S1E05 [EN] - Loaf - Dojo, Onchain Game and Autonomous World 42:48

Welcome to ABCDE. I'm the host, Yiman. Hi, I'm Lobby. We are super excited to have the founder of Dojo to join our podcast today. Love, do you want to say hi to our audience? GM, yeah. Hello, everyone. Thanks for having me. Good to be here and good to chat all things on-chain. Yeah, first of all, I think, Love, you can introduce yourself and also introduce a little bit about Dojo to our audience.

Yeah, sure. Love to. Yeah, so I'm one of the core contributors of Dojo, as well as one of the core contributors of Bibliotheca DAO, which is an on-chain gaming DAO, which...

is really the origins of how I got into this space. A little like quick, you know, a quick couple of minutes about myself. I've been a long time like full stack developer building businesses in kind of Web2 land and a long time crypto investor since like 2014, 2015. And I've always been kind of like, I've always been in this space, but I haven't been a full time builder up until about three years ago.

And that was purely because I was kind of busy with other stuff that I was doing and I was watching all the market and I knew... It was really early, I mean up until DeFi Summer, I mean no one really knew... No, everyone was just making crappy L1s, clones, right? No one was really doing anything. It was only really when DeFi Summer started to happen that real applications started getting built and it started to get really interesting.

um so then i i kind of like um dive back into solidity development back then but i i never really found anything deep in in the defi world that i wanted to like go all in on um and then you fast forward a couple years and um loot happens um and everybody knows what happened with loot um you know loot is was this idea of you know a bottom-up world build and in a way you know it really is like

one of the Genesis autonomous worlds. However way you want to think about it, Loot was really just an idea that anyone could build on this decentralized primitive. And when Loot happened, it really kind of got me thinking. And I thought, well, I think this is something that I really want to go all in and commit and try to build something crazy with. And so like a week after Loot happened, a buddy of mine, Redbeard,

we created a site called Bibliotheca for Loot, which was just a subgraph, basically, that just graphed all the loot projects together. And you could just come and log in. You could see all the loot stuff you had. And from there, we thought, well, what do we want to go next? And we thought, well, let's try and build an on-chain game. And back then, on-chain games were pretty early. There wasn't really anything that existed. There was really like Dark Forest was still doing rounds.

And there was Conquest and there was a couple of other teams doing stuff, but it was really quite early back then. And so we tried to build a game running on Arbitrum and we kind of quickly realized that this was like pre-Martin, pre-Engine or anything. So it was really like stitching together a bunch of the existing kind of infrastructure. And it was clunky and kind of not very performant and whatnot.

But while designing the game, I kind of realized that I didn't think the EVM and Solidity is a very good structure to build expansive worlds. And from day one, really the idea for us has always been, let's build an expansive world that anyone can contribute to. That's really been the number one driving factor. And we've come a long way and, you know,

building everything in the open is a challenge because your ideas change and your initial thesis is always evolved. That single vision of building an expansive world that anyone can contribute to is never swayed.

And yeah, then fast forward. Well, we quickly, when we were building the game, we realized EVM really kind of like, it was struggling. We're getting stack limits. We're getting contract size limits and whatnot. And at that time, StarkNet was kind of starting to like appear. And I kind of went down the rabbit hole of Snarks and Starks. And that kind of really, really flicked a switch in my mind thinking that, you know, we really want to build something long-term here. And, you know, long-term ZK is going to win.

out in the kind of scaling scene. And so we were like, well, let's try and build on StarkNet. It's probably early right now, but it has to promise to be a more performant VM, give us more capacity in what we can do in the future. We won't be bottlenecked by any performance

And so that's what we did. And so like a year and a half has gone by and we're building games in the stock ecosystem using Cairo.

Yeah, and so right now we've got the DAO. I mean, a lot's happened in the last year and a half, but right now we're building two games. We're building Loot Survivor, which is a very efficient single-player roguelike on-chain game that's going to be deployed on StarkNet, hopefully in the next month on mainnet. And we're also building our flagship aspirational autonomous world

starting with this mod called Aeternum, which is really the economic layer of our world. And we're building that in Open and all this stuff is open source. So anyone can come and contribute to it. Actually, it's quite interesting because me and Chao are all a fan of loot. We are the most early adopters in Asia. And actually, I'm one of the biggest loot holders. Oh, nice. Yeah.

Yeah. And is Dojo also a loot ecosystem project or are you positioning yourself? No. Yeah. So Dojo kind of was spawned in about October of last year when it really started off as just a Telegram group, which was

I just created and just basically pulled in a bunch of the other StarkNet teams because everyone in the StarkNet ecosystem is quite close and everybody helps each other and everyone's quite collaborative. And at that time, there was a transition from Chiro 0 to Chiro 1. And for like the last year, we'd kind of been building our game and infrastructure based around this initial engine, which was designed actually by this guy,

named Eeth Worm, which was like a module controller system, which we had heavily modified, and that's what we kind of built our original game on. But by going down that path, we realized that it's not

If you want to build an expansive world that people are going to contribute to, you need a set of standards and you need very consistent standards that people can easily just come and add to that piece of software. It's the age-old software dilemma. It's like you need good documentation and you need good standards if you want people to contribute to it.

And these worlds, like if we really want to see these autonomous worlds reach this, like the scale that we're all talking about,

then like we need this like standardization and like um i mean it's like why we all use erc20 and erc721 right it's like this set of standards that we can build infrastructure around and and the idea with dojo is that it um it was definitely inspired by my muds architecture and it's you know it's inherited like the ecs system um but it's definitely going in its own trajectory um the mud definitely did had an unlock of the like thinking in that very modular ecs um

matter so dojo really just brings this set of standards um to build games um but it's very much like a grassroots open source project it's non uh like i'm a contributor to to bibliotheca down and and to dojo and anyone can contribute to dojo um but it's it's it's a credit it's a neutral uh you know it's a neutral organization it's totally open source in mit right so are those two games that you mentioned before you're building are they using dojo as well

So we actually started Lude Survivor before Dojo existed, well before it was in a stable manner and the idea with Lude Survivor was that we wanted to actually ship something on mainnet and until like EIP-4844 happens and you know cooldown becomes significantly cheaper, it's very hard to run like

like a big world or like a very intensive game on one of these L2s. It's just too expensive. But what we're doing with Lude Survivor, we're basically compressing all the game state into a single storage slot to basically make an on-chain game as cheap as possible. And so...

It's not using Dojo because it's not, Dojo is not audited just yet. You know, it's kind of still in development cycle. But I imagine that, you know, Lude Survivor version 2 will be using Dojo in a couple months. We'll probably refactor it all to use it. But Aeternum, yes. So Aeternum has been building with Dojo since like Dojo was incepted

basically and it's a it's probably the most it's got a lot of if you want to see like a full-blown dojo project that that has like lots of components and lots of systems it's it's it's got a lot of that and it also we've got an end-to-end version running so you can jump into the discord on fridays we do playtests

Right, I see. And I'm always very curious that now as everyone's talking about the game engines of Autonomous World or Fully On Chains, started from mod and Lattice and then comes with Dojos and also Argus, etc. But because B-Man and me, we're both playing games for like 10 or 20 years. And when we talk about the traditional game engines, such as Unreal and Unity's,

I still remember that when the Unreal 5 was published, that game, that showing demo, like the little girl in a cave and then jumping on the mountains and the temples. It was really impressive. It's more about like the show of the graphics

the textures, the rendering, the physics. But I remember when the mod comes up, the setting point is completely different. It's like a ECS. It's entity, component, and system, which is more like a framework of the traditional game engine. So what do you see the fully-automated game engine, how come it's using these things as a setting point? What's actually the core components of a game

gaming engine such as Dojo and the mod? Yeah, I mean, to be clear, like Dojo and mod doesn't bring in, don't bring any graphics, right? Like it's really, they're trying to solve this kind of the architectural challenges of writing a game in smart contracts and the challenges of, you know, indexing the data of the game and syncing it

You really have three things when you're talking about when you're trying to build these on-chain games, excluding even the kind of, well, including the graphics. You have your contracts, which are deployed onto a network, and then you've got your indexing, which is your cache data, and then you've got your client, which can be Unreal Engine or can be Unity or it can just be a web browser running with some graphics or something. And you really need all three of those things to

talk and sync very efficiently. Otherwise you get very bad latency and you get a very bad play experience.

And Dojo and Mudd basically is trying to solve that problem, syncing of your chain state and your index state with your client. And so it's not promising of bringing like crazy Unreal Engine 5 graphics because that's what Unreal is excellent for, right? But Unreal doesn't have any networking to a blockchain. It doesn't show you how to write, it doesn't show you how to architect smart contracts, for example, right?

It's kind of like Dojo and Mudder, it's like the data layer and like the contracts, but it's bring your own client. I see. So if in an imagination that if we have a super fast layer two and very mature infrastructure, that means you can actually...

combining Unreal as the graphic layer and then Dojo or mod as the data layer that make a fully on chain with very good graphics.

Definitely, yeah. I mean, you could do that now. But yeah, it's hard to sync those engines efficiently there. And also, we'll see pretty fast networks. I mean, you can already get really, really fast sidechains. But if we really want security, and we really do want security in these worlds, and we don't want to sacrifice what we're trying to build here and just repeat the old model of games, right? It's like...

like if we go down the path of just you know making a sidechain and it's like super fast and it's you know it's a single sequencer and you know it's like a five person multiseq or something and it's connected to an unreal engine yeah sure you have a on-chain game that's running unreal but like you know like we like is it really what we want like you know we've just repeated we've

we probably just made a worse game it's we've made a worse centralized game of of you know we don't really need model dojo in that in that scenario right it's like well you may as well just use database and this is why honestly i got into crypto in the first place is after you know learning about bitcoin bitcoin's key innovation was you know removing a trusted third party and you know we don't want to repeat that you know just to try to do something we don't repeat the um

you know, just Web 2 world where we have distrust of third party, right? We want to build these new experiences where we can really trust these games that we're playing that, you know, this data isn't going to just get, you know, rugged by the central team, right? We actually have a level of guarantee around it. And we can only really do that when we have a, you know, when we're running on a true L2 or an L3, like a true one, not a sidechain. Yep. Makes perfect sense. Yeah. Yeah.

And recently, I think I've read articles that have might release their V2 version, which is transformed from the ECS structure to different like the user table. Yeah, they've done a really cool innovative thing. They've created a very optimized SQL basically running in using Yule, which is like the low level AVM language.

which allows you to create tables on chain. So you can easily create an SQL-like database in your world, which then you can query in any way you want.

Would this be the direction that you guys go as well, or are you guys trying to do something different? I think we're... In Dojo's backend, the state is kind of indexed in that way anyway. It's kind of... We're using a table structure. But we're really focused on the narrow vertical of gaming. So we're optimizing everything for games and, you know...

It's hard enough to architect smart contracts for games. So we want to really build Dojo and optimize it around the ECS. The way Dojo works is we've forked the Cairo compiler

Well, kind of extended it, probably a better way to think about it. We've extended it so, and we've injected a bunch of shorthand functions into it, which allow you to write shorthand, you can write shorthand commands which expand out into more complex queries.

So we've created a scripting language on top of Cairo. So you can really prototype and write your games significantly faster. So Dojo, what's unique about Cairo, which I think a lot of people aren't quite aware of, is that it was designed by Starkware to compile to these little ZK circuits. So if you write a program in Cairo, you can prove its integrity. And so that's how Starks work, right? And that's how Starknet works.

And so Dojo, and like one of the big things that we're trying to do with Dojo is really lean into that and create what we call provable games. And what we mean by provable games is you can write these really complex games in Cairo and you can run them in your browser and you can kind of like generate as proof.

like a zk proof of you playing that game in your browser using your browser's computation and then you can just verify that proof that you've you know that you've played truthfully

And so if you think about that, it's like we still get the integrity of an on-chain game. So you can't cheat because you can't actually fool that system. You can't cheat in it, right? But when you offload the compute to your browser, well, your browser, there's plenty of compute around the world in your browser and all devices. And if we can just generate these little proofs of you running these games,

then we can really like, we can open the design space of what we can do with these games, right? So like an example would be like Loot Survivor, for example. Like Loot Survivor is like, you know, all the computation is running on StarkNet right now. And, you know, when you do a transaction, StarkNet runs the computation and, you know, updates the state, right?

But there's a path where we could create a loot survivor style game, you know, in Dojo and you could run the loot survivor in your browser using the Kairi virtual machine, which you can run in your browser. You can play that game and, you know, you get, you know,

you play for half an hour and it all... Eventually when you die, a little proof gets generated and you can submit that proof back to Starknet to verify that you've played that game correctly. And so all the computation, all the state changes that you've done is actually done in your browser. It's not done on the chain. You're just verifying that you've played the game correctly and that's the only transaction you need to make. And I think that technique...

in our mind, like the new meta in like six to 12 months. It's pretty hard to do right now. This is very hard to do, but that's where we see this, this, the whole industry going. It's like a roll up. Well, it's not really a roll up, right? It's kind of, I mean, it's not really a roll up. It's not, it's like a state channel kind of. Yeah. State channel, like Latin network stuff.

Yeah, it's the state channel. Exactly. And so you can even do multiplayer state channels. And you play like you and I, we can play a game of chess, for example. And obviously, we don't want to have to do all the moves on chain. We just want to verify that we are doing the moves correctly. And we can do that through a little ZK state channel. And then eventually one of us wins. And we can just submit the proof back to the chain to verify that you've won or I've won.

But all the computation, all the state changes has happened off-chain, so it's practically free.

That is something that Mart cannot achieve at this stage, I guess. I think the thing is Solidity isn't designed for writing. It's not a ZK language. It doesn't compile down to these circuits. The way you do it is you need another language, you need like Sercom or something to write those ZK circuits in, and you kind of use a combination of Solidity and Sercom. There's a couple other languages like Noir that are coming out,

that do specialize in this. But that's why Chiro is interesting to us, is that it's a general purpose language. So you can write smart contracts and you can deploy them on StarkNet and you can run them like that. Or you can run them on other types of StarkNets like Madara, which is a new StarkNet sequencer. Or you can run the Chiro code in your browser in WebAssembly.

And so it's like isomorphic in that approach and you can run these gyro programs anywhere. I see. Yeah, that sounds very cool. It's like moving from the cloud computing to the edge computations.

that can release the design pattern of designing games. Exactly. Yeah. And, Lou, if I want to get some opinions from you about the fully-automated game as well, because the reason I'm asking that is the GameFi, like Axial Stepman, the first generations, it is...

It's kind of like poor playability. So people come for the money, for the earn, not for the play. And even for that, it's still like an entry barrier that people have to know what is MetaMask, they have to know what is blockchain, they have to get some guests, etc. That's already blocked lots of people, so preventing people getting the mass adoptions. And a fully on-chain seems even more hard calls, like Dark Forest, right? It's even harder than playing Axie or Stepin.

So in your mind, how this autonomous world can resolve this mass adoption issue? And also, what kind of game do you think might suit the fully on-chain game autonomous world best, like an RPG, FPS, or SLG, whatever stuff?

I mean, let's be real, crypto isn't nearly adopted by anyone. We're still so early in this kind of multi-decade cycle. I'm convicted that eventually there'll be seamless integrations in basically all of our infrastructure and institutions.

eventually, but we're not there yet. And it's probably actually, side note, I mean, I think in six to 12 months after 4844, we will see significantly more adoption because it does open up a lot of more use cases, like very, very low transactional fees, but you can actually trust it.

but yeah in terms of so there's obviously that's that's a challenge that's like outside of dojo or not and that's like a general crypto thing right um uh but these games will get adoption i think there's a lot of innovation going in into account abstraction at the moment um

So there's like 4337 in the EVM space, which is the big step in the right direction, allowing some magic account abstraction, which allows you to do session keys and whatnot. StarkNet's pretty interesting. It's got native account abstraction built in. So it's basically just there from the get-go. And where account abstraction just allows you to do significantly more

smoother UX when it comes to doing transactions. So you can do custodial, like hybrid, non-custodial wallets. You can do WebAuthn with your fingerprint. You can do session keys built into your browser so you don't have to keep signing transactions, which really breaks the flow of games.

Yeah, I think that's one of the big unlocks that we all need to keep focusing on is just improving the UX. But I think we're almost there, to be honest. Loot Survivor's implemented session keys, so you can go and play without having to sign wallets. So you'll have to see the step change in that. I know Mud's got a built-in burner wallet system that allows you to do the same thing. Dojo's got a burner wallet system

library as well, so you can do the same thing. In terms of games that these Autonomous Worlds types of games,

Autonomous world's kind of a... But really, they're going to be like a collection of games and they're going to be like an ecosystem. And probably, I like to think of it as like... I think Ethereum is the... Ethereum's architecture is kind of the architecture that we all need to aspire to when building these things. It should be... I don't think you can just chuck money at an autonomous world and expect it to exist. Like, look what happened with the whole L1 scene, right? Lots of L1s raised ridiculous amounts of money and they went nowhere. Yeah.

That's because you can't buy social capital. You need to build it and you need to build an ecosystem organically for things to last. And Ethereum has built that over the last eight years and that's why it exists. And that's why it's so robust. It's because it started, yeah, it was the first generalized smart contract platform.

but it really had this kind of like grassroots approach. They only raised like $20 million as Ethereum Foundation. Everyone started building on it. Kind of like all these standards emerged just organically. Ethereum is basically just a collection of all these smart contracts that exist and its functionality keeps getting improved with every smart contract that exists on it, right? It's kind of the same thing with these worlds. They'll spawn organically without any central authority

and they'll evolve and people will build on them organically and then some of them will get escape velocity when that happens. But I do expect that there's going to be a lot of money chucked to trying to build these things, but I don't think it will work. I think you have to build these things grassroots, otherwise those never stick.

In terms of the games on them, I think there's going to be lots of different games, like RTS-style games, pure economic games, just any type of game you can think of, and probably a lot of new styles of games that really lend itself to the technology. Hard to say exactly what, because I don't think anyone's actually cracked the secret source of what makes a really interesting on-chain game yet. I mean, there's a lot of teams trying to explore it,

We have our thesis around what we're doing and we're trying a lot of different things. And in the DAO, we're encouraging a lot of people to build things. In Dojo, we're giving grants for people to build things. And a lot of other people in the ecosystem are doing the same thing. Yeah, I mean, it's just in this experimentation phase right now. We all need to just try a bunch of crazy things. I see. Yeah.

we have seen a lot of fully on-chain games in Starware ecosystems. So Luf, can you describe a little bit about what is the on-chain gaming ecosystem be like in Starware and why many on-chain gaming teams choose Starware to build their games? I think it kind of just happened organically. I think it happened because the, and just to be clear, like I'm,

I'm like team on-chain gaming. I'm building stuff on StarkNet, but if any team is authentic and has integrity, I'll support them if they're building an on-chain game because I think this industry is so small right now that we all need to support each other and just build it together. We're not really competing with each other. We're competing against the whole other industries like mobile and PC and whatnot.

StarkNet's interesting because it uses the Cairo virtual machine, which I was just talking about before, over the EVM. And the Cairo VM has a lot more computational capacity than the EVM. And it doesn't have the same memory design, so you can create gigantic contracts with it. You don't get to stack...

stack trace limits and and all these like things you run into when you're running running big contracts um in the evm i think that was appealing to people um that you know you do have this kind of like new design space um and you're not kind of like shackled um to the evm so i think that's why a lot of people chose it um and there was a couple of early people um in the ecosystem like guilty gyoza um who who did a bunch of crazy like physics simulations um

who kind of like he was one of the driving factors of why i kind of like looked at starkman in the first place um and it kind of just lured this like initial crowd of on-chain game builders um like we i think we announced we started building in like november 21 and that was really early that was like before timestamps or anything so um i think that was that was kind of just it kind of happened organically that people just wanted to build on-chain games on on starkman

Yeah, cool. Another question I'm curious about is that what is your favorite game in Web 2? Web 2? So, I mean, I play...

I play a bunch of different games. I like the Survivor-style games. I like roguelikes. I've kind of been obsessed. I keep picking up Vampire Survivor, and I keep playing it. But I also play AoE, Age of Empires 2. I've done a lot of time playing strategy games. I did my StarCraft time.

But then I also play medieval slashes as well. And I also play kind of slow-paced, turn-based games like Civilization and Turtle War and stuff. My Steam library is ridiculously large because I always download new games and try to

you know, see what, see what new games are about. So yeah, I, I don't really have a single favorite. Probably the most hours that I've put in is probably like, probably strategy, like RTS. Yeah. My, my favorite is also strategy game. I, I have spent about 3000 hours in steam playing strategy game. Yeah.

yeah but let's uh do you play um web 3 games or do you even try any fully on-chain games or etc that you would like to play and can you recommend one of two fully on-chain game to us i don't i mean i don't play any like you know web 2.5 games kind of games with nfts um i don't really play any of those games um that are really interesting to me

And in terms of on-chain games, like it's still, it's still so early, like to be honest, there's not that many that exist because I've been, you know, designing and building Lude Survivor and Aeternum. You know, we're playing a lot of that right now. Um,

There's a lot of cool games that are coming out. There's like, Topology just released their Alpha, which is a fighting game called Shoshin, which is actually using that virtual machine in the browser, which I was talking about. It's probably the first kind of like production example of using that technique, although it's not settling to StarkNet just yet, but it's like an Alpha. It's kind of using the Kairi VM in the browser.

The Small Brains guys are releasing some interesting games. They just released Words 3, which is like a Scrabble game, which uses a variable rate gradual touch auction to auction off Scrabble letters and you try to make words with these letters. That's a really interesting use case of an on-chain game. There's the Playment guys making Downstream.

which is a really interesting game. There's Cartridge, which have developed Ryo also using Dojo. So Ryo is actually one of the first production games, well, it's the first production game using Dojo. And it's like a take on the old Drug Wars game from the calculator, if anyone played that back in the day. It's a game of arbitrage where you go from part of New York City, part of...

New York City's kind of like locations and you arbitrage drugs. You buy drugs in one area and then you arbitrage them in another area. It's really cool. There's not that many on-chain games that are live right now. Yeah, we spend a lot of time playing Web 2 games, especially like strategy games. Do you think that we can make this game into Web 3 in the near future? Like Civilization, for example. Ah, yeah.

Yeah, well, I mean, we're trying to kind of do that with Aeternum. Civilization's one of the big inspirations. The challenge is building like a totally, like an open game that's expandable, that's like

balanced the the actual approach that we're taking with these games and like this is not every on-chain game needs to do this our thesis is like we want to build this like we actually want to build an autonomous world that like starts small and like grows and just like becomes like this self-sustaining organism our thinking is that you know it needs to be kick-started with a like a minimum viable physics um

kind of game. What we mean by that is, uh, Turnum is really designed to be like this first economic layer of the world. And so it's a, um, it's built around the 8,000 realms. And so the realms, the 8,000 realms is how BibliothecaDAO, um, is governed. They're, they're NFTs that were, um, they're called loot realms. Turnum is built around these 8,000 realms and these 8,000 realms have all different resources on them and different properties and you generate resources and you, and you trade peer to peer with one another.

And you can arbitrage different areas of the map. They're all positioned on this giant map, all location-wise. The initial goal of the game will be to build a topology of the map. And so you'll generate resources, you'll trade resources, you'll craft them, and you'll build these permanent structures on the map that will be used in the later stages

later like ages of the world you have to manage your realm you have to manage the happiness level which is like inspired by the civilization happiness level and so there's a bunch of metrics that you need to manage and this is all done this is all on chain but that's really kind of like we're thinking we're trying to make it as like as a minimal kind of game as possible and

Because we feel like these games, like you can call them games, but they're really like, they're kind of like open sandboxes games, whatever you want. It's hard to like shoebox them in like one term because they're so open that people can come and build on them.

And then it's like, well, what is this thing? What is it? It's not like the game that we designed. It's like this network. And this is some of the logic and some of the system components that we've made. But other people have come along and built these other things that extend the game. So at that point, it kind of loses what the original game was but expands it. So it's hard to shoebox it. But our thinking is, let's create the minimal layout possible.

And then let people build on the world with all the tools and all the standardization that Dojo brings. Like unlimited DLCs. Yeah, it's like unlimited DLCs, unlimited mods. And that's the idea. And the original game we designed, and contributors now, we still have this idea in our mind of all these things we want to include. Impossible to have a finished...

like on-chain game that's like balanced because it's open by default. So you never want to start at the end. You want to let the world emerge.

as much as possible and build with the world rather than trying to fit onto the world what you think it should exactly be in every balanced way. Think of a natural ecosystem in the wild. Natural ecosystems have an equilibrium in the balance. All the animals and plants live in harmony. And it's only happened over millions of years. That balance has just formed, but it's emerged over time.

And so we see this as a similar thing happening with these worlds. Like we want to create this like minimal framework and from that balance and like the ecology of the world will just like form.

Yeah, like you said, like Ethereum, you define basic rules and the framework inside. All up to you. So in that case, would that be possible for some stock pair game? Like the first version of your game mentioned that in the V2, it will be using Dojo to re-transform it. Would that be a hard work? I mean, I was thinking... No.

No, not at all. Yeah. I mean, one of the hardest things about the game is just like, what is your game? And luckily with Loot Survivor, we've spent a lot of time iterating on the actual game. Migrating it to Dojo is going to be pretty easy to do because we already have...

the game, right? It's just refactoring and kind of like changing the kind of the project structure. But Dojo is designed to be like really easy to get going and really easy to like just iterate on the project because it has this really nice structure to it. So yeah, it's not going to be hard. Right.

What about the other games? Would that be possible for some other online game on StarCraft which is not using Dojo at this stage and eventually they can merge into Dojo Star or they re... I think if they... Yeah, if they're pre-launched, so if they haven't like...

launched onto mainnet. But even if in that case you still could do a migration, it'd be easy. I mean, I would definitely recommend that anyone that's actually tried to build one of these games understands how hard it is because there's so many moving parts. You want a level of standardization. So like Dojo brings like an integrated tool chain with it. So it comes with like a local devnet, comes with a local indexer that just indexes your world without you having to think about it. It's like an automatic indexer.

And it comes with a CLI tool that helps you maintain and manage your world. And people can extend the world easily through that CLI tool too. If a team is thinking about migrating, then I would definitely suggest spending a day or two even just trying to do it and trying to migrate your game. I would think that you could migrate the contracts of your game and your logic within a couple of days if you already have it established into the kind of a dojo pattern.

Another question, how do you foresee the end game of autonomous world? And can you visualize

what the autonomous world will be like in the next five years or 10 years. Yeah, I think what we've kind of been talking about is that these autonomous worlds really, they act in a way like L1 blockchains. They become this, because they're all based around a singular contract. These worlds will, I mean, I suspect one of them, one or two of them will really get traction and really become this gigantic world

That, you know, just like Ethereum itself. Yeah. So, I mean, I probably, my forecast would be within 12 to 24 months, these worlds are going to be the biggest consumer of block space in the entire crypto industry, just because you consume so many resources. You know, you can think about a game that has 10,000 active users that, you know, people are making actions every minute. Like you can just think of the amount of transactions and the amount of state changes that's happening at that point.

And that's probably like more block space than like exists. Yeah. These things consume a significant amount of resources. Yeah. And I fully expect maybe in like six to 12 months that the idea of these things will be downstream in the mainstream crypto sphere. And a lot of people are going to try and build these things.

because everybody's going to want to be a part of these. And do you think that the autonomous world will combine with XR, AR, VR sectors? We make it like the movie Ready Player One? Yeah.

Maybe with off-chain compute, like doing the car VM off-chain or doing this client-side proof style computation, then maybe we can get there. But with our traditional blockchain approach of blocks and stuff, then it's going to be hard to recreate FPS style and fully... Yeah, I mean, you're ultimately still constrained by blockchain and decentralized...

Asian. So you're, I think it will form the basis of that. You know, it's really what, I mean, it's, it's kind of like meta, right? Like meta tried to buy the metaverse, but like, obviously it's going to fail because you can't buy these things. Like, just like you can't buy an autonomous world. Like you need to let these things emerge and you need to let the metaverse emerge. Like the metaverse is such a dumb thing now, but like,

It will never be like, no one can ever own it. It'll just be this thing that emerges that's owned by everyone and everyone's a part of it. Just like the internet, right? Like you can't own the internet. Like Microsoft tried to make their own internet in the 90s. That failed. Just because these networks, no one wants to be owned by any central authority and it kind of defeats the purpose of them.

Yeah, cool. I think for the last part, what do you want to say to the developers, the game startups, founders in Asia? Do you have anything to say for them? Yeah, I mean, I'll just say get involved and try to build something with one of these engines. I'm on call to answer any questions that you'll have about Dojo or Realms. So, you know, come and pop into Discord, pop into Telegram and ask questions.

But I also think that we're in a stage right now where I would advise against trying to build AAA on-chain games. I would say the best approach now is to try to build something that's something that you can do in three months. The infrastructure and everything around this is changing so fast. You don't want the development cycle of two years. You want to be able to ship something in three months. And so...

And don't just try to think about a clone of Web 2 or something. You could kind of be inspired by a Web 2 game or a traditional game. Really try to lean in to what this kind of interesting technology gives us. The ability to have this kind of decentralized governance, decentralized players, kind of crazy tokenomics without being ponzinomics. And so, yeah, just get involved in trying to build something.

Yeah, I agree. We should build some native game and application in Web3 rather than doing some copycats from Web2. Yeah, exactly. Yeah, sounds great. And yeah, it's really nice to have you today. It's a really amazing talk.

And look forward to talking with you again next year. Let's see what happens. Yeah. 12 months is a long time in crypto space. So yeah, looking forward to it. Thank you, Loh. Thank you for coming to our podcast. Thanks, guys. Bye-bye. Bye-bye. Bye.