Web3 for Cloud Native Developers Why When How Why again David Aronchick

so i'm dave varanczyk i co-lead uh research development at protocol labs um and i'm probably unlike most of you i am very new to the uh web3 space i previously led pm for kubernetes i help found the kubeflow project so i really um put together a presentation on the things that i have been learning over the past few months as i've come up to speed on this space um because i i'm telling you there's a lot and especially for cloud native developers where you are looking at the majority developers in in the web 1 and web 2 worlds um you know this is something that i would have liked to see when we came through now when you hear web 3 often times the first thing you do is here you know it's web3 it's inevitable it's just enormous it's coming um but there's not really a great answer as to why you may throw out that question to the world and you'll get 700 answers but they're all very very specific to the use cases and things like that and so um i think that there's a flow that people need to go through as they build their web 3 apps and understand exactly what they're getting in for and why they're doing it so to start you know i'd like to tell this story why would you want to get into web3 so let's say you had an app out there in the world that you wanted to build it was killer it's globally distributed has unprecedented resources and will affect everyone and in order to do that you have to do all of the things you see here source and provision resources create mechanisms for sharing code use internet standards for backwards compatibility ship and maintain their apps in highly leveraged way so on and so forth and you know there's only centralized services in order to achieve that uh does anyone want to guess the year of the stuff i just described so it was these guys right anyone know who these are so that's google right that's the first version of google and they had to go and build google right what did that involve well the first thing they did was deploy a whole bunch of centralized services and you can argue especially if you know how google is built and they've been very very public about it um they are one of the largest you know quote unquote decentralized apps in the world internally there's no single points of failure every machine every node every compute operates on a zero trust basis so and so forth and over time they built that decentralization but it all started from that centralized resource and they didn't decentralize on day one now the funny part is exactly what you said everything old is new again you know over time things have been progressing from very very centralized single things vertical stacks all of those kind of things into far more decentralized things and there are lots of benefits for that not just from a security perspective from a fault tolerance perspective reliability performance things like that there's benefits to decentralization that's really one of the core things that web 3 allows for it allows for this very low trust environment or zero trust ideally with decentralization and resolving these things on each step now i want to be very clear about these uh these things this is just my take on it you can argue that on any given square any column here there are many many technologies many that continue to pursue to this day but these are kind of like some ideas about what might happen web 3 is really one path for what's next and i think this is something that that really only came um you know to fruition for me or understanding for me as i dove deeper here like there's nothing to say that you need to be a web three maximalist on day one can we get there can everything be web3 sure but that's not going to happen tomorrow and it's certainly if we require that that's going to exclude a whole bunch of web 2 developers so that's why what does it look like well so again i come from the cloud native web 2 world right of kubernetes and other things like that and cloud native apps are great right here in just four commands i can spin up an entire stack of web 2 workloads to provide a you know simple um a simple story simple e-commerce payment transactions everything storing state in a database storing logs appropriately so and so forth four commands that's really pretty impressive uh and i add i can even add that extra money flag there at the bottom to give me you know make it uh extra profitable um but you know are cloud native apps great especially for the things that we care about well you know to do those four things there i had to rely on a whole bunch of external services many of which i will not have audited deeply i have to trust for better or worse these services to provide me um uh reliable things that they haven't been backdoored that they are you know doing things that they say um that someone out there has done security audits on them and so on that's a lot um and so this comes to web three's core thing you know who do you trust and the ideal statement is no one except there's a lot to build if you're going to build a truly trustless app and we're talking about reinventing everything back to the you know root certificates provided you know at your data center i mean it's just so much stuff and so what i think the the essence of web 3 is it allows you to choose who you trust now like i said you can do everything you saw there right this list is real and it's there's probably even more than this but um i can assure you say assuredly there's definitely more than this um but you can choose to say hey you know what i'm okay i'm gonna have to trust intel or amd that they're gonna produce a you know reliable cpu okay that's something i'm gonna trust i'm going to trust my data center provider to not have you know the nsa reading my switches great i'm going to trust that to you regardless what web3 does is allows you to choose who you trust but it puts the onus on you to choose wisely and that's the key right um you need to be able to swap that out if you decide that this is something you no longer want to trust that you're ready to take on that work but like i said that make no mistake the cloud native people who are doing it earlier they're trusting a ridiculous amount of people too but at the end of the day they're able to deliver value very very quickly and they've made that trade off themselves okay so that's what how do you go about web3 well so let me lay out what that app looks like that very simple e-commerce app you have uh dns you have a front-end that might have authentication cdn and search built-in as separate services you have a back-end that has payment processing uh state and uh state storage and log generation this is like 101 stuff this is a you know very very straightforward things now the nice part is every one of those has a web3 solution today right you can go out and do that um but you know if you do that you know you can get very excited about it um if you do that you're taking on all those workloads and many of them may not need exactly what you need or meet you know may not be areas where you really care that much hey you know what for better or worse i i'm going to use stripe because they've negotiated payment processing in fiat you know all over the world that's fine um you know at the end of the day i have their throat to choke if the payment processing doesn't go through they're done or you could say hey you know what i'm in a space where um it has historically been censored or other things like that i'm actually not going to trust them and i'm going to figure out my own solution for that that's something you can pick and choose and the key is to pick and choose based on your needs so like i said every one of those can have a web3 solution but it's really up to you to make a decision on a one-on-one basis so for example let's say i need free immutable image hosting right you can use something like nft.storage let's say i want to improve my content delivery speed i can use cloudflare and their distributed weight web gateway which already supports several browsers brave and others um to download content on your own uh let's say i need extremely large permanent storage i can use ipfs uh publicly auditable db for uh whatever reason you can orbit db or maybe i want to use edge deployments i can use lib pdp and again i'm just selecting in the things that are that i need i don't have to eat the entire buffet at once i can say oh you know what it's it's a brunch i'm going to have a pano shock a lot okay and i like to say this um line as well tech changes people don't and these are kind of my immutable rules of tech we don't need to go through them because it's a very short time but but the truth is is that people don't want to run their own infrastructure this is moxie's whole point people want more control uh they want to save money and they want the easiest solution that solves their needs and no more and these really are in stack rank order in my opinion um for example they will pay more money if it means they don't have to run their own infrastructure and i think these kind of trade-offs are really true and just because we're in this brand new ecosystem i don't think anything has changed i think web 3 is the same thing that we saw before and it's just up to us to accept it or not okay so finally why again um so i meant i showed how to set up a web uh two cluster before a whole web ecommerce shop uh in order to set up for example your own um ipfs cluster you can do that too here you can see i'm going to create uh create a kubernetes cluster on aws here i'm going to install helm on it uh there are actually already helm charts out there for deploying an entire ipfs cluster um i you know i just search for this one there are many out there i'm not endorsing this one or saying it's not bad i just don't know anything about it but at the end of it i actually ran this and i set up my whole ipfs cluster four commands just like before right in the web 2 world that's pretty impressive and there are lots of things out there just like this but again it's really up to you because for you does that make sense does it make sense for me to run my own ipfs cluster i don't know it like do you have privacy concerns do you have needs to like ensure that you have reliability and can restart these things or so on or can i rely on this global network that's going to take care of everything for me i personally am okay like tapping into everyone else and using my own authentication but many people have different security profiles and things like that and so i will leave it up to them to make those decisions um should you endorse or embrace all of web3 um really up to you uh uh many many platforms out there you've seen some already today and there will be many more later today uh have javascript and node and python frameworks that allow you to plug into many of these components in your standard web 2 stack yet still embrace a lot of web web 3 is again that means you're trusting someone else which is critically important some people are okay trusting someone else for these various components using things like http gateways or translations or bridges some people say you know what i need to run this entire thing myself and that's important for me to deliver value to my customer and then you'll run your own nodes and you'll spin everything up and you'll cut the latest builds again completely up to you to make that decision uh i do really strongly suggest that web3 is one tool in the toolbox and it's built around flexibility um trust and decentralization and that's really the the key uh the the number one thing that i walked away when i first began to consume all of the content uh around web3 and people building apps and those kind of things is the idea that um web 3's real innovation is around trust it allows you to decide who you are trusting and why and make that decision on a component component microservice to microservice human being to human being whatever it might be basis and that really is transformative that's something that when we went back and you look at that amazon spinning up an eks cluster and that simple shop you're not making any of those choices you're just basically saying hey you know what service provider out there you just decide everything for me and that may be okay too there's nothing wrong with that it's really up to you to make that decision so um with that uh that's me any questions comments [Applause]