Building High Performance Apps With State Channels - Louis Bellet

Transcript

[Music] So this is the end of the day. I see we have a full room. So let's have some fun a bit. 25 minute. That's even more than I need. Thank you. Thank you everyone for coming. Great to meet you. So let's go on a story. The unparallel future decentralization innovation global empowerment. I don't have the clicker.

This is it. Harnessing blockchain for transparency, equitable and efficient tomorrow. Unlock novel economy model and communitydriven ecosystem on board. The next billion users to web free. Sounds familiar. Should I go on? A truly innovative ZK Snark layer 2 for enhancing throughput.

Let's make a new one. So, we built a token, the AI quantum X, and it powers the meta metaverse for quantum algorithm. You guys still following right? So the unparallel future project seriously bro trust me this is the next big thing this is the architecture we thought about every detail this is not a ponzi this is

recursive yield farming It's just better. It's not inflation. This is emission schedule for ecosystem incentive. Right? We have we heard of that before as well. I think everyone here is too familiar with the pitch sounded really realistic. It's a satir maybe time for a change. And I think with this if Prague

the hackathon coming, let's build some real and let's go. So before we get started, let's ask the audience this simple question. How many million instruction per second is processing Ethereum network? Go ahead. Just raise the hand. Give it a wild guess. Million instruction per second.

MEIPS. It's good old measuring computer throughput. Okay, we have less than one here. Less than one. 0.1 0.1 over there. Patrice two. No clue. Million instruction per second. Okay, let's see. So, let's put some scale. I will get to it in a bit.

Intel processor 412 billion instruction per second. This is a beast. Your grandfather Atari 8bit system or maybe this is your generation. I didn't have one but I was close. One meip groundbreaking etherum EVM45. So we are less than 350,000 instruction per

second. My side note here says well we shouldn't be comparing a CPU and the etherum network. This is a network. Every instruction has been validated by a consensus network which is great. Okay. Well, a word about me. This is traditional. So, I'm a chief architect. I'm the open-source maintainer of state channel.org, or the initiator of a new RC 7824. I'm building with yellow the next

generation traditional finance infrastructure and I spoke more with AI last month than with real pay people. The jokes is that chap GPT thinks I'm his therapist. That's how far I got. Okay. So recently the numbers shows that there is over close to 12,000 data center.

This is basically 5% of energy consumption on the planet. This is the scale of web two. And we being very enthusiastic. This is a great value. Um thinking we're going to on board the next billion people. Well, before we do that, I believe we should be able to leverage the cloud infrastructure that is widely available here. I believe in the next year with AI the 5% may turn into 10% global energy consumption just for compute. This is a staggering amount.

So this is the vision. Web free application should not be web free. They should be web two. They should be capable to scale to the infrastructure we've been building over the last 30 years. I'd say maybe we already have the solution. the state channel. It allows real-time interaction with the security of blockchain, feeling like you're running a web two application with the benefit of

blockchain. Instant of chain interaction sounds pretty good. Natively chain agnostic. Well, sounds pretty good. horizontally scalable. You add more servers, you have more power. You can use the entire 12,000 data center if you had to glitch. I don't know where is the device. So, what's the problem?

Why state channel isn't helping fixing the 1 billion next web free user problem. Our entire industry have been turning around this problem never achieve mass adoption and this is like the elephant in the room. Why? While we have the solution, the technology, the everything, we haven't solved the problem yet. The theory it's simple but complicated.

This is the summary of four five years of state channel research. Oh yeah, it's very simple but it's complicated complicated level of nightmare. So in practice it's simple but um the offchain protocol you just have to figure it out yourself implement everything. This is one thing it's scalable but you need to reinvent peer-to-peer networks. You need to figure out everything here. There is some library but basically you're just starting back from scratch.

It's crosschain but basically there is zero ERC zero chain agnostic protocol talking about this subject. So basically like there is no interoperability with Solana whatever I mean there is the EVM guys there is the other guys and nobody agrees the solution. I don't know if anyone here guessed it. I'm going to drink some water. What's the solution?

Actually, it's more boring. We know state channel can can be can be it's it's simply consensus talking about it agreeing about it debating. This is why we do such kind of event standardization. This is how the internet was born. We did a great job at standardizing TCP IP and it worked.

W3C all of these things hope it rings a bell. So making an ERC is a first step. There is more could be making an af uh could could be more there is like the chain agnostic thing and the SDK shall complement well the documentation I would say the protocol. So this way people can try it, experiment, provide some feedback. So this is what we've been doing. The baby is named nitrolyte. And yes, this is highly explosive material

because see you um we're going to destroy and blow up complexity because for once maybe we should do simple things that works and solve our problem rather than fixing problem that blockchain created in the first place and I think it's very important So our work on state channel, it's a bit of a takeover the open source community with my team. Some of the guys you can see here. You guys should raise the hand.

Don't be shy. I don't know. My my phone wants to Oh, I I need to pick up my son in Ukraine. Okay. So um what did we do? Um so with four years of working on state channel and learning and going through a lot of work paper and work group we understood that it's simple but complicated and we need to simplify it.

Um so we drill down everything. I would like to mention for the public that the only state channel technology that actually works in production is lightning network. And lightning network leverage hasht lock with the stupidest dumbest language that is the bitcoin scripting system has been modernized with taproot. But this is uh I saw it few days ago like 20 lines of code to make state channel on top of

Bitcoin. So it's not very complex to do state channel. Um so we try to stop overengineering and take some great ideas of Nitro state channel.org project and fork and let's simplify everything to the essential what we need. For example, a state channel with two participant is good enough. We have other in this category. So we created a lightweight agnostic RPC protocol because RPC is great. It's been used for many years and lots of legacy infrastructure uses it and so why

not? We created an SDK TypeScript. You install it. You open cloud code. You vibe code and say I want to do PVP Trice in one shot. He's going to make it and you're going to have it blockchain back high performance. So for those who are curious to understand a bit how it works um everyone is familiar with this layer.

We got all of the smart contract capabilities and there is thousands of layer 2 layer one but they all do the same thing and EVM is really widespread otherwise was on top of it we deployed nitrolyte smart contract. So to any ecosystem we want to reach digital asset we have to deploy nitrolytes fairly simple smart contract available on GitHub. We got our first audit. I think we need many more but um 250 lines of code I would say. On top of it, we have the clear node mesh

ledger service. And these are go classic peer-to-peer nodes with API which create a logical layer about ledger accounting double entry accounting for those who are familiar um account number debit etc. So this is creating a logical layer for the app developers so they can develop offchain high performance low latency financially backed application.

You get all of the benefits of blockchain without having to get the bad UX. This is the promise. I wouldn't say it's easy, but I think it's a good battle. So, example, I'm happy the GIF is working. Actually, I I think it's from a PDF. So, a snake is a very good illustration of what we can do with state channel, what we can do with our framework. The game is fast.

Each player needs to use the keyboard and do key strokes in the lowest latency and the quickest. This is how you're going to be a good snake player, right? And here we can do it with money at stake and with a decent level of security. Um, so this is a good example of what type of application that you could never create with layer 2 DAP and that you can create with layer three type of apps. High frequency of chain with zero cost for moving your snake. Um, for the joke, the snake is

hungry for performance. So this is also leveraging our RPC protocol and um basically has some strategies for anti-cheing. So you could organize the world Olympic of snakes with money at stake uh and with referee decentraliz you name it. It's all in there. So how to use neutral light? if you want to give it a shot in the next days. Um, I'm hoping all of you joining the hackathon. So, it's simply an npm

package install. I would recommend going on GitHub ERC7824. Over there you'll find the source code because this is bleeding edge. We still making change as of today. Um and this is really an invitation to come and give some feedback uh like the vision of it community use the default adjudicator. This is a smart contract um which reinforce nitro and RPC because you not going to need to have complex business rule on chain. So you can just go with the out ofthebox

configuration and you can just use the SDK and build your app vibe coded. This is a bit what we're trying to achieve lately. What's important to keep in mind is that when everything goes to sheet, you can call onchain challenge to recover the funds at any given state that was the latest state. So if you were sitting on a poker table, you could probably get back all of the chips you had in the poker without maybe the last hands you were going to lose. If you understand the analogy, this is it

guys. So we're taking a risk here. This is innovative. We're trying to think out of the box. I would say we're trying to make finally something simple um and something different that is not yet another layer two and not trying to build new uh solution using more blockchain and rather leveraging the amazing capabilities of blockchain uh through layer three and I think layer three is really a much better place to be you're going to have the high performance benefit and you're going to have the blockchain agnostic. So if the

contributor are adding an obscure blockchain, you don't have to change any line of code in your app, you're going to benefit this new ecosystem. So probably the message which might be unpopular for a lot of sponsors is like stop choosing a L2, stop choosing their ecosystem and think a bit out of the box, think layer free. Um, this way you'll be chain agnostic and the blockchain guys can evolve and your app will remain doing all of the functionality without having to move it ecosystem or

change it every time there is a new upgrade. That's it. So, thank you very much. Um, our hackathon workshop is going to be presented at the Eve Global Friday 4 p.m. 2 p.m. You don't know maybe Kade is she here? Let's say 4 PM. Look at the schedule on Eve Global. Uh, so I will be presenting the SDK. I think Maxim Andrew will be joining the presentation. We will go a little more in technical depth and well you can look

at the other participant of the hackathon and you're going to find 90% layer two. I think we created a thousand layer tool lately and I'm not sure it makes a lot of sense to build an app for a closed tiny ecosystem where I'm still a strong believer that Etherum is the sheet and maybe you don't need to put your app on a obscure isolated layer two uh and state channel can really change your perspective on things um at this stage of the project. This is truly bleeding edge. So scan the QR, check the

website the protocol. This is submitted on the Etherum ERC's and we're looking for feedback. I think we really need like security review, peer reviews, comment, even naming suggestion for all the things we do. And I think this would be my perfect hackathon to be hacking among hackers. Build an app, give us feedback or give us your opinion on the protocol and maybe how we can like make this a reality. And once again, we're trying to build something simple. Um I guess Leonardo da Vinci says

um simplicity is the ultimate sophistication and I love unique system and unique system has been built a bit on this foundation. Thank you everyone. [Music]