zkEVM L2s Will Scale Ethereum | Scroll

Transcript

this episode is brought to you by pariswap the leading aggregator to find best prices across various dexes you'll hear more about them later in the show [Music] [Applause] this is a really interesting episode of empire i think you're all gonna enjoy it we had the three founders of scroll which is one of the least known but most innovative companies in crypto they're working on zk ups um santi and i have just been talking a lot about what we want with the show and and one thing that's really important to us is we obviously want to bring you the big names we're recording an episode with olaf carlson we later this week we've got kevin o'leary who came on the show we have vance and michael from framework

we want to bring you the big names but we also want to get really into the weeds of what is coming not just in a month from now but in six to 12 months from now these episodes are going to be a little more technical than you're used to it might be with names that you've never even heard of but this is the first one of those episodes it's with scroll they came out of kind of beta and hiding for like a year and they were just heads down building it's it's phd folks it's people have worked on machine learning and ai at amazon they're absolutely brilliant uh they just raised their series a 30 million dollars from bain i know santi is an investor as well and uh yeah because of who santi invests in and just i think because of the show and because

of block works we have access to some of these names and we want to put them in front of you guys so i hope you guys enjoy this and santi i'm curious to get your take what do you think of this episode i think this is sort of gets a peek into what i do a day-to-day uh which is not just podcasting but it really is meeting people that you know from all kinds of walks super smart people you've never heard of um but are doing incredible research um and there's a number of things that i get excited about when i first met them uh you know these people you know sandy's coming from from a pretty successful venture fund she was an entrepreneur before that um one of the other founders was um at aws and their ai team um and uh and another

founder was you know pursuing his phd and doing really cutting-edge cryptography research and so to me it's always really exciting to see that because um there's i think it speaks to web3 and the primary thing that i get excited about whether we're in a bull market or bear market i always look at where are the clusters of human capital and especially smart people want to work on very difficult problems and you we've heard vitalik was talking about scalability since day one um and he's been pretty vocal around zk like it's your knowledge proof so perhaps one of the most fascinating things um that have been talked about over the years it's been very theoretical but i think

we're one step closer and it feels like we're you know a year or two away from actually seeing like really efficient practical implementations of this fascinating technology um and so it's great i mean this the other element that i love is um you know it's a team that's primarily based in china um so i've always kind of been annoyed to some extent that i haven't seen that many chinese projects and i think it's by virtue that i'm not chinese i'm not in the ground uh and friendly and listening i mean we had a great episode with jason who's based there and investing heavily and so i think this is going to be a good episode um you know because i think you'll be hearing more and more about scroll and and what they're doing uh over time

which it's not to say that's a winner take all l2 market i think there will be multiple versions and implementations of them but i think zero knowledge proves credibly really take us to like supporting billions of users i'm gonna go on a limb saying billions uh but i think this is why this this episode is important to get a better understanding of can ethereum scale which continues to be a a recurring theme especially with you know these as more applications get built and they just congest uh the block space and becomes really expensive for users but scroll and other type of versions can take us there yeah yeah like you mentioned i mean vitalik said in the medium to long term zk rollups will win out in all use cases as zk snark technology improves and

scrolls really pushing the boundaries of just scaling and zk tech with their zk evm roll up um and so we hope you enjoy you'll hear from three people here so sandy is one of the co-founders she was previously a partner at fission capital which is a crypto focused investment fund hai chen leeds engineering he was previously a senior applied scientist for the aws ai team and ya ya zhang was uh is now on leave actually from p8 uh he was getting his phd at nyu so they're absolutely brilliant uh becomes a pretty technical conversation santi kept with them pretty well i'd say santi led most of this one but yeah we hope you enjoy it if you enjoy it dm me dm santiago uh follow us follow our account on instagram excuse me on twitter and just let us know if you like

this if you don't like it yeah santi's getting on instagram now so anyways without further ado enjoy the show thanks everyone for the benefit of everyone just give us a very basic uh explanation of what is scroll yeah sure um so so scroll is a zk evm project and we're building um something what we call native zke evm and striving for full evm equivalency and we're trying to deliver exactly the same experience on ethe l1 on in a zk setup so in a much faster and more efficient way and and you know i might be familiar with this jason might be familiar with this we but we want to peel it back for a lot of the audience what exactly is when you say what is zk uh why is it so

important and interesting uh for scalability purposes yeah maybe i can i can just quickly introduce like what's a decay so zero knowledge troop is basically it's a crypto primitive which you know for uh instead of re-executing the mutation itself you can the one who execute this computation can generate mathematically congenital proof which is very distinct for example it's for for a very large computation that your proof may be just kilobytes and you can verify this proof very very efficiently and verifying this proof is equivalent to re-execution of this computation so basically you can get your result in a verifiable manner like you get your result and you get the proof and after verifying the proof you know this result

is exactly the result after you're executing this program and because of this very nice verifiability like it can fit greatly into this blockchain space because blockchain is publicly verifiability for example like if you are you broke as a transaction you know in a blockchain layer one i every node after receiving this transaction or receiving this block you need to need to re-execute all those computations inside and this is very inefficient right because you have that that much redundancy and the more nodes you have the more decentralized you have the more computation you need to do so it's it's very inefficient to do that and so the idea behind the therapy is that we can leverage this thickness of of zika snark or zoonotic proof to compress this large

computation of chain and then on chain you just need to verify this proof so the basic workflow is that you can instead of sending all the connections to a layer one network you send that to a centralized layer to node and this layer to node will generate layer two blocks and for each block it will generate some proof proving that all the transactions inside this block is correct and then it was a business proof unchained and this this proof will be verified by some on-chain smart contract and so it's mathematically equivalent to re-execution of all those all those initial transactions and then after it got verified you know like it's it means like you know it's already included so this can massively include the throughput of ethereum and also like

for each transaction the gathering will be amortized because it will only be the verification c and uh like amortized over i think the ones on the transaction including this proof so hope that okay that's that's a good understanding maybe uh to start like folks might be familiar with roll-ups you know we talk about like optimism and arbitrarily these are some implementations of of types of roll-ups um how does how does your like how does scroll actually work and how does it compare against something like optimism or arbitrarily yeah so basically if you are like so so there are basically two categories for layers one is optimistic throughout and the other is zero up right so they basically all like the

the high-level workflow is very similar like you send the transaction to a layer 2 node and it will generate some proof in some format and some use profound change but the difference is that the keyword relations are another proof basically it's a it's a it's a validity proof which means after you you if you can journey to proof which means those computations are already you know correct because if they are incorrect or you you insert some some bad transaction you can't just generate the proof because your compensation is not not correct so as far as you can generate that it means the competition is correct so it's basically only relax on math and crypto but for for optimistic europe the rest are different proof scheme is called fraud proof

basically it's more like some some economic challenge game which basically the idea is that you can submit your resultant chain and because you you also upload your your road data on chain so everyone who seen this connection they can re-execute this connection on their local machine and they can they can see whether whether we get to the same result or not if we don't get the same result which means they they want to upload the results it's just it's fake or like it's there are some bad transactions and i can challenge this result and then they went into some some challenge period and they they do multiple rounds of such challenge game and then like finally they just like mix all the challenge into just one step and then the unchanged will execute

this step and know like with with practice with wrong and then punish some party so that's also the reason because you know your multi-run interaction or this challenge game takes a very long time so that's why like you need a longer time for withdrawing from from layer two to layer one in automatic one so basically it's through two different assumptions and so yeah but there are more like like in among like uh optimistic there is arbitrary and optimism and in vicki rob there are a lot of players like wiki thing software and right maybe um you know zika rollovers have been sort of all the rage and like with true academics like it is kind of like the most ambitious

and you know when even i remember talking about like the arbitrary guys like i feel like most most people agree that zika rolled should kind of the way the future to to achieve true true scalability because optimistic roles kind of have a limitation at some point um but it's always been like the tech is super fascinating but like computationally it's very difficult to actually get there and so can you just give us a i've been hearing about zika role since like since forever like you know like talking to zuko at um and a lot of like the the leading cryptographers like are really fascinated by this but it's been kind of it has felt like it's been slow uh at times um but obviously a lot of people might not appreciate all the all the research especially in academia that's been going on in this field for a long time i know

you guys come from academia so can you walk us through kind of like what has been the evolution of zero knowledge proofs um how is this this moment in time different than perhaps 2018 now like how far have we come in the state of zika roll up research and how close are we in actually seeing efficient implementations of zika rollups yeah yeah that's a very good question so basically like back in 2018 as you said you know if they are not proof is not as famous as it is right now so the most used library is elipse built by the cache team and the most famous application is do the cache where you can use your note proof to protect your your privacy and you can make some short transactions and which is really cool

and then like i think the biggest breakthrough happens in 2019 it's actually the second half year of 20 2019 where there are a lot of new protocols coming out and the biggest big sprue i think during that time is the appearance or the discovery of polynomial commitment so basically it's a phenomenal commitment is a is a crypto primary which is actually the preference happens like 10 years ago but you're like in in 2019 like cryptographers like you'll have forever find that it can greatly reduce the overhead of of particles and some some other stuff which also makes this happen is that as i mentioned like cooling overhead is very huge but we there are more and more research on ezekiel hardware

acceleration for example there's like we we use one constant asic and ipg acceleration for brewer and we also have the fastest gpu provider indentation and those implementation can be like an order of magnitude faster than previous implementation so like you got you also get very fast cooler because the cooling argument is highly power level and fourth is that you've also got some more efficient construction for recursive proof so the basic idea is that uh so for for one program if it's really large you take a very long time to generate proof right but you can cut this program into several pieces and you only need to general proof for each piece and then you can use one proof to verify multiple proofs and you know recursive miner so that you can reduce the overhead by for

each piece and then you get your final proof how do you differ from the different other implementations and you were touching on this earlier but like starkware and zk sync sure sure yeah uh uh for for software and they think we are we are very different like freedom of software is uh so they are also like we are all building some general purpose they kill up so it's basically in our league you can write a smart contract and you can deploy but we take different technical approaches and we have different compatibilities regarding to the evm for example like software if you want to they have a they have a start net which basically you can deploy a smart contract in stock.net but if you want to develop a smart contract in stock software and deploy them on start.net you need to use their domain

specific language called karlo so this is this language is designed for efficient execution in zonal proof so it's a new language it's pretty different from solidity from those you know those traditional like smart contract developers it's a totally new language with totally new features and you have to re-implement everything there and then like after you implement your program their compiler will compile this down to their their cargo assembly code and then like this cargo assembly code can run on their defined vicky virtual machine so it's a it's a different they also use a different virtual machine which is different from ethereum virtual machine it's a totally new leaky virtual machine for efficient execution for for decay freedom for like you can generally prove it more efficiently and so that's

software like basically you need to use a new language and compile that down to their virtual machine and deploy there and there is an external team who build some external compiler for them to be compatible with solidity so the idea is that solidity is touring complete and there their current is also torn complete so you you can use some compiler to compile this solidity through their current like some readable color code and then you can deploy that on on.net so it's it's somehow achieved some some in some sense it achieves like language compatible because you like for a photoready developer you can also compile but but you know it's like i think developers if they really want to deploy on start net they also need to be very familiar with cargo because it's

very very insecure for developers who don't know like you know what what by code they are deploying there and what's a low level execution detail and to be absolutely clear and fair you would have to learn cairo to interact and and start starkware or starknet you you couldn't like use evm to use the compiler and then like you have to learn this language to to deploy and um is that like a correct assumption uh so i think like this as you mentioned before like there's some a compiler group on third party like they're called warp and they can compile some part of the solidity into the cairo to give like a bootstrap like how to program that and then i think like if you want to go very deep you still need to kind of like learn like what chiral program means and

then you need to also make modification of that and then i think at this point the warp is not 100 compatible with the solidity uh we saw like a solid compiler does so you still need to uh manually look at like the compare the generator code versus your original solidity code and for leaky thing i think they they take approach which is more like evm language compatible and so they're approaching that so they also have their circuit language called called zinc and you can write link which is to write some specific circuit and they also have like because they also want to be com compatible with validity and so they want to be compatible with two languages so they are using a compiler infrastructure called rvm so this is a

very this is a commonly used compiler infrastructure in like traditional like programming language or things like that so the idea is that you can take any high-level language and compile down to some our vmir and then you can do some opportunities in this middle layer and then like you can you can you can compile that down to different like micro architecture like for example like uh like amd and intel like they have different like ether so so that's why like how like you can make different high level language be compatible with different low-light vm structure so they are reusing this and so they basically they have a compiler they also have a compiler but due by themselves to compel solidity or like low level u which is a solidity ir and then compat but but from

very high level you can just imagine solidity to this lvm ir and zinc will also compare to this error and then they will like compile this ir down to their they keep the evm by code so it's they call this bad code they can even bind code but it's different from ethereum virtual machine because it's actually a different vm design it just has this uh feature of being compatible with solidity so basically it's still like compiler and compiled compiled down to a a different byte code or different like low level style and run on different virtual machine which is different from evm so that's that's a another like so but but the good thing is that they are more compatible with solidity than software so they i think they achieve

like maybe 99 of of solidity and like things like that and so it's it's more easier for developer to use that and you they have built some plugin for like in hardhat and for for some other for developer to really easy to to use that but they feel like it's still not evm and uh yeah got it i can also like add like that to some um standpoint like the comparison between like us and then uh the zk thinker and then stockwear so i think one thing like you can take like from like a more high level way to view dsx what your circuit trying to prove like this so there's two level of things like you can prove the solidity language like that you can uh compile the length and then build like a different runtime so you know like the

solidity code like the smart contract they run on the extreme virtual machine is basically the wrong time for the how you execute your smart contracts but you can build like a different runtime in the zk circuits and then you can prove that the solidity code can be run correctly on your virtual machines on your runtime uh that's like the kind of we call that language level compatibility and then that's like how the dk sync and software is doing so they you they different uh they design a different runtime like a virtual machines that says different from the extreme virtual machines but on the other hand like we're building directly like the we're taking the extreme virtual machine specs and then writing the circuit to prove the ac

version machine can be executed correctly so we don't need to care about like how you compile your solidity into your extreme virtual machine's own instruction sets i will just directly prove that all of the uh instructions executed in the ecm virtual machine can be correctly verified inside a circuit as a key circuit how difficult was this um to put together i mean i think maybe if you can give us a an understanding of all the work that's been done in the background how the team came together um and and and how close are you when you say you know a fully kind of evm zk evm like i would assume that took years to build or isn't the process being built like is it is it already complete maybe kind of walk us through the the evolution of scroll over

you know last couple months and maybe years uh yeah i think that we have been building the zk evm for about like uh around like nine months and so far and then we're also working with a community like not only us like building the dke evm also working with the uh some applied tkp team inside extreme foundations and then some other community members uh that's because like that's the the power of like building this like open source uh you know open source way that you can have a lot of people who are interested and who want to contribute to those like they can build together um so i think it's so far like though so we're building like this so we have like all of the high-level designs of like how the zika imagery work and how to uh how you should write the circuits to

prove that but we have like a multiple circuits like a main circuit that proving that state machine uh inside the same virtual machine can be executed correctly and then all of our codes can be acted correctly and there's other like sub-circuit merge uh circuit that can prove the memory and stack all of the ram can be executed correctly and then there'll be also some parts like they're trying to prove the uh the cat check the hash function the chassis hash function can be executed correctly etc so i think that so far the progress has like we have uh i think like around like the six seventy percent of the op codes have been already uh correctly verified and then proved inside the circuits um and then uh and then we i think we're quite close to canadian like

something like they're really working the phone call we can take like a uic 20 transport uh execution trace and then that can prove correctly inside the dk uvm uh so far that's great and so um if if i have this right sandy you were a partner at fission capital you were sort of investing in blockchain um heichand you were previously at aws their ai team i believe and and yeah you were doing your phd at nyu so how did all of you guys come together and decide to build it um and and who else is on the team where are you guys based just give us an understanding of of how you work and um and how to come together sure i think we met about a year and a half ago um kind

of through mutual friends in the competitive math and ethereum community and um and once we got together we started kind of as a group kind of brainstorming what were some of the highest impact stuff that like like zk research could bring to the current kind of overall blockchain environment and to us like you know scaling is by far the biggest problem that exists in the space and you know if we could solve that that's easily like the biggest opportunity for the next decade and so so we started kind of toying around with some ideas and um you know based on some of daniel's research we decided you know maybe we could have an edge in in building something and then and then it it just all kind of spiraled as

snowballed from there i think up till this day like fast forward we have about 25 people now around the world and almost everybody is connected through friends or mutual friends um through a kind of a common interest in in research and building on the cutting edge of what's possible and also just kind of through this like value alignment i think everyone we've managed to gather so far are attracted to scroll not because of you know how high profile we are and and also not because of like you know um how much resources we have but it's more just like strongly aligned in terms of the interest in the project we're working on and also a kind of this kind of common value about building in the open source and then sharing the

code with everybody and kind of um bringing this knowledge to to to everyone i think that that's part of our founding story i think how do you it's pretty how do you think about like developing a moat when you are building truly in the open because other teams are you know are close sourced at the beginning then decide to kind of progressively open source and maybe for competitive reasons but um does this worry you um that it could be a disadvantage for you guys if you're building on the open i think part of our mode is the people we've gathered and the skill set and knowledge we have in-house and that part is very difficult to replicate and also the relationships we've built with the

academic community and i think for now like the the benefit of building in the open source far outweighs the the cons and um you know i think with with more with more people joining this community and the faster we will go and it's very difficult for someone who's kind of starting from a closed source setup to be able to you know move at the move at the speed that we're doing and i also have some experience like working previously in the open source community so i think so i think that's like not different very different from building like a whole like a protocols which you're trying to build some trust with the like a large community so like there's like the community which build

the zika evm and then you can also like even expand that's community to like the developers who like trying to build your uh new applications on your uh iron protocols so i think like the so having the open source cycle you can starting from earning your trust and then like the stuff on starting from the beginning and then you can have like more people who look at your code and who find like that piece of the technology is quite like really good and also people can also verify that can audited that uh that's looking at the piece of code and also this is somewhere like you can check uh those talented with uh developers and researchers who like to look at this piece like a code and they can be very interested and then come like that uh trying to contribute and then even like they maybe can join

you like that in the future so i don't think that's like a disa advantage of that as you said before like i think building the uh ecosystem is like the more of a sex success of the uh to define like a protocol like a blockchain uh a project so that's like the so we don't think like we worry a lot about like people working your code because like that building the community is like the more important pieces of cincinnati uh i want to get your guys's take on something so we had vance and spencer from framework on the podcast the other day uh and they believe that every kind of scaled project will eventually migrate to their own roll up right to optimize the code for their application and end up adding utility to their own token it seems like scroll is on this

kind of mission to have a general purpose roll up do you guys think that apps will end up using scroll for their own roll up or like what is this kind of what does that relationship look like in the future maybe i can explain more like so yeah so they are definitely like you know if you have special needs you can have your specialized circuits and you can deploy your specialized vehicle but you know there are multiple problems inside this diagram like for them having too many applications specifically or later on so the problem one day that you you miss the the most important thing like in d5 and the smart contract is compatibility for example like if you build something for dex you build something for for landing protocol like for each particle you build some music you're up but there is no composibility which means you know

you get some flash alone you can't you can't just you know like spend that on any other application because each proof will only generate for their own circuit execution so it's very hard to have compatibility like you know between different liquid groups so that's you know very important and also like as you as you said like you know you you can have some customized like optimization for your for your program and write some customer circuits but it's very hard and uh like it means strong expertise it needs auditing it takes you a lot more than just using one universal circuit so that's another another issue here uh and the third is that so even if we have built something like very universal we also plan to explore maybe

adding more pre-compiled to our to our zika evm ahead of ethereum and support much more than just evm like at this stage so that's our plan like yeah and uh all right everyone quick break from the show to share a big update from our friends at paraswap the best platform to stake swap trade farm and more paraswap just launched gas refunds based on how much you stake you can now get up to 100 of your gas refunded on all of your swaps on pair swap this is huge for anyone who has spent a lot of time in d fire maybe it's just starting out you know how egregiously expensive the gas transactions can get

the gas fees are ridiculous at some points in time and now you can get those entirely refunded on pair swap to participate all you need to do is stake a minimum of 500 psp big shout out to the pair swap dow for making these refunds possible really it's just it's tough to be pair swap right now they give you the best prices uh they save you money you've got this gas refund if you stake psp they've got a smooth and really user friendly interface fast swapping it's really everything that you'd want from a d5 platform if you don't use them already check out paraswap today at paraswap.io now let's get back to the show on this point um oftentimes we talk about you know the tech can be

fascinating can be superior but there is still a lot of requirement from a bd standpoint to attract projects to attract community obviously you guys building an open source is very much aligned with ethos so i think smart developers appreciate um the the open the true open source kind of ethos that you're taking but talk to us a little bit more about how do you plan a how do you see the competitive landscape and l2 environment obviously we have optimism we have arbitrarily we're gonna have uh you know star queer has there's a few implementations of stark where you have immutable x you have how do you see kind of the broadly speaking the l2 landscape evolving um over the next couple years where is your place there and how do you how do you what is success what is winning

for you guys uh in this in this kind of is it a race uh to become the best l2 do you want to focus on particular applications i'm curious kind of what goes into these internal discussions around kind of winning a market i i think for the time being we're just 100 focused on developing the the kind of the best solution and the most developer friendly and the most user-friendly solution and that's going to occupy the bulk of our time until at least like the end of this year in terms of ecosystem development i think you know a we're trying to we're hoping to attract value aligned like smart developers who recognize what we're doing and they recognize that you know being peer reviewed it means we have hopefully a higher level of security guarantees and

then the third part is that you know our kind of my my personal hope is that you know this this l2 scaling will enable a lot of kind of novel and useful cases and expand beyond the current kind of prism of finance and and game fi and enable kind of more more high frequency transactions and activity on the blockchain and um and i think some of the some of the strategies we're thinking about are are things that will kind of make onboarding a lot smoother or um kind of make the ease of access to l2 interaction a lot easier and um and obviously kind of you know

making the fees as low as possible from from a technology standpoint um but but we're hoping that all kinds of dapps will build on us like hopefully this is going to be a kind of a copy and paste kind of really easy experience to migrate and and this and the experience will be so smooth and seamless that you know everyone will want to come on can you comment a little bit on on how is it working with the ethereum foundation like what does it mean like um i think they've lent their support to a number of other projects i think including optimism for instance um but what kind of resources what kind of experience have you had with the ethereum foundation so basically like when we start scroll like our our biggest motivation is that we want to build something general purpose and research then general purpose liquor

up because you know as i mentioned like applications configura has a lot of problems of compatibility and things like that and also we want to decentralize our prover because we have done a lot of research related to the hardware acceleration so that's the two biggest thing like while we're building scroll it's a like the evm equivalency kilo up and also like we want to support decentralized pooling process to attract even more members to to make this faster and like you know general proof for us so that's two biggest motivation and i sent this draft to to barry whitehead from ethereum foundation who actually actually proposed the crap like years ago and i was talking with him and it happens to to be that the case that he was also looking into this the key ram stuff because you know one one important

road map or other game for you ceramic that it wants to start ricky snuck everything so the idea is that for example like for for even on layer one instrument like you you get some block you broadcast it into your network and every node still need to re-execute the transaction inside this block to to verify that it's correct so if you directly use the key evm in layer one which means it will totally change this you know this workflow so basically for example like you you have some miner and you generate block and you will force this miner to generate proof for this block and it will broadcast this necessary detail for for for this for this block and this proof into their network and every nodes instead of re-execution of this

transaction they just need to verify proof so basically it can scale layer one directly in layer one so that's their very ambitious goal like take one existing ethereum block and generate proof for that and you can even build something which is a minor cell like you know you can get one proof to pull the whole history for for ethereum which is very exciting and they are very interesting to look into this and but what we found more interesting is that we want to build some more efficient layer to which you know we have our own account system we generally prove our layer 2 account and submit this proven chain so that's basically the key even had two important applications and we are pretty aligned with you know because layer two is the infrastructure to scale ethereum so we really want to you know push for the development of ethereum and

so that's you know from that side we are more value aligned so this collaboration actually happened very organically like barry ask whether you want to work on this like collaborator or not and we we think okay so we definitely want to develop our circuit you know open source miner and we you know there are a lot of you know brilliant brains behind applied bkp team and you kind of for now like there were like 20 people working on in this in this repo in this in the open source minor they have around like 10 people we have around 10 people working working with them and so it's you can't imagine a company have 20 like very excellent and experienced vk researcher and working on the same problem so it's a much stronger and much powerful from the community side from the development side like so many great people are

working on the same problem and they have yeah it's even very hard to hire such a strong team like you know for a single company so that's you know it's it's really great you know it's very nice to work with so many you know nice collaborators there and we we talk as hajj mentioned we have some weekly meeting to coordinate different different process and of which we collaborate on the uh developing on this and or not and things like that so that's basically how this how this you know works and uh yeah so we are completing this but we are just leaning towards a more efficient their traversing yeah i want to ask you guys one last question on rollups before pivoting to uh talking about china for a second which is in your guys's minds what is the

maybe for folks who are a little less technical like myself what are one or two things big things that you think get unlocked if you guys are successful with scroll like how does the world look different and how does the industry look different than what it looks like today if you guys are successful and maybe sandy if you want to start here if i can put you on the spot yeah sure um i think the successful case in ten years time is um i think when you know literally billions of people are interacting with with layer twos on it on a daily basis and using it on the back end of tools that they don't even kind of mentally register as you know being blockchain based i think i think that's

the dream i think we're already starting to see some novel applications where um use cases are done at uh where where people are having like very very frequent interactions with it and it's kind of impacting their their off chain life and and we want to be a part of that to enable those things and i think you know a lot of businesses will be built on on web3 and we hope that would be built on there too and um and i think up till now the because of the financial nature of everything that that that that's been built so far um and you know and part of that is due to the exorbitant fees of ethereum up till today and so once we unleash that

possibility with with a much cheaper and faster ethereum we will likely to see a lot more adoption in the developing world and um and and also a lot more use like tools and and use cases that are like beyond finance and and beyond like entertainment but you know i think there's still you know obviously a huge amount to be built within the financial space and huge amount in the game fight nft space and and all of that will just be faster and better and cheaper um yeah so i think we're kind of at the precipice of you know the the the kind of age where where where just before like facebook and the fan companies came about it's impossible to

envisage what would come and how how these like you know applications would manifest themselves that requires a lot more experimentation and imagination and a lot more new talent coming into the space to make that happen and um i think we just want to play our role as an enabler and and kind of you know facilitate and enable those like innovations to happen i love that i think that's great uh high channel yeah would you add anything else there so i would think like they feel like they were successful i think like the effect in the few years we can enable like the 10x like even higher more super like per second on the exam then i think that's kind of defines us like as a success project for the exam i think that's and that's our goal like

also we can lower the transaction fee a lot and then without sacrificing any security guarantee of the econ itself and then enabling more people to use the exam use the crypto currencies so just to make it very concrete um of what you're saying obviously recently you saw the the ape land sale that kind of gray went to 8 000 and obviously puts in a question of course they could have done some optimizations at smart contract level but still it puts in a question can we scale how like how many users in this console say that scroll is fully deployed and it's working like how would this compare to for instance like solana uh or some of the other kind of l-ones that are claiming higher

putting more efficiency in yada yada but i'm curious like how many users how many transactions how do you measure like the the scalability of ethereum with scroll um yeah i think that's a that's a good questions like the how uh so i i think like the first for like the front if you compare the exam uh versus the some solidity like so i think ethereum has like more uh a diversity of different applications and then uh there's also like a lot of more innovations i think that's like a like a stronger uh developer uh ecosystem there and so they're like more different applications there um but i think like that but there's also a problem with the ecm is that it's not very user friendly so not

like either friendly i mean like the uh just like the transaction fee is pretty high and then the uh the waiting time the superfood is not so good for the ethereum so i think those like a problem uh same foundation like they're also trying to solve that and then uh cycle also uh what we are doing um so i think like that if we say like we enable like higher throughput and then the lower transaction vehicles i believe the ecos uh the the whole ecosystem and the developer community of around exam can come up with some more creative ways to build applications to enable like to bridge like more into the daily use cases like not only like the financial applications i was thinking that's like the one thing like i think

uh with like more infrastructure building like the with more you know kind of the this capability of support higher frequency and lower transaction cost so there will be more and more different applications to come up we had jason from folio's ventures who's also an investor in school he branched out and his fund is solely focused on capturing the opportunity in china and it's rare to find i mean perhaps dragonfly but most of the times you know statistically speaking china has a lot of a lot of really good developers and talent i think in engineers at least in web 2. you've had a lot of chinese internet companies build a ton of value perhaps a more specific question is like we talk a lot about china obviously in the context of they've been at different times pro crypto and also

against crypto how is it to to build a team how comfortable are you to be there and what's kind of your thesis of like what is china's role in wep3 and i know there's a whole host of things but just pick whatever you want to go and then we can riff off of that yeah well maybe i can start with um a kind of a first person vantage vantage point from our perspective what it's like to build scroll from the ground up and what it's like to gather a team i think one of the um one of the great things about kind of building in china today is that you know there are some great universities and producing fantastic math talents and computer science graduates every single year and

they're fairly concentrated and it's easy to get access to that talent pool so um the the minute we're doing some working on something interesting um you know we we start to attract like talent within like this core friendship group of kind of you know like math champions and and that community unexpectedly translates very well in the us where there's also a similar community and similarly in like ukraine and eastern europe and and that's how we managed to build like kind of bootstrap the project and and create the team that way and um and i think from a from an application perspective there's definitely a lot of kind of senior and

trained talent and building very user-friendly products and being able to like adapt very well to the user feedback and environment i think that part of the kind of startup science or you know the status of how to use a very user-friendly dapp that's very very um kind of well well entrenched within the developer community in china in terms of the research community there the research community in china is relatively new to web 3 i would say and unlike america where there are so many like professor projects i think a lot of the academics um talent in china have not been so closely engaged with the web3 world until recently and um part of our setup is that we're

we're open source and you know most of our a lot of our researchers are so affiliated with academic institutions and and there are no problems with them sharing their learning and sharing their practical lessons with the wider world and and that reduces our um kind of regulatory risk to of to a very large extent and um yeah and and we're very excited to work with so many kind of experienced talent from web 2 as well um who are looking to build in this to build in this environment so i think you know the last few years has been um has kind of been very kind to us in that sense yeah would you say that um like

are there any web three crypto protocols that have been started in china that you look up to because i know there's there's been a few very great researchers i think one of them would started nervous who was early on in ethereum like a core researcher and then started off nervous which is that i don't know questionable success i guess and um but i'm curious like sometimes you need like uh a figurehead that you look up to uh in your local community is that true for you guys like is there someone in china that you look up to or is it just you know um or other projects that you want to emulate maybe it's web 2 companies in china that you kind of look at as as kind of blueprints to to emulate that

success um i really thought about that question mainly because the way we're set up and the way we're organizing our our project is so different from anything we've seen obviously there's been a huge amount of like startup success and there are lots of products that we really really admire in terms of how usable they are such as tick tock and um i'm sure yeah and heightened have their their version of their kind of you know um projects that they that they look up to um but you know there are so many to name i think you know little red buck and obviously you know tencent which people use every day um and and we want to build a a product that's that that's in a similar sense that's so easy to use

that people don't even kind of think about it yeah maybe the last question for you sandy you were an investor and now a builder um what do you like more and why i guess bear markets you rather be a builder and bull markets rather be an investor until right now i guess you want to be a builder i don't know we're recording this while the world kind of is collapsing which is kind of great and also comforting i i think i was an entrepreneur before i was an investor and having been investing for about three or four years um i i really crave the kind of depth of knowledge that you're able to accumulate by working like focusing on a singular project and and that was the the kind of premises to which we started

kind of you know researching in the zk space and you know and then having met yeah and hai chen you know both being fantastic talents and just you know complete geniuses in in their respective fields you know i think i just decided this was going to be the most interesting and kind of high impact role i can play in the wider kind of ecosystem and community before we wrap it up um i'm curious if you could just comment on when can we expect to see kind of the the first implementation of the beta for squirrel and i know timelines are very fungible in the space and hard to commit to i'm not asking you to do that but just give us a general kind of understanding of of what are the next are you going to see scroll on the wild

like the next 12 months is going to be a couple years just curious yeah i was thinking like that so we'll have like the different phases like a plans like uh i think the the most recent plan we are having like that have some approval concept and then having like the people can see like how to do like the transfer using the decay rub like or even do like some uh some like the unit stuff like the style the swap and on the under the giraffe and that that target like that would be around like summer this year so it'd be like not too far but like afterwards we're also like we need to do like the more like the increment more uh full-fledged uh the k evm that's like something like a quickly like the uh adding like the in the future uh and also

uh also we're also building like trying to decentralize things like the uh citing like increases like first we want to decentralize uh the proving networks which helps you to generate the decay proofs and then like that next one will be like the decentralized sequencers uh trying to make the whole thing like that also uh have like a community like not removing the trash off like us i guess alone like and then have like all have like decentralized things like to avoid some of the censorships and then removing the trust i have like different phases i think i would be like the scene in the like the next uh one or two years okay well uh we'll have to hold you to that then okay um well this is great guys i'm very happy that you

uh came on and gave us a for everyone out there a little bit more familiarity with all the great work that's going behind the scenes um for i don't know if there's any parting kind of comments or thoughts you want to impress upon people if not would be great to understand kind of where people can learn more about scroll i know that you were had some really good resources on you know in east amsterdam but i'm curious uh if there's anything else you want to add before we wrap it up i guess our website is the is the central reference point for all of our information and um and i think one one thing to stress is that you know scroll is a really unique opportunity for for any for people to live anywhere in the world to participate in in buildings ek ebm and literally work on

the cutting edge of cryptography and engineering and also possibly like community building and and business development i think that we're blank slate and it's very um and and you know this is a hugely challenging project and we're hoping that you know we'll have more kind of value-aligned um people kind of joining joining our missed absolutely before we go a lot of times people think that coming you need to be super technical to work in web3 but it doesn't sound like that's always the case like there are plenty of opportunities to get involved something with you guys as well even even as a project that i would think of is perhaps on the more cutting

edge of research and uh and zika rolex it's so you know smart people want to work on difficult problems but you know for everyone out there you know i think it's just a reminder that you don't have to be an engineer to go work for a protocol because you have plenty of other opportunities for people to co-work and and help you guys on a number of fronts that are just not pure technical so it's kind of refreshing to see um well great guys really appreciate the time i know it's late for you out there i know you're in different parts of the world but i really appreciate all of you coming on on and wishing you all the success and i'm sure we have we'll have to host you again once uh you know you make a little more progress and and keep us updated on

um the state of zika rollovers which is fascinating and really mind-bending stuff yeah also thank you for having me thanks san diego thanks jason yeah absolutely yeah thanks everyone [Music] you