Facilitating the Modular Blockchain Revolution w/ Amrit Kumar (AltLayer)

Transcript

[Music] hello and welcome back to the strange waterer podcast thank you for joining us for another episode these days you cannot scroll through crypto Twitter or through warcast without in the phrase modular blockchain and I'll admit I often succumb to using the buzz word casually and without definition on strange water all the time so let's define it we'll first start with a monolithic blockchain the original design of most blockchains we've spoken before about how ethereum uses a decentralized

network of computers coordinated by proof of stake to host the ethereum virtual machine now the ethereum virtual machine or the evm is a Turin complete Computing space think of it like a general purpose virtual computer complete with a processor a hard drive Etc a monolithic blockchain looks to scale by increasing the processing speed and reducing the cost of using the network directly often just by requiring more and more advanced machines to participate in the network and this is effective but it has direct implications for

centralization and so it inevitably limits the value of a credibly neutral system but modular blockchains approach the problem from a different angle a modular approach involves examining the system asking what needs to happen on chain and what can be offloaded into less resource constrained environments for example let's say you want to make a financial transaction maybe borrowing usdc against eth collateral on something like a if you're to do this on a monolithic blockchain all the computation required to make this happen would need to go on inside the network how much is the eth

worth in usdc terms what's your borrow rate what's your liquidation point and on and on and on finally this would all be resolved and the result will be recorded on chain and from that point on no computation is required the position just exists we can recognize that the important part of what happened here is the result and not the actual calculations and so we can move the calculations to an outside environment where it is Trivial to calculate all these answers and then just post it and yes we need to find a way to trust these outside environments but fortunately we've developed

Technologies like fraud proofs and more importantly ZK cryptography to solve this now that's just one example of modularity and blockchain another is something we refer to as data availability and as listeners to this show know more modularity is on the horizon hint check out episode 42 but as you can clearly see even explaining this at a high level is pretty complicated and we haven't even started talking about how to use igen layer to create a restak backed rollup the modularity forces so many more choices and Integrations than the monolithic Paradigm can you imagine how

much more difficult it is to implement fortunately we have people like today's guest building tools to make the transition to a modular future far less painless than most developers are anticipating amri Kumar is the chief operating officer at alt research developers of alt layer you'll hear alt about om's journey through Academia and Entrepreneurship to his current position as a key player bringing the modular blockchain thesis into reality and then we get into alt layer and how they're building tools to make deploying flexible customized blockchains and app chains easy and painless instead stick on through to the

end to hear about some of the big unanswered questions in this space including about data availability and reaking one more thing before we begin please do not take Financial advice from this or any podcast ethereum will change the world one day but you can easily lose all of your money between now and then okay let's bring on M Reit om re thank you so much for joining us on the strange waterer podcast well thank you very much for inviting me of course man I am so excited about uh everything that you're doing with alt

layer I mean of all of the uh darling projects right now like you're definitely having a moment in the Sun but man before we get to that uh I'm a huge believer that the most important part of every conversation are the people in it so with that as background will you tell us a little bit about yourself how you found crypto and I guess maybe why you didn't run away yeah of course uh well I I'm originally from India so I did my undergrad in mathematics in India and then I got a scholarship to um I mean of course most Indians end up being end up in the US because they want to study at MIT or whatever um for some reasons uh I got an offer from a university in Paris um it was an engineering school and the

reason I went for it was mostly because they were offering me a scholarship and so I felt okay the guy is giving me a scholarship I should go for it um so I went to I didn't understand the challenges at that time but anyway I went for it and um so I interviewed the you know went through the whole process of interviewing and everything so anyway I got selected for this interview for this scholarship and I went to Paris and then I realized not then I mean I realized before that but then I think it hit me at that point that everything is going to be in French so uh before that it was like okay you know maybe there'll be some English component you know maybe things lectures will be in English whatever right but it was only when I got there that I realized that it's going to be

hardcore French right so um the good thing was that they had this uh three months uh if I recall correctly three months of a language internship so basically from 8:00 a.m. in the morning until 5:00 p.m. you basically learn French and it was for 3 months so we I went through that process I won't say my that my French went up to I don't know level that allowed me to take Serious science or mathematics or history or whatnot but I guess it gave us enough uh I guess meat to actually go and go to the lectures and at least understand half of what was going being said so it was painful experience I would say um you know reading I don't know geopolitics or or I don't know the history of religion and whatnot in

French was not an easy task uh anyway um so after that I went to so I studied computer science I then did a PhD in France and then um around that time there were researchers uh in Academia they were looking at Bitcoin and and other things that were happening in the space and they were like hey this is something that's exciting or interesting before that so I'm talking about 2013 around that Mark right before that it was like Hey you know it's it's this people doing their thing you know there's nothing I mean it's interesting but nothing serious maybe it will die away or whatever um but you know that was Chang changing so I think there were at least two academic papers that were published um from very big academics in in

computer science and that I think pushed people's interest into Bitcoin and other you know related you know cryptography and and systems and so um some of the people I say senior researchers the lab where I was doing the PHD they were getting interested into or hooked up into Bitcoin and related protocols and so this was I would say the last year of my PhD so I said to my adviser can I can we work on something you know around around Bitcoin and my advisor basically said you know it's it's your your funding is secured by from a certain you know thing and you know you you're not allow you're supposed to focus on this topic and not not basically move away so I didn't have much choice so I said okay fine I'm going to finish one one extra year of my PSD and then I will

try to work on something that is ready to bitcoin um so um so I I got interested already and by the way at that time I was mostly interested at a superficial level so I had not read the Bitcoin white paper uh people talking about it so it made me interested know this is something interesting the fact that you could build kind of money where you don't have a bank which is interesting but I had not you know even though I was I was academic and I had this technical background I did not actually go into you know reading the white paper um and then I said you know what let's finish my PhD and let's go and look for a post doctor research position and this is where I will go and get into my my my bits of Bitcoin and then I applied for a couple of postdoctoral

positions and I went to I got you know an offer from n us which is the National University of Singapore um and by the way this was the time when icos were happening right so you know I was not following that news that much I was more in the academic sort of things but when I landed in Singapore I realized that a lot of people are actually thinking about that you know e was there back then so people were doing icos EOS and tsos some of the other projects were doing iOS at the time and Singapore by the way interestingly I didn't realize back then but interest at that point Singapore was one of the countries that you know was the place for these icos so um it was kind of neutral country they didn't hate crypto and the other country was by by the Switzerland

so surprisingly I applied only for went to two countries that were I guess supportive of crypto and then I got an offer I joined the lab and sadly my advisor um he said there's nothing interesting to do in crypto anymore you're too late to the party and then again I got very disappointed because I wanted to do research while during during my P I got basically slapped there and then when I come back here you know I'm getting the same story so and my advisor said you know we should you should work on machine learning this is this is the new area of of research and you should go for that I worked on that for about about a month and I think my advisor kind of got a sense that you know this is not his thing he doesn't want to do it anymore uh and

interestingly at that point every single person in the lab was working on crypto so by the way there was a guy called L Lou Who was the founder of kyber network which which was one of the first in the ethereum space so everyone else was working on crypto I was the only guy the new guy who was supposed to do work on something new um anyway at some point my adviser gave up and said you know there's this shiny thing called Monero that is picking up you know steam maybe you should take a look at it uh and that was my opportunity I had to seize it because otherwise he would say you know go back to your machine learning thing uh so I picked up Monero and then so I read the Monero white paper first and then I understood okay how Bitcoin would work

because they were es about Bitcoin and all that so um I read Bitcoin Mon's white paper first and then went to bitcoin's white paper um and I found very interesting because monero's idea was simple um and and still quite interesting compared to at that time zcash was there zcash was much more complicated you know at that time uh Monero was a very simple idea um and so I started researching into Monero you know we published a paper with some of my interns and colleagues we show that mono was actually not very private at least at that time i't haven't been following that project anymore and then um you know funnily um zuku from zcash reached out to me um on signal once once and he said hey we have some data set for for um you know Dash and would you

like to run your algorithms on dash and see what what you get out of it and of course there was conflict of interest thing so he didn't want to do it himself you know he wantedd party to it uh and then again uh my advisor said no it's becoming boring let's do something else and that something else became zelica so this was uh 2017 zelica was um basically based out of a paper that L and some of his friends and colleagues in the lab had published and this idea was to build a blockchain that would have sharding built in so this was something that etherum Foundation was also pursuing quite heavily in around 2017 2018 and so um my advisor has this idea of you know let's take this uh paper and try to commercialize it in some way uh

in the form of an L1 and so um me and some of my colleagues in the lab um you know mostly PhD students or who were transitioning from PhD to postdocs we joined hands together and we started silica I would say we the probably one of the last projects that uh finished or went through the Ico and after that basically Ico has died um and so yeah we we I was uh the chief science officer at zelica so my role was to basically look at the problems or interesting problems that exist in the space how to solve them and then um um so I stayed in that role 2021 and then I this is by the way where I met my friend yaoi who is currently the CEO of all Clair he was the CTO of zelica so he had he left zel early on in

2019 when he went to join par and then in 2021 he said you know I'm starting this new thing would you be interested in joining me as said sure so that's kind of the backstory of know me starting as a undergrad to all CL today wow man that's incredible so first of all thank you for sharing your story I think what you're your story is a microcosm of basically like everything that uh all of us in this industry and maybe in entrepreneurship know which is like the path to success is lined with people telling you like no you can't do this you're not like this is not the right time like go find something else to do and what it takes to find success is just to like have passion and the

ability to just say like [ __ ] off like I'm going and um you know like that's what I hear in your story is regardless of all these people who kept telling you not to do it um you just knew that crypto was something interesting and you were going to be involved and here we are in 2024 and you are at the the flag the the Helm of like one of the more exciting projects as we enter into this like new modular uh phase so incredible man I before I let you move on from your background I just I I have to ask um what was it like going through like schooling in a language that you like flat out did didn't speak and I guess like more specifically do you were there any things that like looking back

actually ended up being like kind of advantageous for you like did you look at things any differently or did forcing yourself to learn a new language like give yourself better um discipline like as as an American who we over here are unaware of other languages other than English like what was the experience like of educating yourself um in a language that you didn't speak I would say so I mean it definitely was difficult I won't say I mean okay FR French is relatively easier language compared to something like German right um some of my colleagues were friends were learning ger German at some point at the University um I didn't have the courage to take that so I I felt no I have to go Notch higher so I picked up Chinese later on

uh so I think I think it's interesting one because it gives you different perspectives because I think language is not just a thing to to communicate but it also teaches you culture right um and what you learn from but by just by learning a language for example when we learning French at that time we had to spend I think about a month at a local family so as a as a paying guest you basically pay them and they will they will you stay with them you know you spend time with them they will take you with to shopping and whatnot so they take you you kind of go through that culture it's not just about you know speaking the language it's about understanding uh you know how how to you know how to stay in that Society right basically and of course you know we I

came from India so of course we have a completely different culture right um and so we were taught for example you know it feels a bit bit silly right now but at that time we were taught how to hold knife and fork and things like that this was taught to many of us including Indians and Chinese as well because Chinese are more used to using Chopsticks back in the day right nowadays everyone uses everything so I think all of us were taught how to use these cutleries um because they felt like okay this is this is how you learn the culture they were we were taught how to position different cutteries on your table around you know things like that Al so that's one thing so of course you you learn the culture right and then once you know the language you actually can read things that were previously

completely inacessible to you right so for example I was able to read books in French I was able to enjoy some of the comic books that U you know French have written like DIN and some of the other very popular ones in French and you can see that translations don't work right now you know I had read that in previously in English you know yes you enjoy it but I think there's a certain flavor to it when you read in the native language right um so I say that that was tough but I think it was encouraging and it opened up new ways of uh new ways of thinking and the second thing was you can then assimilate in the society right which is you know if you're going to live in France for seven years I mean you you can't you can't live if you

don't if you don't know the language it's going to be very painful to you you can't interact you can't make friends um but anyway I also P because of that because I had picked French you know I also decided I should pick another language uh which ended up becoming Chinese I learned for about 2 years and gave up up but that gave me my new my wife actually so so it it it also it kind of it went up in a circle where I think you know some people say every work that you put in it kind of accumulates and you know it has a cumulative effect eventually I think that's what happens you know you start something and then it leads to something else and that leads to something else and I think language is is something very similar it opens up New Perspectives so you yeah no I and uh

just a little bit about me I'm uh engaged to a Russian woman who like is from MOS her whole family still lives there and so I'm going through the process of learning Russian right now which um any language learning as an adult is impossible but wow Russian is just like a whole different category but I I do I'm just so struck by uh like they have words that are like entire you know like somebody will say a word and like in order to explain what it is in English you need to like call together like eight like eight other words plus like a feeling and just like try to get get people to kind of get what you're saying and I think that to me that is just like a reflection of like the languages we

choose are like the mediums by which we think you know and I guess let's keep this like crypto related like choosing let's say uh viper versus um Cairo like Viper is all about like readability and just like you like ease of access and it comes from like the python world right versus Cairo is all about just like pushing the boundaries and finding like the most like Advanced interesting stuff and like the the solutions the problems that are even trying to be solved like everything in is different in these worlds because of like how the the thoughts and the Expressions come out so yeah certain things just don't exist in certain languages like you know for example there are languages in in Africa or even India where if you say okay

what's what how would you see this like there's a very common expression for that or word for that in certain languag that doesn't exist and similar extended that to Cairo or or programming language it's very similar which is um their expressiveness could be different there's certain things that you can express very easily in a certain language you know programming language that you can't do you know in other languages I think that also gives you better developer experience right which allows you to express quite easily you can write shorter code maybe less readable but at least shorter code which would be more optimized in some way yeah and so just to harp on the point then we'll move on but like shorter but like less readable but more efficient versus like longer but more readable but maybe less performant like those are actual values choices that like a community has

made and like from that the results will be different so anyway I just I I uh I'm a language my head's on language these days and to find someone who like not only achieved a second language as an adult but like used it to further your education I'm in awe my friend thank you thank you I mean if I extend this point right to to more crypto thing right what what I've also seen interesting is you know there were some experiments on the language side about 3 four years in the etherum space right people were experimenting with um different kind of languages um you know e itself has you know I think couple of languages before before they converge to solidity right um then other ecosystems were building their own language for example teos had

its own language uh melson at zela we were building our own language as well um and at some point you know we forgot or we felt like oh Sol has won the race you know we should forget everything else you know let's just focus on solid and I think Solana has definitely done something different right you know they have shown that you languages you know is can be a good option um we have seen for example um move coming up as a language so I think it's it's a good it's a really good thing maybe it's some of these things are also tied to I guess the modeler world that you know you mentioned because modularity because execution move had moved out of ethereum it has allowed people to experiment a little bit more and given people

flexibility that you know what now you can do it previously it was kind of frowned upon you know if if you had gone with a new language two years ago you would have said you know why you doing this already has gone it you know done it and there's nothing there's nothing interesting that you can do with this now I think the perspective has changed and so I'm I'm really excited to see how move and some of the other language you know get developed and adopted by by the community in general no and I actually think that's a great Pivot Point to talking about alt layer but before we get into kind of like why that's a great Pivot Point to talk about alt layer like can can you take us to this like 2020 2021 time frame and I think like now in 2024 what alt lay are building it's just

like so obvious that it hurts right but and that is a reflection of how much further we've walked down the ethereum road map and how much more well understood like what modularity means in the blockchain context but 2020 2021 like the full full disclosure like that was when I was entering crypto and like during that time like the the idea of Monolithic versus modular was just starting to be articulated so can you take us to that time and like talk a little bit about what is like the Genesis of alt layer and why did you know three years ago that this was like the right play yeah so um when we started 2021 um and I think a little bit

before that um you had you know people are talking about zekas SN you know the idea of you know zekas was there the idea of optimism was there idea of arbitrum was there but it was not kind of said in the model way right the you know we are going to build rollups and that was kind of the idea but what what we what we felt I think very quickly was that hey um and at that time I think orbit when we started really when we set up the company all there at that time I think arbitrum and optimism were doing about 20K transaction day something like that you know very very low number of transactions very very early days and at that time it hit us that you know look um you know if you wanted to have an app specific infrastructure uh you have two

choices you know you could either so let let me let me step back so if you're building a Dap You have mainly two choices one is you can go and Deploy on a general purpose chain so it could it could be ethereum if you want faster and you know low cost and whatnot you could go on optimism arbitrum um and that's the general purpose start so basically you would be competing for Block space with every single app that's deployed on that on that chain the second choice that you would have is an app specific route so you could go and build let's say your own chain using something like Cosmos or or Pok do and they they both had their pros and cons right in a general purpose model you benefit from composability you benefit from liquidity you benefit from user base because it's the same central

place where you have everything shared um so you have all that benefit but the problem that you have in general purpose in you you know you have issues around cast so if once you have because the resource is finite and once you have a lot of people using that Finance resource you will have gas War like situation then we have seen that time and again um you could go to you know a an app speciic route where you could sayy I'm going to build my own Cosmos chain the good thing that I get from Cosmos is I can go and customize my chain in all sorts of ways so I have my own validated Network I would have my own gas mechanism I would have my own gas token or whatnot and it's a chain that is dedicated for you so you have your own block space you can own it and

so no one else would compete uh on that chain but the main problem with with Cosmos or you know app chains at that time was that know ability to bootstrap this so you have to go and find validators first of all and uh once you have found validators you have to go and then pay them in some way right so you have to issue a token you know sometimes if the token doesn't do well no one is happy and then you know the transactions don't do well and then your network is kind of somewhat idle and so we saw uh you know rollups as a way to kind of find something in the middle in some way because the good thing about rollup was that at that time you could spin up a single sequencer you would have a bunch of contracts on on ethereum and you're good to go literally

you know because you're deriving security from eum so you don't have to worry about hundreds of validators but no one was talking about at that time hey we should go and build an app specific instance of optimism or app specific instance of arbitrum so I feel like if you build an app speciic instance of a rollup you could eliminate some of the issues that you know Cosmos and po C and some of the other other apps specific infrastructures have um and one cool idea that we had at that time is maybe not everyone needs app speciic rollups at least for a long time so uh we feel like maybe some people need these instances but maybe just for a short time so uh we came up this idea of what we call disposable rollups so essentially you uh you know you spin up an instance of a rollup and

when you use it for as long as you want when you feel like you're done with this or don't you don't need this anymore or you feel like the transactions have dried up you know there's no longer interest from the community you basically take that infrastructure down so you don't have this recurring cost of operating the network paying the validators and whatnot right so we feel like that that could be a good balance of having an app specific instance of AA but for a short period of time um so I would say that was the journey we felt like that that could be that would meet or give you the best of both words best of a general purpose chain and a best of a cosmos chain because you take away all the the cost of you know running and having validators and also the cost of having a

chain that stays there forever and then obviously you know um over time you know op stack by the way when we started there was no op stack there was no arbitrum orbit we ended up building our own stack by the way so we we built an entire implementation of course inspired by optimism and arbitrum so we built our own stack of optimistic rollup stack that allowed people to spin up these instances now we have almost deprecated it simply because now we have you knowb morbit but at that time there was nothing so we had to build our own St so that that shows you how we were in this in this theory that you will have modularity in the future even though the term wasn't quite there but we do believe that extion layers will be very different new people will people will be new kinds of execution layers and we're

really betting on that totally with you on the vision of like back in 2021 realizing that okay well if uh we have we see the cosmos model of app chains we see what's going on with arbitrum and optimism and ZK sync like we there's obviously an opport to take the Best of Both Worlds and like make app so totally hear you on all that Vision can you just like talk a little bit more about this uh Insight that these like app chains that you might want to spin up and then close down like what are the types of use cases that like were really jumping out to you that um would be met by like a temporary chain yeah so at that time this was I think right after maybe a little little bit late at that time but

this was right after the nft hype you know U you know so you feel like hey look nfts always create issues on theum right you know there's always a gas war and you know people would said you know um one solution to fix gas water is to do this fancy auction mechanism you know Dutch auction and you know whatever right and no one lik that like to adopt that model because they feel like hey if you're running a restaurant right you want a queue in front of your restaurant because if there's no there no people waiting people feel like there's there's the food is not good so you want to have that that queue and that queue basically gives you marketing so when we spoke with some of the marketing and know some of the nft project they said hey we want first come first we do not want all this

fancy auction model because all of that happens back in you know it it it you know dilutes the entire marketing effort but when people are running for it people are you know fighting to click the button first that is the hype that you want right and that's the marketing that that you want so we felt like hey nft could be a good use case and the good thing about nfts was that you only need so you need to solve the gas War like you need something that allows you to prevent gas fors that was the problem and the good thing for us was that nfts don't like sales don't last forever so you know you would start selling it today if you hyped up project you probably will be sold in seconds right you know not even minutes um and and so it was a it was the first I would say

the main driver for us maybe we were a little bit late for that because you know nfts were starting to dry up but that was the main motivation because you felt like hey if you want to eliminate gas War if you want to have first come first serve the only good way would be to have a dedicated chain that' be for nft mint event so basically your users would come so you would spin up an instance of a rollup you would do your mint event on that rollup so you basically will selling all the nfds and once all the nfds have been sold out we would take down that infrastructure because you won't need that anymore and the good thing that you had in this model is that all the NFS that were minted on that rollup will go back to a chain of your choice so for example and

this way you could enjoy all the liquid Equity benefits that comes from let's say using ethereum so you can sell that on openc you know lend that and and borrow money whatever uh so it gives you the benefits of an app chain because it's a it's a chain that is dedicated for the mint event but once the mint event is over you get all the liquidity benefits that comes from having all these nfts on ether so I would say the killer or the initial idea or motivation was nfts but you can extend up to anything right anything like games for example of course people play games there are games like people play for years and years like counter- strike and whatnot but there are many games I would say 90% of the games have a certain lifespan right you know people are excited about a game for like maybe a

year maybe 3 months 4 months and after that kind of game dies away or the there's not enough demand for that game or daily active users drops quite quickly so basically the idea would be very similar you would spin up a um an instance of this rollup you would host your game on that rollup and once your um you know game um has dried up or your user base has dried up you'll take it down and what you can do is any nfts that you had on the rollup you know coming out of the game would go back to ethereum and then you could go to your drawing board and say hey let's go and build a new version of the game and once you have built a new version you can break that uh new version of the game on a rollup and then you would bring all

the nfps back from ethereum to this rollup so it also kind of gives you a little bit of composability as well in your own setting but yeah nfts games uh airdrops for instance um voting anything that is shortterm that hu drives a huge transaction volume in a short span of time would be ideal wow you just like really unlocked like a new like Paradigm or way for me to think about rollups because yes everything that we all talk about about you know uh like removing execution from the expensive L1 keeping like the settlement guarantees cryptoeconomic security like all that stuff but a thought just occurred to me that what rollups do is allow you to create a business or protocol or something and like users can know that

even when this business is like no longer supporting this thing like I keep my uh like ownership in perpetuity and like maybe that ownership is not going to be worth much in the future but I guess there's something for me profound like when I use like my Google account and I like to track my location history like there's part of me in the back of my head that knows that I am only privileged to access this data as long as Google's interested in like allowing me to and I guess what rollups do is allow people to get just as invested as they want into like new projects that may or may not be around in a couple years but without assurity that like well even if the business fails like I can still take away what I earned doing

it and by the way not every like there's a reason why disposable as a term exists in the real world as well you have disposable cups you have disposable plates and whatever and and and and that that you know applies in the in in the you know World in this world as well because not everything is persistent know um even you know maybe we'll talk about data availability right the data availability like Celestia and even ethereum in the in the D charting that's coming up data doesn't stay there forever right so I think um there are a lot of things in the in in in the real world um and in also in the digital world which are not persistent um in a true sense of it right and there are use cases which do not even require that right so and and

you rightly mentioned dation layers when when when you move it out of eum you can imagine different things right and that's exactly what we're seeing where people are building for example eclipse is building Solana VM attached to with him you would have done that before because in in a completely you know let's say more ethic world you only see what the L1 offers you right um and and you are offered a certain VM and you're stuck with that you're offered a certain gas model you're stuck with that um for examp today it's it's ridiculous to be honest with you to when I transact on eum today or any L1 for that matter I pay a fee to the validators and I pay a fee to the underlying app so for example if I'm Trading tring on Unis swap I have to pay a fee to Unis Swap and I also pay

a fee to the validator this is kind of like saying u in the in the web 2 world that if I send a if I send a let's say tweet on on Twitter I have to pay Twitter but I also have to pay the internet service provider every single time I do that which is which is doesn't make sense right so I think this this idea that you could have um a model of rollup where gas is completely free there's no notion of gas you could do that for example you could say hey my rollup has Z Z gas what I mean by zero gas is not that the fee is zero it's like there's no notion of gas at all um and basically you would send a transaction to this to this to this rollup and the sequencer will take the transaction will process them and if

there's any fee that has to go on ethereum or Celestia or some other you know third party Network the sequence will bear that cost on behalf of user so there would be fee of using the application but there won't be two fees there won't be a fee that goes to network and the one fee that goes to uh you know the app and then somehow the user is has to deal with all of that mumbo jumbo right so I think all of that goes away if you abstract some of these things and it becomes much easier if you do that in the r upite man that's like so exciting and I think we can go for the rest of this conversation on a tangent on um like how the design space opens up but I'll try to keep us focused here so like totally hear you on where

the Insight was what the thinking was in 20121 but let's fast forward to 2024 like can you just at a super high level like walk us through through what is the alt layer product today and I guess like maybe first start with just the rollup as a service just packaging everything we've been talking about but then I really want to like dig into what modularity means and what the different kind of like choices are and and what we can build from that yeah so um a as a base it's a rollup as service so we make it easier for people to deploy rollups because you could imagine having a game builder um that who doesn't understand SE how sequencer Works they do not understand how to pick and choose let's say different da layers and whatnot so our goal is to run and manage um you

know those rollups for them and the benefit is three one is it saves takes away all the complexity from you so you don't not have to deal with all of that number one number two is um you it helps you um focus on your code expertise so for example you can if you're a game builder you would focus on game building you won't have running like in for example where uh you are still also building infa but also building app and then you don't know what infa do and then you are stuck because your Bridge got hacked and whatnot right so it it allows you to focus on your which is your code expertise and it saves you money because um you you would have to go and hire people who understand the right things you have to hire devops all of that goes away and lastly out say is

it allows you to get some extra services so for example if let's say you building a game you may want or def5 project you may want to have an oracle service now you know you would want to go to chain link and say hey chain link guys can you support us and it will take you maybe 6 months to realize that it's not going to happen so it also seemes you BD work right because um what R service can do is they have been speaking with these third party services and providers and they would say hey these guys uh would be the one you should talk to and we can hook you up with them or maybe we have already integrated that service into our offering as a package and come to us we'll Sol them so roll up service basically offers you that um and so as a service you and by

the way before I said you as I said one it would have been just ethereum or arbitrum or optimism things like that but nowadays now you can pick and choose different things right so you can pick and choose okay which stack you want to pick so you have choices for op stack arbit mbit uh you know polygon cdk ZK sync ZK stack and a bunch of other ones will come for example scroll will probably have their own stack maybe Tao will have its own stack uh so you you basically pick your um you know the first thing to pick is your your stack that you want to have then you pick your da layer so you could use ethereum you could use valadium mode so basically your own own set of notes that will serve and host the data um you could pick um you know Celestia or alternative

D layers uh you could also pick for example it's not a core component I think it will become core component very soon which is U interoperability layer so uh once you have this rollup how do you connect with some other rollup right if you go back maybe two years ago it was basically called a bridge right um and the way you would the the way Bridges would add a new chain is that you have to go to a governance you know portal and you say hey guys you know here's my fancy new chain uh could you support my chain u in your Bridge service and then there will be a governance process and maybe last for two weeks um you have to go and you know convince people and Lobby people to support your chain and then once your

chain is supported maybe at sometime you have to actually then do a voting for the asset that you want to support right so it could be oh now that I've supported polygon now let's go and see which asset on polygon will be sear on the bridge it was a painful process and so nowadays people are building these permissionless modular Bridges as well so that can automatically connect with any chain uh and you have the service that you can run yourself so I say that's also a component that's becoming much more interesting so it's not just about rollup stack but also these middleware services that would be useful right we for example also building this service where which uh you would have around decentral sequencing or shared sequencing right you could also now pick okay I don't want to just run a sequencer or single node sequencer I

would also like to pick maybe between um you know espresso or Astria or some of the other you know projects out there so I would say at a high level there are two main I would say um layers of modularity one is your U up stack and the other one is the da layer but I think there are other services that are attached for example shared sequencers Bridge providers and so on yeah and we didn't even talk about just like some of the basic stuff but even like block explorers are are one of those things where it's just like day one if you go build it yourself like good luck like maybe to call Ether scan and hope that you can get their code but I mean you might get you have to pay a huge sum of money to to get them to support you and so um I think this is

also something you know this this you you you may not know this as a dab Builder you would probably spend maybe three four months talking to all these vendors and then you realize that ether scan is not the right thing because it's it's designed or the business model surrounds around bigger projects and you may not be the right for them so yeah you know you could imagine indexes you could imagine uh nowadays people are building new won't say model but new uis for for Bridges um so you may want to go and build your own UI which may have some issues or you you know the design may not be good so now people are offering Bridge UI as a service as well right people are offering wallets as service account abstraction as a service so many of these things are also part of this broader you know modular service

and R Services help you pick and choose between these services for you yeah and I really like what you said is that like the abstraction of rollup as a service or maybe alt layer allows like your team to focus on building your app right like it abstracts away all of this like blockchain nonsense so that you can worry about like providing the service that your users want to use and like that's uh definitely like the step towards making blockchain like a core part of technology and not just this like weird little side project for people like you and me indeed indeed so okay two things I really want to make we hit on one is about um in this like multi- rollup world where there's thousands of rollups you can spin them

up on your like with with ease I want to talk about like the fragmentation of liquidity and like how we're going to solve that and then the other thing I want to talk about is um more about data availability and um and some of like the longer term stuff with that but let's stick with um fragmented liquidity for a moment so before we like really get into what what like some of the rollup uh so something like a polygon or arbitrum or a like ZK sync with their super Nets how they're looking at how this is all going to play out can you like speak a little bit on what's your personal like thoughts on is a like multi- rollup World a fragmented world and um if yes

then what are the types of things we need to build to fix that and if not like what is the fundamental misunderstanding I think so I would say that yes it's a fragmented world right um um yes you could say that everything is on ethereum if you're attaching to ethereum um but now that you know your da layers could be could be non ethereum Your settlement layer could be non ethereum people are building Sovereign rollups I think it definitely is a fragmented World um and and so there there is a need and people are actually building it there's a need to build better Bridges better liquidity mechanism where there's a shared liquidity that can be tapped into uh different rollups I think each one of the rollup stack Builders are actually doing so as you mentioned you know you know hyper chain or super chain or lxl wi

brid in polyon cdk they all are focusing on a common layer or common middleware that would serve the rups that are attached to them um uh to to share liquidity right and but what what is still not being done right now is in in a more generic way is let's say for example optimism is building its own Super chain thing right which will have let's say Shar sequen potentially maybe liquidity provisioning again Zing is building own thing cdk is building own thing I mean there's no there's no common thing that connects all of them in one way right people you have options like hyperin you have options like XLR and some of the other things but I think there's a lot more design thing that needs to go to connect these otherwise

it be kind of like Cosmos but in in in in a weird ethereum way of things which is oh here here are all the here's the cluster of all cdk chains or here's the cluster of all the op stacks chains or here's the cluster of ZK chains you don't want that yes it' be easier for you if you're op stack rollup it' be easier for you to talk to other op stack rollups but it won't be so easy if you want to connect to let's say cdk rollup from an OP stack rollup perspective so I think it's it's still early and many of things are still not built yet uh you know people are still most of the stack Builders are building some of these liquidity systems but there people building um you know systems that would do these cross domain or cross roll up stack ecosystem U you

know liquidity provisioning but I think there has to be something that needs to be built and again not in a model where people have to go and Bridge Ares from one place to other it's we have gone through that process painful there have been end number of hacks we have seen and something else needs to be done using maybe Z KCK lines you probably know better than me so correct me if I'm wrong but my understanding of what like the model that is emerging as the way this is going to get solved is like whether it's polygon cdk whether it's arbitrum um orbit I guess or the op stock like it it sounds to me like the solution is is that if you're building one of these chains on Bas layer ethereum what it's going to look like is even though you're your own Sovereign rollup you're all

settling to the same smart contract and that will allow like the this the virtual accounting which like is doesn't require expensive you know like gas transactions on Main net so like yes you're in your Sovereign cdk rollup but you can bridge over to polygon ZK evm without actually triggering an onchain uh main net like transaction and then like that's what solves the problem and so I'm wondering if like as you say that there um there needs to be some sort of solution that Bridges all of these like let's call them like hyper pools together like is that an opportunity for someone like alt layer to like walk in create large Accounts at each of these

smart contracts and then create your own virtual like accounting system to you know do a kind of similar thing but across ecosystem so we thinking about we have not built that to be honest with you but we are thinking about I think in the next two three months or maybe six months uh you would see something built you know either by us or someone else because there's a problem that you you will see and at the moment there are some op stack chains there are some arbitrium chains that are live uh but many of them are still on test net so I think in the next 6 8 months you would see many of these test Nets will go live and more will come which means that you will have to build something like this way and you don't want to touch ethereum right you know I mean you want to touch

eum in the sense that you don't want to send every transaction through eum and then come back to the rollup side as you said there has to be something that stays on the rollup side because it makes it easy and cheaper right so yeah we we we do see this as an opportunity and we are working on this but we don't have a concrete design quite yet uh what we are doing at the moment is this idea of what we call a fast finality layer so basically using I layers EVS mechanism so basically the idea is we'll build a network kind of like a middleware that takes any rollup so for example you could have cdk or you could have op stack or you could have arbit morbit they all post data or something on the CD on this Mac layer and the ABS operators will then check

and verify that um the data looks correct the state look commitment looks correct across all these things and then you would have uh liquidity mechanism built in there on that on that system uh maybe that's not the ideal design maybe we need to refine it maybe someone else will come up with a better design but I feel like there has to be a common framework that allows the cdk chains to connect with other chains including Sovereign rollups you know for that matter yeah the The Sovereign or sorry the like kind of um Master pool style like rollup as a service like that's clearly a good step forward but it's really just taking the same problem we have and like consolidating it into just smaller bigger pools and it's also not so so straightforward because and this

is by the way one of the problems with modularity right which is sometimes the system becomes so heterogeneous that you will have to think about n number of things right which is for example orbitum orbit doesn't work in the same way as uh you know optimism even though at a high level the design the same but when you go into the details of it it's not the same thing right similarly you know zking doesn't work in the same way as as uh you know cdk I da will not work in the same way as as Celestia so I think there are lot of you know these finer details that makes a little bit more complicated so that you know because in the end what you're building is kind of like a a universal system that takes heterogeneous systems and kind of builds a more homogeneous system

right and in principle in theory it looks good and you know looks easy but when you start going into the weeds of things it's it's rather painful uh not so not so straightforward no makes a lot of sense so I want to Pivot over to the data availability side of the conversation and so like we have to caveat everything that we're saying with data availability is like so new as a concept that like I guess Celestia is live but other than that like Ian da is not live even ethereum D sharding or Proto dang sharding is not live so like the Paradigm is obviously like completely up in the air but earlier you even touched on this that like once we um are in this like da world like we are looking at a world where the data that

is posted to these layers like does fall off after a certain amount of time so with Dank sharting I think it's supposed to be about a month I'm sure with the like private Solutions it's going to be more configurable but uh my first question to you is like how do you see the market shaping up to like service all the needs in this new like expiring transaction Paradigm so for example like yes I understand the transactions are going to expire from these decentralized networks but like what happens when I need that data post XPR yeah so that would still be available but that' be through let's say external you know full nodes or um other you know providers for example let's say

you might have a roll up using Celestia Celestia might delete the data eventually but you may have ether scan that is incentivized to offer that service because they are offering this ether scan like service you know where you can search so data is will still be there but it will be hosted and I guess saved for eternity you know by other providers but Celestia and the ders and ethereum for example will not store that for you and the reason main mly is you want the data to be available uh during the time you know you could challenge for example if let's say if you're optimistic R up you want to make sure that while the challenge perer is over is is still on you have the data to be able to go and use that to challenge right and so as long as that you know

your your deletion period is is longer to cover that buffer you're kind of fine and after that it's mostly on storage providers or data storage providers to going to take over and say hey now if you pay me for long enough I will keep that data forever uh and this is kind of the incentive for the likes of ether scan indexers and you know other third party Services out there you could still you the UN is not going away but someone has to store that data and that they these people have to have incentive to do so so as a role up Builder you have to incentivize the likes of ether scan the likes of I don't know graph to make sure that the DOA data is there forever for you well I guess to me like this is one of those moments where like the the

technology and the design is like so cool but some times I worry if it gets like a little bit ahead of the overall system and it just seems like almost counter ethereum to me to say like this like fundamental part of like making this system works like the the data that is in data availability layers like we're just going to hope that like the market creates some sort of incentive to like make sure that it's available when it's needed and I I feel like part of the mechanism design of ethereum is like we're never hoping anything like if it's needed we're kind of controlling it and so I I think this is one of the problems of modularity right um so I would say there two things one is um you know I

think siram from a DA was or aan Le was expressing this quite quite you know vocally that From ethereum's perspective every D is a DC right which is which is a valid point because um in the end ethereum cannot verify what goes on this all da so someone else has to basically a test that okay I have checked something and that something looks good to me and here is my my signatures and eum will only check that signature because eum has no capability to actually check external systems which means that when you're using let's say Celestia and some of the other modular components you are kind of trusting you know some of that component which so for example you do not get the same guarantees as etherum when you use an

external network right so one is this modity brings that you have to understand the different components that you're using and what security entails right um that's one thing the second thing is that it's it's it's it's complex right you know previously as from a develop perspective you are just looking at hey e offers me everything let's put my stamp on it now you have to worry about okay I have to analyze what does celestia give me uh for example if you're looking at latency for example right you could say hey I want to use my shared sequencer um I want to use uh da layer but when you actually look at the transaction life cycle it has to go through Shar sequence first then go to the executor for then the executor will then execute your transaction then it it

will put the data on some D if you add up all that latency it's going to add up it will be probably become the you know the the worst roll up in the market so you don't want that right so I think there are certain aspects that I think people need to consider um and that's one of the I guess other disadvantage of you know having a modular system because if you don't pick them correctly you will end up with a not so performant system at least and and on the one side right this is like like just straight up a business opportunity like within the realm of alt layer like you guys could very easily create a paid service that is like we will be the cold storage for data availability on like XYZ Network and like that's great but I I do wonder

if like the the way we architect this is to get as modular as possible and at each like Realm of modularity is a business opportunity like are we just like leeching value out of the system and like we're f forcing the user to like um to to pay for all of this modularity I would say so and so I said latency is one component the other one is fee because every single layer that comes into play you have to pay a fee You could argue maybe that's cheaper than if paying to ethereum in the if you add them up but we don't know that yet you know um maybe initially some of these ders will uh will subsidize the cost in different ways maybe the you know the demand will be cheaper and so

sorry the the cost will be cheaper so so maybe in the end initially at least the cost will probably be lower than ethereum but it does add a cost to it right it's not it doesn't come for um eventually so um there's there's definitely a cost to it that will again will come come from a user side you know there's only a certain period of time and you could say okay fine you I have enough money or it's it's VC funded so you know it will it will come from come from the Investments that we have gotten and that will be our user acquisition cost but at some point it has to go it has to go to some of these you know you know some of these third party vendors that will have to take take some of the cut out so yes there's a cost to it we

don't see that quite yet because again um you know it's it's all subsidized but at some point you will yeah got it so with the last few minutes here I want to just kind of shift the conversation to this new era we're entering and like how I kind of DeMark this new era is let's say like the release on mainnet or whatever but it is Igan lay and I think that igen lay was this like we can talk about how this happened but it was the moment that the market realized that there is going to be this huge opportunity for uh basically a compute that is um augmenting ethereum and so since then we've just seen so much hype and so many points and so many

reaking services and um I think like a lot of the like the vision and like the the impact between behind why this is such a meteoric has been lost uh in the conversation but I guess like my my question for you to walk us out on is like can you reflect a little bit on like why igen layer has um like what it means for the ecosystem how it's transforming things and like how you foree ethereum changing because of this like new resource of I guess like almost tradable or movable trust yes I think movable trust is is really or Bor trust is really the right term for it right um and again you can you can look at it

from a simple perspective perspective so let imagine I want to do my own chain okay that chain hosts some defy or maybe that chains is a Oracle for you know offers you know chain link like service whatever you're building some peer-to-peer system and you want to have nodes in your system that offers and offers that service uh one usual way that we have been doing so far is to say hey I'm going to give you my tokens you take those tokens and you'll get more tokens um and then you would you you have the incentive to run that what happens when the token goes down now no one has the incentive because the incentives that you're providing is not good enough right that's number one and when that token goes down the security goes down you know because in the end everything is secured by the you the

dollar collateral that's that's sitting there in the form of tokens I think I li basically says that hey um you could use ethereum because ethereum is one of the most let's say stable asset in some sense stable in the sense that you know it's not going to go down tomorrow zip to zero and it's a network that exists and you have providers there in the market that can offer these services so for example you have Valu eum Network which are huge in number by the way at the moment they could come and offer extra Services right secured by eum so this idea of borrowing security from withum to bootstrap external networks I think that's powerful right so what I has done is they building or that's kind of the vision right you you would build a marketplace where you have these stakers

that can bring their Stak assets they provide the stakes asset and those Stakes assets then get reused uh for security so that's the first uh stakeholder the second one is the validator that are serving ethereum at the moment they could be the one that could do other services and third one are these uh builders that are building new systems and new middleware services and they would say hey I want to uh I want to secure my system and can you connect together can you bring your assets together and your services together to secure this so I would say it's it's a kind of a three-party kind of model that brings these three parties together and allows you to secure your system so I'm I'm definitely very excited about IG of course there are certain risks to it I think we have debated and I think Foundation you know

have debated all on that I'm not saying there's no risk to it of course there's a risk to it I think I is also very careful about this right they're thinking about one to roll out in a model you know in a phased manner where for example uh you won't have slashing initially right so imagine if there's a big AVS tomorrow let's say Unis swap becomes its own network let's say they their AVS gets hacked or whatever someone gets slashed and then unisoft because it's Unis swof they may have the right power or the author or the you know the people behind it will basically at the social they would like to push and say hey guys can you save me like the way you saved uh when the Dow hack happen and that's dangerous PR and so there there's certain risk asso with it and I think this is what I

are careful I Le team is very careful about so for example initially uh most likely there won't be any slashing conditions uh so even if this a misbehavior you as an ABS operator you will be blocked and Frozen so you won't be able to operate anymore but your assets will not go go down so your assets will not be seized basically so they they're taking PR you know precautions uh but again we have to see how some of these things unfold when you open on the pendora pendora box in a much more open and permissionless way yeah and so to me like the the magic inside of IG layer is like I'm a node operator I'm sure many people are node operators like we're all running computers that are so much more powerful than needed for ethereum right like on purpose ethereum is supposed to have low

requirements and the idea is like okay I have this computer it has extra resources like why don't I like offer them to people who need these computational resources and like that is a match made in heaven like that is so beautiful to me that you know me as a East Saker can support future Unis swap Chain by like all that stuff makes sense I think where things like are a little bit make less sense to me and I wonder if it's much more about just like the the casino of crypto than it is about like hard economic principles but like I I feel like when when we're talking about people with stake delegating to other people that are going to be actually

operating the compute we have taken a cryptoeconomic system that where the whole point is to remove trust and directly introduce trust and the trust is not between the AVS and the operator it's between the operator and the person providing stake and so like point is is that if the operator screws something up like guess who suffers it's the person who gave stake and reputation right and reputation that's just trust so I don't know if you have any thoughts on that or Reflections on that I mean we will have to see how these um let's say LST or LRT providers are going to select the ABS right um you could say hey we are going to take avss or support avss that are not risky at

all right then someone say it's not risky there wouldn't be enough incentive right because historically and we have seen that you know the the more riskier the the asset the more um upside you potentially have right and so I think there's again go to be a balance where we have to see how these abs uh you know or at least LRT providers are going to pick the ABS that they're going to support because in the end it as you said it it it boils down to users assets so if I am if I'm taking your asset and I'm picking up a certain ABS in a way we are taking that providers is also taking a risk like they need to know what what and this again this risk modeling incentive modeling is something that I guess it has it has just started right

you know you you have I haven't seen any any any real work quite yet so I think it's going to take um at least a couple of abss to go live and then you would see people coming up okay now that this abs of is life here are the risks that you're taking here's what could happen potentially and how to mitigate them at the moment there's nothing life so you're kind of speculating without actually going deeper into okay what's the real risk because the risk would depend on the ABS right risk would depend on what these a op actually do and we don't quite have that on minute quite yet so I think it's it's still quite a bit early so I think in in then once once I goes on minut it after that maybe three months after that we'll start to see a little bit more activity

on okay yeah the rest with this ABS you should do this you should not do that it's not quite clear quite yeah and I think that's the like the game we play when we're like dancing literally on the bleeding edge of Technology that's true that's true that's true cool well I'm I could go literally four more hours I want to do one on igen layer I want to do like two more just on alt layer I want to talk about like the the new design space that's opening up but I think for the sake of all of us I will uh wrap us up here so first of all I just want to say thank you so much thank you for your generosity and thank you for for speaking with us um incredible conversation thank you very much of course and Before I Let You Go um can

you just share with the audience like where they can find you where they can find ALT lay and if they're interested in figuring out if you know customized modular rollup as a service is right for them like what's the best place to start yeah so um do follow us on Twitter so Twitter is the I guess the central place where um we push out updates and also Twitter basically retweets our account officially retweet some of the so you could also find me there um and mostly if if you if you're building in the space and if you're thinking about even even a thought of hey maybe I should think about building my own Ro up come and talk to us you know um just DM us and we would be happy to hop on a call even if you do not want to have a roll

up you would actually have an understanding of you I even wanted right because I I believe that you know BD in the crypto world is more about doing consulting rather than doing sales you know you have to understand if you're just selling it's not going to work you have to come and understand people's problems and see if what you're offering is the ideal solution for them so if you're building the space and if you're thinking about having an app specific rollup or not thinking about it and you want to say Okay can someone help me out come and come and talk to us we' be happy to have a chat uh and and hopefully we'll be able to help you out awesome man and everyone like please like take that gem that he just gave us that was so incredible that in crypto but probably in life business

development is much more about like Consulting than it is as sales and I think that's beautiful and something I'll definitely take with me so am thank you so much thanks for the time and have a good rest of your day thank you very much you too n