so you guys are doing a good job um I'd love to hear it yeah so I'm Nick I'm the CEO of Celestia Labs um I got into crypto back in 2018 and actually originally in in college I studied electrical engineering specifically like neural networks but I left AI because crypto just kind of grabbed me as this thing that was actually going to make the world a better place and that it's going to

empower people to sort of be more self-sovereign and I I've co-founded this project called Harmony in 2018 my whole Vision was basically this technology is so powerful but it's not scalable essentially in the way that it's currently built it's not going to be able to have the impact that I think it could have because the reality is that not everyone will be able to have access to it and so that's kind of been my mission since since those days and um originally with Harmony we were trying to solve that using sharding but I ended up realizing that sharding was not the right approach when I read this white paper called lazy Ledger back in 2020 which you know gave this whole new vision for how to build blockchains in a

modular way that in my mind actually solved all the core problems of blockchain infrastructure and could be the foundation on which we could build truly scalable you know decentralized networks and so that's kind of been my my thing since then and now fast forward to 2023 we're really close to launching mainnet so uh stoked to be here so can you give us a background on the main problems that Celestia is actually trying to solve and I'm also curious like how far away are we from some of this stuff coming to fruition like what do you actually see the typical architecture of blockchains looking like and let's call it one to two years so to understand what problems Celestia solves we have to first back up and talk about what problem do blockchains solve right so blockchains are like a verifiable

computer meaning that uh anyone can look inside of what's happening on a blockchain and verify that the rules are being followed and this is unique it's something that you can't do in web 2 in web 2 like you know for example Facebook runs a Facebook application you don't actually know what they're doing you're just kind of a consumer and they can have they can make rules unilaterally and change things on you and in web 3 the thing that's different is that users actually kind of Define the rules and enforce them and so why is that challenging is that if I as a user want to verify everything that's happening I end up having to run more and more like have a bigger and bigger computer the more things I need

to verify so in a standard monolithic blockchain as the amount of applications increases and the amount of transactions increase I have to verify every one of every single one of those transactions every single one of those applications and that doesn't scale so the the problem that Celestia solves is that it allows you to actually have a ton of applications and have a bunch of transactions and users are still able to verify them so you you preserve the core value proposition of what a blockchain is while still being able to increase the capacity of the network so I can talk a little bit about how that works if if that's interesting yeah please do yeah so uh the way that it works in a

monolithic blockchain is all the transactions right get packaged into a block and then attached with that block is a commitment to the new state of the chain so new let's say like it's a it's Bitcoin it's what kind of balances does everyone have and obviously if there's billions of people sending billions of transactions the block size gets bigger so there's more data in that block and there's more transactions to verify one by one and so now if I'm an end user I need to verify all of those things I need to download the whole block and then I need to re-execute every transaction and then compare it to the the state that the the block commits to right and in a modular blockchain you're able to break that problem into two

separate things so the first thing is that rather than re-executing and verifying every transaction individually you can actually use something called a roll-up which means that one computer one node can do all the verification for you and just show you a proof that the work that they did is correct so you can tell for yourself without actually having to re-execute every transaction that the state is correct so it's like okay I know that all these are all the right balances I don't have to actually do the work myself but the thing is that in order for that whole construction to be secure you still have to know that the block data was there was available so like if in the previous monolithic world you would have downloaded all

those transactions now instead of downloading them you just have to make sure that they're there that they were published on the internet and So that's its own separate problem that Celestia solved which is that the Celestia block data is made available in a way that nodes can check that very very uh in a very very light way so they can download less than one percent of the block and verify with 99.99 security that it is all available okay great that's super helpful context so if we look at the current state of Roll-Ups on ethereum today though um would you say that they're kind of missing the forest for the trees by just trying to build out more low-cost uh evm block space and that data availability

will become a bigger problem and that's the problem you're essentially trying to solve for is that like a fair take um I think yeah at the end of the day the ultimate bottleneck will always be data availability like just with a raw block space that's like the raw material on which every decentralized application will have to consume um but I think uh in terms of like the difference in in the vision between sort of ethereum and and I guess Celestia is that um in Celestia we were able to start from the beginning with this Insight in mind which is that actually blockchain should be modular this is a way that we should design them and so we were able to basically throw everything out about how blockchains are typically built

you know and and start from scratch with a purely maximally modular design essentially so Celestia does no execution on the base layer it's literally well the original paper was called lazy Ledger because it's a lazy blockchain it almost looks like it's pointless because it doesn't do anything except for making data available data available and so um it's it's we're kind of with this we have this privileged position of being the first modular blockchain Network where you know we were able to start from first principles and and design from the ground up so I think that's kind of the main difference hey everyone big announcement from the blockworks podcast Network we're launching a new show called Lightspeed and hiring Two

Hosts to come build it with us Lightspeed is a show for Builders focused on the use cases that will onboard the next generation of crypto users by taking learnings and inspiration from the garage days of Silicon Valley we really want to capture the perspective from Builders because that's what the ethos of crypto is content experimentation and Relentless Innovation to build products that users can't resist if hosting a show like this sounds exciting to you then head over to the careers page on blockworks.com which we have Linked In the show notes you can also reach out to mirror Sam on Twitter to talk more about the opportunity but overall we're stoked about Lightspeed so if you think you'd be a great host please please do not hesitate to reach out and then in terms of cross-chain composability when you think about chains that utilize Celestia for for

data availability what are the main bottlenecks there is it required for example that any chain that wants to be composable with another chain that's using Celestia also has to be using Celestia for da as well that's a good question um there's a great blog post that our co-founder Mustafa al-basam wrote called clusters and in that he paints this sort of like mental model of how interoperability looks right and there's an impossibility result that is that two separate chains that don't share a common data availability layer can only interoperate with trust assumptions so they basically have to like either assume the other

chain is honest or you can use some kind of committee in the middle but there's no way to actually have like trust minimized uh cross chain interoperability and um one of the beautiful things about having a data availability like Celestia is that two Roll-Ups if they share uh Celestial or any common da layer they can interoperate in a trust memorized way so what we call that the mental model that Mustafa uses is is clusters so you have like these clusters of chains that share like this kind of I don't know border of of security and then whenever you want to bridge outside of that cluster you have to end up making new trust assumptions and this is really important for it's not that

important at today's scale of blockchains let's say and you only have maybe like 10 that are out 10 like legitimately used secure chains but in the future in the long term future that we believe in there's going to be millions of blockchains and when you get to that scale you're going to really want to make sure before you connect to a chain that it's within your you know cluster so to speak right or you're going to be very selective about the chains that you you know bridge out to outside of your cluster and so having like one way to think about Celestia is not just that it's like this consumable resource of Da but also that it's like this this network right that's why we call it a uh the modular blockchain network is because it's it's sort of like all these modular blockchains that

are plugged into Celestia are all part of one one unified ecosystem okay that's that's interesting so that begs the question then is this kind of a Winner Takes most space would you say like I know there's a lot of other providers like you have eigen Larry you have polygon of veil but when I'm thinking about it intuitively in my head it's kind of like the more chains that are using your service the more value that's accruent to that token which is Ultimate securing the data and ensuring that it's available so would you say that this is kind of a Winner Takes most Market or do you think it's a a plenty of room and there'll be enough activity where a lot of different solutions can Thrive that's a good question I think long term you know I I you see a lot of like power law

distributions and lots of different like areas of technology I think there are network effects uh to data availability networks um I don't think security is the most important thing I think it's more about sort of the ecosystem or I don't know there's a lot of different dimensions but I don't think Security will end up being the one that that sort of like causes that flywheel necessarily um but in the short term uh it's such early days um and there's so much growth that I don't think like it's important to like think about it in that sort of negative some mindset I think it's very like positive some and I think the more experimentation and the more different versions of data ability layers that

people build the better and also frankly there's just going to be a differentiation in terms of different features so like you know data availability layers might have different consensus algorithms or they might have different data availability schemes or different interfaces and all those things might you know be useful for different kinds of applications and and so role of developers might choose a data availability layer you know for other reasons essentially than just these these Network effects and so I think it's important that we kind of explore the the trade-off space here and the interesting thing is that this trade-off space basically has not been explored at all because we're really entering this new era and so I think and I hope that there's going to be a lot

more projects you know choosing a new like interesting designs essentially and and that's also one of the reasons why we have the same modularism not maximalism because we really believe that like what we should be building are a bunch of different like components for developers to choose from and and that like we're building this for the end users and the and the developers and the more like variety there is the more like trade-offs they have to choose from the the better for everyone so that's that's what we believe yeah it's an interesting point but from the larger standpoint that does bring up a good point around security though um how does celestia guarantee the the

accuracy of the information that uh other chains using Celestia for their data availability uh can rely on that data what is there like some sort of economic guarantee like a standard proof of stake system uh or how do you guys think about you know kind of providing that security yeah so security is a topic that a lot of people in blockchain like get wrong in that um so there's there's different dimensions of security right so you can't really bundle it into one thing and just say this is security right so there's um there's a few different things that blockchains provide uh and and all these kind of get bundled into into security one is sort of like consensus or or I would say I would call this it's

something that Nick Carter wrote this piece on way back it's called settlement assurances it's like hey once I've done sent my transaction and it's included in a block how settled is it like how hard would it be to rewrite that transaction like how costly would it be essentially so there's an economic cost if you want to revert the chain and re-org it and get rid of that transaction um and so that's one element of security and and proof of stake or any consensus algorithm what what it does is it provides you those settlement assurances right like once something some proof of work every time there's a new block added it's like well if you want to mine a new block you have to pay all this energy to to re you know make a new block right and if you want to reorg the

chain so there's actually an economic cost and the deeper it goes the bigger that cost become in proof of stake it's it's similar um it's that once that's in a block if you want to Fork the block someone's going to have to basically burn a ton of their stake um and so there's that side of security and that is that is dependent on uh on the the value of the token the value that's staked on the network right but really like the the more important side of security is what I would call validity essentially which is like is this block valid according to the rules that we've set out right and so this would be like in Bitcoin if I try to send a transaction that spends coins I don't have like that's invalid and this

is really I think even more important than like reorg resistance because like ultimately if someone can just mint infinite coins I break the rules and steal all your money or something then like the whole point of the blockchain is is is basically pointless versus like a reorg it's it's bad it's like someone could do a double spend but to me that's orders of magnitude less bad than like completely Breaking All the Rules and so validity uh Security in a blockchain is actually independent of the consensus mechanism or the value of the token at stake it's it depends on basically people running full nodes so how easy is it for someone to run a node that fully verifies the chain and if you're running a node that fully verifies the chain

then then you have like true validity security if you will and the beautiful thing about Celestia and modular blockchains is that we really really really emphasize that which is we want all users to be first class citizens so we want people to be able to verify the chain themselves and they can do that of the modular blockchain using something like a smartphone because you can run a data availability light node on your phone you can run a roll up night node on your phone and so and you can't do this like obviously with ethereum or Solana or any of the monolithic chains because you actually have to run you can't even run on a laptop in a lot of these cases you have to have like a cloud instance and that's the only way

you can verify the chain or even a lot of people in ethereum just use an RPC endpoint they just talk to someone else running a full node so like the whole security model of blockchains is like completely broken in the monolithic model and that's something that we want to fix and so I like and so in that sense you know people were talking about like oh well the security of ethereum uh is going to be bigger it's true that the consensus Security will be bigger if it has like more value at stake but like the actual security from a user standpoint will be much lower unless they have these like fully verifying light nodes interesting I think that's actually a good time to hop into really what a light node is and what exactly you can do with them I saw you tweet

earlier actually definitely wasn't stalking your Twitter before this but I saw you tweet something along the lines of a smartphone with an app store where you can pick different like clients to actually run so can you kind of like um I guess just paint the vision of what clients are what like clients are and like how you see them playing a role in the future yeah well I mean I I'm really passionate about this but um like the future of of crypto the way that I see it is that everywhere in the world with a smartphone literally can connect to this verifiable network of applications and uh they can that can be payments it can be defy it can be literally anything and it's like this peer-to-peer Network that they can

fully verify and transact with anyone else in the world in a trust minimized way and um the way and I think modular blockchains are actually Paving a way to get there and but what we need to do like we have now the technology like it's possible already to run a Celestia light node on a phone Mustafa has done it like many people have done it you have to have an Android phone unfortunately um and maybe they'll put this in the Solano phone uh I I would love to see that um so it's possible but now we have to make it really easy so we have to like I don't know make like applications that people can just download and install and have them running we need to do the same thing not only for the celestial light notes but for uh like the roll up like

clients because Celestia like running it so if you want to fully verify an application running on the modular stack you have to run the data availability client so you're sampling all the data you also have to run the rollup client which is that you are listening to the rollup chain and verifying the fraud or validity proofs and so you need to we need to make both of those um very very easy for people to do and my vision for that is that there's a there will be some sort of App Store essentially like an app you can download and it's running a celestia-like client and then at the same time when you're like oh I want to start using this D5 app chain or roll up you you click download all of a sudden you start sampling and listening to that chain too and so like you can you can sort of like

choose what parts of the web 3 ecosystem you want to participate in and verify and like I I hope that people will be able to verify everything that they're that they're part of that they can be first class citizens of every Network that they want to be part of so I know um you guys are pretty tightly uh tied to Cosmos of course so how do you see Celestia's role playing a role in Cosmos for app specific chains versus um I guess something like ethereum or something like a sovereign roll up like where do you think most of the activity will occur and like what is going to be I guess the most dominant architecture and I know that's really hard to predict but what's your best shot at it good question so as people know we uh you

know are very aligned with the cosmos ecosystem we've even built a lot of the Celestia blockchain with the cosmo stack so we use tender mint and Cosmos SDK to build like the base of the Celestia chain and I also just want to make a like a side note here because I keep this keeps coming back up but um there's a difference between modular software and a modular like blockchain protocol so like Cosmos for example invented this idea of like modular blockchain software meaning that tendermint as a library in a library but as like a piece of software and Cosmos SDK as another sort of like software stack our modular they separate consensus from execution in the same way that Celestia separates consensus from

execution but in the protocol stack and there's a big difference which is that with the cosmos SDK and tendermint a lot of people build monolithic blockchains that don't actually like the blockchain still is monolithic and not modular anyway I just wanted to insert that for for people's awareness um now the the beautiful thing about where Celestia and Cosmos intersect is that we share this this vision of sovereignty which is that not everyone should be building their applications on one shared monolithic State machine and there's a lot of reasons for that one of them is that you want to have customizability right and it's not necessarily like all the way to the app chain level but like let's say you just

want to add new OP codes to the evm it's so difficult why is it so difficult for them to add something like account abstraction or there's been so many improvements of the evm that people want like parallelization why is it so hard because there's only one state machine that everyone has to share and any changes could break everyone else's applications so they necessarily have to move really really slowly and so the beauty of decoupling applications into their own chains is that you can actually experiment more and you can optimize your like virtual machine to the kinds of things that you want to run and I think that's heavily underexplored and um something that like

we we strongly believe in the other thing is sovereignty itself which is that if you're embedded into the like ethereum State machine then you're kind of beholden to the ethereum consensus like social consensus so if something happens to your application you want to fork or upgrade or do something different you have to convince all of the ethereum community to go along with you which now ethereum is so large and diverse that it's like basically never going to happen and so having your own chain is is a really powerful thing that I think people will start to realize is like uh uh so much better vision for the future of blockchains than like having one monolithic State machine that

everyone is like wrapped up inside of because it's just everyone's going to be like tied up and and kind of constrained from that um another thing that's different about us and Cosmos is that um with with Cosmos the challenge is two things in order to launch a new chain you have to bootstrap a new consensus Network and a new proof-of-stake token and that's a huge amount of work like people you know you think that oh the hardest thing about launching a blockchain is software it might use might have used to be that way but now with the cosmos 60k and tendermint is actually pretty easy but you still have this High overhead of coordinating that the social and financial side of it and Celestia basically eliminates that by

being this self-service consensus Network so you come to Celestia and you say hey I want to launch a chain oh great you already have this set up for me I get consensus and data availability out of the box I just have to write the logic essentially of my chain and then you take care of the rest I don't have to issue a token I don't have to worry about securing my network and then the other thing is in Cosmos um you know they have probably the best I would say interoperability standard today which is IBC and it works extremely well um but as we talked about like in the future when there's lots like thousands and millions of chains um IBC still makes that trust assumption you're always kind of bridging outside of your cluster and so Celestia also

solves that problem which is that it can be the security layer for the interchain and allow these different Sovereign chains to talk to each other in a trust minimized way so yeah that's that's super interesting about the broader Cosmos there and being that that interoperability layer I'm also curious so last week we had Preston's at Preston Evans on who came and talked to us about Sovereign Roll-Ups and was very excited about using Celestia as the da layer for Sovereign Roll-Ups and a couple weeks back we had John Adler on who's building Fuel and he was super excited about kind of plugging and playing that execution layer with uh with Celestia and potentially another blockchain for settlement and so when you look at all these options now I'm curious which ones are you most

excited about I think have the most potential uh to be like very exciting uh or maybe where more users are more I don't want to say users but what is the most promising come configuration of these modular pieces to make the best blockchain you know it's it's a good question and I don't I don't want to I don't want to cop out but I what I will say is I I truly do believe in modularism not maximalism in that like there isn't like people have been searching for a long time like what is the ultimate perfect blockchain you know what I mean it's like no it's Solana it does all this stuff or no it's Avalanche or you know it's like in this mentality is like there's there's one design to rule them all and that is basically Celestia is is

like taking the the opposite approach as like the antithesis to that which is that like there actually is no best design the same way like if you asked a web developer like what's the best you know I don't know like setup for your thing it's like well it depends right like well you know I I want to choose what is I'm going to first think about what am I trying to build and then I'm going to go back from there and say okay these are the pieces I need and the components I want to hook up right and so that's kind of the way that we see it um and and so but within that like there's a lot of different like components to explore that seem like they will be very useful depending like for different kinds of trade-offs obviously you know the Sovereign Labs

team I think is is really brilliant and I'm really excited for for what they're building because uh I do think that in a lot of ways like what Preston was saying in the episode like ZK Roll-Ups have a lot of advantages when it comes to interoperability and uh and all of that and um they're they're really easy to make like clients for um and I think that's so I'm very excited for that I'm excited for um more like Roll-Ups as a service kind of thing I actually think that um there's there's a huge demand for people to be able to sort of like one click deploy and have all that sort of taken care of like that is our end vision is that like at my Envision so similar to this like app store Lifeline thing that's like the use the user

experience I want to see the the developer experience I want to see is like you go to like a portal and it's kind of like AWS and you choose all these different configurations like okay I need this much I want this kind of VM I want to have this kind of sequencer model I want to have you know these kind of like like clients or whatever and then you just press go and then like the chain like magically spins up you don't think about anything else and and you're Off to the Races and uh so that's what I want to see and I think there's going to be just like a huge amount of demand for some someone who whoever like owned like really really makes a developer experience like just super crisp is going to do

extremely well and I guess another thing that I want to say is like one of the most promising things the thing that was like most worrying me about this I like making this possible was just the fact that like spinning up a new sequencer set is obviously everyone knows that like a lot of change right now are centralized sequencer um optimism basically every roll up in existence right now is still centralized and that's not good from from aliveness and censorship resistance standpoint and so we want to move towards more decentralized sequencer rollups but the problem is if you want to launch the centralized sequencer roll up it seemed like the only way to do that is to basically uh like create a new sequencer set every time and that kind of becomes

almost the same problem as launching your own blockchain and so I was very worried that like okay this is going to make things kind of Impractical or kind of decrease the value proposition that we're originally trying to go after and um the beautiful thing is that like now I feel very confident we're going to solve that there's all these new projects tackling decentralized sequencing and there's like some very compelling ways to to actually solve the problem in a way that is decentralized and Trust minimized and um so yeah we can talk about that but as tree obviously is one that recently announced you know there's espresso there's a handful of other projects I think even flashbots might be looking into something similar with Suave so um that's that's what I'm also really

excited about I think that's the next most important layer of the modular stack to get right yeah strong agree with that we definitely need higher accountability I guess for decentralizing sequencers but to play Devil's Advocate a little bit like every like so arbitrum and optimism have their native tokens ARB and op and everyone's like oh like the little crew a lot of value as soon as the sequencer is decentralized with their token as like a proof of stake Economic Security model or um you know an Mev will be extracted that way so I guess my question to you is why wouldn't L2 want to use let's say a data availability service or a um a decentralized sequencer when their tokens not involved in the value accrual

process at all yeah that's a good question so the um Jesse from variant put out this really short blog post and tweet the other day which is basically about like networks versus libraries right and um the point was that networks in in web 3 are fundamentally kind of new in that we have a way to sort of like um generate economic value from from networks and sort of like have them be owned and they have a reason to sort of have a token and and so forth whereas there's also libraries where for example like ERC 721 is a library right but it doesn't it's very important and it's very useful but there's not you can't

really build a network around it right and I I found I thought that this was a really good framing because it kind of articulated something that I've been thinking for a while and why this is relevant to to Roll-Ups is that um when I think about what what a roll-up is at the end of the day there's two two sides of it one side is the Library part which is like hey here's the execution environment and here's like sort of like the proof system and model that we're using to secure the roll up right and then in theory anyone could take that like they could take arbitrum's code like the library side of it Fork it and deploy their own roll up the only thing that's stopping people right now is the business license for example but and then there's the network

part which is like hey I'm going to launch a roll-up like the arbitrum main chain that is sort of like the the network behind the technology that I built and so anyway like I think that it's true that if I am building a a Roll-Up that I and I want to like sort of own this network I guess I wouldn't I might want to sort of own like I need to justify having this token and maybe the way to justify that is to sort of like also run the sequencers and be a little bit more vertically integrated but I think in the long term like the average developer is going to be building like they're going to take this like the libraries like things like

what Sovereign Labs doing The Sovereign SDK like it's a library someone's going to be able to take that launch their own chain and um I don't think they might not even want to have a token they might have a different kind of business model or use case right or maybe the token is useful but not it doesn't need to be justified for the sake of decentralized sequencers it could be part of like the built-in application for example and that's the network that they're building so I think it makes sense it's more scalable and secure long term when we pool resources right and I think having a shared sequencer network is another sort of like natural sort of like pooling of resources into one common Network that people can launch on top of another a few other things just while

we're on the topic is like you get a lot of benefits on using a shared sequencer like cross chain composability between these different Roll-Ups can be insured because they're sharing the same set of nodes that are ordering their transactions so like there's a lot of reasons why I think shared sequencers will end up accruing a lot of the sort of uh like roll up ecosystems yeah I agree with you too and it's almost like once one of those or two of those kind of gets that Lindy effect then you're almost at a disadvantage not being on that same shared sequencer right if you don't have that easy composability with all the economic activity that's happening uh you kind of put yourself at a disadvantage so that'd be an interesting state to see if we actually reach but I

want to back up a little bit there and you know you mentioned that and state that you described with like the developers logging into this portal and kind of like building your own chain with everything that exists and I'm curious to get your take like where do you think ethereum stands to to really fall into in all this good question I mean I think uh where ethereum is is going is is actually very modular um they are you know basically moving to to make the ethereum base layer just like a DA layer um and so I think like in that portal right like you'll be able to probably choose between different data availability layers to plug into and I think ethereum Bank charting will will be one of those one of those layers um I think the only difference again

going back to we touched on this earlier between for example Celestia's vision and ethereums is that ethereum does have this baked in execution layer which which is an advantage but also a disadvantage and so it's again it's a set of trade-offs the uh the advantage of having you know the eth L1 execution is that you can have settlement right and there's this Rich you know pool of assets and users that are already on the ethyl one and now when you launch a roll-up you can have a a bridge down to that ethyl one and it has like you know it it can be sort of like this the root of the ecosystem if you will um the disadvantages is that then you now have a more stateful sort of like uh

complex layer one that's not as minimal or modular and so Celestia's kind of choice is like to eliminate that and to be maximally minimal and maximally modular uh and so we don't have any execution and so a lot of people are like oh well then where am I supposed to settle my roll up and actually we should we should talk more a little bit more about like smart contract versus Sovereign Roll-Ups um but basically Celestia thinks that in the long term um yeah settlement should not be on the layer one it should not be coupled with data availability and um instead should should be sort of a separate thing you know and that's either settlement via Sovereign Roll-Ups or settlement on specific Roll-Ups built on top of

Celestia that specialize in settlement rather than kind of enshrining something uh from from from day one So speaking of settlement Celestia some people will argue one of the downfalls is that you can't have a two-way trust minimized Bridge um how would you respond to that what are the trade-offs and what like what are the benefits really yeah so I actually agree and I've been like kind of it's something that that bothers me right now about Celestia right now is that we don't have a trespondized weight of getting the Celestia token up a layer above to the ecosystem we do have a way to do that we can do we can use IBC or again I don't want people to get mistaken that like there's no way to get

the token up it's it's very possible but you're going to use it's not going to be this trust minimized bridge but which Anyway by the way none of the Roll-Ups really have trust minimized Bridges right now like they all are have multi-cigs or like for example even arbitrum you know they are like a lot of them the fraud proofs are not actually active so I also just want to say that like there aren't very many trust minimized bridges in the wild anyway so it is not really that different but like at least as of now there isn't a trust minimized way to get the token up however we've actually come up with some really really interesting ideas around how to do this one is like for example we could do basically interchange staking which is the the cosmos version of restaking uh to create sort of like a

parallel like settlement chain or like a bridging chain that Bridges up to two Roll-Ups or there was another interesting idea that we had recently which is that we could um in the same way that like what Bitcoin is doing we could have specific um sort of operations within the state machine that can verify ZK proofs and by doing that it's like a stateless way of us basically being able to bridge tokens in and out of of the Celestia base layer and I think that's actually a really cool uh cool one it will obviously take a little while to to be built um but like there there are it's it's it's going to happen it's just that we we're the reason why we haven't committed to anything is that we want to do it in a way that preserves the like

the minimalism of Celestia because we think that that's sacred you know and so um we're just trying to we're trying to figure out what is the right design that makes the right trade-offs essentially um so yeah hopefully that answers your question yeah it does I'm curious is the reason that you and your team is thinking about this is because you want to ensure there's enough I guess economic value around the Celestia token so that it can secure the chain from an economic standpoint I know you already mentioned earlier that you think there's other um reasons to prioritize for security over the proof of stake value at you know that can be slashed but I'm curious if that's why you guys are thinking about a two-way Bridge yeah I mean I think of the Celestia token as being a very critical part of the the overall

modular ecosystem on top and I think that um you know it's it's going there's going to be demand anyway to I think Bridge it up and use it as collateral or maybe even rollers may want to use it for fees or you know there's a there's a I can't even imagine all the different ways that people might want to use the token aside from just staking and so just it seems very silly to me to limit the utility of the token simply to staking and like in the future sure let's say there's like liquid staking derivatives it's like okay well I've staked my token now I have this you know sticking derivative I want to actually like you know I don't know use an ND Phi or what have you um and so uh yeah I do basically what I'm saying is yes I think it's important

that the token can be free right and like um I mean this is true even in like traditional Finance like if I have a bunch of Apple shares right like I I also might you know borrow against them or what have you like it it's it's it's ultimately it is an asset that I think like should be you know able to be tapped into in lots of different ways great that's that's super helpful um this one's really in the weeds here so hopefully it makes sense but so I've read about you know obviously in the ethereum roadmap you're going to want to use Erasure coding in order to um sample the data and ensure that it was actually extended correctly so you need these light clients to be able to

to check that um in in the case of Celestia so everyone always talks about oh you need to use a 2d Square so that way you only need to download either a row or a column can you kind of dive into that how that process works and uh how you have the exact same security guarantees in terms of data availability under this kind of system yeah so it's a question about sort of Celestia's Erasure coding in general or is it just like social versus like compared to other da schemes yeah Celestia compared to other da schemes because I know as well you guys are using fraud proofs instead of ZK proofs so I'm just kind of curious how the whole system ties together yeah sure so um so there's there's uh basically to so

first we should frame how data availability sampling works so um the way that it works is that you have like this block data right which is the actual transaction data and you want to make sure that all that data has been published but again you don't want to have to download all of it because it could be a lot of data and you you might only have a phone like I'm not going to be able to download a 10 megabyte block every 15 seconds um and so what you do is you take this square and you you use something called Erasure coding which basically adds redundant data in such a way that I can reconstruct the original square with a very small subset of this extended data and so what that means is now if if the

validator wants to hide even a small amount of the square like literally just like one bit of information they need to hide a huge amount of the overall data that includes the extended square and so um that's what makes it so now I can sample the block at different spots and if the validator was hiding something from me chances are after 32 samples or so if I never hit an unavailable spot that means that like with very high probability the whole block is available that's in short in summary how data availability sampling works now the the kind of main difference between Celestia's approach and ethereums is

that is is the kind of Erasure coding technology that we choose to use so in Celestia we use read Solomon Erasure encoding and in ethereum they use what's called kzg commitments the difference there is that in a ratio coding it's possible for you to basically Forge to pretend like you're you're you have to when you extend the data you can do it in an honest way which is like to actually do the real like actually extend the data or you can do it in a malicious way where it looks like you extended it but actually if someone then tried to use the data to reconstruct the original data they couldn't and so what you need to do is basically make sure that the ratio coding was done correctly

and um and so what you need to do uh is if every time like a block gets generated the other full nodes download all the data and they check their ratio coding and make sure that it's correct If it's incorrect they generate what's called a bad Erasure uh coding fraud proof and they can send that fraud proof to all the like clients who can then say okay I'm going to ignore this block because the Erasure coding is done incorrectly because if you don't do that they would sample the data and think the blocks available but in reality it's not because they couldn't reconstruct the original data if you will now in kzg's the cool thing is that um it's you can't like do it incorrectly like if the correctness is kind of there already from the beginning so you don't

need to send around a fraud proof so it's kind of like it's very similar to the difference between an optimistic roll-up versus a ZK roll-up if you will so and and there's also this is where the trade-offs come in the kzg openings are very expensive and slow to to compute and so like it's gonna it takes a shitload like a huge uh like block Builder to do that and it can also the latency could be really big obviously I think they're working a lot on parallelization it's very similar to DK wallets where you see a lot of like hardware optimizations and like trying to make this faster and in the future if kcg becomes viable and scalable um then it's something that we could

absolutely switch to but for right now it just doesn't seem like the best trade-off basically uh and so we're going with the the Erasure coding approach the other thing is that it's also possible to actually ZK uh prove Erasure coding potentially so Something That We're actively researching and so in the future we might actually get the best of both worlds and by the way like the Erasure coding is like doing read Solomon uh so that ratio coding is extremely optimized and very very fast and very scalable so if we want to scale to like much larger blocks we know that we can do that in efficient way whereas if you want to scale up with like larger blocks with kzg it's like holy crap like that's um you know like a let's say 100

megabyte Block in kcg is going to be super slow like so anyway those are the trade-offs that's probably a little bit too in the weeds no that was that was great I really appreciate it I I guess I had it backwards so I thought that posting fraud proofs you know you're assuming all of it is valid until there's actually a fraud proof submitted and I thought that ZK proofs were more instantaneous but they're more gas intensive like you mentioned so you kind of have to wait a longer period of time before you actually submit them is that what I'm getting like is that where the actual delay comes from on the ZK side of things yeah so I mean um so with the fraud proof scheme you have to wait a certain amount of time to

elapse that you didn't see a fraud proof to counsel me as final right and and so that's where the delay comes from on that side it actually could be quite quick to generate that fraud proof but you want to just want to be safe like from from your perspective that like a fraud proof is generated and it got circulated too from the on the ZK side it's a little bit the opposite it's like once I get the ZK proof or once I get the proof that it's valid like I'm good to go um but it could take a while from for them to generate the ZK proof to begin with because it's just a lot of it's very like expensive to generate these proofs okay I understood thanks for clearing that up for me so I guess my next question would be post 4844 on

ethereum what is the data availability capacity of ethereum versus what Celestia will be uh come launch in either the end of this year 2024 sometime yeah so um I'm actually I saw some threads about this on Twitter about the the uh Proto dang charting like 4844 capacity I didn't really understand it as far as I understand they're adding one megabyte of capacity it could be totally wrong to each uh ethereum block um which is which is a good amount of data and I think it's gonna like go very far to decreasing the cost for Roll-Ups on ethereum um and Celestia's case where when we launch we're going to have a maximum block size of eight megabytes

um and so uh that's gonna be I I I I don't know like it's hard for me to compare the other thing that I think the main thing that people need to realize right and this is because like we get into these arguments also with people about like from the Sonic Camp which is like oh Solana has all this like throughput and like even and then they take the eight megabytes every 15 seconds which is Celestia's initial sort of like block size and block time and they say like oh that's only X number of megabytes per second of throughput and um the the thing is that it's like apples and oranges in the sense that the whole point of building a data availability layer and doing modular blockchains is that there is no fixed throughput you

can increase the throughput as there's more nodes online so we can actually flip a switch and say actually you know what we're going to change from eight megabyte blocks to 32 megabyte blocks right we we can we can continue to increase the block size um and so like even just thinking in that mindset of like oh what is what is the capacity is kind of a misnomer and and the thing about again getting back to like 4844 is that it's not modular there's no data availability sampling so they can only this is like a one-time increase as sort of like a Band-Aid sort of like fix to like add a little more capacity but it's not like they can just wake up tomorrow and be like okay there's more nodes in the network we're going to increase the size again they can't do that until a full dank sharding

is shipped where they actually have true data availability sampling and so that's that's really the core difference is that Celestia's block size is uncapped and you mentioned there the parameter the limitation was the number of nodes online and can you just dive into why that's the case sure so in the way that these data availability scan sampling schemes work is that um you so you as as yourself a bunch of different points of the the of the block right um and then you know that if there was if he was hiding something then you probably would have found at least one of one of the ones you sampled would have been unavailable um now the problem is that what the

block producer could do is basically only let the uh sort of light sampling nodes sample data and not let anyone actually download the full block themselves and so then what you need to do is you need to be able for those light nodes that sampled to be able to like re basically put all their separate samples back together to reconstruct the original block so you need to have an enough light nodes that sampled enough data for you to reconstruct the original block and so that's that's sort of like the one of the security like Properties or like requirements for this scheme to to to work so that hopefully that

answers your question so then as you have more and more light nodes more and more people like sampling on their smartphones for example all of a sudden you can start to increase the block size and so the beauty of this is that you know monolithic chains have this negative Network effect where the more users there are the more applications there are the the slower everything is or basically the more expensive everything is it becomes like crowded and congested and in a modular setup we actually have these positive Network effects where the more users are more applications the more people running nodes the more block size we can have so we can actually have this positive feedback loop where you know we can continue to increase the block size to uh basically support more users and more applications no that was

perfect I think that really helps paint the picture of why the number of light nodes sampling uh matters in this case right and it's beautiful that it's like perfectly scalable like this too but I want to change the tune just a little bit here and uh you know we one thing that was really hot on crypto Twitter over the past maybe two months ago one month ago was roll kit and really the idea of using Bitcoin as this da layer um when you think about using Bitcoin as is that something that's practical to you you know obviously the 10 minute block times are uh the first thing that comes to my mind but I'm curious about how that that whole framework works in your head yeah I mean so this again comes back to well let me give a background on what rolekit is so

like back in 2021 when the project was still quite in an early stage no one so there were a bunch of roll-up uh teams building on ethereum right but they were building specifically they weren't building roll-up software that was generalizable that like anyone could take to launch their own roll up they were building their own kind of proprietary Stacks essentially and so we saw that as a problem because our vision is like everyone should be able to deploy roll-up so we started with uh started building something that was meant to achieve that and kind of borrows a lot of inspiration from the cosmo stack and that's now what rocket is uh but now fast forward 2023 there's a bunch of people like Sovereign labs for example a ton of other teams building these sort of like role of sdks that share that

vision and so and because we really believe in in being like credibly neutral right modularism not maximism we've spun roll Kid out into its own separate thing um so that we can be independent and neutral um but anyway we'll rocket uh made a bit of a splash as you said when it basically integrated right now it started with just integrating Celestia as a DA layer um but it recently integrated Bitcoin and showed that you could deploy for example like an evm based roll up on top of Bitcoin and in terms of using Bitcoin as a data availability layer I think the interesting thing about this is that um Bitcoin has a ton of security it has a really big ecosystem and I also feel

like Bitcoin is as an asset is kind of like like we said like kind of Trapped and on a chain with not much functionality and if it could get sort of like if you could add more functionality it could become like a more interesting ecosystem and I think my hope is that this can kind of like spark more Innovation and conversation in the Bitcoin community but I still think there's a long way to go like do would Bitcoin actually be useful as a DA layer compared to other Solutions like Celestia or ethereum like like you said that the block times are really slow there's really the block capacity is also quite small frankly the data throughput is like pretty minuscule um but I still think it's it's worth doing and in the future like I think

basically the fate of every L1 is to eventually just become a DNA layer because that is really like the root level of blockchain is just consensus and data availability so in the future I could see Bitcoin kind of moving in that direction too potentially I don't I don't know right now I like your points there I just quickly checked and there's about four and a half billion dollars of rap Bitcoin on ethereum so uh there is seems to be some sort of demand to to get a more interesting use case for Bitcoin um but no that's that paint's a great picture there um and I I kind of agree though like I I it's almost interesting because right so like ethereum would be doing less than it is today if it was just a DA later whereas Bitcoin would be doing more right today it's just a proof of work

chain but kind of like adding that data availability function would be an exciting development and I'm definitely in the camp of uh seeing more innovation in the Bitcoin space would be super exciting yeah strong agree there Dan I have one more question for you before you go Nick and that's uh oracles have you thought at all about how Celestia could improve the current Oracle landscape I know that's super Niche question so no worries if you don't have an answer for it but I just kind of view that as like an Achilles heel of defy and I'm curious if if uh Celestia and data availability Services would help at all on that front good question I don't think about oracles that much I mean itself's own like rabbit hole I feel like a technical problem to solve um I do think that uh one thing that

could be interesting and the way that oracles like intersect with the modular stack is is sort of just on the app chain side of of things which is that like what I would love to see is someone who builds like an oracle chain or roll up I would say on top of Celestia that anyone when they launch their like roll up can easily like tap into and integrate and and have like sort of a trust minimized Bridge with um and I think that that's sort of like the future of where I think roll up should go it shouldn't be like you know a new roll up on every chain or or like deploying a contract on every single chain that needs an oracle it could be like some kind of like out of the box sort of like roll up like Oracle as a service kind of kind of roll up

um and I think that's a lot of like I think the future of the modular stack is going to have a lot of chains like that that are sort of like they specialize in one service that they sort of provide to the broader ecosystem and um I think Oracle is obviously like a very natural fit for that kind of um uh sort of like use case super interesting well we're really excited to see how the the project progresses I guess you want to give us a timeline and then also where people can find out to learn more about you and Celestia yeah so as people might be aware we are currently running our incentivized testnet the block Space Race um we are you know a few weeks into the process and and there's like another month or so left and uh that's really

exciting because we have you know hundreds of nodes uh running we have like I think 800 light nodes so all over the world uh it's the biggest live data availability sampling Network to ever be created uh and um so it's really exciting too because basically this is the last step before we actually launch the the real Network like mainnet and so I'm excited to see how the blog Space Race goes and and what what kind of like bugs we find and and so forth and then depending on that hopefully we will be able to launch mainnet later this year so um that's going to be a really historic moment because to me it's like the first modular L1 to to really be birthed and I'm hoping that

this is sort of like a turning point in which the broader blockchain ecosystem starts to you know we've been talking about modular blockchains for like years now and I want to see them actually be live and and be used and I think uh there's definitely a lot of Kinks to work out like you know this is a whole new paradigm there's a ton of stuff to be built not just on the celestial layer but all the the execution settlement shared sequencers like wallets all this stuff there's so much to build but uh it's exciting to to know that this is kind of like the first step and there's a whole movement of people you know building projects that are that are aligned with this and um it just kind of feels Unstoppable at this point because enough

people are modular pilled and there's just so much momentum behind this new sort of mental model and Paradigm of of blockchains and uh and I think that the mainnet launch is really going to be like one of the Sparks that hopefully like lights this this flame um and and so I'm excited to see where that goes um in terms of following us like I would say Twitter is the best place I'm Nick white but instead of the I it's an eight uh and then we have celestia.org is our main account and we're always sharing things I encourage people if you're interested to learn more about modular blockchains we have a modular learn modular section on our website so celestia.org learn um there's a lot of really good articles

there and diagrams um you know I try to also sort of share my insights when I can um and if anyone has any questions like we we love when people hop on the Forum and ask questions or just you know tag us on Twitter because really this movement depends on people kind of like learning and understanding the philosophy and the design principles of modularity and we're still so early on in this whole process of sort of like you know activating people's minds in this new way of thinking because it's it's it is a little bit if you come from the monolithic standpoint it doesn't really make sense at first so you have to put in the effort of like learning a new way of thinking but it's so fun and

addicting like once you go down the rabbit hole and you become modular build like you never come back so uh yeah I'd love to love to like modular pill more people basically absolutely we'll definitely play the our part in helping you with that you mentioned two really great blog posts in the middle of this uh interview and so we'll definitely link those in the description but Nick thanks for coming on man it was a pleasure talking to you uh and honestly just hearing the passion in your voice about what you're building gets me so fired up to keep doing what we're doing as well so Nick thanks a lot for joining us man well thank you guys very much for hosting me