DAPPCON 2024: Building An Onchain Knowledge And Collaboration Layer -A.Tsamados and V. Krishnavanshi

Transcript

hello everyone okay beautiful how are you so for those who were here in the morning panel I assure you this is uh completely different we're not talking about AI a little bit at the end but I want to mostly introduce a thesis to you uh problem that concerns all of us and the solution that VJ and I and our team have been building at fers so I want to start very simply and clearly as much as possible internet Freedom cannot be

guaranteed behind World Gardens I think this is evident for most of you here and there is something even worse to say about those W Gardens which is they limit the type of global coordination and collaboration that we can imagine and let alone realize and crypto is all about solving Global coordination and collaboration problems games now the problem speaks for itself I'll give you some details but enj just two minutes of research on Google you realize that your tools your favorite

tools are leaking data constantly they're censoring you your account is not really your account it's an entry on somebody else's uh uh database and this leads to a bunch of known problems like the one that happened yesterday where uh huge company lost all their data that they were storing on Google and the backup data that they paid Google to backup on top of it so just to give you a bit of an overview of the main issues that come with those centralized collaboration and storage tools that you're using uh you're at the mercy of mass

surveillance and the behavioral manipulation that comes with it arbitrary decision from somebody in the company or a dis gruntled employee that wants to up before uh they leave you have insufficient of or complete lack of encryption let alone into an encryption you have link rot which is very related which uh fun fact 38% of all content that you had on the internet in 201 13 doesn't exist anymore so this is a huge problem for academic Journal list Etc you also have the problem the the possib ability of the company going bankrupt getting bought

leaks server outages user logins from the design of those apps and finally constrained ownership configurations meaning I want to share a Google drive with uh the people I interact with most on my favorite subreddit impossible but we can do that now with multi for bank accounts for digital Bank bank accounts so I want to keep those uh those reference points and explain how we're approaching them in order to explore not only a way to solve all those issues but actually also open up a design space that is

completely unique and allows for group collaboration now 3.2 billion people uh around the world are dependent on two companies and a few servers in the US mostly to create and share information for crypto teams individuals communities it's the same issue we should know better but we couldn't do better at most what you could do is use a file sharing peer-to-peer Network like bit torant or maybe use a privacy enhancing centralized collaboration Suite but they were each solving different problems one is a open peer-to-peer Network the

other is centralized technology it provides privacy but it has a lot of the issues we we just went over now to avoid capture altogether and the abuse that comes with it uh we believe that our most essential activities and data should leverage open protocols and networks instead of closed products managed by centralized companies thankfully we're not starting uh from zero I think that it's been now 40 years 45 years give and take that the open- source Community the cipher punks the

all those people that we admire we're thankful for for everything that they they established to create a crypto industry on top of them have been trying to create the types of tools that I'm describing to you today or the type of infrastructure that I've been describing so far but they've been depending mostly on distributed systems on peer-to-peer networks some Access Control Management and inevitably some centralized servers to do uh essential tasks or essential um permissioning that you could simply not do with a tech that was

available until now and then Community computers stepped in uh I like to use that term it's not mine it's Ethan bman from the cosmos community that U I first saw using he describes it as uh personal computers Unleashed unimaginable levels of so sovereignty and interoperability for people think the iPhone and Community computers will do the same for groups so now we're talking about a architecture that is a bit more developed uh there's something missing here um that is a bit

more developed than what we've had until now so it's not just peer-to-peer networks anymore uh with you know people could have their own uh uh public and private key which is a great addition to to peer-to-peer networks if you want to be self- Sovereign but there is a last element that we consider extremely important to create a more Universal public good infrastructure for any type of collaboration between people and management of knowledge and that's public blockchains so on this slide there should be a second column where I describe the the properties and the

advantages of both public blockchains and peer-to-peer networks um I'll just go over a quick quickly why we think that the combination of both basically allows us to create completely new types of applications unlocks completely new types of interactions between people and what we like to call self- sovereignty so uh public blockchains provide you with global consensus strict Global consensus it enables uh objective identity it's a one of the most important uh properties it's is the credible

neutrality that comes with public blockchains you have computational and data integrity and it allows mostly for some type of intera composability through smart contracts for example okay we're missing uh one uh column so this is the thesis I'm going to pass on to VJ to explain to you the architecture and how we're putting all that together but I want to leave you with that mental model we've had peer-to-peer networks since forever good cryptographic tools as well they weren't enough for us to Simply replace Google Suite notion all those uh amazing apps

we're using all day but that are basically working against us and now with public blockchains we think that by merging the two we can create a new type of infrastructure that can unlock a bunch of new uh activities uh um coordination Etc all right let's see next yeah and thanks Andreas for doing a perfect job of explaining the vision of file and what we are trying to do now what I would like to go into is a bit of a detail of what CS that vision and as in like why we Flex or Tech uh in general and to highlight different components

that we have uh the one component that we have is the offchain component where we uh mainly rely on peer-to-peer Technologies for example ipfs nodes gun nodes and uh different uh like uh uh peer-to-peer systems that we are using for powering this offchain infrastructure and then we have the onchain component uh the onchain component is the component which uh where we are storing things uh like let's say if you have the pointer to the storage if you have the uh like a registry for identity and if you have the um like address of the multi that you're using to control your file wor stack in general and the

permission registry in general that's the onchain component of fers and then we have the app layer and this is the most worked upon layer if you will uh from our side and this is something uh where we are working on different uxs uh where we are working on different uh uh patterns that we we are working like we have developed for allowing users to interact with uh the above mentioned onchain and offchain stack seamlessly uh and yeah so after working on the previous tag that I was explaining uh we came up with the different uh what do you call different

common Primitives that we are using for our own self meaning while we're building the file verse uh infrastructure we figured out there are few common pieces that can be used by other teams as well uh and since they have been developed by us for our own use case why not make it available to everyone else and the main work on the file ver protocol that we are doing there are you can categorize it into three different layers uh and the first layer that you have is the identity layer then you have the storage layer and then you have the collaboration layer we'll go into the details of each one and what are the specific Primitives that we have developed to power these uh layers so first one is the identity

layer in this uh we have uh you already know about EAS and multisig we're using them to allow you to control your file verse uh stack and file verse uh smart contracts this is uh the new work that we are doing with respect to program programmable identity it's something that's based on uh safe safes work that they have done for ERC 4337 wallet but we are combining the cryptographic access using encryption keys and the signing keys with the recovery uh um multiple recovery strategies that can be automated and uh are not assuming uh let's say that it's the person that's

going to use that and they can be used within the code without relying on uh human interaction then we have the storage layer and in the storage layer there are two components uh like two high level categories one is the onchain storage and the other is offchain storage in the offchain storage side what we are doing is like we're relying on peer-to-peer systems and uh on the onchain storage side where we have this uh we have worked on this collaboration smart contract which allow for these peer-to-peer systems to agree on a state and uh what this offers is like it allows you to take your data move to a different person or different peer or

different network alog together without compromising uh or losing any data in the process then then we have the collaboration layer and this is the most interesting one uh because there are there is one big piece uh that we recently figured out uh with the help of different uh teams that we are working with uh the first one is the collaboration smart contract which is what I was explaining with respect to storage permissions uh and uh this is something which is giving us the data neutrality for moving for giving right to exit to the users then realtime collaboration in this we have uh like uh this piece we have figured out recently

and in this we have uh peer-to-peer uh uh like uh peer-to-peer collaboration with end to an encryption and also adding the offline first support with the browser uh like with with the users being able to open that document into the browser as well and continue their work I'm going to explain these things uh in the workshop uh which we'll have uh uh like later at the Dacon and one at the safe con then I let the looking for yeah so just to summarize uh you've heard the thesis you've heard elements about the

architecture let me put all that together once again so today we're all depending 3.2 billion people are depending on Google suite and Microsoft Office for their daily activities preservation of knowledge Etc for everyday people that's fine there's still a lot of issues that are affecting uh people uh seriously for crypto teams it's catastrophic for communities don't even talk no need to even talk about it those two are simply not made to serve the needs of a decentralized community or a group that gets together maybe some of them are unnown they want to collaborate but they

cannot co-own the storage space the documents that they that they that they share the combination of peer-to-peer networks and uh public blockchains allows us basically to serve the needs of all those people not just crypto uh teams and communities but everyone that depends on those tools for their daily activities it gives them the right to exit no need to use our app if you don't want to use our app you can continue using the access permissions that we've perfected for you you can continue using storage that we've perfected for you you can continue live collaborating with 10

100 people on the same document without ever needing to hit one of the fers servers so really it's it's that right to exit and composability that we've finally put together and there are a lot of uh uh really cool use cases that are opening up with those I won't go over the the ones you're familiar with I'll just cover uh a few uh for example the F stack at the moment can be used permissionless with foraster two protocols interacting with each other you can put all your documents favorite

files links Etc on a file portal smart contract and even without using our our own uh app you can create uh for example a frame that goes through your uh your uh your content that you've stored the somewhere directly on farter it also means that you can plug F smart contracts in Crypton native social media without ever needing to request permission uh to anyone this also means that we've created new primitive almost on chain which we call collaboration smart contracts but that also act as

Registries for Content hashes that you've stored somewhere of chain um and this has been quite useful for teams that are experimenting with onchain AI agents actually so I think you're you're all very much aware that most blocks produced today in public blockchains don't just have Financial transactions and interactions occurring they also have a lot of content hashes and uh um pointers to certain media to certain types of interactions Etc so those AI agents can also interact with this type of content uh store and give a representation of their activity on

chain and have access to their own storage space of chain without ever needing to get a user involved uh go through a kyc process sign up on a centralized app Etc so just to make it extra clear I want to give you a a quick uh quick demo of the first building block we're releasing uh today it's uh it's a decentralized uh app it allows you to do everything you do today with uh link tree but instead of just having links you have a decentralized web page you can add your favorite links documents of

yours files images videos and all that without ever needing a server and on top of that you can invite people give different access permissions to different people again without ever requiring to interact with one of our servers or fearing that you might be locked out of your account at some point because of X or Y reason just doing a quick sign to the tech team is it possible to to get the video going hello is it possible to get the video going super

[Applause] [Music] [Applause] [Music] [Applause] [Music] [Applause] [Music] [Applause] how okay I hope you enjoy that I'm sure some of you recognize the music that was in the background but this is just a first view of uh what we're releasing uh

um the first building block that we're releasing you can scar scan this QR code if you want to try it out it will only be open today for a few hours to get an initial uh uh wave of uh of um first users get feedback be able to interact directly with people and not just uh uh on board uh everyone in crypto before we ready to actually make this completely permissionless voila thank you so much thank you VJ and Andreas we have a few minutes for questions right here in the middle I was

interested in like the storage you were using you had ipfs and like R weave and then Etc is there any option for like local storage so say if like you and I want to collaborate but I don't want to put on a public network I wanted to store it locally and yeah how does that work uh yeah so yeah uh yeah so what we are doing is like we uh the way that we have written the architecture for our uh storage layer as you will as you can say is uh by writing adapters and different adapters can store different uh store things to different places one of the adapter is ipfs other is RV the one adapter that we really like to use is uh index DB as well like meaning you could

do things on your uh browser side if you need so yes local storage as well but we wanted to make sure that we didn't just reproduce types of peer-to-peer apps that you know you would download the a local app which are great provide a lot of security privacy Etc but that are simply not accessible to most of people uh around the world most people use Google Drive or Google uh uh docs by going directly on their browser accessing that yeah uh and yeah that's it all right give it up for the faor team one of the most important projects on nosis chain