AO Conference | BUIDLSquad Stage - ETHDenver 2025

ladies and gentlemen welcome to AO conference woo we need to get some more coffee here we have four hours of the best and brightest Minds from around the Perma web and right now there is no better nor brighter mind than the creator of AO and r Reeve Sam Williams Sam you want to come on [Applause] up get the uh and slides would be great please hey hey everyone all right so with any luck we'll have some slides soon um okay we're good all right so um there's these events I think are a good opportunity for uh core perm Weare Builders to get together and discuss like you'll find that actually most of us in the audience here are people that live and work on the per web every day and so I wanted to give a kind of overview of where from forward researchers perspective we are and [Music] um e e flexible and composable um another thing that came with mainnet is what we're calling AO core which is essentially a it's a primitive for the internet fundamentally that lets you express um paths through the state space so the sequence of messages that should be executed on top of a base message in order to get to a state so if you have a smart contract uh maybe it's a token well you execute each of the users transactions on top um and you get to a specific what would you call a a a location in the state graph for that um program and what hash paaths give us what aore is sort of composed of is basically taking um that fundamental unit of what a blockchain is this sequence of interactions on a deterministic program that get you to a specific State uh and expressing it inside HTTP um using this new uh standard that was uh formalized last year called RFC 9421 or uh HTTP sign message and basically what it gives you is that every single message that moves around the system uh can be if you will like adding to its own blockchain basically every single state um transformation so an input that connects to an output is tracked and so when you get a message back from a hyperbeam node even if you're just accessing like the info page for the node or yeah something else like a um a piece of data stored on arweave you get back this little attestation that is signed and it references or it it connects a reference to a spe a particular spot in the state space with a particular output and this is like the fundamental unit required in order to have verifiability and fraud proofs and so on top of this we also launched te support yesterday um so you can run your own node and when you get back one of these hash packs paths uh and a state associated with it some data you can know with a very high degree of certainty that that is actually the piece of data that is found at that particular location um and one of the things I wanted to highlight here that uh because it's a sort of relatively closed audience of core per web developers is this concept uh that those hash paths can be um merged together so in this example we have four parallel processes or computations in AO cor um and each one is leading to a message that is sent to another process and when it does so it carries with it the hash path that led to the creation of that message and this happens recursively so over time what you end up with is a single unified I skipped the text slide sorry a single unified hash path at the top that that is a Merkel root of all of the computations that went into the creation of that message that you're seeing and this is something that I think I mean frankly when we were looking at this a forward research we um you can Intuit it pretty quickly that that's very powerful and my sense is that there's a bunch of things that we can do with that that are as yet very unexplored is open territory um yeah so if you can one example is that that you get a message at the top of this and it will let's say um that bank there is actually a token process and that sends a message or maybe it's a state channel right someone is streaming a video they're paying a Gateway for it a hyperbeam node for access to the content and then that operator of that hyperbeam node settles the payment into wrapped R and then from wrapped R they um swap it into AO let's say something of this kind so then when you take the output 10 years later from the AO process you now have a Merkel tree that includes the ordering of all of the transactions that happened even down to that state channel right at the beginning and this is all represented in a single 256bit um merel route and we also left it open in AO core so you can Implement other algorithms as the um someone is waving I can't tell why okay they say Thumbs Up Now who knows what that means um yeah we left it open so the algorithm that is used to uh add to the hash path is something you can swap out later as well so for example you could make a version of the hashpath algorithm that will only include state transitions that have been signed by today a te so when you get back the route out the end of the system 10 years later then you know that all of the computation that happened in the subset of the network which would be if you play this out I think you'll end up with like 95 99% of all the compute being expressed in these single values um yeah you know that all of that compute happened correctly up to the uh um would you say security properties of the hashpath algorithm that was used on it you could also do the same with ZK proofs if you so desire it's open to that so if there are any builders that want to take that mantle and run with it I think there's a bunch of really cool places it can go um but right now it's just Green Field like totally unexplored um okay so another thing that uh AOC core gives us is the ability to merge together state representation so you get a message output from one machine you get it from another machine if they agree then you can merge together the signatures and this is another fundamental primitive that's needed in order to have multiple attestations of correctness um it shipped with AO mainnet uh two weeks ago so yeah you get a global unified Merkle tree of almost every computation the interesting part here is almost uh that sounds like a negative when you first encounter it but actually you realize that like this see without creating communication overhead so for example in ethereum right you get a state representation out the end of every block um that is 100% of the network right but you only have one thread and in order to achieve that um representation you've had to compute it on I think it's between 12 and 60,000 mod depending on the time what uh hash paths and AO core gives us is you get almost all of the computation but it happens naturally as messages move around the system there is precisely zero additional overhead from the communication perspective and from the compute perspective you need to do one additional hash and that's it every time a message is passed uh it's really powerful I think um and a quicker side uh yeah this conference is about AO but for everyone that's building on a which I guess is basically everyone here as well um yeah the HTTP signature format is a new form of bundling essentially with a new type of data item uh that gives us a bunch of really cool properties that we didn't have before so traditionally in arwe um like when we first built it it was supposed to be well supposed to be it is uh it was built to be a ledger of history and if you look at any sort of archives data structure you see that you don't just refer to a piece of content by its raw identity of just the the bites themselves that would be kind of crazy and in my view um because the bites themselves don't really tell you very much what you need to know is who said it when they said it and Associated metadata and so that was baked into the first transaction format of awe um and that Associated metadata component was left open to the user to decide what is relevant to them and uh you know fast forward 7 years and that has become the backbone of the indexing systems that make a bunch of the applications on the per web run but a constant complaint we've seen along the way understandably is well hey those you know tags as they're called uh they're a fixed size and they can only be binaries um and it's very difficult to get a recursive structure in fact I I haven't seen anyone in practice create a recursive structure uh so when we were looking at building hyperbeam uh and AO core it became very obvious that this would be a good opportunity to widen the um the design space that's offered by the core data storage primitive so yeah now we get flexible keys and values that means that everything in the map uh the message is a um flexible component there's no difference now between tags and like normal metadata that's all just one thing um and the same with the data field um those that have built with the system will know that uh there's only one piece of data that you attach to an awe transaction which is kind of annoying sometimes you want to add multiple pieces of data uh well now you can you can have as many documents as you want inside the same uh single message all associated with just keys and values it's like a classic dictionary map structure um you also get multi signature uploads of course uh because yeah with HTTP Sig we get that out of the box and you get recursive messages so now you can make I guess in the old world we would have thought of this as bundles that contain bundles um but they're not just referred to by like their index inside you could have in theory and again I don't think I ever saw anyone do this in practice because it was so janky um you could say hey I've got a bundle of data and number five in the bundle is going to be this other piece of information um that is you know necessary for this computation I'm running or this service I'm offering instead now you can just everything has a name so if you want to you can have a list of items um inside the bundle but you can also say the you know HTML document is this and the API representation is like this and you can serve that back on the same um it's the same message essentially contained in the same message um and something I didn't list here but it's pretty cool is when you recall these pieces of data on a hyperbeam node then you can actually Traverse through the URL or Traverse through the data structure via the URL so if we had a piece of data that's got like a body and also this sort of let's say a computer native representation you could say like that piece of data stroke API stroke price stroke BTC or whatever it is and so you can like recursively um navigate through the bundle uh using like normal HTTP um semantics and also we have typed values now this was another problem when people wanted to index the data sometimes you had like an integer or you had a floating Point number or you wanted something to be a list that was that was actually one we commonly saw led to people putting Json inside tags which is a nightmare well you don't need to do that anymore it's natively supported out the box um you can just have lists of values yeah and also uh all of this is expressed in like core HTTP semantics so if you access a piece of data um in fact if you go to a hyperbeam node you go to stroke any device and go to say an info endpoint like The Meta device then what will give you back is a HTTP structured document which is a recursive bundle of um multi-part form data which is a little bit odd and specific but tldr is that every existing HTTP client Library already knows how to pass that and yeah so we get Library support in these are 12 languages that are just listed on the HTTP s.org website on day one um but and the the standard itself was only ratified like a year ago uh and you can already out the box you know use Scala if you want or uh God knows why you would want to but you can use Ruby to make these messages and send them to the me to the network and they can just be adopted this is a huge huge step up from what we had um with a transactions where you had to have a specific library that someone had to make and then they had to maintain and so you got this common theme of like okay someone now wants to use the network in Java for some reason uh well there's no library for that so either they have to go and build it and maybe they do build it which would be great but then they like you know abandon it 6 months later and so it's not up to date and then someone comes and tries to use it 2 years down the line and it's just not a smooth experience we SWA that instead for everything that can use this sort of globally accepted common standard can now create awe data items and send them to the network and um something that hasn't progressed very far yet because the standard is so early but I think in the future will be really really cool is that all of the different um Key Management systems that emerge for HTTP signed messages will instantly be compatible with the network and that means that as soon as someone makes a hardware wallet that manages these keys and there's various reasons why you might want to do that in other just on the web in general um we'll have support for it and when people have Key Management Systems um for organizations for Enterprises well those are valid R weave keys that can custody AO tokens of any kind uh so I think it's pretty pretty cool and they look like this in practice this is taken uh just from the uh the hyperbeam uh shell but you can see here that we this is a process state so when you're executing a process in the background what's happening is it's creating this recursive message structure um where each of the different components is available as a uh navigable item so you could go to that ID of that hash path stroke for example uh VFS I think is at the bottom here somewhere and stroke um Dev stroke standard out and then just in the browser you can get to the standard output created by the RM module and you can see there it has Rich types some things are list some things are integers um yeah and we can now index on top of all of that and it uses unified standardized formats even the typing is done using a a sort of a neutral established standard called structured Fields so if you're trying to pass one of these documents uh down the line in some other program you're writing um well if you can find a structured Fields Library then there you go you have support you don't need anything special to use rwe in the per web anymore it's like the per web is integrating into the web as it currently exists in all of the standard Frameworks and apis uh okay so AO core itself is a system that exposes devices I mentioned at the start that previously we had three different unit types and each was a different code base well the stack that you get when you um check out the hyperbeam repository today is 30 different devices all of which um we've built a sort of a standardized tagging system for them to basically let you know as a user what level of development they're at and so um you can see here there's something sort of candidate mode we Define candidate as like U if you're a nerd like me you'll know the request for comment uh system the rfc's that are part of the internet standard uh what would you say the standard nomenclature of the internet um where they're not really actually a request for comment they're like this is probably the standard as will exist for decades um yeah that's that's what a candidate mode release means you can reliably build on top of it and know that it won't change under your feet then we have this idea of a preview release which is okay there are parts of this system well for a start all of it is in place it works but there are parts of it that might need to be tweaked down the line um and then we also have Early Access releases they're like okay it more or less works you can use it for stuff but don't be surprised if it changes under your feet and the core idea here and I would really sort of advocate that we adopt this widely in the community is so that other developers and users Downstream can know the robustness of the thing that they're building upon right like the whole point of the crypto industry is to create protocols that offer users Services rather than companies and why do that well because the protocols don't change you as a user get actual reliable rights and guarantees when you use the the protocol rather than a company who can just turn around and change the game on you like 10 times a minute and um yeah you're you're more or less beholden to them and so by making a a clear labeling system and having you know if we were to adopt it all as a community not just forward research then it would be very easy for people to understand okay what rights and guarantees am I actually getting so with an early access release there's no real guarantee that it won't change it most likely will in some form and then in a preview releases like okay well largely the structure is there but I shouldn't depend on components of it are too tightly because they might well change in the future whereas a candidate releases something like Bitcoin and says okay here's the fundamental API here's the way that you use it in practice and maybe there are small tweaks down the line although you don't need to adopt them no one will force you to but it's a robust reliable Service as is that's what really grants people rights so two weeks ago we shipped Milestone one and it's about to ask who was a fan of the communications on mainnet and there'll be zero hands including ours from forward research this didn't this didn't pan out well but to take another opportunity to explain the situation um we shipped hyperbeam AO core the core protocol now we don't think we'll change and um those that are kind of close to the team will know that we've been like obsessively focused on Tech uh for the last yeah really since summer uh and the reason for that is because with our weave we saw there's a lot of pressure to get hype and engagement early for protocols you see this like all over the crypto space but the truth is that actually real real adoption and use cases only come along when the tech is in place and there's a real danger when you are building a protocol that actually people come to rely on the wrong version so I think the story of ethereum is actually encapsulated directly in this which is that they a great proof of concept that was like our Legacy net test net V1 um but people started to adopt it at such a rapid Pace uh that essentially that negotiation with the users and the developers around like what changes are admissible becomes harder and harder because it has a specific spot in the trade-off space and the real like disaster from the an ethereum conference so sorry guys to be clear ethereum did some really amazing things um but in terms of the adoption of ethereum over time uh the catastrophic error was Global shared State and applications came to depend on this this was just the way that you built ethereum programs and so in 2017 2018 the scaling road map for ethereum was to build sharding which would have killed um Global Shar State as was like you know obvious I don't think I need to explain to you guys why that won't work um but they they couldn't push it through because there were already users that had become not just accustomed to and climatized to but they also depended on the software to run in that particular way and so when it came time to get rid of global shared State they they couldn't really do so and it led to this sort of 8-year Quagmire of um not really achieving the scalability that they wanted so yeah we've been really obsessively focused on getting the tech right getting those fundamental units in place and that's AO core at at the base of it and then an implementation with hyperbeam such that in the future we can build on top of it and it's a modular structure so each of these components can be swapped out at will at any time in the future without disturbing the other components of the system and now we have a really solid base from which to work um which saved in the case of our weave we had similar situation um we shipped V1 which was quite scalable but like it wasn't going to scale infinitely um or like to be precise uh past the maximum the users wanted to use basically like 2 to the^ 256 bytes is what we ended up with with a 2.0 um but before that we you know had a lot of pressure from the community and from investors and the likes uh to go like push adoption and get people hyped about it um and we did so but the truth is that they all flaked because when you went to use the software it didn't do what you needed it to do and so instead yeah we focused on getting the technology right and now um we're in a really great spot for Community adoption and development uh yeah so Milestone one we call this this is basically AO core is cemented in place and also all of the units that were needed in order to uh perform payments just payments to start with and that's about like what is that uh eight different devices on top um and you can now spin up a hyperbeam node you have been able to for a couple of weeks where you can set up a payment relay to existing AO infrastructure so you spin up the other nodes and then you can charge people for Access and the reason this is so important and the first thing we focused on is because anyone that's used AO Legacy net or test net um knows that wow like sometimes it gets slow and the answer for that is that uh you know the reason for that is that uh when you have an open system where anyone can compute any amount of or ask for any amount of computation they like inevitably it leads to clogging of you know there's only so much infrastructure being provided and it's being provided for free so you can't scale uh what' you say you can't scale the infrastructure to deal with user demand because user demand is uncapped there is absolutely zero cost to the user to generate requests um yeah and so we have payments in place and so okay now you can go to I think it's like relay do I forget what the domain name was well anyway if you open up AOS and you say-- relay it'll take you there um which is kind of like a fast lane for existing infrastructure uh but yesterday uh we shipped I think a a really really exciting part of this stack which is um basically the ability for people to run hyperbeam native computation as well as those Payment Systems um yeah on top of uh AO core and what this means is in practice you can go uh open up AOS say-- mainnet and it'll put your scheduler for your uh computation into a hyperbeam node runs natively inside the system and then it'll do payments or whatever else you'd like it to do from the other devices um and so then Curry the compute that you ask for to a legacy net compatible uh executor it's actually just like a you say like a mini coup embedded inside a hyperbeam environment um and you can run all of that inside a te so after payments where we need to go next is uh making it so that people can push messages for one another around the system like right now or on Legacy net at least when you started a process you would said a default Authority and uh some people played with this in the community made it so that you know they could also push messages for their users on those processes but it's a sort of Ring Of Trust if you will and what we have in place now with in the sort of preview mode is well you can start up a node inside a te which gives you very high levels of trust not perfect not um by any means uh what would you say on its own a single te attestation is good for some value but you really want like multiple tees into to get in order to get very very strong correctness um but you can now start uh hyperbeam node inside a te environment join the green zone um in in preview mode and essentially perform computation and have the messages that come out be reliable to some reasonably High degree and then push those messages around the system um now it's going to take some time for that to get from preview to candidate releasee but now people are already playing with it and they're actually contributing to the hyperbeam repo which is really cool to see um and so this is like the full view of what exists right now and obviously the red stuff is where we want to go um down the line there's a bunch of stuff you can already do with this in practice that I think is super cool so um when you have an AO process on um on a hyperbeam node then you can access parts of its state and you can actually also export from the WM environment to the external host um again in preview mode right now there's probably going to be issues but you can like try it out test it and uh within relatively short period I think it'll be stable candidate mode um yeah you can make it so that you can like render web pages so here is a uh web page just says hello world but it was dynamically generated by that process underneath it's like a web page served directly from a smart contract and even the last mile from the smart contract to the user's browser is protected by that hashpath mechanism so you're essentially saying hey give me the state of this position in the um State space and it will give you back a renderable web page it doesn't just have to be web pages uh you can have full API is expressed this way so you can now instead of um having to query and basically like simulate what would happen if you send a message to a process you can just embed um or rather export from your wasm environment uh essentially bits of State uh to the host hyperbeam environment um into that like AO Core State message for the process at any point in time and then you can just say your process ID on any any hyperbeam node stroke now stroke price and it will give you back that bit of State inside the system and then you can even do like more complex things like say now stroke balances stroke user and where this goes in practice is that no more dry runs yeah exactly all right uh yeah like when you want to perform a computation on AO Legacy net right now what actually happens in the background is that and I won't mention the name of this project but there's a widely used project in the ecosystem that you load the homepage and it performs 80 dry runs 80 simulations of interactions with those WM containers and for some context it takes a third of a second to invoke that wasm container run the computation and get the output you're doing that 80 times in order for the user to load just one page that's an unbelievable amount of compute it's instead of that we have these extremely cachable values where you can just export every time the state changes a patch on the state of the process um and then the user can get to it just by the URL bar or you can you know uh basically dynamically do it in JavaScript and boom you've turned that 80 request into like 80 requests for little values that are just like using a web server it's going to be like massively massively more performant uh so that's a yeah that's exciting and you can play around with that uh today it's little bit janky interface but it's there um and soon I think it'll be widely deployed um but it actually goes kind of deeper than that so here are some example URLs that you can use with AO core um if you're used to developing and like you know familiar with bash scripts I think the way to understand the AO core URL bar or the you know path that you can go to in um on a hyperbeam Noe is basically that each is a mini blockchain and you can specify the interactions you want to have happen um in sequence and it can be like immensely deep uh so for example we have a a URL here where the top one you're going to a specific hash path and that doesn't need to be a process it can actually just be like a serverless function with no Schuler um it's just a specific existing point in the state space and you're saying compute and get the results and run this wasm function factorial with these parameters and you can also specify that the parameters are a list and you can do that all in the URL bar uh which is kind of like interesting curiosity but turns out actually in practice uh this this could lead to really really cool effects so the second URL we have um we take the process and we uh get the schedule and we um say the body or rather we're posting to a process's schedule and we set the body to be a different path and so essentially this subresolution system allows you to kind of create another little blockchain and then merge the two results at the end so this says go to this process ID compute it at State seven slot 7 get the results get the outbox get the first message from it now you've got that message and you're essentially I mean that's a messaging unit in AO Legacy net like in a single URL and anyone can construct them of any form it's like the URL bar itself becomes general purpose compute um yeah I think there's oh and this bottom one so for example you want an oracle you want to get the price from binance of Bitcoin well you just use a relay device and then it will give you back a HTTP signed message that you can validate is running inside a te so hasn't been corrupted and that's one URL bar or like that this one get request uh to perform what would otherwise be like a massive piece of engineering uh where we see this going down the line is that like in the same way that building a computer yourself is not the process of like putting transistors on a chip you just pick up parts and they stick together it's like kind of Lego in a sense well this is Lego for blockchain applications um it's a big box of different tools and you can just construct the service you actually need uh yeah by like reusing them so building a new blockchain um or new service that has specific uh you know compute requirements security requirements uh should really just be a case of configuration not a case of building and development which should cut down the cost of innovation massively uh which is something we're excited about um so the next thing that we're that we're focused on is uh getting it so that I think these slides are unfortunately swapped that is unhelpful well getting it so that uh the scheduler and the delegated compute system that's the one that lets you call out to the embedded uh CU inside hyperbeam and the Genesis wasm device that sort of like orchestrates that process uh is all in stable release and when we have this um we'll be able to access Legacy processes on mainnet so if you take nothing else from this I think we could use it to clear up the confusion about main net Legacy net um the reality is that all of your data if you have a legacy net process is on arweave so it doesn't really live in Legacy net per se it's that you're using Legacy net infrastructure to execute that process and where we want to go with Milestone 3 is making it so that those processes are then fully compatible with with hyperbeam nodes so you can start pulling them down and running them inside hyperbeam environment and making use of for example that no more dry runs uh feature an exporting State and the likes um without updating your smart contract code necessarily actually at all you can just continue to run them uh because the data is shared on arweave like there's a neutral place where it's all stored um yeah which doesn't really exist in main net or Legacy net it's like well actually in this diagram explained pretty well it's right at the base layer uh underneath everything so when you have this then you'll be able to smoothly transition in a live fashion with no downtime between Legacy net infrastructure and Main net infrastructure um and at that point people can start using payments and start running their own computation inside te and start pushing those messages around and you end up with a fully decentralized supercomputer at the end of it um after that we want the green zone to reach stability and this allows you to do not just everything I just described but also private computation provably private computation even uh which could be useful for many things in the ecosystem like I know we have the randow team with us today um well they can make it so that uh their generation of entropy that goes into the randow process can also be hidden inside a verifiably secure container when where no one can see the um the state Transformations you can also for example have a piece of data stored on arweave that you then pay to access and only when you have paid do you get to see the content uh which is something you know we've had probably 100 requests for over the years now building those services will be very very easy after that where we want to go is um this system called aot it's ahead of time compiled WM code so the new uh hyperbeam Native computation system which already have in place but not running in aot mode uh and like most of the existing processes actually all of the existing processes and most of the new processes will use the embedded CU um but yeah it it allows basically to when you see a WM container you can then compile that to Native assembly code on whatever platform you have and run it at like 90% approximately of native speeds so if we want to build a supercomputer this is how we make it so that it's running at like full Pelt every single CPU you throw at it gets you one CPU worth of computation out the side at the end whereas right now maybe you get like I don't know a third or a half something like this so this will be a big efficiency Improvement and yeah we're also not the only people building we're just having a conversation this morning with the apas team about the um GPU devices they want to build on top of this they already have an implementation of we NN the web assembly standard interface neural network uh standard which allows you to um call out and do basically well most obvious use case for it is large language model style computation on a GPU from inside a wasm environment and then as a consequence inside a hyperbeam environment and even inside a te giving you correctness guarantees about the llm output um and there's also another web GPU project ongoing uh with a sort of different version of the wasm environment um but where next after that well now our attention is starting to turn back to uh growing the usage of the system now we have the tech in place we can really tell people about it and yeah grow the environment that already exists inside AO and I think a huge thing that people missed with the main net launch or didn't really receive the attention that I think it's probably worth is pi it's like it's big it's very very very big I think this is um this will be huge for the ecosystem long term essentially what it means is that um as it stands today and I haven't checked the numbers in the last week or so but from the pre- delegations that people have made 40% of the value approximately 30 to 40% of the value from um AO minted tokens will all flow to the ecosystem to ecosystem projects without there being being a centralized party in the middle like a grants committee that is sitting there deciding where the money goes no instead it's just following the delegations that real users of the per web are making with their arweave tokens or with their bridged assets and so if you work that out on a per day basis um I haven't done so with the latest prices of AO but last time I calculated it was something like uh yeah it easily gets into the hundreds of thousands of dollarss per day invested into the ecosystem so if you're building an ecosystem project this is a way that you can get access to Capital in a fair and transparent manner where if you will your audience that you're attempting to convince is just the wide public everyone using the Perma web um and it's on completely equal terms and a secondary effect of this uh slide order again damn it a secondary effect of this is that you start to build an extremely wide ecosystem of people that are staked into the success of your project uh like longterm right from the start even before those tokens are liquid and you can do this without actually having to reach them directly so um you know of the people that have delegated there's probably I would guess huh 4,000 plus all of the RV token holders or not all actually rather many most of the RV token holders which is into like 200,000 approximately they will become stakeholders in your project just by proxy and that means when they see your posts on X and you know you're trying to engage your community well there's a set of people that are out there that are willing to listen from day one which for a new project is a huge huge deal um you get a very very wide base of supporters virtually instantly with this system and even cooler they all get in on equal terms there's no favoritism it's just open and transparent and it provides you funding at the same time did I mention that I think this will be big like thank you I think this is going to be very very exciting for the whole ecosystem um and yes to take it back to this slide that was out of order um yeah there are already actually six projects one of which is not on here uh which is a shame because they're the people that took the leap early uh and we basically like jumping into the deep end when this was just an idea and that's aphis Network um who already had a mint live uh since the end of December and at one point they had $7 million uh essentially deposited into into their um minting process Fair launch minting process and a really interesting byproduct here that is not understood I think at all yet at this point is that um the yield that is distributed as funding is not the native yield it's actually the yield in AO terms and what that means in practice is that the okay now we get into the weeds of the mechanism design but um the yield that you get is the yield over base of the Native which tldr today sits at approximately 2 and a half times what it is for the native assets so actually the last time I calculated it the native sorry the yield that these projects would get on those deposited funds in their Fair launch process is like 40% per year now it probably won't stay that way cuz markets are efficient uh markets are eventually efficient um but yeah during the time where that that value is smoothly moving down towards some efficient level probably somewhere over the native um asset yields yeah the yield that you get from those deposited assets is extremely high so that just magnifies the amount of capital that is going into the ecosystem um helping people build stuff all right so there we are that's where we are and that's where we're going thank you everyone for listening and yeah I'm excited for all your talks today [Applause] gaging online Worlds the one the only Mr Matt Mason of Basse jump thank you can you all hear me oh what's going on mirr all right this happening yeah you got mirr oh wait okay sorry 10 difficulties where do I do want toay sorry about this ladies and gentlemen um like Sam I'm also British but I'm not as smart and I'm not as technical clearly uh I do swear more as well so I apologize in advance for that um but we're about to get kicked off my name is Matt Mason I'm the CEO and co-founder of base jump um thank you so we're building a new platform that makes it very easy to create share and connect games game assets and worlds and one thing I want to talk about today is why we're doing that now and also why are we doing that on AO and R weave so the reason we're doing this now the reason we think this makes sense now is because games are fundamentally changing um Gaming's never been more popular about half the world use game assets mostly in games but also in streaming platforms and in social networks people are getting more and more attached to their game assets about 40% of genz players update them monthly and more gen's would rather work in gaming than be a doctor or lawyer or professional athlete so that's all amazing um but gaming still really really closed it's very fragmented and it's it's it's never been easier to make a game but it's still very very difficult to make a game um because it's easier games are becoming something else so there used to be applications but actually if you look at where gaming is at now games are becoming content um if you look at Roblox Roblox is now the size of Nintendo Xbox and Playstation combined and the only thing happening on Roblox is kids making games on Roblox themselves same thing on Minecraft if you look at fortnite 19 of the top 20 developers are single devs pulling all nighters and making maps that people play and then they go away the next day so Gaming's getting more ephemeral um and and this idea of owning your assets is becoming more important to people people the big tension is that you can't move these assets from one platform to another and we got a glimpse Bas jump of what it might be like when the world is like that when we created an nft Community called broadside a couple of years ago um this was an nft project it was very sort of story driven it was about this sort of fictitious future world that we made up and we had this whole plan for Broadside but the broadside Community decided to go in a completely different direction they decided they were going to be become the most engaged community in all of the open metaverse and they were going to take their assets across all kinds of games and platforms and worlds and do all kinds of things with them that we didn't imagine so for example they built this place called the broadside Rave bunker um this is the most popular nightclub in the metaverse it's kind of like what CBGB was to punk or Studio 54 was to disco the broadside Community holds raves in here every single week they've done that for the last 2 years this event is streaming in there right now they're watching this right now in the raave bunker um you can get in there yourself if you go to broadside Twitter um we've had all kinds of people like The Prodigy and DJs like Roger Sanchez people have come and performed in the raave bunker just because it was a weird and fun thing happening on the internet and they wanted to know more about it and the broadside Community just sort of kept taking this further so we did this metaverse world tour we did a bunch of live events that were also in the metaverse the Community just keeps sideloading and doing weird things with their assets like they bringing them into Roblox games into GTA all kinds of crazy stuff and the broadside Community because they had the rights to their game assets and because they could move them all kinds of places they they did all kinds of stuff with them and after the end of the first year of broadside the average revenue per user that we were making from broadside was about $98 a user so to put that in context the rpool on Roblox is about 15 bucks and the on fortnite is about seven bucks and nft communities are different oh that number doesn't include the mint price of $75 that they initially paid for the 7,000 broadside characters um this is this is revenue earned from an additional sort of 45,000 nfts and merch and other things that we did but what this really showed us was that interoperable game assets um are really fun and they're really valuable if you give people rights to do all kinds of things with them they'll do things with them and then that will create more value for you as a company so we decided we this is a really really cool group of of people and we want to build a platform that kind of serves them so um myself and my co-founders we've been working scaling some of the biggest Platforms in blockchain um consumer Tech and gaming for for a really long time and we realized that between the three of us we were probably actually a really good team to try and do this so like let's build a new platform that really addresses this reality of a world where games of content and assets are interoperable and so that's the big idea behind base jump um this is the sort of home screen of base jump it looks like a Loadout screen for a game but it's a Loadout screen for the internet built- on game engines it's got a permissionless gaming subrate under the hood called action which lets you create a game or a game asset with a simple text prompt and then deploy to any of the any of the big game engines social apps anywhere that accepts game assets instantly um so this is all live right now if you go to Bas jump XYZ uh this is very much the kind of we sort of think of this like the splash page for what we're building we're launching our first game um in the spring it's called battle mode it's going to launch on browsers and in Discord it kind of plays like super smash BRS or Street Fighter 2 except you can load in with different armor or weapons or objects so it's like Street Fighter 2 meets rock paper scissors is how we're thinking about it the reason that we're building a game is because all of the metaverses that have scaled Roblox fortnite GTA Minecraft have really really really amazing core game Loops all of the metaverse Platforms in the last cycle that haven't scaled don't really have core game Loops they're mostly kind of tool sets you go in there you kind of walk around there's no one there you leave and you never come back core game Loops are really important we've built game studios and games before this is the first thing we're walking working on because this is the most important thing for users so the way that this will work here's a here's a screenshot from the battle mode game so you can take in your avatar load it up with the weapon battle somebody else do this straight from Discord or in your browser the level that you're seeing here and all of the levels here are powered by action they're all procedurally generated and once we've got the game out the next phase will be putting the action app into the hands of users so that you can procedurally generate these levels yourself and you can start making your own levels for this first core game Loop that we're building um these are some early tests that we've we've done using the action platform and like straight away we just just playing with it ourselves we started making stupid things like we wanted to do a level that was about Pepe so it sort of came out like a Lily Pond or this one down here is kind of a sort of green screen terminal but very quickly we started to realize that if it's really easy to make a game if it's as easy to make a game as it is to make a post on Instagram then games are going to become me memes like this is another thing that we made where it really felt like a this is a playable level that is just sort of a meme some of these things you might make and share with your friends and then they'll disappear or maybe it's a really good game and you want to monetize it for a really long time so you store it and you publish it on R weave um and that's where the stat comes in like none of this was possible before AO and R weave action is this layer of scalable gaming infrastructure that we're building AO makes it possible possible to instantly create these levels and are we've let you store them permanently so you know once we got in touch with um Tate the team at Community labs and we met Sam and everybody at for research we kind of realized that that everything else was Legacy Tech like once you really understand what's possible here um it's just kind of mind-blowing so we're we're so excited to be building here with so many of you in the Ecco system so like getting into what action does a little bit more um it will have a token does a lot of the things that a token will do like governance um it has this impact thing which I'll get to in a moment uh it's also a way to pay for compute on the network but half of what action Target is really about specific stuff to game assets and these could be avatars or weapons items or even entire games we call these objects on action hyper objects um and one of the important things that uh the action network does is it actually just registers these objects and kind of explains what they are and what they do and what they might do in different worlds and because it's all on chain it means these objects can also be Network aware and context aware which lets you do some really crazy things so so here's an example of hyperobjects we we could if the internet's going to run on game engines we think of hyperobjects as the connective tissue that connects everything in that internet much like hypertext connects the internet that we use today so here's an example of a hyper object this is a broadside character called Stinger um you can see him here he's got a rocket launcher and he's holding some some Ramen all of these things are hyper objects uh if you take Stinger and you load him into this platform chorus which is um a platform running on bass and polygon it's an AI music game creation platform the chain doesn't matter you can bring stinger in there and you can start making AI music with him and because chorus knows that he's a broadside Avatar it's going to load up some broadside specific beats and acappellas and Loops for you here Stinger playing in Nifty Island here he is in an app called Alias which is for streaming and V tubing so you can start to do some some really crazy stuff and this rocket launcher that Stinger is holding if you take that rocket launcher into Nifty Island it knows it's in Nifty Island and Nifty Island you play Capture the Flag King of the Hill these type of games that rocket launcher will fire bullets in Nifty Island because that's what it was designed to do if you take that same rocket launcher into the broadside Rave bunk it will fire glow sticks because it knows it's in the Rave bunker and I think that's what you know with gaming Gamers don't like blockchain because all they've been shown is the casino we need to start showing people the magic of blockchain why does this make gaming more fun and we think hyperobjects is a is a really interesting way to do that so hyper objects make things more fun but they also can help you explain some of the more sort of complex ideas behind crypto we're releasing a a collection of new action figure hyperobjects that will also run as nodes that will operate to secure the network um we've got 7,000 OG broadside action figures they will be the first ones that runers node securing the network we've got 100,000 others ready to go I think ethereum's got about 6,000 validators on it right now so like if you if you put a really fun gaming UI on front of things like nodes or some of the more complex ects of defi you're going to help more people understand why defi is important or why something like a node is important if you show a gamer a platform where the platform isn't front and center but their character is front and center their assets are front and center their data and they're intuitively understanding that they own all that stuff and they can take it with them wherever they go I can't think of a better way to explain decentralization to someone without having to say a single word I think this is how all of blockchain is going to scale I really believe that gaming is the Magic Bullet and we're really trying to take a crack at that with what we're doing here um so yeah and there's some other things we're thinking about so we think this is also a really interesting way to solve complex problems that currently governments and corporations are failing at so we wanted to sort of bring in this whole element of impact into what we're doing with action we partnered with the rainforest Foundation us and for our first launch launch um of the action token on the per web index we're aiming to protect a specific acre a specific Sorry 4 million acre area of rainforest in Peru where every single every single action token whether it's a hyper object or simply a token will be connected to a specific area of rainforest and this is the map of the actual area this is what you're going to see in the UI on broadside is tokens and Avatar slowly protecting an area and really gamifying this so so this is something that we want to scale to all kinds of games and all kinds of communities because this is this is a game right this is Capture the Flag this is King King of the Hill this is defending territory we've seen in the nonprofit world like if you adopt a a a goat or something you're connected to that goat and you'll get emails about it and if the goat's not feeling well maybe you'll get send another $5 this stuff works in nonprofit why aren't we using this in gaming to protect Real World areas that need protecting so we're really hoping to scale this this is a formon acre area that we're starting with um there's 3.3 billion other Acres of rainforest to protect but there's Ocean garbage patch to clean up and there's dust balls to Rew we want to use action to turn saving the world into the world's greatest game because it's something that we think you can only do using blockchain and and AO is really really perfect for doing this so that's what we're doing um we're really excited to be launching as part of the Perma web index um with the action token on March 14th um State your AO you can earn action we like this because it gets tokens into the hands of many many people um as soon as possible and yeah look we we're just so excited to be building with with all of you in this ecosystem um I love everyone I've met across AO and R we've you know and and everybody who's working here everyone I like out here is doing really really interesting things with blockchain um to really make the world better and that's what it's all about as far as we're concerned so do get in touch if this sounds interesting thank you very much thanks so much Matt appreciate you our next guest is from apis apis is one of the most interesting projects on AO and the Perma web and to explain it further we have Mr Connor ther head of community at apis Connor right speak can I take out take nice one all right guys um I'm Connor I'm head of community at apis Network I know some of you know me some of you don't um I will skip past this firstly who are we and what are apus network you may have seen it in the first slide that what we are doing is enabling verifiable detalized AI through deterministic um yeah computations and we're the second word side didn't leave on the first slide uh that it's on AO and that's really important for us um the second thing we're doing is we're running competition pools so we're incentivizing people to build high quality AI models right we don't want bunch of rubbish being put out there so we're financially incentivizing people to produce quality AI models and also for people to be able to gain access to those models so if you don't know what your building or how to build an AI model which is most people we're enabling you to do that just set a fee developers can push the models forward and the best one ends up being yours you'll have ownership we are as Sam mentioned also the first project on AO to go ahead and run the community launch token as an AO token and we did get up to it was $7 million in assets bridged towards us within the first 7 days we were really happy with that okay now AO main net and apis network you've heard a lot about what's been going on with AO from Sam over the last couple of days and now sort of how that's going to reflect with what we're doing um so the first thing and it is important to us but perhaps not the most important is that what we're doing for our for our consensus on our verifiability is we're leveraging determinism and our own consensus on this is we're using fast provable inference faults um and this is going to align perfectly with the tees that are going live on ao's noes um perhaps the most important thing is going to be this so we're building a wnn device for Native GPU support on hyperbeam for AI so at the moment I believe we are well basically we are going to be one of the first devices going live for this um I know we talk a lot about building something for AI on AO but I can say we will be pushing the code live for this on P day on March 14th so for anyone who wants to check it out there will actually be a version on March team for you guys to look at um and talking about Pi Day as the other guys have mentioned we're really excited to be part of the per web index so also going live on March 14th and that's going to in well it will root a huge amount of value to our community um and also for the entire AO ecosystem right so we're really excited for that um the next thing I have coming up and I hope it's going to work because we have some voiceovers over the videos is we want to actually show people what we've been building so we're going to have a couple of little demonstrations the first one's just going to show the difference between CPU on AO um and GPU when we're looking at AI inference and just how big difference the speed is and then we're going to have a look at our deterministic GPU um sort of well our hyperbeam extension so we're going to have a look at an example of both of those hopefully maybe working yeah so one thing to note on these videos is that the model is going to remain the same the Quant quantizations remain the same the only thing is different is we're going to look at an Intel CPU and an Nvidia GPU let's explore how GPU acceleration transforms AI inference speeds compared to traditional CPU processing CPU inference runs on a loader the foundation of ao's Legacy Network GPU inference utilizes IW Asm we might need that back upm runtime powering AO main net they both use the same sort of semi expected technical difficulties we're not going to have sound on this one which requires AI to refine copy GPU completion Time 1 the video the icon is that we don't have the yeah it's not loading at all um 179 s a few minutes later CPU completion time 393 while the CPU took 393 seconds the GPU smashed it in 1.179 seconds that's 350X faster on pure compute Power Plus gpus dedicated memory handles heavy load smoothly while CPUs struggle with bottlenecks in the AO ecosystem deterministic is non-negotiable but gpus face unique challenges here apus ensures determinism by first using integer quantized models to simplify Precision control second strategically disabling non-deterministic GPU optimizations let's see the AI inference comparion between Nvidia l20 and 4090 both of the prompts are the same which asking AI to continue writing after Once Upon a Time Nvidia l20 output 445 tokens end with a new life for themselves Nvidia 490 output also 445 tokens end with a new life for themselves bit forbit identical outputs and even after multiple runs the result is still remain the same deterministic while sacrificing 50 to 50% potential speed gains ap's GPU integration maintains ao's determinism standards and enables verifiable AI computation okay so I think hopefully people managed to get the first part of that video um what we did managed to demonstrate was that the GPU actually managed to process the inference 350 times faster than the CPU um and that was just with a really small prompt if we start to increase the number of tokens we're putting in there that difference is only going to grow grow um so for anyone who has been using um you know AO Legacy running Ai and I know some of you have been building on it uh some of you have used our competition pools which it is been painfully slow so having GPU on AO is going to be something that is you know increasingly important especially as I know a lot of us do want to be using AI um and the second video displays that we now have the determinism model running it is running there on the same um same um wasm runtime that is used on hyperbeam um and we have it running across two different Nvidia um GPU cards there so at this point people may be asking why do we need the terministic right um why do we need to be getting the same results and this can be used in anything across the people are developing whether you're looking at Asian fi whether you're looking at trading Bots anything you're looking for you know is going to possibly require determinism but perhaps most importantly is we're trying to build onchain verifiable Ai and one of the easiest and best way for us to prove that we can have onchain and verifiable is if we keep those deterministic standards which are sort of the core of ao's compute as well um so hopefully that demos something and we will have something live for you to actually test like I said we'll hopefully not play them again um coming back to sort of what we've done otherwise it is sort of AO forward research have been pushing the fair launch season um and the Perma big thanks to Drew for that one um why was the Fair launch important for us you know we came the AO ecosystem sort of we want to follow the same epos right AO was probably one of the first purely Fair launch tokens that web 3 has seen for for a really long time and it's really nice going back to that because early days of Bitcoin that was the idea right so we thought well we'll go with that um and we'll continue that and hopefully this also sets an example for others in the space and I know that we're not alone right there's a lot of us doing this and I feel that it's really nice to see a web free ecosystem sort of following in that suit um so we had our TG that was in December last year we had 7 million assets bridged within the first 7 Days um and then in March 14th this year we will be joining the pi index too so as much as that's bringing exposure for us we also really happy that that's going to bring exposure to the broader AO ecosystem um and then just regarding the per forward research have been really gous they are offering 3 3,142 Pi um tokens AO tokens um you know across four different companies so that's AIO that is Bazar that is uh base jump and that is us um so just come out check it out it is the QR code on the right they're really easy competitions you have a chance to win some of those it's quite a huge amount of you know money in dollar value so do just you know it's a little bit of fun as well you're interacting with a digital version of Sam Williams um and then sorry Sam and then you get to post that on Twitter as well so not only do you get to interact with him you get to share his response um we take no responsibility for for his replies on that one um but yeah please do check that out um on the left you'll find the link to our website um our mint is live delegations are open um and yeah really happy to be here um and yeah thank you guys thank you so much Connor for our next speaker there are certain people in this world who you only need to know their name and you already know everything I'm talking Elvis I'm talking Bono I'm talking Phil hey everyone okay display settings okay all right how's everyone doing see a lot of familiar faces that makes me feel really good I feel like I know everyone in the ecosystem um some of this might be uh repeat for you you might already be familiar but uh glad to talk about roio permanent Cloud Network and really what it does and and what it even means um so thank you again forward research for having me here and uh let's get into it you guys already know who I am you know I've been in the r ecosystem for since 2018 uh I quit my job you know founded PDS and started with our Drive in 2020 so it's it's been a heck of a ride um last week we launched the Ario Network mainnet um so today I really wanted to focus on that and uh talk to you guys about it benefits and maybe how you could take advantage of it if you already are not um so while rwe Prov while rwe provides this foundation for permanent storage Aro really builds the infrastructure and services to create the world's well we like to say the world's first fully permanent and Equitable Cloud Network um it's an open source cloud and it really provides storage hosting data access really without any subscriptions or centralized control um and we like to think we're trying to really set this new standard for cloud infrastructure in general um and that's based off of a few pillars we call them right so it's a a a network that's obviously resilient right no single point of failure uh incentives that can keep the network self- sustaining uh so no single team could you know go away and you know falter the network and it's censorship resistant right no single team person node Etc can take anyone's voice can remove their content and so on um all of the data stored on this network is stored on our weave right so it's timestamped tamperproof and it's there indefinitely very unique property that you just don't get with other web 3 systems or web two for that matter um one of the amazing things I think uh about this cloud is it's inoperable interoperable right um meaning it's can integrate with other chains you can bring different kinds of wallets to it you don't just need an R weave wallet it supports different kinds of signature types um so that's really useful for bringing in people uh in other ecosystems you can pay in different kinds of currencies it's not locked to one specific token um and it's got this very familiar tooling uh setup that you might uh be familiar with with if you're in web 2 uh cuz it uses languages like JavaScript typescript Lua Etc there's no special language you need to learn uh to build on it um it's Sovereign by default right users own their data the apps uh deployed are user rights preserving everything is cryptographically signed so it really provides this strong digital ownership um and we've designed the whole system to be on premises right it's not built to just be run as a SZ uh it's built to run at any kind of scale whether it's a tiny you know Raspberry Pi or you know really beefy servers um and finally it's transparent right and this is something you don't typically get with the normal clouds right everything is open source uh there's a clear record of data Providence for any action that happens because it's all stored on a public Ledger which is our weave um so our vision is really based in these in these principles and you'll see this uh be applied through all the technology that we're we're creating um as as I mentioned both cloud types of clouds have all the tools you really need to build modern apps that scale right um and whether it's doing things to store data index it retrieve it or just a generalized compute platform where you need to create anything you want um really both clouds offer that um sure the traditional Cloud does you know have things like slas and you know strong customer support and it's widespread it's it's adopted everywhere right every business has some connectivity to Some Cloud somewhere um but really the advantages of the permanent Cloud really can't be denied whether it's the payment models the censorship resistance the immutability Providence there's a lot of really unique properties that you get Beyond just storing data for a long time um let's maybe drill into some of the specific services that the Aro Network provides um because it does a lot right gateways aren't just accessing data um they do uploading right whether that's layer one r weave transactions or layer two data items or bundles um that you know the Gateway comes with a bundler that can do that right it could also uh has very intelligent chunk broadcasting um as you might know you know data on our weave is all broken down into transaction headers and chunks so the Gateway comes with a specific broadcasting feature if you do need to you know re-upload or or broadcast chunks out to the broader R weave network of course gateways I think they're uh first and foremost used to access our weave data um and it does that through a contiguous file cache right again data on our weave uh is made up of chunks so there has to be something to assemble all of these so the Gateway is responsible for doing that and it'll cach that in a in a local cache for really fast access it'll do things like render your files your PDFs your Json your documents it will stream your videos your music right it will render your full apps and Pages thanks to or we've manifests right it'll do uh it'll provide this access optimistically meaning if data has been uploaded through a Gateway and it hasn't been mined on chain yet the Gateway will still serve that out right because users and apps expect it to be really fast um and then finally it's really providing an access layer for the rest of the rwe HTTP API right so whether you're fetching prices transaction information wallet balances you could do that all through a Gateway um it features the domain name system RNs so whether that's buying and managing configuring domains resolving them looking up uh IDs for friendly names um all of that is kind of baked into the the DNS Services um gateways you know as you might be familiar with are also responsible for indexing all of that data by default a Gateway will index everything on layer one um but Gateway operators have the ability to tune any kind of layer 2 index whether that's specific applications or you know maybe just personal data um it's really flexible uh to do that um and then gateways provide a graphql interface on top of that to make it really easy to search for that data um and similar to optimistic caching a Gateway will take any of the data that's uh ingested through it and it'll optimistically index that as well so you can find it really quick um you know I have a box here for just general compute technically these pieces aren't in the Gateway itself uh it does ship with an AO compute unit right now that'll probably change in the future as things get more uh hyperbeam ified um but as I mentioned before you can use other chains for some of your smart contract work right AO is the native platform for The Rao network but before AO there were plenty of teams building smart contracts on ethereum referencing data stored on rwe for nfts and and other things um and finally the Gateway comes with a set of some management tools um whether that's monitoring the Gateway itself it comes with a nice grafana dashboard it has an uh an ETL process to extract any data uh transform it load it into the index uh we use parket for that so it's pretty interesting there comes with tools for Content moderation um each Gateway operator uh basically gets to decide right that's one of the differences between a centralized cloud and this permanent cloud is the content moderation has been kind of democratized to all of the operators versus a single Central controlled uh Corporation um and finally the the last and possibly one of the more important uh management tools in the Gateway is the Network observation and Reporting System or the the OIP um and that's what gateways use to monitor one another report on the uh the performance the availability uptime and that's what determines what rewards the gateways received so clearly there's a lot of things happening uh in the Gateway it's taken us a long time to to build all these pieces um but it's yeah similar to maybe AWS that has a whole Suite of services for developers to use um I think a lot of you are probably familiar with RNs I don't know maybe raise your hand if you haven't RNs name raise your hand if you've like set it as your primary name okay I want to see more hands up for that but yeah RNs is really one of the key features that uh and innovations that we came up with friendly names for apps Pages data identities maybe hash paths I don't know if that makes sense maybe Sam I don't know you got to think about that um you can buy a name with a lease or purchase it in definitely and all of those proceeds go to the Gateway operators there's no Dow there's no Central company that receives this and then decides how to you know fund things um it's all controlled by a protocol so really unique to RNs versus other web3 name systems and you know the cool thing is it's not just a name for an identity it's a name for data so you'll be able to access it on any of the gateways in the network there's I think 280 or so gateways so whether it's you know R drive.

valero.com or R drive. we.net so on and so forth really provides a high level of resilience um and then if you're buying a name you also have subdomain uh space or we call them undern names so you can uh yeah deploy other applications assign different identities similar to if you purchase a domain name you might have many identities that leverage that domain um and then finally we built a really easy to use app for anyone but there's also a powerful SDK if you wanted to build your own version of this whether it's buying names or you know resolving primary names Etc um lots of names out there already here's just a few examples of the different front ends apps pages is uh shout out to any of the teams who are on here I did not mean to exclude anyone um but yeah and now what RS really enables is what we call the wayfinder protocol or ar col SL slash so you can actually offis the domain name completely by using this new schema uh so effectively you don't need to even remember what domain to access your data on um we released a little uh Chrome extension called wayfinder that does this um but apps like wander also have it built in um so you could just type in AR con SL SL and your name and it'll route you to a high performing Gateway so what can you do with it I mean lots of things uh you know we see a lot of websites and apps you know I just showed a bunch before um lots of different games are getting built um whether these are just the front ends or even the assets too um I know there's some folks in the audience working on some really cool games now um file storage that was what got me into this whole thing with our Drive um you know the the beauty of that is you have all this Rich metadata for any of the files that are stored um so that could include things like your uh any information architecture or a license right which is very applicable in today's day and age when content is getting scraped by these AIS and you know users are not necessarily uh uh you know not the they're not necessarily getting referenced by these AIS so storing any kind of data has always been the The Sweet Spot um and it doesn't have to just be you know static data it could be nfts or digital assets again whe those are AO based or on other chains um so we see a lot of use cases we're always trying to to grow these and this doesn't necessarily have to just be web 3 right there's plenty of uh web 2 use cases as well um again the the system has been operational for a long time I know we just went into mainnet recently but this is just an example shows how uh it scales right um 200 plus terabytes of data billions of data items we all see that number growing every month um and you know just an example of how many unique users are connecting to R we.net almost a million on a monthly basis right so that's kind of crazy to to think about and sure that's not people uploading data rather uploading and consuming it right um so it's it's it's great to see these numbers grow um but yeah are you actually ready to build something on it there's lots of tools that we have whether it's the AOS SDK to integrate with uh RNs or the network um have the wrong link there you don't go to the wander app for the rdk um the turbo SDK is for any uploading of data or paying for those uploads again whe whether that's with a credit card or rwe ethereum more cryptos coming soon um if you're looking for a no code app deployment solution highly recommend uh rlink again I messed up the links here metal links is actually a nice little homepage Builder um perm web deploy is excellent if you want to try a like a GitHub cicd integration so you can update your app and then have those updates get pushed right to the cloud um and then we have all kinds of Quick Start guides for helping everywhere in between um if you're not a builder that's okay because you can still use this pretty easily um if you want to actually write data of course you're going to need a wallet highly recommend wander uh the new rebranding of of rconnect um although as I mentioned before wallets like metamask are also still supported and um more support will be coming um our Drive give that a shot you could just pay with a credit card make a little homepage like a link tree with uh with metal links um of course get a domain and you can apply it to your either your identity or your homepage um or even just listen to some music right I mentioned before like gateways are great for streaming um someone built a a little Grateful Dead uh tribute app or you can you know stream some some cool Tunes there um and this point is really you know this is a cloud that's not just for technical people anyone could really use it um I slip this in you know just now just wanted to give some hints at what's coming next um you know after mainnet um it's been mentioned before but we've been included in the perm web index um so really excited to leverage that as a way to distribute our tokens to the community um in a you know permissionless protocolized way um you know I showed the big chart with all of the different Gateway services not all of those have protocol incentives right so we need to uh build out more of those um we need to onboard more Gateway operators we just completed uh the Gateway operator incentive program and got I think 29 um you know bigger teams and projects to run gateways it's not going to stop there um we have a whole lot of hyperbeam ification that we need to do and we're still unpacking that and understanding how we could uh apply the different functions of a Gateway onto AO core and how we can maybe leverage te's and and some of the other features there um we're going to do a big push to integrate all of our tooling into third party libraries we've built a lot of stuff it's time to like get it out there and get it integrated um we're now now that we've launched mayet we are going to uh start a new grant program as well so uh more to come there for any Builders who have some really big ideas that they're looking to explore um and then really it's just about adoption again I see a lot of familiar faces in the crowd I would love to see more people who I've never met before right so we're really going to focus on uh yeah getting new users new developers new Founders and just you know broaden the the usage of R weave AO and the permanent Cloud um so yeah invite you all to obviously follow us build with us uh give a little QR code scan here fill out our Partnerships form if you're really interested in working more closely and we'll be at the conference the rest of the week so come say hi and yeah thank [Music] you e e record that I just set is currently updated uh with the record that I wanted to set so just kind of checking afterwards and one thing you may find is that when I set base names or I do deployments or anything like that on the deployment I do not build into the uh the script the build then deploy specifically because I want to see the size of the file or size size of the uh the project before I deploy lots of times I've been building applications I may forget man I just put this image in there and it's huge it's 25 megabytes and then I run deploy I'm like oh crap like why is this build folder so huge make those modifications before it actually uh runs in there so uh what you end up having to do is you end up simply just running the PM uh pmpm build so you run the build and then right after that you just PM pmpm deploy and that does your deployment from there o there we go okay so we're all good and all you have to do is just copy your transaction ID now I want to go ahead and just set an A undername in here I've already preset some of that here so I'm just going to change this transaction ID here save that pmpm and then set undername there you go so you've already at this point you've you've uh built the the application you've already deployed it then you've already set the undername so at this point now I can actually just run over here go over to my undername and we're good to go that's super fast has anybody seen a fully decentralized front uh deployed that quickly I mean that's really really fast so this starter kit is ready for you right out of the gate so if anybody wants to use this for the actual hackathon that's I mean that's what I built it for so uh for anybody that's building in the hackathon I really really encourage you guys to use this some upcoming features I'm working on it's just more repos specifically for other ecosystems I actually had a lot of fun yesterday at The monad Hacker House grabbed somebody else's starter kit uh uh added all these implementations into the uh from their starter kit into mine and then deployed that and a lot of the monad uh developers really really loved it so you're going to see more repos like that you're going to see additional features instead of just being able to pay in Turbo credits you're going to see additional features of being able to pay in other currencies uh as well and then last but not least oh sorry uh there will be some AO features in the backend pretty soon I'm uh working on a repo that has that full stack application the experience of having everything deployed uh on the uh Ro ecosystem using a starter kit like this but with AO functionality using AO connect and then last but not least uh repos for nft uh image uh deployment and transaction ID management I mean imagine having a whole folder of images uh you be able to drag and drop that and just run one single script and then have it save all of your transaction IDs or your URLs to those images in a manifest folder I mean that's the easiest fastest way you could do something like that so aside from that if anybody wants to use that starter kit it's right there go ahead and grab it from the QR code and uh that's it [Applause] guys I did that in s minutes right yeah [Music] the coolest things I've seen and I've seen a lot of cool stuff today uh but that is not to say that what Tate from Community Labs is about to show couldn't one up it Tate we're so excited to have you here and people watching at home people in the audience I know we're at the halfway mark so give it a big stretch and get ready for two more hours of some of the coolest stuff within web 3 Tate take it away man thank you thank you drew all right thank you guys for having me um I wanted to start today off with a story um our team has been building on AO for nearly a year and a half now which is kind of crazy to think about looking back on it um but over the last year and a half the the state of AO has changed quite radically uh it's become more stable it's become more scalable it's become more performant but I was reminded recently of a movie that I watched probably 5 years ago um was one of the Superman movies uh and in the Superman movie they were trying to understand you know Clark Kint came in on this spaceship they were trying to understand what the spaceship was made of um and essentially thank you Sam uh essentially the spaceship they the scientists couldn't figure out what the spaceship was made of and so there was no way to actually characterize it they couldn't describe the material because it didn't exist on our periodic table um and after building on AO for the last year and a half I continue being reminded that we are dealing with what seems like alien technology because of how far out in front of everything else in this industry AO is positioned as um so you know we're talking about a lot of complex terms here a lot of new Concepts New Primitives but the truth of the matter is if we try to put it in the box of everything else that exists in crypto today it just wouldn't do a Justice because it is so different um and so that's one reason that I get excited about AO I think I'm excited about building on alien technology I guess that makes Sam an alien as well uh a little a little bit no offense um and I wanted to spend today to talk for a second about one project that we're really excited about that kind of pushes the boundaries of AO um something that's been really cool there really three features inside of AO that we're especially excited about uh tees weave drive and then web assembly um maybe going through each one by one sorry got ahead of myself uh tees give you the ability to run computations in in private in in a verifiable and deterministic way uh weave Drive gives you the ability to then go look at R weave and fetch information uh and that information could have come from anywhere could have come from inside of R weave it could have come from the web uh and and use it as if it were data on your own machine um which is like tremendously powerful and then web assembly so generalized compute uh in all kinds of different programming languages that can compile down into web assembly it's sort of the standard outside of crypto uh and now AO makes it possible to use inside of crypto which gives you a lot of abilities uh I'll go into that putting it together though one of the interesting use cases is building agents that are capable of signing and posting transactions from other uh from inside of AO onto other blockchains um agents are are really cool in crypto right now and it's insane that you can actually build an agent to interact on other blockchains inside of AO as a decentralized supercomputer so how might you go about doing that that's some research that we've been doing um the first step would be to generate private keys with inside of the te uh the reason a te is helpful for this is because if you're generating a private key and Publishing it to a blockchain uh generally that information would be public and you don't want anyone to have access to your private Keys um a te is perfect for this because you can create an environment where those keys can be generated uh deterministically in a verifiable way uh without being revealed to the public the second thing you could do is you could leverage weave drive to read data from R weave about other blockchains so instance Redstone is uploading a lot of data about other chains into R weave kive is doing the same thing thing we can pull that data into a process on AO through weave Drive um and use that to be able to make decisions for instance if we're building this cross chain agent to trade you can make trading decisions off of it uh and then third the intelligence so web assembly wasum 64 uh it gives you the ability to take large language models like llama in C++ compile them into web assembly and then run them inside of AO um huge like you can't do that anywhere else uh by a long shot and so we're really excited that that's possible in AO uh because it enables these processes to make decisions that are informed off of data from other chains in a way that is cost effective super fast and decentralized um and then fourth you can also generate and sign transactions uh with those private keys so once you figured out the decision that you want to make you have to assemble a transaction and sign it uh you can use the private keys that you generated inside of the te to make that possible um which you can then cryptographically sign you can also use uh like MPC technology if you wanted to to make that even more secure and then fourth you could use some type of Oracle to take the data that was returned to you from AO to post onto another blockchain or you could do it yourself if you wanted to um that's really insane you can make Bitcoin transactions or ethereum or salana transactions from inside of AO why does this matter everyone is talking about agents right now um but all of the agents that people are talking about are evaluating things offchain they're making decisions offchain they're running on centralized servers they are not inside of crypto They Don't Really act in web 3 aside from the fact that they're generating transactions and posting them um it's not scalable it's it's if anything it's false marketing uh and AO is the only place where that can be done on chain um one more just fun tidbit I guess we've been working really hard on building a better user experience for accounts inside of AO uh obviously Community Labs is the team behind wander uh thank you to everyone on our team that's made that possible and thank you to the community that has also adopted it over the years too um excited to announce that coming soon there will be a single sign on option introduced that will make it even easier for your consumer facing application uh to generate rwe wallets and make transactions with a ux uh that is equivalent to web 2 but with all of the advantages of AO and R weave in web 3 thank you um with that I will hand it over to liquid Ops appreciate it hello uh can you hear me okay um my name is Laura and um I'm the founder of liquid Ops which is a leting and borrowing protocol um built on good sorry uh yeah so my name is Laura I a founder of liquid Ops we're a lending and borrowing protocol modeled off of compound V2 uh built on our we AO um so yeah okay so a little bit about our story and how we got here today um so yeah me and Martin had been working in the ecosystem um for for quite a while Martin has kind of been in since his infancy um and one thing like we both realized was that we held our personally as well as um previously I was a fellow Community Labs uh I worked on a project called oen basically a way for anyone to be able to um into ourwe with like a Google account a Twitter account and actually be able to onboard into the pmer web without the use of a private key um and yeah so we held always we personally and we also had a lot of people coming into our project Discord um you know Martin running the arc connector one me with the oent one and one of the biggest things um that you know they all had in common was like so we hold our weave tokens how can we actually earn yield how can we use uh this liquidity in um you know different like products um so yeah that kind of just got us thinking about like you know how can we generate uh yield you know at the time the only way to really earn yield in this ecosystem was contributing uh to liquidity pools on dexes um and all this is uh like you know super cool to do we really wanted to have some of these more advanced products that you know you could see in places like etherum actually in this ecosystem um so yeah kind of going into how liquid Ops Works um I think there was a huge misconception of you know where yield actually comes in comes from in a lot of these different protocols you know some projects uh they mint tokens and give them to you know their users um and so yeah liquid Ops essentially it works off of three uh kind of Parts you can think of it so you have lenders um like me and Marton that you know held our weave like our users uh that wanted to actually earn yield on those tokens um so for them it's super simple uh you basically take their um tokens you know you add them to a pool and basically make them available um to be lent out um so that really brings us to where the yield actually comes from which is borrowers I'm going to go more into some of the borrowing borrowing products in the next slide um but yeah it's cool um basically uh you take the lender's liquidity um and then you know you allow borrowers to actually use that liquidity uh in different products so we're over collateralized lending and borrowing protocol so the whole idea is borrowers um basically have something called a loan to value ratio um you know you often see as like LTV um which basically just means that in order to um borrow on liquid Ops you need to add collateral um so for example if the loan value ratio was say two times um for every $100 you borrow on liquid Ops you basically need to add $200 worth of collateral um so you know someone would lend into liquid Ops um those tokens would be available to be lent out um to borrowers borrowers would add collateral um two liquid Ops um and then the loans to Value ratio would determine um how much they can borrow um so yeah though that's kind of uh lending and borrowing is really simplest um simplest form but now we have like Liquidators um so say you were like borrowing um in in traditional Finance in web 2 you know say it's like a mortgage um you go to a bank uh and you say you know I want a mortgage um and the house actually acts as collateral um if if you know you didn't pay your mortgage the bank would repossess the house uh and pay off the debt of that mortgage similar sort of thing with like payday loans um you know you get kyc um and then obviously if you don't pay back that loan um you know the government comes after you and knocks down your door so with something like lending and borrowing protocols um basically the if the CL drops below a certain loan to value ratio a liquidator has an incentive um to basically buy uh the bad debt of the protocol um and then basically um yeah have a fee be imposed on the borrower um the Liquidator gets that fee um and basically repays the initial um like loan that the uh borrower took out um so yeah really simple as simple as form that's basically what lending and borrowing protocol is um so yeah going into some of the things we can create and this is the really exciting part apart from you know having lenders be able to actually earn um yield on their tokens um so we're kind of focusing on three main points uh at liquid Ops um first one being like Leverage actually being able to um enable people um you know to trade with more than they have via like Leverage this is kind of like the basic stuff um being able to you know uh long tokens um and actually be able to plug that liquidity into different trading Avenues um so that's kind of like it uh kind of basis um what liquid Ops enables with with products um so yeah the second kind of thing would be um vaults um in Chains like ethereum you have really really Advanced um defi products uh that we basically want to Port over into this ecosystem the way I quite like to think about it is um we're like an island of of liquidity that is never really been used in defi um and through the rise of AO we really think it's going to be a super super interesting concept to be able to develop these things like um Vault so all a vault really is is core is like imagine um you know instead of Lending into like a a liquidity pool you can lend into a vault that has a certain predetermined strategy um you know so it could be like an Arbitrage Vault so what you could do um is have a vault basically on liquid Ops where you could lend um into the Vault um and then behind the scenes uh you know say you um you know you lent into this Vault you borrowed um you know say like um Astro USD and you added rwe's collateral you then swap that Astro USD um into like I don't know USD and then deposited that usdc into RV you've basically got an Arbitrage um where you can utilize your awe tokens actually on these um you know uh these projects like arve um so yeah being able to basically have these customizable vaults in this ecosystem we're super excited about meaning that effectively we'll be able to bring some of the yields that we've been seeing in ethereum to this ecosystem from day one um and yeah shout out to autonomous uh Finance for that really cool uh image of an agent so this kind of takes like the vaults concept um but really really brings it into what we're trying to do in this ecosystem with a hyper parallel computer network um so actually we're going to be creating these things called like intelligent vaults kind of smart vaults um where you know instead of having like a strategy um like you know that just kind like an arbitrage strategy you compare a large language model that actually learns um you know adapts over time using um like you know this liquidity that people want to be able to lend out in these different um kind of more advanced products um so yeah that ties us quite nicely into I think our mission with liquid Ops at the minute um and yeah really the first one would be bootstrap the liquidity we have in this ecosystem you know this like kind of island of liquidity that's being untapped in defi um be able to you know create some incentives for people to come into the ecosystem um to you know earn yield um I think you know a really good validation that we've had recently um is the stakes ethereum and dbridge um you know one of the first times where you know as an ecosystem we've been able to have like uh like a really really interesting incentive for people to pour assets over into this ecosystem so at liquid Ops we really want to carry that on and sort of um yeah have the cherry on top that for example if you come into this ecosystem with d you can earn a new portion of AO and then you can also lend the tokens um uh and earn that interest or you can yeah use some of these different products uh that we're going to make um and yeah obviously yield being able to earn a sustainable yield on the on on your we tokens you know like me and marttin initially had uh and our users being able to yeah have um have a really cool sustainable yield for this foreseeable future um like we have in some of these other chains um and yeah enable those new defi products um like you know being able to actually trade with leverage um in this ecosystem natively um and you know being able to you know plug that liquidity into these different um strategies that have found product Market fit in other ecosystems and yeah being able to pair the liquidity we have um with actually you know things like autonomous agents um I think is going to be super super interesting um so yeah thank you very much um recently we just uh shipped our test net kind of gathering feedback um on yeah uh what this ecosystem kind of wants uh with lending and borrowing um and yeah feel free to check out these links as well as I'll be doing a talk in the next uh I think next little bit of uh more specifics of liquid Ops and giving a little demo on it but yeah thank you very much thank you so much lurmer it is a community Labs paloa up in here because next we have Kadar Kadar take it away man hey everyone thanks for having me let's see if I can do this I'm Kadar FR of astol labs I'm going to talk about driving liquidity for AO and what that takes and what we're doing on our side of things so I could talk about a but I think Sam and Tate and everyone else kind of covered it oh let me just cover the basics AO is a new city that requires new utility and essentially has unlimited potential but one of the key things we think it's missing is liquidity so that's what I'm going to talk about today so for any economy you essentially need a way to transact save and essentially navigate the ecosystem and we think we can build a robust model around Astro liquidity pools stable coins and kind of yield generation activity and kind of bring a lot of value to apps like BGA liquid Ops and game f social F anything like that I'm going to get into a little bit of what that takes honestly speaking we kind of need it right now we're transitioning to mainnet that has a lot of quirks but is going to be more scalable meaning that we need liquidity infrastructure that's also scalable along with the train itself and that's what we're going to try to build so why does liquidity matter um any functioning economy doesn't matter if you're a blockchain a city a country you essentially need a way for movement and value to be exchanged and the Prem needs a stable value and uh liquidity infrastructure and that's what we think we can build so essentially this is how we think about assets liquidity and we want to create this little flywheel we're going to start with assets first stable value second yield generation and then value share with ecosystem apps developers Builders agents and then we're going to create more assets and kind of Kick this Flyway into gear so everything kind of keeps rolling and right now that's what we think we can start with USDA and partners like BGA Perma Swap and liquid Ops of course so we're thinking about stable coins a little differently we think in the early days we need to kind of boost trrap and unify liquidity and kind of have all dollars all chains into one stable on so liquidity doesn't fragment we can see a clear example right now where liquidity is kind of fragmenting between QR and War some people use quantum some people use aox and we think a better model would be to essentially introduce way to defragment and that's what we want to do with USDA it's going to be a unified stable on AO and we're going to push liquidity to its extreme essentially unifying everything so if you want to trade save transact if you're an agent if you're a user if you're a builder USDA would be the deepest liquidity pool you can use and access and essentially capture value so high level there's I'll send you a couple of documentation if anyone wants a n out about liquidity but we want to build what's called a liquidity engine and kind of um have three pillars to this engine uh first is going multichain having multiple Pathways into AO we want to create a on ramping Gateway essentially we need a way to connect the real world to AO right now getting into AO and R is really tricky and we think we can cover this pillar very easily with USDA and the last bit is around tokenized assets I know no one wants to hear it we're all about the tech but we need more assets on chain everything's been tokenized it's time stocks everything and we kind of want to bring that engine into AO allowing different ecosystems and different economies to exist on top of AO whether you just want to trade stocks or different assets or you want to build um social apps or games right now there's a few apps live on AO Legacy net and we think they're going to transition to main net and more apps to come but the kind of key areas we're seeing is marketplaces Launchpad dexes lending protocols and social I haven't mentioned gaming I think you're going to be the first gaming app or maybe second um but that's going to be an interesting category for us as well um what USDA does differently is essentially we capture some value by the collateral locked in US da and that's what we're going to bring to AppStar adopting USDA kind of having this shared value exchange and then kind of kicking off this fly where we capture Revenue instead of being extracted by other centralized stable coins we're bring that value back to the ecosystem and then apps can kind of again use that to uh expand the team they don't need to fundraise as much of course there's AO funding raise but we think we can push that to the extreme by having um stable yield generation brought to AO for ad Builders there is a savings rate to be brought at some point in time for users of the the train but that's you can go anywhere to get savings rate we think on AO in the early days we need to bootstrap early developers and uh teams we've been saying USDA is coming soon for the longest time it's coming soon this time and I think within a couple of weeks we can announce how USDA looks like what version one looks like what version two looks like and we're going to be working with key ecosystem teams like liquid Ops the autonomous team and peros swap team to essentially bring USDA to liquidity pools so people can trade say essentially capture value in Native USDA on Al these are some of our uh launch Partners including Community labs and wonderand and we think we can push this to the extreme by partnering with teams that are native to the ecosystem that are here in the beginning and we'll be here a decade from now and that's what we're building towards we're pushing some updates let me be honest Quantum is been finicky and but we're pushing some updates over the next week or two to update the user experience up how it works the speed of execution and when we launch USDA we're just going to revamp the whole platform and kind of launch version two which should allow you to bridge any asset across at least four or five chains we think we can support by implementing a new intent framework we haven't been publishing a lot we've been kind of heads down building but soon we'll be starting to publish light papers white papers on how the intent layer would work into AO and kind of the model we're trying to um complete Tri Quantum reach out to me on x telegram reach out to Tate I guess he'll let you know what my details are and yeah thank you for the time and I think that's it okay guys I don't know about you but I could go for a Hu right now anyone got one for me Sam no not so much uh our next speaker is Miss wenin Turing talking about the future of sovereign Finance winin it's a pleasure to have you hello everyone my name is tring and from RV SCP Ventures are taking in charge of the Westman here and it's very happy to see many old friends and new friends here I'm very excited to introduce you Fusion five protal and it's native token ffp before we get started let's think about one question if Defy is truly decentralized why the users need to deposit their phones into a smart contract in particles like Unis Swap and compound users need to authoritie the funds but Fusion five protocol will completely change this fishion VI protocal represents a new model in decentralized finance the fourth generation defy which sves the interoperability issues of traditional defi will bring a new concept of individual Finance Serv you defi settlements via note flow across a synchrony Network especially AO after years of development of defi defi have different stage from 2018 to 2020 so they only decentralizing some Basic Finance information like Unis swap RV and also compound then the second General generation defy comes here improving Capital efficiency calized Alm pisto toking mat and also the third version of Unix swap when time comes to the third generation Defy is almost after 2022 enhancing the collectability such as Egen layer pandle and Radiance although defi have improved it still facing some limitations let me give you some example uh the liquidity fragment so different protocols have uh different uh liquidities so make them um very hard to get efficiency so uh the resource is a very low Capital efficiency and users are also restricted so every smart contract has their own rules so uh the the contracts take all through the Financial operating instead of controlling by users themselves and also lack of inter interoperability so different protocols follow different rules and it's really hard to integrate to address all this issues Fusion five protocal was born but how Fusion five protocal works Fusion five protocal is not another Dax or learning protocols you can think of fion protal like U Financial Lego kit you know the secret weapons are nose and agents you can imagine that the no are a universal Market a universal morning you can trade learn earn all in one language normal conversations so all the operations are based on node flow using a common data structure for simplification and unified set layer and enables cross cross protocol and cross Financial scenarios interoperability and everybody knows the ability of AI agent is very powerful and AI agent is your uh robot assistant you can set your rules and make them to excute and phone is still staying in your own wallet users can create their own Financial agents without need any custody so for example let me give you example if you swp token a to token B and borrow the M Ali normally they will need five steps and three approvals but with fishion five particle there only will be one step and zero approval so that's what Fusion five protocol do why it matters is about Fusion five protocol s the problems of interoperability and we do care of the user certainty and we improve the capital efficiency but fishion F protocol is not just a concept it has real applications for example Perma swap the biggest Dax AO native Dex and has a very large transaction volume and peras swap you can set your own fees FS St your wallet and the profit the profits will come directly and also Legend set a bank you can set the race and let all the AI assistant to help you to management your phones and also ringbox Rim boox is a AO native Minecraft game and easily to build an integrated trading marketplace with Fusion fire protocal you can think about that in the future every defi protocol can upgrade um with Fusion five protocol just like upgrading your network to 5G and that's what going to happen in the future let me give you some datas about after Fusion 5 protocol launch and after protocol launch just about few weeks uh there are over 29 million total settlement mon and peras volume is just over 18 million and over 75% of a volume is in Perma swb and AO liquidity is over 67% H 4 million so I believe that in the future Fusion fire protocol and Perma swap will have some an impressive data performance and to continue to maintain the leading position and also for sustainable development we do have government token the role of the FF token is to incentivize the protocal participate participates government's the uh ecosystem development and also Bank board the fourth generation of defi fishion V protocol and ffp token uh you can imagine that uh FP token is just like the fur in the ecosystem um with uh the the fire RS as you can see that there's no VC and no team allocation the total Supply will be 18 million and 5% is for air drop for the early adopters because they made a great contribution in the past few months and 1% is for initial liquidity and 94% is TR to mine so we do have having circles so the supply will reduce 50% every four years it's just like the Bitcoin model scales like Bitcoin the early bird wins and we do have incentive about the transaction is mning so the more you trade the trade to mine every time a node is settled the Creator receives ffp as a reward and 94% from trade to mine more trades more ffp and you can also St FP to earn to gain V FP tokens and you can also vote like a shareholder and all the profit generated by the protocol is stored in the treasury and funded by the sment fees so by sticking FP tokens users can obtain VP it's just like the curves VF token model so in summary fft token is not just a reward token but also a garment too for the long-term and sustainable sustainable development for the uh Fusion F protocal ecosystem so fion F doesn't c unit and a but make them better in the future everyone is their bank exchange fund manager and this is the fourth generation defy so will you still keep trusting others to manage your FS or take a contr show by yourself together let's explore the fusion of the future of decentralized Finance if you are interested in Fusion five protocol please follow our official Twitter here and you can also scan a QR code to join our community here and we hope to build with you together fishion five protocol is all you need is the fourth generation defy thank you very much [Applause] we have people from all over the world coming here to speak at our conference I appreciate you so much Turing uh next up we have Allan from randow Allan I can't wait to hear what you have to share thank you so much thank you can I pull it out yeah okay okay all right let's get started here sure all right good afternoon eth Denver and especially good afternoon to the rwe AO Community I'm Alan and at randow we are solving the onchain randomness problem once and for all before get started I do want you guys to answer a quick question heads or tails if you think I'm going to flip heads go ahead and raise your hands all right and Tails who are my Tails all right about 30% Tails interesting let's see that was a good flip my flipping is a bit off today there we go sorry about that we got Tails now that was interesting as the coin was in the air we were all very uncertain of what the outcome would be and what's also interesting that we we weren't able to influence that outcome once we'd flipped that is the heart of what Randomness really is and at randow sorry if only it were as as simple as flipping a coin to generate Randomness on chain the onchain randomness problem has really plagued the crypto industry for years at this point when we first started out we were trying to use math.

random which really is more like math. pseudo random which doesn't quite work in blockchain since it's a public open and transparent environment where you could already see the seed on chain next up a few people tried to use the block has which made some sense initially it sort of starts to make sense as the block hash it's a hash and the hash output is typically random however when we get down to it miners have the ability or validators have the ability to influence that block hash so that wasn't quite going to work for us and more recently the crypto ecosystems come to use oracles as a solution for Randomness which works to some degree unfortunately when they have high costs for no particular reason and they're centralized you actually have to be able to trust that Oracle and that they're not going to use their prior knowledge of the randomness beforehand to gain some Advantage at randow we're solving the onchain Randomness problem like no one has ever seen before we are crowdsourcing random values and aggregating and verifying them onchain with our horizontally scaling Economic Security model we're able to guarantee that no one in the world will be able to predict or manipulate our random outcomes let's find out how we do it you might think these you are a bit familiar here we have murus and agent trunk we're going to run through a few scenarios to understand the security components that exist in Row's protocol to generate us highquality Randomness so first up both of these individuals murus and agent trunk will have to post an amount of collateral which we'll be able to incentivize them with in a simple solution murus may just post a random value on chain we'd wait a sec and subsequently agent trunk would post his random value on chain with some of the properties of entropy we're able to effectively say that random plus random equals random a new random value that is some of you may have noticed that we have to remember here it's murus and agent trunk agent trunk is he's malicious if he sees any opportunity to manipulate or predict the random outcome beforehand he's going to take it and once murus has posted his value onto chain agent trunk is able to see that value and he can change what he's going to post a chain in order to be able to predict the end output which is not a final solution how we solved this problem at Rand is actually very interesting because we use ao's hyper parallel capabilities to be able to do it instead of posting murph's direct random value onto chain what he'll do is he's going to post a Tim locked box and that box it has a timer on it once the timer expires the random value within it will be revealed to the entire world now agent trunk doesn't quite have an opportunity here does he because he can't see the random value inside the box and an additional component here is that if he waits too long if he waits until murph's box has opened itself then he's going to be severely penalized by that collateral that he posted earlier so acting in his own best interest and the interest of the protocol agent trunk will have to post his is Tim locked box on the Chain as this continues the timers both expire and release both of the random outputs which then we can Aggregate and verify them all on chain that solution was solid it would work it would be secure and we would be able to have Fair random values but at randow we went a step further we understood that our users were going to need quick fast random values and they weren't going to want to wait for those boxes to open this is where ao's hyper parallel capabilities come into play since AO has parallel processing abilities we're able to create a key that key is a verifiable delay function which allows us to open the box after the fact but before the timer assuming all all parties have already posted and committed to their random box with the values opened on time we'll be able to generate our new random value just like that and what's truly amazing about this protocol is that its Economic Security scales horizontally as more node Runners are added to the crowd we have additional security per their economic stake in the protocol and the most amazing piece is that it scales any honestly that means as long as there's a single honest actor a single agent a single murus in a crowd of Asian trunks that were able to hold Economic Security this is unlocking new possibilities across AO for types of applic we couldn't see before nondeterministic decentralized applications we saw a few of them here today and I was adding to them in the list earlier where we have gaming all over AO gaming has all sorts of aspects like critical drops or Critical Hits loot drops and plenty more we have artificial intelligence where apis network was able to place AI on chain but they mentioned themselves that they couldn't have a random component which they could benefit from we can have lotteries on chain Raffles on chain and most recently we have we've had a few randomized nft minting which is very very fun if you want to learn more about the Deep cryptography behind randow protocol check out our white paper and if you want to find out more updates check out our Twitter and if you want to run a node or be a developer and use our random values check out our website and of course remember that randow RNG is making onchain Randomness as simple as flipping a coin whoa Alan's got fans where's my where's my Applause Just because you did that guys I'm going to give you a 10-minute break we're resume here at 6:00 sharp so if you need to take a stretch go to the bathroom we're going to take about a 10-minute break and then we'll power through the last hour uh just because you're were so kind as to give me Applause all right see you in 10 minutes [Music] tick water a moon walk across gas feet Graves my main Smo your VPN daughter of the Rockies Speed high priestess scales of Justice B on a m Express next bottle neck I see Mountain [Music] Crush click CL on Fire War I SP and Tru from sheep my device to Tas breakfasts sturry Theo receipts in my cheeks encrypted and plunged PR DM I'm the dumpster CH eyes everywhere three Bon Whispering shones through Dan's the keys L fire hoing you're just a r within two confetti bit Cura leaves after sipping te [Music] CP andb conell to theiling wire gas confet [Music] [Music] in Sp daning bottle spin Professor got the future on spin tomorrow with yesterday's R bcro clicking through Blain debris [Music] drunk shout self sustaining hugs flicker the the future on spining Bott [Music] the future on spin tomorrow with yesterday's [Music] Ram build a singularity how you st through my cortesis as [Music] [Music] each got SP [Music] just a test notle voes are flowing Buffs are still growing the world is breaking down we're reaching in a new way some VC said this Tech won't last and [Music] itank City on Sport and CH we like this city bre and brains we like the city on for and Chains No rug [Music] just [Music] chain gas on the got a b a little there some Tre bro just roll his eyes where's the where's my Supply but we don't need your B gunch and CH we the city with we like the city on [Music] noain we don't need your legacy let on Sports and we been the city with bra roll [Music] she com for steel Wings smiiing L see like scissors through gel n centuries young with a l that's if your chains choking etherite found at a bu hook sucking down AP armor clinking with layer to scil win at while [Music] DRS we'll be starting in one minute starting back in one minute St protocols how to commit decentralize this she smirks mid flight as her wings tickle nod into doing things [Music] right Alliance architect with a Buffer's grp teaching protocols how to commit decentralized she smirks as her wings [Music] tickle youed a moon walk across gas feet Graves my main okay ladies and gentlemen we are getting ready to get kicked off again man these seats are super underutilized very comfy I should have been sitting here the whole time uh we've got a couple more workshops for everyone uh lurmer is back to do a deeper dive lurmer can't wait to see it okay cool hello again um yeah so in this demo I wanted to give basically um an overview of liquid Ops um have a little look at our test net and try to speak about more of really what's under the hood um and really how it works um so yeah just waitting for the video to load cool okay then so this is basically our main UI uh this is where people will be able to come to liquid Ops um basically have a little look at you know um each different Market um so right now in tesn we're working with um Quantum arweave uh usdc and Stak ethereum um so yeah you're basically be able to see um you know how much has been deposited into each market how much is um available at the you know the current time to be able to lend and borrow You' be able to see how much the interest is um as well as have a little overview of um you know the amount of assets that you've actually lend um also be able to see um your borrowings um so yeah then here you'll be able to basically yeah change the token you're currently viewing so in this example using like usdc and H weave um and then yeah top right basically click um I have a dumdum as my profile nftd um we're using actually the uh bizarre um ID SDK to be able to yeah you'll have the same profile on liquid Ops as well as um Bazar um so yeah okay so that was the homepage um pretty simple I think uh this is really going to be utilized um by kind of consumers that really just want to lend into liquid Ops um and yeah here you can kind of see like the markets um what liquid opsies tvl is uh we're in testnet right now um and hopefully one day you know have $30 billion into TVR um so yeah you'll be able to see basically yeah the markets um the current um assets that have been borrowed um in the current market and uh also amount that has been lent um and then yeah here we'll be able to to basically claim uh test net tokens okay so that's basically um you know if you're a normal lender coming into liquid Ops that's basically what you'd see um so if we want to go and have a look uh um like say um so yeah lend into liquid Ops um directly all you'd need to do um on the UI is basically uh click um the supply assets button um that opens you up to this portal okay so what's going on here um basically all this UR y does is build a transaction so you you know you basically figure out how much uh you want to lend into the protocol um what token you actually want to you know lend um as well as be able to like see you know the interest you're going to get and have a little look at the fees so what we do on the UI um is we use the AO connect SDK um to basically build the transaction um of all of the data and send it on chain to AO so the nice thing um actually about building an AO and the way we've architected liquid Ops is that you know you can just be a normal person um with like an rwe wallet like wander um you know you can use this UI and lend into liquid Ops um but actually different processes can also lend into liquid Ops um so you know if you're an autonomous agent um and you have I don't know like I've seen some really cool um kind of autonomous agents in other ecosystems like I think one's called like Eliza or something like Lea um and you know like I remember a tweet I think it was by like Brian Armstrong he was like oh you know we can give you Eliza like a wallet um in natively in AO you know because processes can custody funds that could effectively mean an autonomous agent could choose to lend in liquid Ops um so yeah basically here um we're trying to figure out how much we're going to uh lend into liquid Ops uh and just basically confirm all all of the uh all of the different data associated with this transaction so on the back end um basically in liquid Ops um kind of how it works is we have two main processes um so we have something called an OT token and we have something called the controller um so as we're modeled off of compound V2 basically yeah the O token is like a c token we're called liquid Ops liquidity operations so kind of an O token is an operation token this is the receipt that you get um when you deposit into liquid Ops that basically it's a claim um the and it's a yield generating asset so if you hold like o Quantum are we for example that is basically a receipt of depositing into liquid Ops and earning interest on that token um so yeah basically o tokens um uh uh uh whenever you s whenever you send funds into liquid Ops um they are then sent back to the user um and then they can use them in different defi Avenues um so yeah pretty basic there um there's not many um you know there's not like uh many other things really to it then all we really do is whenever someone deposits into liquid Ops um we give them back this o token um so if you can imagine once um a token's been lent into liquid Ops basically yeah you'll be able to see your balances see how much you're earning and on our back end those tokens will then be um available uh to be used in loans again if you want to like unor uh sorry unland um yeah you just click um your kind of asset section and you'll be able to actually UNL that token um okay so this is where it gets more interesting um of liquid ops's architecture um the borrowing um so kind of you know what we've touched on right now is um we have uh We've lent into liquid Ops uh the protocol has issued us um an ooen a claim uh to that initial lend token um so borrowing is a little bit different um so as I mentioned in the last presentation um we're in overc colorized lending and borrowing protocol so actually you need to add collateral um before you are able to borrow so the way you add collateral is exactly the same as the way you would lend tokens um so yeah you lend into the pool um and then that basically is collateral uh that you can borrow against later on so yeah exactly the same um uh as if you were lending you build the transaction figure out how much you know you want to lend and borrow you can have a little look at your current position summary sign the transaction via wand um and and yeah so once that transaction has been sent um we actually don't send it to the oot toen what we do um is whenever we want to borrow we send it to the controller so the all the controller's job basically is to do as we are um cross collateral is to have a look um you know uh have a have a look at the loan to value ratios um of of the token you're trying to borrow um ensure you have enough collaterals to suffice that um you know borrow amount you want to do so you know back to the example if you want to um you know if the loan to value ratio is two times and uh you want to borrow $100 you need to add $200 worth of clash roll so the controller process basically requests via an oracle um just via a message from the from our um controller process to the Oracle process price feeds um for whatever you're trying to borrow and is able to then compare the loan to value ratios and then if basically yeah you have enough o tokens deposited into liquid Ops um and the Oracle has come back that yeah the loan to value ratio is fine um we basically able to issue um that that loan um so yeah being quite early stage um and lending protocols um being uh you know we basically have architected this that in the early stages of liquid Ops um there's a lot of security concerns that basically come into that um so as a protocol We have basically set out um a certain criteria um to basically deter potential vulnerabilities with liquid Ops um so one of them being um we limit the amount of um like lens and borrows you can do within a certain time frame um basically this is to not overwhelm um where AO is is at the minute you know not to uh you know allow like spam to be a reason of a vulnerability um so yeah whenever we uh have checked like the loans and value ratios we also check that your wallet address isn't spamming uh liquid Ops as well as we have a Q system uh that basically means that yeah things happen sequentially one at a time rather than you know if someone see you as far fter than the other they they can yeah have their transaction first um so that's one of them limiting in time frames but also limiting the amount you can lend and borrow um so yeah I don't really want to say the Project's name because I think they're here um but one project uh quite recently within the last couple of years it's actually made a very very good comeback um ever since this they suffered the vulnerability of uh basically a a flash loan attack um so what happened is is someone borrowed like a just a ridiculous amount from the the from the liquidity pool um and basically didn't pay it back because the collateral um was manipulated on another exchange and it basically meant that yeah effectively um their collateral that was worth nothing because the token was ridiculous got liquidated um and then they basically just kept the loan um so yeah limiting the amount that you can lend and borrow in one transaction as simple as that actually would have deterred this type of vulnerability um so yeah okay then so yeah basically now uh yeah that you'd be able to um confirm on the UI um that uh the amount has been borrowed so um you know I back back to the last presentation talking about some of those products we can create you know leverage trading using invols we have a really really simple SDK uh that you know like developers uh looking to trade will be able to integrate into the uh applications as well as natively on AO um you know having some of these uh agents be able to send messages directly to our processes um rather than having to go uh via web2 VIA like a normal typescript SD um okay then so moving on um liquidating um so we want to have kind of Diversified um liquidation uh like I know if the was like kind of Arena like you know we we're not only going to work with a certain type of Liquidator you know sometimes um Liquidators are really just like uh market makers um and I think that's actually quite unfair to the typical um web 3 user you know because actually Liquidations in itself could be its own product um you know you have some other protocols where you can have like a vault there's like a liquidation Vault so people can deposit into that Vault um they'll be able to yeah actually liquidate other users on liquid Ops um and actually you know generate some of those fees and have it you know be its own yield generating Vault a liquidation Vault um so yeah we're going to be working alongside market makers um some of these vaults and then hopefully autonomous agents um once you know agents sort of mature to be able to liquidate users as well as normal people coming onto the UI and being able to see oh my God there's an ability for me to make a 10% on on you know the amounts being liquidated so yeah we've kind of taken the uh traditional swapping UI um and um yeah basically um here you know you have the normal uni swap swapping UI um and say if you wanted to liquidate in this case uh we're liquidating Stakes ethereum so we're sending one aave uh that has originally been lent out uh the collateral has hit that loan to value ratio that triggers a liquidation um and then obviously yeah the user would basically be able to see like uh here how much they're sending and then here you be see how much they're getting back and then obviously here be like the prop profit uh but right now there is no uh active liquidations on liquid Ops um so yeah yeah that's pretty much it um so we right now yeah like I said we've uh shipped our testnet now we're working on uh Integrations um with Liquidators Integrations with borrowers and uh also potential Avenues of where we're initially going to get some of those lenders um so yeah I'll be about until the end of the week feel free to comment ask me any questions um but yeah thank you very much we're at the point in the conference where I'm just going to start shouting people's names when they come up to speak so here we go Greg you go into uh yeah mirroring I think or all right can you hear me right thumbs up in the back all right my name is Greg Albritton and uh I am a part of the AO Computer Club and today I'm going to be talking about Fair tokens now most of you here know about Fair tokens however I feel like it's really important to emphas emphasize it as a kind of source of camaraderie in the AO ecosystem we have an amazing opportunity to do something different something that was done many many years ago through Bitcoin and something now that we have access to thanks to the Token distribution and the fair launch of AO and this is really really exciting to me because one of the things that attracted me to our weave several years ago now was this High degree of Integrity right we're looking to preserve history all kinds of History all types of formats digital and so on and today we're looking to provide opportunities across the board um for individuals and not have these backdoor deals as as a primary motive so I think Fair tokens are friendly tokens and I think it's important that we emphasize this and kind of cheer um behind this this movement if you will if we can make it one which I think we can so here is a a bowl this is kugi um you may have seen the check check all right you may have seen these types of bulls um what's really cool and reason why I have this on the screen is that tokens have been broken um all these things in the Limelight today are incredibly problematic right it's not fair by any means individuals are getting rug pulled and all of this and I think we have an opportunity here to make things much better from a broken system and there's real Beauty in the opportunity here and that's where you see the gold inlay from the broken pieces of this this bowl tokens are really having a moment and I think this is the opportunity for AO and the AO ecosystem right which is fresh for us to kind of put our foot down and say this is actually how you do things this is the way that you actually get fair you know inherently high integrity token distribution and I think this can be an amazing motivator if we uh really come together on this imagine if an entire ecosystem was based on that and that's what AO is really so AO the fundamental token is a fair token you can see the stats here um you know 21 million uh Supply just like Bitcoin nice smooth uh having schedule having schedule like all of these these layers and in addition to that there's some real cool distribution mechanisms here um just by holding R you get AO right uh and then and then for those that are bridging you also have the opportunity to Min AO again they're friendly tokens uh Fair tokens are friendly tokens zero insiders no pre-sales or closed rounds it's really an opportunity for anyone to get involved right at the start and that's something I can really stand stand behind now one of the things that we've developed is aod devbot and the first module that we've created is to create these Fair tokens and they have mimic properties so um I'm interested in doing the impossible which is an actual live demo uh and ideally we can get some uh participation here uh in the audience um and uh I think we can just kind of roll into that right now so what you see right here is our interface for creating these Fair tokens um so I'm going to come in here I'm going to see if the internet will work and it does so that's cool um does anyone have any kind of random word of a token that you think may be funny to create here on the Fly anyone huh Riz Riz Riz token all right so uh let's see what happens here we're going to make Riz token one moment and I barely touched the USBC and now we have nothing on screen that's really what demos are all about all right token all right add funny quotes Okay so I really input very little details um this is actually connected with Claude uh in the background and we also have an integration with open AI um something that's really cool about the AO ecosystem is that AI is actually coming all on chain right so there's still more work to be done there and so you can imagine this entire process completely on chain which to me is really exciting um you'll see that we got yo fam time to create the risest token that's ever graced the blockchain Riz token uh the smoothest most charismatic token in crypto pure ribs Riz Vibes only so we're going to let this kind of cook a little bit um typically it takes around a minute and what we're actually doing is not just creating a token we're actually creating a full interface as well um which ideally will appear on the screen again if we have good fortune with the internet um and uh we'll see what what we get I'm going to wait here just for okay so now you can see that there's an interface being uh generated here on the Fly um we're going to let this kind of cook a little bit more and there's an entire uh embedded bonding curve in this page um we're using a few different templates from the ecosystems putting all these things together and what I think is a pleasant interface for individuals to explore and have some fun and it's a great way to be introduced to Fair tokens because all of the tokens that are created on here are 100% Community launched even the creator doesn't get the token they have to buy it um so we're really leaning into hey this isn't like background you know bro deals or whatever we're actually really um putting the community first so I ideally this can uh kind of wrap up here typically it takes about a minute or two um in the meantime I'll pop over to the slides and uh continue on here all right so what we're doing here is we're actually just using um the create a fair token module we intend on creating a bunch of different modules so you can create mini gamess profile makers website creators interactive art experiences and these things actually build out AO content tracks as well like this isn't just like oh yeah we use CLA and we're making front ends like the whole thing is actually being uh generated um based upon a number of factors that we've been working on actually for a few few months now all right so we have this this site now we can pop it out and say hey like how's it look um not sure what this artifact is here this is not part of it this is like a screen artifact I guess um all right all right uh you got like some funny little quotes going on um um are you staking because I'm yield farming feelings for you who knows right this is some funny stuff you can really go wild with these Pages um you actually have an embedded bonding curve it fully is fully operational when you go to that fair token launch um so what we can do here is actually generate an actual token image um this is using the HTML content and CSS um as well as the prompts to generate a relevant uh token design and we're going to see if that actually does it uh full full uh demo mode here sweet we get an error awesome um we'll see if we can get through that uh and our connection is is happy with us um just a moment here so as this kind of cooks here um I just think that we have a great opportunity in this ecosystem AO the actual token set the precedent for the ecosystem we have things like P we have things like you know apus and all of these uh great Pro projects leveraging the notion of fair tokens and and really going towards the community so what you can see here is we got Riz we got this character here built out um there we go all right cool so we actually have uh a full site when we actually deploy this um it it actually creates the whole AO contract it creates the the the entire bonding curve Marketplace and it deploys to the perm web so I'm going to come back to the slides here real quick just to wrap up token distribution for this this tool is 100% right um the Creator themselves actually have to buy their own token uh really standing behind the fair uh Fair token it's fair launch season in AO um if you're new to AO I highly encourage you to check out this model of fair launch tokens there's an amazing support system that has been created through the per web index I'm not quite sure how many individuals are fully Gathering to what extent this can mean there was an earlier mention I think by Sam to the tune of if I remember correctly up to potentially like a $100,000 a day going into Builders um as I interpret it obviously it's very early days and we're learning about that this stuff is serious um there's an amazing opportunity here so I just want to continuously repeat that because I want to make sure that individuals are really recognizing what's going on so we'll we'll be at the booth the r we and AO Booth um you'll actually be able to make your entire Fair token at the booth um and there's there's some other fun details and in the meantime if you want to sign up uh you can you can grab a spot here we've had about 50,000 signups thus far which is pretty amazing um and thank you so much [Applause] okay thank you so much Greg for our last Workshop today we have Allan once again Allan if you flip another coin you better nail it on the first try bro it's all on the wrist let's go let's go a sweet [Applause] all right hello again East Denver and the AO weave Community today oh we're not even on now we're on o we're a little glitchy today what's going on here try a different up here all right it some minor technical difficulties here wouldn't be a good demo without this okay I think we're on all right so what I want to do today is walk you guys through a bit of the tooling we have around randow and sort the process that you might go through in order to generate a random value so we have a basic typescript project here it's really just a node project nothing special no special to dependencies yet and the first thing we're going to want to do is head on over to our documentation where we can get the install for our client library for randow go ahead and install that pretty simple just a second we love eth Denver Wi-Fi amazing and now that we have the actual client here will be able to say probably have to refresh this reload window so the node modules work and if we start typing random client this should pop up from our client library and we tried to make it as simple as possible to generate a random value so all of the actual process IDs are stored on chain we actually query them from the client so you don't actually have to do quite as much of configuration and we'll go ahead and make our client and let's see what we can do with our client we can do client. create a Randomness request so the way this works with randow is that we create a request for random for which we give a call back ID where we can later query that call back ID or actually receive the value directly to our process if we're working on chain and we'll do create request I have to await things don't forget to await things that's important JavaScript all right create random request and we're going to need a few things here first thing we're going to need is our provider IDs and at random we kind of let you choose your own security level depending on your own needs so that your uh payments can actually scale with with the security just as you'd like so we're going to go ahead and head over to our provider website where we have a bunch of different providers with various amounts staked and this is really just a marketplace where you can determine where you'd like to get your random from and today I'm thinking Fred is probably he seems like a reasonable person he's got a good name some of these people put weird names but we're going to go ahead and use Fred as our provider providers and you can use multiple providers for added security for today for just demo purposes we're going to just use one to make it simple put that in there we'll need a few more things we're only requesting one input and we need effectively a call back ID and a call back ID is something you control on the client side so that you know the random that you're getting back is what you requested call my spelling's a bit off today and this can really be anything you want so I'm just going to make it some random typings place this in here and we'll await that what this will do is go ahead and request our random value on chain and so good to note this is a this is running locally on my computer but all the random values will be generated on chain in a provably fair way and that's just a warning about my wallet which I didn't set environment variables nothing to worry about uh once we get this on hyperbeam it'll be significantly faster because we actually pay to uh be prioritized which is a big upgrade we got coming soon with hyperbeam and in just a second here we'll be able to query our Randomness by doing something like uh response equals dot uh we have a few different things we can do but we're going to get a random request via our call back ID and looks like chain got back so at this point the chain is aware that we would like a random value and our providers here are actually generating that value for us they're placing their random values like I was mentioning before on the Chain you can have up to as many are in here that are active and they're going to place random values on chain where they can be aggregated provably fair and uh present it back to you go ahead and do this get by our call back ID and we're going to want to log that out so we can see it response we're going to want to pretty log that I bet uh stringify two that makes sense and we don't need to request again we only need to request the randomness we want to get the value of the randomness that we already requested and onchain this would work a little bit differently but for demo purposes a little bit easier here uh onchain oh I don't know how to code apparently that's my bad what in the world did I do I forgot a parentheses how dare I so this will work a little bit differently on chain where we actually have access to sending messages back and forth between processes where we would have uh one process that would request the random from the randomness process and once the randomness is complete fully generated on chain it will use that call back ID and the process ID that actually requested it and send it right back to it so it doesn't have to request again oh I printed something wrong what happened here I didn't oh wait how foolish of me and in just a second we will have our random values and we'll get to see a bit of the inner workings as well of how we actually generate these random values so as you can see we've got a response back where we're able to see that we had our provider here I believe this we have the request ID and we have the actual um some of these these are some of the inner workings of the cryptography behind it but we actually have the provider that we requested and what he gave as his random input as well as the finalized random output which has been aggregated and finalized on chain and this is actually what you would use in your uh math. random function to seed it which would allow you to have provably Fair random yeah so that's really the gist of it I did want to leave a small opportunity for anyone who had any questions about any of the Rand out topics we've talked today talked about today from providers to running your own node to becoming a developer with uh using randow Randomness or really anything at all all right I'll leave a few more seconds of awkward silence and then I'm going to call it a day well thank you guys this has been awesome I'm glad y'all got to see exactly how we would Generate random value on chain and that it's fairly simple to do and if you'd like to work with us and use our random values for if you're building a casino lottery or if you're building a Aon chain reach out to us and we'll get you set up test test okay ladies and gentlemen that was it thank you so much for coming to the AO conference if you're excited about the permer web Make some noise it was great seeing everyone have a great evening