Mass Adoption Infrastructure | BNB Chain Labs | ETHDenver 2023

my name is Dylan I'm from BNB labs and today uh we wanted to talk about Mass adoption in web 3 a huge wave is coming participation we're uh discussing the next wave of adoption coming to web three uh we're going to be discussing the infrastructural improvements um with our guests uh we have blast API chain stack node reel and remind me where uh storage quick node and quick node so we have all the infrastructure here uh ready for you to listen to awesome so there's going to be some new use cases that of course we're going to be able to talk about but I want them to tell you how we're going to actually improve the performance of web3 first before we start understanding how applications are going to run a lot faster and smoother and cheaper awesome so I wanted to point the first question at um uh the big question that we have about layer two applications and layer one applications is uh app chains or blockchain applications right so node reel I wanted to ask you specifically because I know that you do this um can you distinguish the difference and tell us what you favor over application chains or blockchain applications yeah definitely um so we first of all we don't really favor one or the other um so our heuristic with supporting certain changes you know where the users where are people developing on and I would say as of now most people are developing on evm compatible chains um but app chains do have a very good use case because gaming for instance requires a lot of throughput uh so app chains allow you to control the trade-off Matrix of centralization and performance so you can create a blockchain that you know is controlled by three nodes or thousands um so that brings a lot of opportunities um yeah opportunities for games and speeding all the games up as well awesome guys um and this is actually a general question for everybody um you know do we all believe that every application should have its own blockchain um and form its own modular network of application chains what are your opinions I'll grab that one uh Ryan here from chain Stacks so I'm going to start off by saying it depends um I think the whole idea of modular blockchains is fantastic the ability of having or the idea of having app specific chains or dedicated chains is incredible it comes down to a number of things and it really comes down to what the goals of the app being built actually are um for example if you're a game you're going to want to take as much of the resources from a CPU perspective a memory perspective Etc and you really don't want to have the impact of anything else running on the Chain so having an app specific chain for that definitely makes sense now on the other hand it is great to have a bunch of complementary services and or dapps that can run together on the same chain that may add to each other it could be for cross communication purposes uh it really actually doesn't impact the performance because quite often they're just lower performance applications so I'll come back to my original statement it depends on what's being built it totally depends um anyone else Avenue cool all right that was a great um so do we do you think that the uh the way that blockchain scale today is similar to how traditional data centers scale in traditional web 2 systems um I wanted to actually bring this over to quick note because we we know the adoption of Google we know the adoption of Microsoft we know the adoption of all these things and they set the entire thing on storm so what are we doing differently what is the decentralized Internet of Things doing differently how are we scaling differently can you take that yeah I think probably the most exciting part of how web is scaling is that we can all participate right we can all take part in the upside of that scale uh when you have the ability for new people to be able to spin up infrastructure participate in the network um and participate in the upside of the rewards right that makes it something where we can all monetize our um our technical capabilities our infrastructure right as opposed to it just being someone like AWS or someone like gcp that's taking that money home right and I I think that's great especially some of these chains that really make it something that we can all manage on our own infrastructure and it doesn't have to maybe come into a server farm with you know uh these gpus that maybe some of us can't afford or some of us can't keep up with so I think that's quite exciting I think that really is a huge part of the future and that we can all participate in the upside and monetize its growth so data data can be shared but it also can be monetized as well because we're representing the same data so there's a more of an incentivization correct right and as there is more traffic on the networks right and as there is more rewards to be earned right uh that's just transaction volume that we're able to all capture and monetize together as a community awesome so now I actually wanted to get like a bullet point from each one of you um there's various levels of decentralization It kind of goes across the board across from UI across from having a solid governance model across from having better rpcs against from better file storage right uh and the way that you're Distributing these systems so in your opinion what is the major pain point for uh for for you or for the rest that you've seen and that you think needs the most attention so that we can decentralize and have mass adoption uh if you wanted to actually take this first yeah sure hi everyone I'm radhu I'm from viewer labs and blessed the way we think about it that viewer Labs is that [Music] is it is a difficult problem to to tackle actually in my opinion the two most important parts would be having the UI part and making web3 applications uh seamless to use by people outside of the web Tree in order to get Mass adoption into web3 and then the second part that I think is also important is the way we think about RPC and what you're trying to solve that blast is keeping the performance of blockchain apis and in the same time decentralizing the infrastructure so I think that is an important part because at some point uh you need the whole flow of data from a web 3 application to the blockchain to be decentralized reliable and performant but the problem until now is that any uh Enterprise solution any application would always choose performance over digest decentralization so basically what we need to do is find a way to give that performance in a centralized way that's what we try to solve at last with blast API so faster UI well with responsive rpcs creates a faster UI experience and then we'll have this yeah I mean we already have pretty high performance apis I mean everyone here allows for pretty good Solutions I mean really good Solutions uh what we try to do is keep in line with the performance they have and do it decentralized we've been doing a pretty good job until now it's still a long way to go but we think that would be the solution in the end basically awesome so I'll add to that I'll say one of the most important components uh for onboarding the masses especially as we look at the devs and the projects that are being built is going to be Simplicity and choice um if we make it too complex if we limit the choices and when I talk about choices I'm talking about access to the types of RPC endpoints uh dedicated or elastic type nodes and a choice of blockchains in particular as we look at a lot of the dapps being developed today a lot of projects are actually looking at the concept of multi-jone or multi-joon as such and if we can provide a simple easy solution for developers to be able to get access to the resources that they need across multiple chains we make the onboarding and the adoption of these projects and platforms a lot easier simpler UI and simpler interfaces that makes more sense for people yeah I I have like one of these questions with whether or not we do or don't pay gas for people and I and we just have erc4337 or something and I know that you have some pretty cool opinions but what do you think uh Julian is is is the best thing that we need to do for decentralization making it easier yeah thank you for bringing that up um so you look at the most recent plague that's affected our um Community which is the FTX you know counterparty risk with you know holding your coin on centralized exchanges um so but at the same time you know normal users are not going to want to have to do with the archaic process of generating seed phrases and doing all that and managing their own keys um so ERC 4337 token standards is going to enable account abstraction uh which I suggest everyone really look into it's really a game-changing uh primitive that is now in the crypto space um so this is going to enable two-factor authentication um key backups and a variety of other things that is really going to simplify the user experience of you know finding that equilibrium of being able to custody your own keys while still not dealing with you know uh the archaic process as I mentioned of you know dealing with all that uh you know key management and everything so it's going to be huge I like that you mentioned like 2fa um I mean I I have like a general question about 2fa if that can actually be generated in the cloud maybe I can actually start enabling 2fa maybe all my security permissions are are done by me but in a decentralized manner I think if we have more Security in a sense um then that would have more of this like trust formed between us and our users if we have also more trust it's like that different sense that we're not always thinking about or also maybe granting uh authority to other you know trusted people uh to then you know being able to be able to authenticate backups and everything so that's actually a really really good point so uh I've had a few talks here at Denver and recently we've talked about this this difference between trustless and trusted um but it always comes down to who you trust first and this is a really interesting point uh that I was speaking with a couple of Partners and I I don't know if you wanted to allude to what this trusted idea is um on your end and how do you establish more trust uh in a decent centralized application you know you can obviously do it through the UI or whatever you want but how do you establish trust between you and your user yeah I can talk about that from the perspective of storage which is a decentralized open participation cloud storage Network we're basically user trust derives from the diversity or the statistical non-correlation of the endpoints that participate in the network so you can start to create layers of trust for users by having non-correlation between the types of Hardware that's run the type of software the operating system the network connection and most critically the owner of the underlying infrastructure just like you would input all of your money into one stock or all of your eggs into one basket you don't want to put all of your data with one storage provider you want to distribute that risk across a diverse set of endpoints that are statistically non-correlated through that through that core level of of trustlessness we can create trust for users and we can start to underwrite things like a service level agreement in SLA the the durability the availability the performance of data all of that can be derived through proper incentive construction across a distributed or decentralized or open network and so what we see when it comes back to the user experience perspective is that it's critical to abstract the complexity so while the complexity drives the key benefits of performance security it underpins the economic Advantage versus centralized competitors like Amazon or Google by abstracting that complexity we're able to bring more developers into the network ecosystem so from a usability perspective we play with the S3 API so if developers using AWS as their object storage layer they just have to change their configuration file change their API key and they don't have to rewrite their file system or learn how to use new systems I think abstraction is is very key and critical when it comes to usability as a general theme data abstraction yeah not necessarily account abstraction but abstraction of the underlying mechanism of the storage service through a consumable SDK API and if you can make that compatible with the industry standard the more the better yeah so how how is it how is it important to have a diversified portfolio of of an SDK that can reach to the masses um for greater adoption like does it have to be in Java does it have to be in go does it have to be in like an Android SDK like what does it need to be in order for this yeah it needs to be compatible with the S3 API but broadly you need to meet the developers where they are so you need to support a broad range of libraries and that can be accomplished through binding your core functionality of your application of your network to a specific Library like Java like JavaScript like python Bodo 3 is a key library that powers on the AI workloads that we're compatible with there are many others and so by uh having uh more access points into a decentralized system you can increase usage scale performance and make the experience better for everyone yeah I was thinking about iOS and mobile as well it seems like I don't actually see enough of the iOS and mobile and it's like this is the only thing I use so um but but what about yourself um is it is it the rest of what we're kind of alluding to yeah absolutely so I I think of it more from a developer experience perspective right because if we put the tools in the hands of the developers of our customers at quick node to be able to make it as easy as possible for the end users then that's the net positive for the industry right so um you know one of the biggest things that we see you know even at the low cost of RPC Services is that people really struggle to be able to reduce their cost especially in times like this bear markets Etc right so they're looking to be able to access data very easily not have to go and index the whole blockchain build up an index on their side you know how many just be able to say hey what nfts does this address hold right that is not something that's easy that's you know hundreds of thousands of API calls to every single smart contract it's just completely unreasonable and that's things that we're starting to see as to how we can get this data to users more easily and also be able to get them the data that they need without maybe them having to make the calls themselves uh let us know hey we want to be fed this information we want a day to feed and us just pipe the information that they need at that time directly to them in real time awesome there's um it kind of reminds me of uh whether you're going to have a monolithic or a micro service architecture so I I there's another open-ended question but I I have a really important question on where our applications are actually stored and how they're actually going to be serviced because I think that's the misconception is I have to know how it's going to actually get distributed before I actually start writing something so the the question is and and I want to lead it with uh with blast API because I know that you really focus on on decentralization and speed so most of our applications are running on a centralized service system um and we can already think of a name of a few so um what what what excites you about what you're seeing from BNB you know that we have Greenfield coming up which is a massive decentralized storage Network um a mesh of decentralized storage networks actually so um how do you believe that we're solving for for the problem of centralized applications and moving into a more decentralized service-based approach distribution approach hey actually the question is really good the answer shouldn't be too complicated right because we've been looking a lot about what what is the paid path of data and requests throughout the whole blockchain infrastructure and infrastructure providers so we think that a good solution or an improved solution for distributed or decentralized data storage was kind of missing so far of course we had ipfs but that was from our experience and from our discussions with our customers that was pretty difficult to use still required you to run your own infrastructure so if Greenfield I'm currently not that familiar with the project just read the available public information but if Greenfield managed to find to give us a way in which to access this decentralized data the state naturalized storage easier than it might be the missing into the whole flow of the information so so making the access point more easily accessible um how do how does blast API make the the endpoint more easily accessible what what do we do at the blast API we basically Leverage The Power of independent node providers that can come into the protocol expose their nodes and we basically take their nodes and expose it to customers in the same time we have an entire protocol that make sure the nodes have the necessary performance required for any applications to run on blockchain so basically we are just trying to build on the many node providers that are out there and expose their data into the poor web web tree developers to use and Greenfield comes one layer below that with exposing decentralized data storage easily for web 3 developers and eliminated the heart also running difficult infrastructure services and and just so that if anyone's not aware of what Green Field is because you mentioned it a few times is So Greenfield is a decentralized storage Network and it's made up of many many different primary storage providers that have many different data storage secondary providers and so we have a single endpoint and we can have any uh infrastructure or any storage provider service clients directly with their data and then we have a data stored economy and that is sharing uh this data across all of the storage providers uh equally so it's using the BNB token as a utility uh for for this to happen and so we have interesting correlations with what all of us can do together as far as how we're actually servicing a large cloud of data and how that's going to actually be serviced to the rest of the D5 Builders and the application Builders and how we're going to make data more accessible and also cheaper and faster so um I know we're almost out of time but I wanted to point another question node real uh Julian we were discussing before about how you're supporting Greenfield and what exactly is nobrael doing to provide service uh to D5 Builders using node real definitely um So Greenfield is incredibly exciting and I suggest everyone go out and read the white paper when you get a chance but so uh node real is a One-Stop blockchain infrastructure provider um we do this one way through our core product which is called Mega node which is a suite of apis for RPC endpoints but we also have an entirely separate Services business where we're building out application specific blockchains ZK Roll-Ups and optimistic Roll-Ups for a lot of web 2 companies and Enterprises bringing them into the web 3 space so we're supporting Greenfield through of course our suite of apis and we'll be doing that but I think there's going to be a lot of opportunity uh to use Greenfield as kind of this Lego block uh in our services business to implement it into these blockchain solutions for web 2 companies so a lot of opportunities with you know enabling users to take ownership of their data and being able to monetize it so that's kind of like this dream we we have with web3 you know and uh I think Greenfield's really actually is going to enable this and the the possibilities are are endless be seen so New Primitives are going to emerge uh from this and yeah it's incredibly exciting well we have like now we have cloud and crud along with a smart contract so you can interface with Greenfield through a smart contract and be able to change the data delete the data move the data and it's still Cloud so it's I can finally do open AI if I wanted to I'm assuming so it's interesting of the things that we can do but then all the infrastructure actually has to be there in order for all of this to work so there is um many different use cases that I'd like to see come out of Greenfield and what we can start doing with smart contractual Cloud applications and I think that's a very exciting new topic that I don't think anyone's really kind of discussing as far as the use case is concerned uh what would you do I mean there's social fi and then there's D5 but then I actually have an idea identity that's stored in the cloud and then I can actually interface with this D5 application on my did then we can get into really exciting use cases on ZK because ZK allows for you to be either you know Anonymous or not so I would say what are some of the exciting Cloud applications that we can start thinking of whether it be games or whether it be web hosting and how you think about social public good and now I can start doing you know crowdsourcing with the open source since my my website can be on the cloud and everyone can use a Smart contract so Greenfield opens up a whole opportunity with Amazon and cloud and all these things were allowing us to do but now we actually have a decentralized way of doing it with a smart contract so I my question to you guys before we end this is what excites you about the cloud and smart contracts and what is something that you would like to see yeah I can kick that off so I think the most interesting architecture right now across software broadly is the artificial intelligence generative AI workflow uh and it's extremely well suited for decentralized cloud and for web3 you have massive large pre-trained data sets for example lion 5B which consists of 5 billion pictures 250 terabytes of data that are basically it's compressed into model weights like stable diffusion like Dolly too from open AI that enable new kinds of creative applications that drive workflows across advertising gaming all different kinds of Industries and obviously those models those machine learning models they need to be accessed and democratized public Way open AI has gotten a lot of flack for not actually being open well with the blockchain we can ensure that these models and the data that train them remain open and remain accessible through decentralized infrastructure layers and we can create architectures that are not only more globally accessible and available in terms of numbers of nodes and points of presence but are also way more economical for businesses that are creating applications around these tools so today right we see that decentralized cloud can be a tenth the price on both storage and bandwidth as compared to Amazon you see the same thing in the market for gpus for compute that's training large models and machine learning data sets and that's gonna I think in combination create a new kind of workflow that is public blockchain driven more ethical uh and more economical for for businesses across the world imagine that with the layer two yeah sub second finality and no no cost attribution right so then you can start to have hash based attribution for the inference output of these models uh it's going to create all new kinds of ways to monetize creators and help creators monetize as well as drive more efficient AI driven applications and I think we're just at the start of the the crest of the wave of what's gonna happen in that world brilliant well I think I'm going to end it there thank you so much guys for talking about your ideas and being creative with us and uh and explaining everything that you've explained and I think there's a lot that we can learn from where we're going and the improvements that we have to make and I think there's a lot to talk about and to think about so thank you so much and thank you everyone for uh attending thank you so much thank you thank you thank you everyone [Applause] [Music]