{"id":"7QHkAmaPpeE","title":"Unlocking Scalability, One Piece at a Time","description":"Blockchain scalability suffers from limitations in data storage. This session explores how modular data availability breaks down data for efficient storage and retrieval, paving the way for a more scalable future. Join experts from Nuffle, Valence, and Automata as they delve into the power of modularity and its potential to unlock a new era of secure, decentralized networks.\n\nPanel: \nAltan Tutar, Nuffle Labs\nMatthew Wang, Valence\nDeli Gong, Automata Network\n\nModerator: Dylan Hunzeker, Stanford FDCI\n\nTakeaways:\n• Modularity in Blockchain: Modularity allows different components of a blockchain system to be developed, replaced, or upgraded independently, enhancing customization and scalability.\n• Importance of Collaboration: Successful modular systems rely on collaboration among different players to create uniform standards and interfaces, reducing friction and improving user experience.\n• Challenges and Trade-offs: Building modular systems comes with challenges such as security risks and fragmentation. Understanding these trade-offs is crucial for developers.\n• Chain Abstraction: Chain abstraction simplifies user interaction with blockchain by hiding the complexities of underlying infrastructure, enabling seamless cross-chain transactions and interactions.\n• Future of Blockchain Compute: The evolution of blockchain infrastructure towards specialized, modular systems will be key to supporting complex applications and achieving wider adoption.\n\nTimestamps:\n0:00 - Introduction and Moderator Welcome\n0:48 - Panelist Introductions\n3:01 - What is Modularity and its Importance\n8:02 - Technical Advantages of Modularity vs. Monolithic Systems\n12:27 - Real-world Examples of Modularity in Blockchain\n17:22 - Incorporating Chain Abstraction in Modular Systems\n22:01 - Importance of Collaboration in Modular Blockchain Development\n26:40 - Advice for Building in the Modular Space\n30:53 - Final Thoughts and Future Outlook\n\nChain Abstraction Day at EthCC 2024 in Brussels on July 10th, bringing together NEAR and Agoric for our third installment in the year-long series.\n\nJoin NEAR's community:\nWebsite: https://near.org/\nNEAR AI: https://near.ai/\nTwitter:   / nearprotocol  \nBlog: https://near.org/blog/\nMedium:   / nearprotocol  \nGitHub: https://github.com/near \nDev Docs: https://docs.near.org/\n#NEAR #nearprotocol #NEARAI","channel_name":"NEAR Protocol","published_at":"2024-07-16T00:00:00.000Z","thumbnail_url":"https://img.youtube.com/vi/7QHkAmaPpeE/hqdefault.jpg","transcript_text":"hi so we're going to get started with our next panel I'm Dylan Olivia huner I'm the moderator I'm currently a researcher at Stanford future digital currency initiative um and we're going to chat unlocking scalability one piece at a time so today we have uh Deli gong from automat Network we have Matthew Wang from bance and Alton tutar from nephl Labs we're going to have a really Lively discussion um yeah so I'm just going to sit and we're going to get started so why don't we actually begin by everybody introducing themselves a little bit about their backgrounds and then what your firm does sure um can you guys hear me yeah okay um this is Deli I'm the co-founder of theom we've been building a lot of middleware in the past using trust execution environment So lately we're actually running a modular taxation layer trying to uh get something called like proof of machine Hood uh verified and uploaded on ching and people can verify it so um um we've been actually building this uh since uh April because we are now running everything as a uh kind of um ion layer as ABS we have a couple of our partners who are using our uh Stacks like multi over there uh so yeah it's really great to be here with a great panelist over here as well so yeah yeah nice to meet everyone uh I'm Matthew co-founder of veillance we're still semi selfish but what we're essentially trying to do is build infrastructure for bringing model driven workloads on chain so things ranging from you know data pre-processing feature encoding transformations to different flavors of model inference uh to you know data post processing all of that available from you know your smart contract by simply just calling simple functions uh my personal background is in uh software engineering and ml research spent most of my you know professional career before this in uh Quant Finance hey everybody uh good to see a lot of friends here uh I used to be at the near Foundation before just spun out of near uh 3 weeks ago so it's it's quite new Venture uh but we're we're enough for labs uh I'm the co-founder uh yeah so we buil two products the first product is the near da layer uh it's pretty much like a stateless smart contract built on near that you can post data too uh and uh actually Henry who's one of our partners is here uh they they're using Us in uh on mainnet today uh they're probably one of the top 20 uh blockchains on ALB today uh and the other product we're seeing obviously da is not enough uh the is not enough actually that's kind of the meme um we're building a ABS on top that does fast penality what you can do is essentially dump all the sequences to near you can take operators take that sequence uh run a full note or get that testation block headers at stations from uh from infura and then kind of run a node and and provide State aggregation so that's a second product we're building which is essentially allows you to uh build cross chain applications let's say automata would use it to basically do something on their rollup and you can interact with it on op so that's kind of gives a rundown of uh what we're building okay great so I'm going to start with a question that has been on my mind which is I think we can all agree that modu it means designing a system in such a way that different components or modules can be developed replaced or upgraded independently without affecting the overall system why do you think modularity as a narrative in crypto has become so popular over the last 18 months um yeah let's let's start there right uh to be honest I think it's driven by the market because like people saw the uh opportunities like in many of the verticals uh so for maybe for a roll up you have a sequencer appr over uh you have maybe a relay to do the pay Master for your like smart contract transactions and now with daer available I mean these are new really new spaces so I mean naturally people want like build their own Solutions and sometime along the way they might figure out okay I have a better way to do this so I think it's become very natural for for like for creators so they they think this is something that I can do maybe have a better solution I will just do it so that that's why we see a lot of like competition going on in these different vertigos so uh yeah I think that's that's basically the reason why modular is so popular because it's so uh like like it creates so much rooms for Innovations yeah I think if you think about modularity like the idea of modularity and composability has existed in software engineering for quite some time it's not really new or unique to web 3 and um I think when you scrutinize exactly its impact on web 3 it's very easy to tell why it's become so popular it's because it's made infrastructure customizable hot swap able infrastructure so much more easy to to build and access right as an example if I wanted to build uh some rollup that had some custom da or you know the da da was stored somewhere or if I wanted to have a custom VM these are all things that you know would make sense to to tailor design infrastructure to suit certain use cases right um without I guess quote unquote modularity you'd have to spin up your the entire stack by yourself figure out you know how to to to set up your roll up you know uh have your own or or if it's an L1 have your own validator nodes you know what VM you want to use you know where you would store data what the trust guarantees around you know each part of the stack are whereas nowadays now that you have specialized blockchains some that are designed to do you know data storage some designed to do you know like settlement you can kind of like uh you know hop on conduit or or Caldera choose uh you know what how you want to compose your stack and just like launch a roll up instantly for whatever application you're building and uh I think the the the fact that it's made so easy and accessible that we've seen this kind of cambering explosion of of uh new rups entering the ecosystem which ultimately um going back to what Deli said kind of led the market to kind of uh you know become very excited about this technology yeah I think like we started at near with uh mic Vision so we we had a pretty much built on L1 that's scalable with sharding and one of the things that we've seen while you know the op stack was taking off and last year there was the base launch uh we looked back and said you know ethereum is not going to scale on itself it's going to become a settlement layer uh and execution is going to move out of out of ethereum why not just look at our Tech stack and say hey what we can build for this uh sort of like growing ecosystem so that's how we started uh we started last year in August uh in uh within within near and said hey like what can we provide uh pretty much seeing the trends and like show like looking future and saying hey this is something that we can build uh and the other thing I guess like Matthew mentioned already everything is modular Within within software design you're using a certain API to do something you're using a microservice in AWS and you kind of will see the same sort of Trend in in in pretty much like blockchain design uh and it's great to be part of like taking what uh near built over time uh with the with the sort of like its own da data availability layer and providing that for people uh and that's why we're just kind of like hey let's let's just build something cool here and then if you build that what do you build on top it's been always the case of like um modular design brings sort of fragmentation and how do we solve that fragmentation problem uh in a very unique way so that's that's how we came to the space so I'm going to build on that Alton so what are specific technical advantages to modularity versus monolithic right and I think we've listed a few of them and I think that that's kind of like the question do J of like crypto right now cuz we're seeing salana just blow up we're we're seeing other monolithic chains doing Rel well and I know that people have their own thesises for why that will or not will not sustain yeah iing your block space is the one right so basically it's always been the it's always been the case I mean when near started it just went through the same sort of evolution um and the second part is I guess like you basically if if somebody else is using another application somewhere uh actually near design is quite nice because you can eventually uh with future with stateless validation you can take all the apps that are like blowing up with transactions and split them own chard and like we'll use the same sort of uh sort of design at uh in nle but essentially just like having that as owning that sort of block space and controlling that is very important and then essentially there was an interesting discussion today actually about s yesterday an AG day about synchronicity versus asynchronous um sort of model uh we always went through this like asynchronous model is going to be a big important step and uh kind of modularity fits nicely around there where you know different blockchains have different VMS and you can combine them differently in a very unique way with sort of like Chain absortion Solutions that obviously uh near protocol folks are still working on and uh with the with the fast finale ABS rebuilding yeah I think block space is definitely one of them um and having like your own Sovereign kind of control over your own stack right and if you think about it it's just it's just it's pretty straightforward it's just customizability right I can have my own roll up I can choose where I want to settle I can choose where I want to store data I can choose my own VM it's like buying a new car right I you know if I want to to drive fast and race I can get a nicer engine if I want you know more I guess space into the car I can get you know like a larger car right and the same idea kind of applies to rups and now you have all these specialized rups that are designed to do specialized things um you know every application can have their own you know roll up and uh I think it's obviously kind of shed a lot of light on just how important chain abstraction is in in the kind of like modular thesis where sure it's really cool that you have all these specific applications on on customized infrastructure but it really fragments and and is is detrimental to the user experience if they have to figure out oh what canonical Bridge do I have to be using to to to kind of bridge uh tokens between these these two chains or you know how would I want to kind of like uh you know if I if I wanted to interact with multiple protocols they got to like do all these things and then and have multiple wallets and it becomes difficult for uh the regular user um to to interact with with all this so definitely think uh yeah train abstraction is is quite important to kind of making sure that that all of these specialized infrastructure rups can uh you know live together in harmony right I think the question actually reminds me of the the other panel I just did yesterday uh also about like the modularity train abstraction so I think um where it same time we're entering some Cycles so uh if you like look at before we are um it's it's always like fragmented then we figure out some way to aggregate to make sure it's accessible to users if you look at like how we have also many layer ones back then and and also like when def5 was booming we have a lot of forks like that creates like fragmented liquidity as well So eventually we have aggregator try to help user to um like make it easy for user to access the liquidity and now we are like into that stage again we have uh rollups uh serving different purpose uh it's really built for um the biggest page I think for rollup is really is is nice that people can decide on their own stack they can choose what what they want for their particular applications or even they can optimize for different uh aspec for your maybe security more maybe latency maybe performance so um so I think it's it's really essential that um like coming back to the chain abstraction uh angle it's it's now right it's right now the the time for us to really figure out what's the interface that user wants uh is it really like that important for them to understand the difference between chains difference between different wallets or how you manage the keys or it should be just we should just go for like a way better interface look at what's happening in in web to right now now uh people are stop using uh like Google search uh more more and more people are start using like chat GPT for stuff right so maybe this is uh perhaps the right way to go we should like figure out way for users to um like uh have better approach to access blockchains are there other real world examples of um modularity improving efficiency and performance of blockchain networks like I know that you you mentioned um fragmentation because of defi Etc but is there any in recent memory that just someone could attach the their thinking to so that they could understand this in a more concrete way you mean the aspec on like the performers and things or you mean exactly like a time when like the the data availability did yeah right I think pretty good examples is on execution layer like when people are and maybe I can also touch a little bit on the da layer as well so on the execution layer I I've seen some development on the parallel execution right uh actually we had this kind of discussion long before when we're like when people when ler ones were trying to solve the scalability issue using sharding but now we're we're coming back to that topic again because like we really think think that the single threat model for execution is not enough we really should like try to find a better way to do the do the uh paralyzation but I also remember that U looking at the data on ethereum the paralyzation Factor isn't really uh that viable because like some analysis already showed that maybe you can at most paralyze U by by the factor of two even reaching three is difficult because it's just not available but I mean this is already a Improvement of course on rollups now you have other choices to use different VMS that might have better optimization at the execution layer they have this kind of Justus in time uh like com execution uh component so that uh so maybe more and more execution we will just guide the uh uh guide it like better optimization time uh also on the da layer there are really like really different approaches and also different guarantees like how when the data will be available and how long the data will be stored so uh these are the things I think that that's already available um and we eventually allow developer I mean actually as a as a developer who's going to use this layers they have to really understand uh what's the pros and cons and trying to figure out uh what's the best feed for the a stack yeah I think a good like mental model would have around like modularity to think about it as like you know modules that you Import and Export or some commodity that you Import and Export right um as an example if your some rup that settles on ethereum uh you are kind of importing security from uh you know uh from from ethereum you know as a settlement layer right and you use uh kind of the the merized route that you post to ethereum as a way to determine okay what is the economical chain and that's how you inherit security right when you post like da to to to to near da or or Celestia you're essentially you know saying hey I trust you that you know this is the data that the block data that we have on our rollup and you know trust that you will store it and that it will be available when the nodes want to query this data and be able to access and you know reexecute transactions for increased verifiability right and what we're working on at balance as well um as an example the kind of like commodity we want to export is kind of model driven compute right uh you know being able to pre-process data transform data feed it through different models to get some results offloading that comp computation from existing RS that aren't tailored to do the kind of compute to um you know our Network that it that can you know essentially be this co-processor sleng engine for you and do all the heavy lifting um and that's kind of the the module that we we export to them right that's the good that we export which is which is uh this sophisticated computes um that is you know secured by you know again other blockchains such as a settlement layer andda um yeah and I think that's a good mental model to think about how you are able to kind of take uh you know the good things that certain blockchains are designed for and kind of have a best of all worlds type approach where you know you pick and choose and ultimately use these you know to to your advantage uh does that apply to real world examples within blockchains or outside of blockchains or both both but I would love like outside too okay yeah so outside I mean it's pretty simple like in Industry Revolution like you basically have to split the uh production line into different specializations and eventually people sort of like evolve to have that right um for us obviously you know um Deli and uh Matthew mentioned uh modularity in a space of like rollups but I'll take it to the reaking example right I mean reaking is quite the similar uh sort of like design space uh you use the restake uh use the restake eat to do something outside of ethereum uh to pretty much like design a system to be able to do one thing uh and that's what like leaving that da aside uh we're also working on something that does specifically fast finality for uh for rollups to have access information for um pretty much is not not pre confirmations that a lot of people are talking about more like using another external validator set that uses restate eat to um to provide uh State de station so you see that example like coming from history where everything started to modulized to web 2 where AWS did the same thing to pretty much now to uh to new design spaces as well cool so um we just talked a lot about chain abstraction um how do you know do other modular systems other than agoric um incorporate chain abstraction into software architecture like do your companies do that is that like a general blanket kind of term for for what you're doing Etc I mean we are at chain abstraction day so yeah I think I think it's actually it's quite difficult actually what we are doing is trying to be multi- Ching but we felt that it's pretty difficult to to be uh ch abstractive uh simple reason is there just so there's just so many so much difference like among different chains uh often you really have to think about if it make sense for you guys to just support yet another stack uh for maybe for scalability or for for Market reason um but what we are doing maybe I have a great great example over here we have RPC product called one RPC it actually help user to U protect their IP address when you're accessing the uh different like RPC endpoint um but but the best we can do over is just like uh simplify the uh this of access uh like end point so that people can just take like access multiple like different networks from from one single endpoint but still like if like when things come to like keing management and stuff they still need to like have their own way of doing that but we do see like some development right now on like smart contract wallet and even some layer one chain already support this uh it's kind of like reduce a requirement and those reduce the like overhead on like user side to to maintain the key but still it's it's kind of difficult also I've seen like things like pass keys right these are relatively available on mobile devices or even on laptops uh it could be the it could be a new er for like people to like using it as a key to manage like your accounts across different chains uh but it's just a matter of time like when this like all the different chains can support the uh crypto scheme used by Pass key uh so far I think not many chains have the support but there's some changes over the year yeah yeah no I think uh as we've been kind of developing our you know MVP to devet to testnet we've really starting to realize just how important chain abstraction is to uh you know really make our product kind of uh work well and and interrupt well with other existing blockchains um and and we kind of like consider it like compute abstraction a a little bit uh internally where you know our end users are not so much kind of like retail folks uh trying to move money around but rather they're kind of decentralized applications that want to Leverage The compute on our Network and we've been yeah sitting down thinking about what makes sense what kind of design decisions um you know Integrations we should make to make this a universally accessible uh kind of piece of of technology that can be leveraged by any blockchain anywhere or even just in web to as well and I I think the kind of primary ways we're doing that is you know really straightforward but just integrating every single kind of cross-chain messaging application cross street messaging protocol that we can integrate with um building in like pay masters building in like regular web 2 HTP style apis um and having very uniform interfaces to to make just like our our uh software interupt well with with um these protocols yeah this is an interesting question so um I actually run the chain abstracted podcast that's sponsored by near and when you ask this question to everyone everybody has a different answer um so for for near folks I'll start with near I'll go down uh sort of t nuffle uh that's chain signatures is to be able to drive Keys uh from from a near account to let's say an evm account or anything then you be able to sign that uh sign of transaction with an NPC uh NPC network uh for us it's a bit uh unique in a sense that uh we when you give those uh State uh State attestations then other people can use that to build uh a multi-chain app right so in the front end it does it actually using a couple of different chains but um the front end should just be like unified even though the um the system is very asynchronous so uh but first some folks right like I mean if you go to Espresso and other ones that do shared sequencing they will have a different answer like they would say oh you know we we're building a shared sequencer so that's kind of their chain abstraction for polygon folks it's uh it's the agare that that allows to do that so there's a lot of teams that are working on different solutions I don't think there's one answer to that it's just like uh depends on what the developers are looking for cool how important do you think collaboration is among different modular SL da kind of layers and would you say you're competitive each of you are competitive to some of the biggest ones that aren't sitting on stage that we all know yeah it's a funny question so we get asked a lot like are you guys competing with IG da uh and uh just because we're in the igen ecosystem um the way we see this is very collaborative so uh obviously um NDA was the first thing that the design space that we knew so we've built that uh eventually what we'll do for example for our P finale layer is to integrate other da layers right so every system we would have prefer actually just using NDA because it's everything is like uh in the same design fashion but uh some some folks will use different da layers just by Design and just by alignment with the different teams uh so we basically start to kind of like build out and add uh let's say a DA eventually Celestia or others to be able to for everyone to run that and run like basically give those statea testation so um the other thing is like I guess it's more collaborative like modular design space uh allowed us to kind of like work with tier one teams uh that wasn't the case before uh so while we were actually at the BD team in the near Foundation layer one's almost competed uh like to get those applications um we don't have that design anymore uh so which we don't have that like sort of um the first thing that comes to mind is not competition it's collaboration and we see that with everyone that we work with like igir even with like polygon folks that historically were kind of considered competitors now we work on things together yeah I mean uh I I I think uh I'd be surprised if anyone kind of uh disagreed with the fact that collaboration is important for modularity but uh the entire like kind of ethos of modularity is that you know it's only useful because you have collaboration right it's like you know a DA layer by itself is is not particularly valuable settlement layer by itself is not particularly valuable it's only valuable when you are able to take kind of the best of each stack and piece them together right and in that way we need to kind of collaborate together um in an introver vertical fashion I would say primarily to create just like abstractions and interfaces that are uniform and make sense and allow people to kind of piece these things together so they can pick and choose whatever you know stack or infrastructure they want to to kind of create their rule so yeah um I mean at least BU as someone building a network that's kind of how I've seen the space from the outside I would I would also add to that that it's totally necessary to have a very large Network effect for these things to work and that requires collaboration for the state that web 3 is in right now so you um yeah actually the Tipping Point is really like when um when there's a standard that people can really stick to so then the competition become more I would say fair but before that uh it's really just uh kind of wide space like because like as a first mover you don't really know like sometime what's the uh what's the customer what customer want right so you have to talk to them get their requirement and and this kind of design often is driven by that and then uh when like multiple players come in people are start having more collaborations they will also start doing more user education so that their customer can have better understanding on the stack they're building and eventually there will be a standardization Pro progress so maybe a couple da will like just eventually work together to just figure out the common standard so that their upstream and downstream could just work more smoothly with them otherwise it will be a lot of frictions yeah speaking of Standards what you said actually reminded me of a meme I I do completely agree that like having these standards makes sense and you know it's quite important but there's this uh like Stickman meme that I saw uh that's like quite popular is like you know two people are having a conversation one's like you know there are currently 14 different standards to do a certain thing and then the other person's like no this is crazy we must create a standard that unifies all the standards and then the last slide is kind of like oh there are now like you know 15 competing standards um and I think it's kind of an interesting way you know if you apply that kind of thinking to what we've seen in the kind of modular ecosystem you know you have all these like rollups right and now you have uh you know multiple players trying to do some degree of you know shared sequencing or or aggregation and now you have like all these different players trying to launch their own like agers of sorts and you have many different competing standards um which I I think is is is in some degree it's a good thing because you know it's a healthy amount of competition people are always trying to improve um and and and build better technology but uh it's also just quite funny to point out that um I I do think collaboration is important but you definitely see a lot of uh competition to try to kind of aggregate and create this like Network effect for themselves as well what advice do you have to people trying to build in the modular space I know this is General but it's just it's useful yeah I would uh to start I would say um I think so as someone who's kind of like more recently gone into two up3 and and have come from a more uh I guess like software Tech uh Finance focused industry um it's interesting to see how short of a halflife a lot of crypto companies have um you know it seems like every cycle the the kind of like companies that are around and and are considered relevant it's almost like a new set every time um and in that way I think one piece of advice that you know I'm you know I constantly remind myself and and I would give to people building in this ecosystem is uh you know you your technology and your infrastructure is only as strong as its weakest link so I think it's important to choose the right partners that you know have withstood the test of time um have withstood you know like multiple cycles and to keep the Lindy effect in mind right it's like cuz if you have some You Know da layer if you have some settlement layer or if you have some part of your stack that ultimately soon becomes unmaintained becomes obsolete or you know wor potentially exploited or whatnot you know then your stack immediately becomes vulnerable and susceptible to to to everything so I think yeah choosing your partners choosing what kind of infrastructure to include within your stack I think is actually something extremely vital that a lot of existing players don't think too much about because um oh this person's giving me a grant I'm going to build on them right uh that's like the kind of major of the poll that I've been seeing in the space um so yeah just something to keep in mind I think in general um I mean a suggestion to the like new boers no matter you're building a stack or using the stack is just to you really have to understand the tech itself whether it's truly helpful to you or you're just like getting aligned in some way right so uh on that actually uh you really have to spend a lot of time researching like different vertical also like if you are deciding which vertical to enter you probably have to really see like which stage you are at or what kind of stage this particular vertigal is at so if it's really early stage you have a lot of risk because no one has explored this area it might not like worse the risk of like just doing like spend capital or spend time on it to to try to create a standard or even like like create some Market options um you could sometime you could just follow a kind of do a follow approach so you can just uh try to like fight a space that there already maybe one or two players but there still maybe your room so you can get in and try to just uh find your particular Edge like compared to them so um for for users who are using uh this kind of mod get stack I think I agree with what me that so um like some some of the points especially um uh choing the stack Wis because like there will be a vendor locking risk if you if you find find that okay you probably can't change it afterwards then you should really like make the right commitment otherwise it would be quite difficult for you to make any like um like move like move after like Market the condition condition change as so yeah um yeah so challenges uh I think I think Matthew and uh Delhi have been doing a great job here just like looking and studying all this because we've been friends for a while um but essentially understanding like um what what do you need a rollup for uh and then saying like you know you look at op stack you look at arbitrum you look at uh ziki sync uh you know polygon and understanding like uh do do I really need a you know optimistic roll up or do I need a you know zare roll up a lot of folks can't afford like if you're early startup can't afford a ZK roll up currently uh so they might just go on up with like an OP um sort of op and arbitrum stack uh that's the one Challenge on the on the technical side and the other side is I guess like uh just understanding the security uh parameters of a system that you're using uh so let's say um we for for near for example like we've been running for a while so that's been kind of the that's been kind of the security sort of uh pitch but like understanding sort of uh what are the tradeoffs I'm making if I'm using it one system or the another or maybe if I'm using ethereum uh itself uh versus uh versus other Solutions so there's always trade-offs uh as a as a Founder as a builder and really understanding those parameters and making that decision accordingly perfect so um any last kind of comments that we want to make I think we're going to wrap up any anything that we're excited about or anything yeah I can just mention I think we we chatted with Del yesterday about this as well I think one thing people don't realize like people think about train abstraction is from a user persective user experience perspective but a lot of people don't realize the security assumptions that comes with right uh with you let's in our case as well like if you're using rollup a to roll up B what ends up happening if something happens on rollup a uh so that's something to definitely think about uh just like because a lot of things are built on multi- sigs right so like understanding the teams you know building there's still it's trust Li but you're still building trust with some organizations right I mean you trust op sequencer and you trust like everything so still need to understand like these uh sort of security assumptions of of of chain abstractions pretty important yeah something Dean always talks about is that like when something is modul it's almost always less secure and people just don't realize that yeah yeah I think like one thing is uh if you look if you like zoom out and think about what the blockchain compute space will look like in you know not a year or two but rather five years 7 years 10 years from now um I I think throughout the the past couple years you've really seen this trend of like all these like applications coming on you know like there have been a lot more applications being developed and to support these applications and to support kind of new users people been developing customized infrastructure or you know their own rollups right to have their own like Sovereign block space and whatnot um and I I think in the next couple years right if we really want to see blockchain computes if we really want to see decentralized applications slowly gain market share and take over um you know or you know eventually be a competitor to uh our trafi counterparts we really do need a a lot of kind of like customized infrastructure to do many different things right infrastructure designed to do uh sophisticated comp computes whether it's AI or ml or some degree of optimization things like this right convex optimization for routing you know a lot of these things could be very important for applications that are developed in a decentralized fashion and to do that we need yeah like a lot of uh players and and people who want to build this stuff uh both applications and infra so ultimately to make everything work together in a conservative fashion uh I do I do think chain abstraction is going to be an increasingly important topic that people talk about uh both you know from a user standpoint and from a developer standpoint um so people can leverage what's already been developed to kind of Empower uh their use cases on the application layer yeah actually building in like building Ching abstraction is quite a fun Journey because along the way you find there's a lot of blockers especially coming from like standards coming from like some kind of incompatibility with with layers but I think as a developer I think what you could do is um just like speak the word out like make sure that people hear you and uh they will be Chang is going on and sometime you don't really have to build a perfect solution you just need a solution that works maybe for maybe over 50 of the chains or standards then along the way there will be like other people joining you and trying to like work with you collaboratively and to figure out like what what's the eventual uh like outcome or like the result of this so um yeah it's it's it's going to be really interesting and it's necessary as well all right well thank you so much this was a great panel thank thank you thank you guys thank you thank you","transcript_synced_at":"2026-11-03T05:44:55.351Z","segments":[{"start_seconds":8,"text":"hi so we're going to get started with our next panel I'm Dylan Olivia huner I'm the moderator I'm currently a","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=8s"},{"start_seconds":16,"text":"researcher at Stanford future digital currency initiative um and we're going to chat unlocking scalability one piece at a","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=16s"},{"start_seconds":24,"text":"time so today we have uh Deli gong from automat Network we have Matthew Wang","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=24s"},{"start_seconds":32,"text":"from bance and Alton tutar from nephl Labs we're going to have a really Lively","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=32s"},{"start_seconds":38,"text":"discussion um yeah so I'm just going to sit and we're going to get started so why don't we actually begin by everybody","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=38s"},{"start_seconds":48,"text":"introducing themselves a little bit about their backgrounds and then what your firm does sure um can you guys hear me yeah okay um this is Deli I'm the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=48s"},{"start_seconds":57,"text":"co-founder of theom we've been building a lot of middleware in the past using trust execution environment So lately we're actually running a modular","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=57s"},{"start_seconds":65,"text":"taxation layer trying to uh get something called like proof of machine Hood uh verified and uploaded on ching and people can verify it so um um we've","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=65s"},{"start_seconds":74,"text":"been actually building this uh since uh April because we are now running everything as a uh kind of um ion layer as ABS we have a couple of our partners","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=74s"},{"start_seconds":83,"text":"who are using our uh Stacks like multi over there uh so yeah it's really great to be here with a great panelist over","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=83s"},{"start_seconds":91,"text":"here as well so yeah yeah nice to meet everyone uh I'm Matthew co-founder of veillance we're still semi selfish but what we're","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=91s"},{"start_seconds":98,"text":"essentially trying to do is build infrastructure for bringing model driven workloads on chain so things ranging from you know data pre-processing","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=98s"},{"start_seconds":106,"text":"feature encoding transformations to different flavors of model inference uh to you know data post processing all of that available from you know your smart","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=106s"},{"start_seconds":114,"text":"contract by simply just calling simple functions uh my personal background is in uh software engineering and ml research spent most of my you know","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=114s"},{"start_seconds":122,"text":"professional career before this in uh Quant Finance hey everybody uh good to see a lot of friends here uh I used to be at","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=122s"},{"start_seconds":131,"text":"the near Foundation before just spun out of near uh 3 weeks ago so it's it's quite new Venture uh but we're we're","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=131s"},{"start_seconds":139,"text":"enough for labs uh I'm the co-founder uh yeah so we buil two products the first product is the near da layer uh it's pretty much like a stateless smart","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=139s"},{"start_seconds":147,"text":"contract built on near that you can post data too uh and uh actually Henry who's one of our partners is here uh they","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=147s"},{"start_seconds":155,"text":"they're using Us in uh on mainnet today uh they're probably one of the top 20 uh blockchains on ALB today uh and the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=155s"},{"start_seconds":163,"text":"other product we're seeing obviously da is not enough uh the is not enough actually that's kind of the meme um we're building a ABS on top that does","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=163s"},{"start_seconds":171,"text":"fast penality what you can do is essentially dump all the sequences to near you can take operators take that sequence uh run a full note or get that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=171s"},{"start_seconds":179,"text":"testation block headers at stations from uh from infura and then kind of run a node and and provide State aggregation so that's a second product we're","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=179s"},{"start_seconds":187,"text":"building which is essentially allows you to uh build cross chain applications let's say automata would use it to basically do something on their rollup","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=187s"},{"start_seconds":194,"text":"and you can interact with it on op so that's kind of gives a rundown of uh what we're building okay great so I'm going to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=194s"},{"start_seconds":203,"text":"start with a question that has been on my mind which is I think we can all agree that modu it","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=203s"},{"start_seconds":210,"text":"means designing a system in such a way that different components or modules can be developed replaced or upgraded independently without affecting the overall","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=210s"},{"start_seconds":218,"text":"system why do you think modularity as a narrative in crypto has become so popular over the last 18 months um yeah","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=218s"},{"start_seconds":227,"text":"let's let's start there right uh to be honest I think it's driven by the market because like people saw the uh opportunities like in many of the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=227s"},{"start_seconds":235,"text":"verticals uh so for maybe for a roll up you have a sequencer appr over uh you have maybe a relay to do the pay Master","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=235s"},{"start_seconds":243,"text":"for your like smart contract transactions and now with daer available I mean these are new really new spaces so I mean naturally people want like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=243s"},{"start_seconds":251,"text":"build their own Solutions and sometime along the way they might figure out okay I have a better way to do this so I think it's become very natural for for","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=251s"},{"start_seconds":259,"text":"like for creators so they they think this is something that I can do maybe have a better solution I will just do it so that that's why we see a lot of like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=259s"},{"start_seconds":267,"text":"competition going on in these different vertigos so uh yeah I think that's that's basically the reason why modular is so popular because it's so uh like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=267s"},{"start_seconds":276,"text":"like it creates so much rooms for Innovations yeah I think if you think about modularity like the idea of modularity and composability has existed in","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=276s"},{"start_seconds":284,"text":"software engineering for quite some time it's not really new or unique to web 3 and um I think when you scrutinize","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=284s"},{"start_seconds":292,"text":"exactly its impact on web 3 it's very easy to tell why it's become so popular it's because it's made infrastructure customizable hot swap able","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=292s"},{"start_seconds":300,"text":"infrastructure so much more easy to to build and access right as an example if I wanted to build uh some rollup that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=300s"},{"start_seconds":307,"text":"had some custom da or you know the da da was stored somewhere or if I wanted to have a custom VM these are all things that you know would make sense to to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=307s"},{"start_seconds":315,"text":"tailor design infrastructure to suit certain use cases right um without I guess quote unquote modularity you'd have to spin up your the entire stack by","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=315s"},{"start_seconds":322,"text":"yourself figure out you know how to to to set up your roll up you know uh have your own or or if it's an L1 have your own validator nodes you know what VM you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=322s"},{"start_seconds":331,"text":"want to use you know where you would store data what the trust guarantees around you know each part of the stack are whereas nowadays now that you have","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=331s"},{"start_seconds":339,"text":"specialized blockchains some that are designed to do you know data storage some designed to do you know like settlement you can kind of like uh you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=339s"},{"start_seconds":347,"text":"know hop on conduit or or Caldera choose uh you know what how you want to compose your stack and just like launch a roll up instantly for whatever application","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=347s"},{"start_seconds":355,"text":"you're building and uh I think the the the fact that it's made so easy and accessible that we've seen this kind of cambering explosion of of uh new rups","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=355s"},{"start_seconds":364,"text":"entering the ecosystem which ultimately um going back to what Deli said kind of led the market to kind of uh you know become very excited about this","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=364s"},{"start_seconds":373,"text":"technology yeah I think like we started at near with uh mic Vision so we we had a pretty much built on L1 that's scalable with sharding and one of the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=373s"},{"start_seconds":382,"text":"things that we've seen while you know the op stack was taking off and last year there was the base launch uh we looked back and said you know ethereum","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=382s"},{"start_seconds":389,"text":"is not going to scale on itself it's going to become a settlement layer uh and execution is going to move out of out of ethereum why not just look at our","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=389s"},{"start_seconds":396,"text":"Tech stack and say hey what we can build for this uh sort of like growing ecosystem so that's how we started uh we started last year in August uh in uh","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=396s"},{"start_seconds":405,"text":"within within near and said hey like what can we provide uh pretty much seeing the trends and like show like looking future and saying hey this is","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=405s"},{"start_seconds":413,"text":"something that we can build uh and the other thing I guess like Matthew mentioned already everything is modular Within within software design you're","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=413s"},{"start_seconds":420,"text":"using a certain API to do something you're using a microservice in AWS and you kind of will see the same sort of Trend in in in pretty much like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=420s"},{"start_seconds":429,"text":"blockchain design uh and it's great to be part of like taking what uh near built over time uh with the with the sort of like its own da data","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=429s"},{"start_seconds":437,"text":"availability layer and providing that for people uh and that's why we're just kind of like hey let's let's just build something cool here and then if you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=437s"},{"start_seconds":444,"text":"build that what do you build on top it's been always the case of like um modular design brings sort of fragmentation and how do we solve that fragmentation","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=444s"},{"start_seconds":452,"text":"problem uh in a very unique way so that's that's how we came to the space so I'm going to build on that Alton so","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=452s"},{"start_seconds":460,"text":"what are specific technical advantages to modularity versus monolithic right and I think we've listed a few of them","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=460s"},{"start_seconds":468,"text":"and I think that that's kind of like the question do J of like crypto right now cuz we're seeing salana just blow up","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=468s"},{"start_seconds":476,"text":"we're we're seeing other monolithic chains doing Rel well and I know that people have their own thesises for why that will or not will not sustain yeah","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=476s"},{"start_seconds":485,"text":"iing your block space is the one right so basically it's always been the it's always been the case I mean when near started it just went through the same sort of evolution um and the second part","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=485s"},{"start_seconds":494,"text":"is I guess like you basically if if somebody else is using another application somewhere uh actually near design is quite nice because you can","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=494s"},{"start_seconds":502,"text":"eventually uh with future with stateless validation you can take all the apps that are like blowing up with transactions and split them own chard","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=502s"},{"start_seconds":510,"text":"and like we'll use the same sort of uh sort of design at uh in nle but essentially just like having that as owning that sort of block space and","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=510s"},{"start_seconds":518,"text":"controlling that is very important and then essentially there was an interesting discussion today actually about s yesterday an AG day about","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=518s"},{"start_seconds":525,"text":"synchronicity versus asynchronous um sort of model uh we always went through this like asynchronous model is going to be a big important step and uh kind of","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=525s"},{"start_seconds":534,"text":"modularity fits nicely around there where you know different blockchains have different VMS and you can combine them differently in a very unique way with sort of like Chain absortion","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=534s"},{"start_seconds":542,"text":"Solutions that obviously uh near protocol folks are still working on and uh with the with the fast finale ABS rebuilding yeah I think block space is","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=542s"},{"start_seconds":551,"text":"definitely one of them um and having like your own Sovereign kind of control over your own stack right and if you think about it it's just it's just it's pretty straightforward it's just","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=551s"},{"start_seconds":559,"text":"customizability right I can have my own roll up I can choose where I want to settle I can choose where I want to store data I can choose my own VM it's like buying a new car right I you know","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=559s"},{"start_seconds":568,"text":"if I want to to drive fast and race I can get a nicer engine if I want you know more I guess space into the car I can get you know like a larger car right","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=568s"},{"start_seconds":577,"text":"and the same idea kind of applies to rups and now you have all these specialized rups that are designed to do specialized things um you know every","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=577s"},{"start_seconds":584,"text":"application can have their own you know roll up and uh I think it's obviously kind of shed a lot of light on just how important chain abstraction is in in the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=584s"},{"start_seconds":593,"text":"kind of like modular thesis where sure it's really cool that you have all these specific applications on on customized infrastructure but it really fragments","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=593s"},{"start_seconds":602,"text":"and and is is detrimental to the user experience if they have to figure out oh what canonical Bridge do I have to be using to to to kind of bridge uh tokens","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=602s"},{"start_seconds":609,"text":"between these these two chains or you know how would I want to kind of like uh you know if I if I wanted to interact with multiple protocols they got to like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=609s"},{"start_seconds":617,"text":"do all these things and then and have multiple wallets and it becomes difficult for uh the regular user um to to interact with with all this so","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=617s"},{"start_seconds":625,"text":"definitely think uh yeah train abstraction is is quite important to kind of making sure that that all of these specialized infrastructure rups can uh you know live together in","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=625s"},{"start_seconds":634,"text":"harmony right I think the question actually reminds me of the the other panel I just did yesterday uh also about like the modularity train abstraction so","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=634s"},{"start_seconds":642,"text":"I think um where it same time we're entering some Cycles so uh if you like look at before we are um it's it's","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=642s"},{"start_seconds":649,"text":"always like fragmented then we figure out some way to aggregate to make sure it's accessible to users if you look at like how we have also many layer ones","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=649s"},{"start_seconds":658,"text":"back then and and also like when def5 was booming we have a lot of forks like that creates like fragmented liquidity as well So eventually we have aggregator","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=658s"},{"start_seconds":667,"text":"try to help user to um like make it easy for user to access the liquidity and now we are like into that stage again we have uh rollups uh serving different","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=667s"},{"start_seconds":675,"text":"purpose uh it's really built for um the biggest page I think for rollup is really is is nice that people can decide on their own stack they can choose what","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=675s"},{"start_seconds":684,"text":"what they want for their particular applications or even they can optimize for different uh aspec for your maybe security more maybe latency maybe","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=684s"},{"start_seconds":692,"text":"performance so um so I think it's it's really essential that um like coming back to the chain abstraction uh angle","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=692s"},{"start_seconds":699,"text":"it's it's now right it's right now the the time for us to really figure out what's the interface that user wants uh is it really like that important for","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=699s"},{"start_seconds":707,"text":"them to understand the difference between chains difference between different wallets or how you manage the keys or it should be just we should just","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=707s"},{"start_seconds":714,"text":"go for like a way better interface look at what's happening in in web to right now now uh people are stop using uh like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=714s"},{"start_seconds":721,"text":"Google search uh more more and more people are start using like chat GPT for stuff right so maybe this is uh perhaps the right way to go we should like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=721s"},{"start_seconds":729,"text":"figure out way for users to um like uh have better approach to access blockchains are there other real world","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=729s"},{"start_seconds":738,"text":"examples of um modularity improving efficiency and performance of blockchain networks like I know that you you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=738s"},{"start_seconds":745,"text":"mentioned um fragmentation because of defi Etc but is there any in recent memory that just","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=745s"},{"start_seconds":753,"text":"someone could attach the their thinking to so that they could understand this in a more concrete","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=753s"},{"start_seconds":760,"text":"way you mean the aspec on like the performers and things or you mean exactly like a time when like the the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=760s"},{"start_seconds":768,"text":"data availability did yeah right I think pretty good examples is on execution layer like when people are and maybe I can also touch a little bit on the da","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=768s"},{"start_seconds":776,"text":"layer as well so on the execution layer I I've seen some development on the parallel execution right uh actually we had this kind of discussion long before","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=776s"},{"start_seconds":784,"text":"when we're like when people when ler ones were trying to solve the scalability issue using sharding but now we're we're coming back to that topic","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=784s"},{"start_seconds":792,"text":"again because like we really think think that the single threat model for execution is not enough we really should like try to find a better way to do the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=792s"},{"start_seconds":800,"text":"do the uh paralyzation but I also remember that U looking at the data on ethereum the paralyzation Factor isn't","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=800s"},{"start_seconds":807,"text":"really uh that viable because like some analysis already showed that maybe you can at most paralyze U by by the factor","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=807s"},{"start_seconds":815,"text":"of two even reaching three is difficult because it's just not available but I mean this is already a Improvement of course on rollups now you have other","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=815s"},{"start_seconds":823,"text":"choices to use different VMS that might have better optimization at the execution layer they have this kind of Justus in time uh like com execution uh","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=823s"},{"start_seconds":832,"text":"component so that uh so maybe more and more execution we will just guide the uh uh guide it like better optimization time","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=832s"},{"start_seconds":839,"text":"uh also on the da layer there are really like really different approaches and also different guarantees like how when the data will be available and how long","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=839s"},{"start_seconds":847,"text":"the data will be stored so uh these are the things I think that that's already available um and we eventually allow","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=847s"},{"start_seconds":855,"text":"developer I mean actually as a as a developer who's going to use this layers they have to really understand uh what's the pros and cons and trying to figure","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=855s"},{"start_seconds":862,"text":"out uh what's the best feed for the a stack yeah I think a good like mental model would have around like modularity to think about it as like you know","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=862s"},{"start_seconds":870,"text":"modules that you Import and Export or some commodity that you Import and Export right um as an example if your some rup that settles on ethereum uh you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=870s"},{"start_seconds":878,"text":"are kind of importing security from uh you know uh from from ethereum you know as a settlement layer right and you use uh kind of the the merized route that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=878s"},{"start_seconds":886,"text":"you post to ethereum as a way to determine okay what is the economical chain and that's how you inherit security right when you post like da to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=886s"},{"start_seconds":894,"text":"to to to near da or or Celestia you're essentially you know saying hey I trust you that you know this is the data that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=894s"},{"start_seconds":902,"text":"the block data that we have on our rollup and you know trust that you will store it and that it will be available when the nodes want to query this data and be able to access and you know","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=902s"},{"start_seconds":909,"text":"reexecute transactions for increased verifiability right and what we're working on at balance as well um as an example the kind of like commodity we want to export is kind of model driven","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=909s"},{"start_seconds":918,"text":"compute right uh you know being able to pre-process data transform data feed it through different models to get some results offloading that comp computation","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=918s"},{"start_seconds":926,"text":"from existing RS that aren't tailored to do the kind of compute to um you know our Network that it that can you know essentially be this co-processor sleng","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=926s"},{"start_seconds":935,"text":"engine for you and do all the heavy lifting um and that's kind of the the module that we we export to them right that's the good that we export which is","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=935s"},{"start_seconds":943,"text":"which is uh this sophisticated computes um that is you know secured by you know again other blockchains such as a settlement layer andda um yeah and I","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=943s"},{"start_seconds":952,"text":"think that's a good mental model to think about how you are able to kind of take uh you know the good things that certain blockchains are designed for and","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=952s"},{"start_seconds":960,"text":"kind of have a best of all worlds type approach where you know you pick and choose and ultimately use these you know to to your advantage uh does that apply","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=960s"},{"start_seconds":968,"text":"to real world examples within blockchains or outside of blockchains or both both but I would love like outside too okay yeah so outside I mean it's","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=968s"},{"start_seconds":976,"text":"pretty simple like in Industry Revolution like you basically have to split the uh production line into different specializations and eventually people sort of like evolve to have that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=976s"},{"start_seconds":985,"text":"right um for us obviously you know um Deli and uh Matthew mentioned uh modularity in a space of like rollups but I'll take it to the reaking example","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=985s"},{"start_seconds":993,"text":"right I mean reaking is quite the similar uh sort of like design space uh you use the restake uh use the restake","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=993s"},{"start_seconds":1001,"text":"eat to do something outside of ethereum uh to pretty much like design a system to be able to do one thing uh and that's","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1001s"},{"start_seconds":1008,"text":"what like leaving that da aside uh we're also working on something that does specifically fast finality for uh for rollups to have access information for","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1008s"},{"start_seconds":1017,"text":"um pretty much is not not pre confirmations that a lot of people are talking about more like using another external validator set that uses restate","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1017s"},{"start_seconds":1024,"text":"eat to um to provide uh State de station so you see that example like coming from history where everything started to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1024s"},{"start_seconds":1031,"text":"modulized to web 2 where AWS did the same thing to pretty much now to uh to new design spaces as","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1031s"},{"start_seconds":1039,"text":"well cool so um we just talked a lot about chain","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1039s"},{"start_seconds":1045,"text":"abstraction um how do you know do other modular systems other","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1045s"},{"start_seconds":1053,"text":"than agoric um incorporate chain abstraction into software architecture like do your companies do that is that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1053s"},{"start_seconds":1061,"text":"like a general blanket kind of term for for what you're doing Etc I mean we are at chain abstraction day","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1061s"},{"start_seconds":1068,"text":"so yeah I think I think it's actually it's quite difficult actually what we are doing is trying to be multi- Ching but we felt that it's pretty difficult","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1068s"},{"start_seconds":1076,"text":"to to be uh ch abstractive uh simple reason is there just so there's just so many so much difference like among different chains uh often you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1076s"},{"start_seconds":1085,"text":"really have to think about if it make sense for you guys to just support yet another stack uh for maybe for scalability or for for Market reason um","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1085s"},{"start_seconds":1094,"text":"but what we are doing maybe I have a great great example over here we have RPC product called one RPC it actually help user to U protect their IP address","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1094s"},{"start_seconds":1103,"text":"when you're accessing the uh different like RPC endpoint um but but the best we can do over is just like uh simplify the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1103s"},{"start_seconds":1110,"text":"uh this of access uh like end point so that people can just take like access multiple like different networks from from one single endpoint but still like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1110s"},{"start_seconds":1119,"text":"if like when things come to like keing management and stuff they still need to like have their own way of doing that but we do see like some development","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1119s"},{"start_seconds":1126,"text":"right now on like smart contract wallet and even some layer one chain already support this uh it's kind of like reduce a requirement and those reduce the like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1126s"},{"start_seconds":1134,"text":"overhead on like user side to to maintain the key but still it's it's kind of difficult also I've seen like things like pass keys right these are","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1134s"},{"start_seconds":1141,"text":"relatively available on mobile devices or even on laptops uh it could be the it could be a new er for like people to like using it as a key to manage like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1141s"},{"start_seconds":1150,"text":"your accounts across different chains uh but it's just a matter of time like when this like all the different chains can support the uh crypto scheme used by","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1150s"},{"start_seconds":1158,"text":"Pass key uh so far I think not many chains have the support but there's some changes over the year yeah yeah no I think uh as we've been","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1158s"},{"start_seconds":1166,"text":"kind of developing our you know MVP to devet to testnet we've really starting to realize just how important chain abstraction is to uh you know really","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1166s"},{"start_seconds":1175,"text":"make our product kind of uh work well and and interrupt well with other existing blockchains um and and we kind of like consider it like compute","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1175s"},{"start_seconds":1183,"text":"abstraction a a little bit uh internally where you know our end users are not so much kind of like retail folks uh trying to move money around but rather they're","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1183s"},{"start_seconds":1192,"text":"kind of decentralized applications that want to Leverage The compute on our Network and we've been yeah sitting down thinking about what makes sense what","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1192s"},{"start_seconds":1199,"text":"kind of design decisions um you know Integrations we should make to make this a universally accessible uh kind of","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1199s"},{"start_seconds":1206,"text":"piece of of technology that can be leveraged by any blockchain anywhere or even just in web to as well and I I","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1206s"},{"start_seconds":1213,"text":"think the kind of primary ways we're doing that is you know really straightforward but just integrating every single kind of cross-chain","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1213s"},{"start_seconds":1221,"text":"messaging application cross street messaging protocol that we can integrate with um building in like pay masters building in like regular web 2 HTP style","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1221s"},{"start_seconds":1229,"text":"apis um and having very uniform interfaces to to make just like our our uh software interupt well with with um","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1229s"},{"start_seconds":1237,"text":"these protocols yeah this is an interesting question so um I actually run the chain abstracted podcast that's sponsored by","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1237s"},{"start_seconds":1244,"text":"near and when you ask this question to everyone everybody has a different answer um so for for near folks I'll start with near I'll go down uh sort of","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1244s"},{"start_seconds":1252,"text":"t nuffle uh that's chain signatures is to be able to drive Keys uh from from a near account to let's say an evm account","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1252s"},{"start_seconds":1260,"text":"or anything then you be able to sign that uh sign of transaction with an NPC uh NPC network uh for us it's a bit uh","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1260s"},{"start_seconds":1267,"text":"unique in a sense that uh we when you give those uh State uh State attestations then other people can use that to build uh a multi-chain app right","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1267s"},{"start_seconds":1276,"text":"so in the front end it does it actually using a couple of different chains but um the front end should just be like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1276s"},{"start_seconds":1283,"text":"unified even though the um the system is very asynchronous so uh but first some folks right like I mean if you go to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1283s"},{"start_seconds":1290,"text":"Espresso and other ones that do shared sequencing they will have a different answer like they would say oh you know we we're building a shared sequencer so that's kind of their chain abstraction","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1290s"},{"start_seconds":1298,"text":"for polygon folks it's uh it's the agare that that allows to do that so there's a lot of teams that are working on different solutions I don't think there's one answer to that it's just","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1298s"},{"start_seconds":1307,"text":"like uh depends on what the developers are looking for cool how important do you think collaboration is among different modular","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1307s"},{"start_seconds":1316,"text":"SL da kind of layers and would you say you're competitive each of you are competitive to some of the biggest ones","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1316s"},{"start_seconds":1325,"text":"that aren't sitting on stage that we all know yeah it's a funny question so we get asked a lot like are you guys competing with IG da uh and uh just","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1325s"},{"start_seconds":1332,"text":"because we're in the igen ecosystem um the way we see this is very collaborative so uh obviously um NDA was","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1332s"},{"start_seconds":1339,"text":"the first thing that the design space that we knew so we've built that uh eventually what we'll do for example for our P finale layer is to integrate other","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1339s"},{"start_seconds":1346,"text":"da layers right so every system we would have prefer actually just using NDA because it's everything is like uh in the same design fashion but uh some","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1346s"},{"start_seconds":1355,"text":"some folks will use different da layers just by Design and just by alignment with the different teams uh so we basically start to kind of like build","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1355s"},{"start_seconds":1363,"text":"out and add uh let's say a DA eventually Celestia or others to be able to for everyone to run that and run like basically give those statea testation so","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1363s"},{"start_seconds":1372,"text":"um the other thing is like I guess it's more collaborative like modular design space uh allowed us to kind of like work","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1372s"},{"start_seconds":1379,"text":"with tier one teams uh that wasn't the case before uh so while we were actually at the BD team in the near Foundation","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1379s"},{"start_seconds":1386,"text":"layer one's almost competed uh like to get those applications um we don't have that design anymore uh so which we don't have that like sort of um the first","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1386s"},{"start_seconds":1395,"text":"thing that comes to mind is not competition it's collaboration and we see that with everyone that we work with like igir even with like polygon folks that historically were kind of","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1395s"},{"start_seconds":1404,"text":"considered competitors now we work on things together yeah I mean uh I I I think uh I'd be","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1404s"},{"start_seconds":1411,"text":"surprised if anyone kind of uh disagreed with the fact that collaboration is important for modularity but uh the entire like kind of ethos of modularity","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1411s"},{"start_seconds":1420,"text":"is that you know it's only useful because you have collaboration right it's like you know a DA layer by itself is is not particularly valuable settlement layer by itself is not","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1420s"},{"start_seconds":1428,"text":"particularly valuable it's only valuable when you are able to take kind of the best of each stack and piece them together right and in that way we need","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1428s"},{"start_seconds":1437,"text":"to kind of collaborate together um in an introver vertical fashion I would say primarily to create just like abstractions and interfaces that are","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1437s"},{"start_seconds":1444,"text":"uniform and make sense and allow people to kind of piece these things together so they can pick and choose whatever you know stack or infrastructure they want to to kind of create their rule so yeah","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1444s"},{"start_seconds":1453,"text":"um I mean at least BU as someone building a network that's kind of how I've seen the space from the outside I would I would also add to that that it's","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1453s"},{"start_seconds":1463,"text":"totally necessary to have a very large Network effect for these things to work and that requires collaboration for the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1463s"},{"start_seconds":1470,"text":"state that web 3 is in right now so you um yeah actually the Tipping Point is really like when um when","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1470s"},{"start_seconds":1478,"text":"there's a standard that people can really stick to so then the competition become more I would say fair but before that uh it's really just uh kind of wide","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1478s"},{"start_seconds":1487,"text":"space like because like as a first mover you don't really know like sometime what's the uh what's the customer what customer want right so you have to talk","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1487s"},{"start_seconds":1495,"text":"to them get their requirement and and this kind of design often is driven by that and then uh when like multiple players come in people are start having","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1495s"},{"start_seconds":1503,"text":"more collaborations they will also start doing more user education so that their customer can have better understanding on the stack they're building and eventually there will be a","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1503s"},{"start_seconds":1511,"text":"standardization Pro progress so maybe a couple da will like just eventually work together to just figure out the common","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1511s"},{"start_seconds":1519,"text":"standard so that their upstream and downstream could just work more smoothly with them otherwise it will be a lot of frictions yeah speaking of Standards","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1519s"},{"start_seconds":1527,"text":"what you said actually reminded me of a meme I I do completely agree that like having these standards makes sense and you know it's quite important but there's this uh like Stickman meme that","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1527s"},{"start_seconds":1536,"text":"I saw uh that's like quite popular is like you know two people are having a conversation one's like you know there are currently 14 different standards to do a certain thing and then the other","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1536s"},{"start_seconds":1545,"text":"person's like no this is crazy we must create a standard that unifies all the standards and then the last slide is kind of like oh there are now like you know 15 competing standards um and I","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1545s"},{"start_seconds":1553,"text":"think it's kind of an interesting way you know if you apply that kind of thinking to what we've seen in the kind of modular ecosystem you know you have all these like rollups right and now you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1553s"},{"start_seconds":1562,"text":"have uh you know multiple players trying to do some degree of you know shared sequencing or or aggregation and now you have like all these different players","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1562s"},{"start_seconds":1571,"text":"trying to launch their own like agers of sorts and you have many different competing standards um which I I think is is is in some degree it's a good","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1571s"},{"start_seconds":1579,"text":"thing because you know it's a healthy amount of competition people are always trying to improve um and and and build better technology but uh it's also just","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1579s"},{"start_seconds":1586,"text":"quite funny to point out that um I I do think collaboration is important but you definitely see a lot of uh competition to try to kind of aggregate and create","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1586s"},{"start_seconds":1594,"text":"this like Network effect for themselves as well what advice do you have to people trying to build in the modular space I","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1594s"},{"start_seconds":1603,"text":"know this is General but it's just it's useful yeah I would uh to start I would","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1603s"},{"start_seconds":1609,"text":"say um I think so as someone who's kind of like more recently gone into two up3 and and have come from a more uh I guess","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1609s"},{"start_seconds":1618,"text":"like software Tech uh Finance focused industry um it's interesting to see how short of a halflife a lot of crypto","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1618s"},{"start_seconds":1626,"text":"companies have um you know it seems like every cycle the the kind of like companies that are around and and are considered relevant it's almost like a","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1626s"},{"start_seconds":1634,"text":"new set every time um and in that way I think one piece of advice that you know I'm you know I constantly remind myself and and I would give to people building","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1634s"},{"start_seconds":1643,"text":"in this ecosystem is uh you know you your technology and your infrastructure is only as strong as its weakest link so","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1643s"},{"start_seconds":1650,"text":"I think it's important to choose the right partners that you know have withstood the test of time um have withstood you know like multiple cycles","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1650s"},{"start_seconds":1658,"text":"and to keep the Lindy effect in mind right it's like cuz if you have some You Know da layer if you have some settlement layer or if you have some part of your stack that ultimately soon","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1658s"},{"start_seconds":1667,"text":"becomes unmaintained becomes obsolete or you know wor potentially exploited or whatnot you know then your stack immediately becomes vulnerable and","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1667s"},{"start_seconds":1675,"text":"susceptible to to to everything so I think yeah choosing your partners choosing what kind of infrastructure to include within your stack I think is","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1675s"},{"start_seconds":1683,"text":"actually something extremely vital that a lot of existing players don't think too much about because um oh this person's giving me a grant I'm going to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1683s"},{"start_seconds":1691,"text":"build on them right uh that's like the kind of major of the poll that I've been seeing in the space um so yeah just something to keep in","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1691s"},{"start_seconds":1698,"text":"mind I think in general um I mean a suggestion to the like new boers no matter you're building a stack or using the stack is just to you really have to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1698s"},{"start_seconds":1706,"text":"understand the tech itself whether it's truly helpful to you or you're just like getting aligned in some way right so uh on that actually uh you really have to","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1706s"},{"start_seconds":1715,"text":"spend a lot of time researching like different vertical also like if you are deciding which vertical to enter you probably have to really see like which","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1715s"},{"start_seconds":1724,"text":"stage you are at or what kind of stage this particular vertigal is at so if it's really early stage you have a lot of risk because no one has explored this","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1724s"},{"start_seconds":1732,"text":"area it might not like worse the risk of like just doing like spend capital or spend time on it to to try to create a standard or even like like create some","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1732s"},{"start_seconds":1740,"text":"Market options um you could sometime you could just follow a kind of do a follow approach so you can just uh try to like fight a space that there already maybe","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1740s"},{"start_seconds":1749,"text":"one or two players but there still maybe your room so you can get in and try to just uh find your particular Edge like compared to them so um for for users who","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1749s"},{"start_seconds":1758,"text":"are using uh this kind of mod get stack I think I agree with what me that so um like some some of the points especially","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1758s"},{"start_seconds":1766,"text":"um uh choing the stack Wis because like there will be a vendor locking risk if you if you find find that okay you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1766s"},{"start_seconds":1773,"text":"probably can't change it afterwards then you should really like make the right commitment otherwise it would be quite difficult for you to make any like um","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1773s"},{"start_seconds":1781,"text":"like move like move after like Market the condition condition change as so yeah um yeah so challenges uh I think I think Matthew and uh Delhi have been","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1781s"},{"start_seconds":1790,"text":"doing a great job here just like looking and studying all this because we've been friends for a while um but essentially understanding like um what what do you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1790s"},{"start_seconds":1798,"text":"need a rollup for uh and then saying like you know you look at op stack you look at arbitrum you look at uh ziki sync uh you know polygon and","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1798s"},{"start_seconds":1806,"text":"understanding like uh do do I really need a you know optimistic roll up or do I need a you know zare roll up a lot of folks can't afford like if you're early","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1806s"},{"start_seconds":1814,"text":"startup can't afford a ZK roll up currently uh so they might just go on up with like an OP um sort of op and arbitrum stack uh that's the one","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1814s"},{"start_seconds":1822,"text":"Challenge on the on the technical side and the other side is I guess like uh just understanding the security uh parameters of a system that you're using","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1822s"},{"start_seconds":1830,"text":"uh so let's say um we for for near for example like we've been running for a while so that's been kind of the that's been kind of the security sort of uh","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1830s"},{"start_seconds":1838,"text":"pitch but like understanding sort of uh what are the tradeoffs I'm making if I'm using it one system or the another or maybe if I'm using ethereum uh itself uh","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1838s"},{"start_seconds":1847,"text":"versus uh versus other Solutions so there's always trade-offs uh as a as a Founder as a builder and really understanding those parameters and making that decision","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1847s"},{"start_seconds":1856,"text":"accordingly perfect so um any last kind of comments that we want to make I think we're going to wrap","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1856s"},{"start_seconds":1865,"text":"up any anything that we're excited about or anything yeah I can just mention I think","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1865s"},{"start_seconds":1873,"text":"we we chatted with Del yesterday about this as well I think one thing people don't realize like people think about train abstraction is from a user","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1873s"},{"start_seconds":1879,"text":"persective user experience perspective but a lot of people don't realize the security assumptions that comes with right uh with you let's in our case as","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1879s"},{"start_seconds":1887,"text":"well like if you're using rollup a to roll up B what ends up happening if something happens on rollup a uh so that's something to definitely think about uh just like because a lot of","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1887s"},{"start_seconds":1896,"text":"things are built on multi- sigs right so like understanding the teams you know building there's still it's trust Li but you're still building trust with some","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1896s"},{"start_seconds":1903,"text":"organizations right I mean you trust op sequencer and you trust like everything so still need to understand like these uh sort of security assumptions of of of","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1903s"},{"start_seconds":1912,"text":"chain abstractions pretty important yeah something Dean always talks about is that like when something is modul it's almost always less secure and people","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1912s"},{"start_seconds":1920,"text":"just don't realize that yeah yeah I think like one thing is uh if you look if you like zoom out and think about what the blockchain compute space will","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1920s"},{"start_seconds":1929,"text":"look like in you know not a year or two but rather five years 7 years 10 years from now um I I think throughout the the past couple years you've really seen","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1929s"},{"start_seconds":1937,"text":"this trend of like all these like applications coming on you know like there have been a lot more applications being developed and to support these","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1937s"},{"start_seconds":1945,"text":"applications and to support kind of new users people been developing customized infrastructure or you know their own rollups right to have their own like","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1945s"},{"start_seconds":1952,"text":"Sovereign block space and whatnot um and I I think in the next couple years right if we really want to see blockchain computes if we really want to see","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1952s"},{"start_seconds":1960,"text":"decentralized applications slowly gain market share and take over um you know or you know eventually be a competitor","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1960s"},{"start_seconds":1967,"text":"to uh our trafi counterparts we really do need a a lot of kind of like customized infrastructure to do many different things right infrastructure","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1967s"},{"start_seconds":1975,"text":"designed to do uh sophisticated comp computes whether it's AI or ml or some degree of optimization things like this right convex optimization for routing","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1975s"},{"start_seconds":1984,"text":"you know a lot of these things could be very important for applications that are developed in a decentralized fashion and to do that we need yeah like a lot of uh","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1984s"},{"start_seconds":1993,"text":"players and and people who want to build this stuff uh both applications and infra so ultimately to make everything work together in a conservative fashion","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=1993s"},{"start_seconds":2001,"text":"uh I do I do think chain abstraction is going to be an increasingly important topic that people talk about uh both you know from a user standpoint and from a","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2001s"},{"start_seconds":2009,"text":"developer standpoint um so people can leverage what's already been developed to kind of Empower uh their use cases on the application layer yeah actually","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2009s"},{"start_seconds":2018,"text":"building in like building Ching abstraction is quite a fun Journey because along the way you find there's a lot of blockers especially coming from","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2018s"},{"start_seconds":2026,"text":"like standards coming from like some kind of incompatibility with with layers but I think as a developer I think what you could do is um just like speak the","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2026s"},{"start_seconds":2034,"text":"word out like make sure that people hear you and uh they will be Chang is going on and sometime you don't really have to build a perfect solution you just need a solution that works maybe for maybe over","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2034s"},{"start_seconds":2043,"text":"50 of the chains or standards then along the way there will be like other people joining you and trying to like work with you collaboratively and to figure out","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2043s"},{"start_seconds":2050,"text":"like what what's the eventual uh like outcome or like the result of this so um yeah it's it's it's going to be really","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2050s"},{"start_seconds":2057,"text":"interesting and it's necessary as well all right well thank you so much this was a great panel thank thank you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2057s"},{"start_seconds":2065,"text":"thank you guys thank you thank you","url":"https://www.youtube.com/watch?v=7QHkAmaPpeE&t=2065s"}],"youtube_url":"https://youtube.com/watch?v=7QHkAmaPpeE","url":"https://near-pedia.xyz/videos/7QHkAmaPpeE"}