AO: The Decentralized Hyper-Parallel Virtual Computer built on Arweave

GM GM everybody! Welcome to what's going to be another great DevNTell. So if you didn't know, DevNTell is a 30-minute podcast for folks who are passionate about Web3 and have been building really cool products. So whether you are a established Web3 company or weekend hacker or anything in between, this is your platform to demonstrate your product to the Web3 community. And today I'm ecstatic to welcome Patrick Skinner, who will be giving us an overview of AO. If you didn't know, AO is a decentralized hyper-parallel virtual computer built on Arweave. So if you stick around for today's show, you'll learn what AO is, what you can build with it, why you should build with it, and how to get started. All right, let's get going.

GM GM, welcome Pat! Love to have you back.

Hey, how's it going?

Good man, good. You?

Good. You know, it's been too long since you and I talked or anything like that. I feel like I kind of missed you during hockey season. We didn't get to talk too much then and I've been so busy, it's hard for me to pop back in the DAO all the time, but I'm here and yeah, I'm excited.

Excellent, excellent. Yeah, man, that's just how it goes in Web3 sometimes. You sometimes don't even get the time to chat with your homies about hockey. But we'll take whatever we can get from you. So welcome to the show. I'm ecstatic to have you on and thank you for taking the time out of your day to talk AO to us. But before we get into that, and if folks aren't familiar with you and your background story, perhaps would you like to enlighten the crowd here about who is Patrick Skinner?

Yeah, yeah. So my background's obviously, you know, for those that may already know, a bit unconventional in the Web3 or even tech space. You know, I came from a background where I worked in, you know, the US military. I was a US paratrooper medic and worked in special operations a little bit and specialized in mass casualty coordination as well. And during that time, I actually was mass casualty coordinator for the earthquake in Haiti back in 2010. The first quake hit like in early January.

So, you know, me and my unit, we responded to the earthquake on like a humanitarian mission called Operation Unified Response. And it was then that like, as I was doing a lot of the mass casualty coordination and I was doing everything from like surgeries and search and rescue and all kinds of craziness, we actually had this boy named Oresto. We found him in the rubble. And he was roughly about three or four years old, lost both of his parents during the earthquake and we ended up having to amputate his right arm and do over a hundred stitches on his head. It was crazy. Well, he kind of became like our adopted son in this whole craziness for the whole platoon. So we all took care of him.

We were all like, even though the unit was rationing, we were like rationing our own food just to be able to feed him and everything too. So we took great care of him. And I used to go and take care of his surgical wounds and everything on a regular basis. And I would put my military cap on him every time because he loved it. And one time after finishing, you know, I left and realized, oh man, I left my cap. Went back and he was just gone. Didn't realize what was going on.

And I turned around, realized some humanitarian guy, or at least a guy in a humanitarian vest kind of thing, was just like speed walking with him away kind of thing. And I was like, man, okay, what's going on? So I run over, get him and this guy is claiming, oh yeah, you know, we're taking all the kids from the hospital area and moving them to another location so that we can, you know, take care of them in a better setting, better location. And at that time, especially when all the chaos is going on and you've got thousands of patients to take care of and you've got a handful of supplies and a handful of personnel, you're just trying to take anybody that will be willing to help.

And so, you know, you assume they're there for good intentions and you're like, hey, you know, totally understand, but you do got to get verification from US Army Command. It's just back there, all you have to do is give them your ID, that kind of stuff, sign some paperwork and, you know, no big deal, we'll get you in here in the pipeline so you can help us out. And I was like, and Oresto, he just got to stay here because, you know, I'm still taking care of his wounds. He gave me this like really awkward smile and gave me back Oresto and instead of like going this way, he went that way out of the hospital.

And I was like, that's weird. Seemed really awkward and he kind of seemed very nervous at that time. Low and behold, you know, something was off, so I told US Army Command, I was like, hey, something just didn't seem right about this conversation. Let's have an eye out for this kind of interaction. Make sure none of the kids are being or, you know, being taken out of the hospital area without some sort of approval. And they're like, okay, cool. And low and behold, two weeks later, he and like three other Americans, mind you, were caught at the Dominican Republic border with about a Conex of like 40 kids. Like basically trafficking them out of Haiti.

And so at that point I was like, man, I can't believe, you know, that could have been Oresto at that point. And that was when I was introduced to human trafficking. So I left the military because I tried to get into like working into human trafficking in the US military, had a really difficult time doing that but I did get some offers in the private sector. So I started working, you know, for basically teaching combat medicine and stuff like that, but also like working with human trafficking interdiction teams in like Northern Africa, Middle East and stuff like that for a little bit. And that's when I realized where the money was coming from as far as like who was doing all the human trafficking and everything and also like how funded they were, you know, how well-funded they were.

And I realized that number one source of income for human trafficking actually is surprisingly counterfeit goods. A lot of people don't realize that. And when I say counterfeit goods, people think of like Gucci bags and fake Rolexes and stuff like that. That is part of it but it's really like fake pharmaceuticals, fake electronics, fake car parts, like all kinds of craziness. And the money is just, wow, it's nuts. It's massive amounts. And I told myself at that time, if I could build something that could like cut off the revenue source for human trafficking, I would definitely do it.

And so roughly about 2014, my wife and I were like, you know, if we want to have kids, this is not a great career for you to be in. So what are you going to do? And I remember I literally Googled what's the safest thing I can do while still making like $100,000 plus or something like that. And of course software engineering is like on the top of the list and I'm like okay, all right, that's what I'm going to get into. So I started learning software engineering right about 2015 to 2016. Had a couple of different startups and stuff like that.

But then, you know, it was like 2017 to 2018, I saw an article about how NFTs could be used to authenticate physical products. And I didn't know what an NFT was at that time, so I was just learning exactly what it was and I'm kind of like dabbling a little bit in like the Web3 space until roughly about late 2020. That's when I fully aped in at that point. I was like, okay, this is what I'm going to do. And so like, I was on this show what was that, two years ago, for my previous startup, for Agami. And Agami was intended as a proof of authenticity for physical products. And it's also when I discovered Arweave, which is really, really interesting too, because I was building with Arweave on Agami.

I needed a way to have some sort of proof and like permanent proof. Something that couldn't be manipulated, something that wasn't going to suddenly disappear for any reason whatsoever on a physical product. Because let's say, you know, you can authenticate with an NFT or some sort of like UDL or something like that, we'll get into that a little bit later, with a physical product, but the problem is is like, let's say that product outlasts you and that company. That could be several things. It could be certain car parts, it could be a Rolex, it could be like vintage goods of some sort you're trying to authenticate. And, you know, if nobody's home to keep the lights on and somebody's trying to scan an NFC chip or something like that, then those that NFC chip isn't any good anymore.

You know, nobody if you're using some sort of centralized system and that profile that you're trying to use isn't readily available anymore, then what's the point of authenticating that product anymore? And as a matter of fact, in my opinion, if you do that and that product's one of the product's feature is being able to authenticate it somehow with an NFC chip, as soon as that link is bad, I think it actually devalues that product because now a particular feature of that product no longer works. And I think that's a huge problem. So, you know, that's when I started building with Arweave.

And that was late 2020. This is when I really started kind of building at that time and I think it was about mid 2021, maybe end of 2021 is when we had that big NFT dip, kind of the market drop and everything. And I was trying to get some funding and everything, got a little bit of funding and then but then things just kind of went really sour after that and I just couldn't find anything. So I was like all right, let's back this up a little bit. You know, I'm still kind of new to the Web3 space, but so I still have a lot to learn and I, you know, at that point I had joined Developer DAO but I wasn't super active.

And that's when I think it was oh Billy put together that hackathon for Streamer and I ended up competing in that hackathon and I won that hackathon and which was, you know, huge surprise to me. But that's when I realized okay, I love engineering pretty much in any aspect and in any angle of Web3 because even that project was kind of like a feel-good project as well. And you know, much more of like a real-world benefit. And so the other aspect I learned too, because obviously you have to do a lot of writing and the documentation for your hackathon project and you'd have to do a video and everything, kind of fell in love with that process too.

So I was like, you know, I loved being educator in medicine and you know, why can't I be a good educator in technology as well? And so that's when I discovered Billy started working as a DevRel, I believe it was for API3 at that point, yeah, and he still works for API3. And now I, you know, I was like, you know, this is kind of what I want to do. So I joined DevRel Uni soon after and I was like the first cohort of DevRel Uni. Shout out to DevRel Uni by the way, an amazing program if anybody wants to become a DevRel, highly recommend jumping into a cohort with them. It's an amazing experience, amazing instructors.

You got people like Nader and Patrick Collins was in the last cohort and everything as well as an educator. Steph Orpilla, like a lot of like DD members and stuff like that are mentors in that in that course and I'm a big fan of it. So if you want to be a DevRel head over there. After DevRel Uni kind of dabbled into like a DevRel-esque role inside Developer DAO, worked a little bit at Intuition and then Pocket Network and now, you know, early this year in January, I got offered a role as a lead DevRel for Arweave and Forward Research. So that's kind of how I got here. Sorry, kind of long-winded but yeah, kind of crazy route and a little bit different than what you would typically get in the typical like CS degree kind of thing.

No, man, that that was beautiful. I love how you basically segued each part of the story into each other. That's fantastic and I think that that whole intro could be a podcast in itself. Maybe yeah, maybe we get into that later. Yeah, one day, that's right. But yeah, man, that's amazing and I love the story and I even learned some things I didn't know as you were telling that. So really awesome to see where you landed and just more evidence of hard work can get you anywhere you want to go. So definitely very ecstatic to see you where you're at now. And I guess in the interest of time, let's get into today's topic of AO and who you work for now. So if you want you can share.

That's right, that's right. And folks, as Patrick is going through his slides and and presentation, please feel free to post any questions you may have and we will get those answered for you as we go. Here we go. Let me see if this hit the play button kind of thing here. Oh wait, it's doing something weird here. Sorry. All good, all good. And there we go. And here we are. Yep, you're good. All right, you can see the slideshow now, right? Yep.

Okay. So let's, there we go. All right. So first and foremost, the biggest thing that I kind of want to break down is like, you know, what are we going to freaking talk about? And big things that we're going to obviously talk about is like, you know, we're going to be talking about AO, but I do want to do a real quick refresher on like Arweave because not a lot of people are always familiar with Arweave in general. We're also going to talk about like what's being built, you know, how do you start, and then just overall opportunities and everything that's going to be coming up.

So my desired outcome in this case is just to make everybody bullish on Arweave and AO. Because I'm definitely bullish. I know I sound biased and everything, but I've been a big fan of Arweave for a very, very long time. But AO is just kind of honestly is just I feel like has been dramatically, you know, catapulted Arweave's capabilities as an ecosystem. So you know, what the fuck is Arweave? Let's just put it that way. All right. So Arweave, for those that don't know, it's just known as the storage blockchain.

A lot of NFT projects have relied on Arweave for storage of metadata or even images. I'm sure everybody has seen somewhere where like an NFT project suddenly has all these dead images. And yeah, that's a real pain in the ass and, and you know, that doesn't necessarily happen with Arweave. And the reason why it doesn't happen is because of how nodes and data providers are are incentivized on Arweave to make sure that the data is available for a minimum of 200 years. It's a wild notion to think of of like a minimum of 200 years. But it's true, the incentive program is completely based to make sure that those incentives are very, very long-lasting.

So it's because of this that not only NFT projects store data. We're talking about like financial institutions, we're talking about all kinds of different verticals that use Arweave for storage of some sort. And now, before there was AO, what kind of data were you able to store? Well, a lot of it was like kind of back-end data. You can use SQL or no SQL databases built within Arweave. And so like WeaveDB is a good example of one of those. And so you can have your database stored, you know, permanently. So that's the back-end side. Now the front-end side is you can actually have your static and somewhat dynamic front-ends deployed on Arweave as well.

What that means is like I can actually build out like a static page and I can actually deploy that whole page on there. And that means I do not have to pay monthly fees, I don't have to pay, you know, for hosting provider or anything like that. I don't have to pay for domain services, especially with services like RNS now, which is the Arweave Naming Service, provided by the same team that actually provides ArDrive, which is kind of like a permanent version of like Google Cloud or sorry, Google Drive or even Dropbox, anything like that. Super intuitive. I think ArDrive is one of the best Web2 friendly applications in Web3 in existence, period. Like I think they've done bar none probably the best UI example of what Web3 should be as far as like a user interface and user experience.

But yeah, so that's pretty much Arweave. You had this like really great front-end side and you had this great back-end side, but there were some things that were lacking. So transactions can sometimes be slow, that communication layer between the front-end and the back-end can be slow sometimes too. And so then, you know, one of the major fixes to that was, you know, the release of AO. So what the fuck is AO? AO is it stands for Actor Oriented, which is based on actor-oriented modeling and it's the hyper-parallel computer which is basically a decentralized virtual computer that isn't owned by anybody.

You don't have any crazy DAO or anything like that that is doing some sort of voting mechanisms or anything saying who can do what on AO or anything like that. The consensus models and the voting models and everything on it are very, very light. This is what makes AO super duper fast as a virtual computer. So now those things that were lacking, like slow transactions, we're talking about difficulty in the communication layer from front-end to back-end, stuff like that, that is completely blown away and gone as issues when it comes to using AO. So let's just kind of jump into some of the things that AO is now kind of unlocked.

DeFi, on-chain games, on-chain AI, decentralized middleware, decentralized data streaming, and autonomous agents. So okay, this is so cool and this is why I love working with the entire Arweave ecosystem. And this is where I feel like this is where I've kind of found my home. Long before AO was even a thing, I came to Arweave specifically for the technology, but I stayed because of the community. And why do I say that? Because DeFi didn't exist and you would think yourself, well, well if DeFi didn't exist then why the hell are you there kind of thing? Because DeFi is kind of the big unlock of Web3.

Well, that's not what I came to Web3 to build. You have to remember in my backstory, I came to build, you know, authentication for physical products to kind of try to kill the whole idea of counterfeit products, counterfeit manufacturing, counterfeit product trafficking, and things like that. So I didn't have really like a DeFi bone in my body at that time because it wasn't really necessarily important to me. And so it wasn't a big feature of Arweave as well. And so why does that make me bullish on Arweave long before AO was even a thing? Was that if you came to Arweave and you stayed to build and you became a team member or anything like that in the Arweave ecosystem, you believed more in the mission than you did believe in like just trying to make a quick buck kind of thing.

And that's, I mean, let's be real, a lot of people come to places like Ethereum and Solana and stuff like that because they want to make a quick buck. And so it makes it sometimes kind of hard to see, okay, you know, do I want to work with this guy? What are their intentions? Are they here to just make some quick cash or are they really here to, you know, really solve a genuine problem? And for me, you know, DeFi, you know, it makes it kind of muddies the water a little bit. And so for me, when I was working with everybody in the Arweave ecosystem, I honestly I just fell in love with the overall community because everybody was so mission-oriented. And so I loved that about the Arweave ecosystem.

And so now AO has really unlocked DeFi. Now, is that a bad thing? No, because everybody that was here before is still very mission-driven, very mission-oriented, so it's very easy to see who was here before that and then, you know, and who has really come to kind of become part of that original team or that original community. And so I think Arweave is one of the best places to network and become part of a community that is really driven towards a genuine mission and not just, you know, here to make a quick buck. So that's my spiel on the DeFi side. But DeFi is a huge unlock now. Transactions are significantly faster, especially, you know, now we have like wrapped AR versus AR and you know, there's various different tokens that are available now, tokens are like you can deploy a token extremely easily in AO as well.

On-chain games. On-chain games is cool because in AO, it's not just like, you know, I a lot of people say, oh Web3 games this and that. Most of the Web3 games out there are really about Web 2.2. Let's be real. Because it's really just like the currency and the you know, maybe some like items in the game like NFTs or something like that. But the overall backbone of the game, the graphics and all that stuff, the actual play, it's all, you know, it's all centralized and and it's not on-chain. Now, so to me that's like okay, that's really just about Web 2.2. And so with AO, everything, everything is on-chain. I'm talking up, down, left, right is on-chain.

And a lot of people immediately ask the question of like why would you want up, down, left, right on-chain? Well, it's interesting because I'll break it down a little bit and we'll have a an announcement about a feature called Arfleet pretty soon, which actually creates a layer for temporary storage, which, you know, really scales modularity and everything inside of the Arweave ecosystem. So parts of the game you can actually run sections of the game that could be up, down, left, right and things like that, the actual gameplay, you know, in the Arfleet layer and then you can take the results and act of that game and store it on-chain. Which is pretty freaking awesome. On-chain AI, man. This is such a cool feature too and it is growing rapidly. We have so many different teams and features and stuff like that building on on-chain AI.

And people I think are sleeping on what is truly capable with on-chain AI in the Arweave ecosystem. We already have Llama 3 fully operational inside of AO. And we even have games, actual games where the NPCs are ran by that by that LLM. Which is so freaking cool. And so, yeah. So I mean you can the fact that you can have NPCs controlled by AO and the entire game itself is on-chain and everything as well, as well as what is controlling those NPCs, the LLMs and everything as well. It's super impressive to me. Decentralized middleware. This is something I'm pretty excited about because obviously the big thing that people kind of griped about was the static front-end communication layer to that the data back-end inside of Arweave and slow transactions or slow communication. It's gone now. And that's because of that decentralized middleware feature that AO kind of unlocks.

And then decentralized data streaming. We're very, very close now. We even have a few bounties out there, I believe, where if anybody wants to build decentralized data streaming and I when I say data streaming, I mean all the way up to like video streaming, audio streaming, stuff like that. We actually have bounties for that are available to people to be able to start building with decentralized data streaming inside of AO. And then autonomous agents. Last but not least, this is so cool and I you know, if and when we have time, I'll actually put together a really quick autonomous agent for everybody where it will basically get a price feed for you as well as a news feed and then, you know, later on in your own time you can actually build that so it feeds that into the on-chain LLM.

Oh, what happened here? Hopefully can you still hear me? Yeah. Okay, cool. Had a little bit of weird actions going on but we're good. But yeah, so like you can basically build these autonomous AI agents on-chain where they can get the price feed, they can get the news feed, feed that into LLM, that LLM can decide, okay, you should buy or you should hold or you should sell. And then, you know, it'll spit that information back into that autonomous agent and then it will basically make those transactions for you. And so there's all kinds of different things you can do. The the NPCs I mentioned previously with on-chain games, those are autonomous agents in their own rights as well. So there's all kinds of unlocks.

Next, okay, this eco overview for AO is it's dated. I actually think it's like three times as big now. Probably and but I mean we have everything from like new infrastructure layers, we have huge unlock for block explorers, we have new bridges, investment a lot of new investment opportunities, some big wallet unlocks, I'll give a couple hints to that in a little bit. AI you can see like DataOS, Pointed AI, Outcome. Yeah, I think that that section alone I think is like four times bigger now. The communities are much, much bigger now. Weavers is a great organization that just recently started.

I got to meet a lot of PermaDAO people in the Chinese community. Man, PermaDAO is a great community. So if you are part of the Chinese community I highly recommend hitting them up as well as Arweave Oasis. Amazing community as well in the Chinese community. Only Arweave does some of the best ecosystem updates for Arweave and AO, which is ran by Longview Labs. Highly recommend following them. It's @OnlyArweave on X. DeFi, I mean we got Permaswap, Arswap, Astro is really cool because they're the first provider of a stablecoin in the Arweave ecosystem. Really freaking cool. Bark, Harvest, and Dexi are all in there.

A lot of new developer tooling, lots of social because of those fast transactions and everything now we've completely unlocked social there. The gaming oh man, AO effect was just kind of the beginning of things. But now we have like Permaverse, we have AO games, we have Lammaland, Reality Protocol. I mean so many things have completely just been unlocked by AO and it's happened in such a short amount of time. We have to remember guys that I mean the ecosystem is roughly three times bigger than this and on top of that, which is, you know, mind-boggling to me, it's like we've only been around for about seven-eight months at this point. I mean we released in February 28th and we're still on testnet and yet we've just got a flood of these teams and developers and engineers and even business people, investors, stuff like that just really pouring into this because this has been such a big unlock to the Arweave ecosystem.

I like to show this slide because this is basically the outcome of a three-day hackathon at a for Arweave India. They did a three-day hackathon like hacker house thing. And this was the outcome of it. And what they built in three days is really freaking cool. And I like to show this slide because it also shows how fast and easy it is to build with Arweave and AO as well. So we've got SAM which is smart contract auditing, Loom interactive 3D worlds on AO. I mean imagine building something like that in three days. That's freaking cool.

Data ownership and content monetization, USDX which is a fractional algorithmic stablecoin as well, Arkey MPC-based wallet on Arweave, AOVest that team I've actually been talking to recently really really cool, Forever AI decentralized and on-chain AI inference platform. That's I mean the fact that people were able to build something like this in a matter of three days is super impressive. And then FundArs, permissionless ecosystem for funding platform. Really really cool stuff to be put together in only three days. I mean I'm very very impressed, you know, I do have to shill the Arweave India community. Very very impressed by that community as well as the engineers there. I'm a big fan of everybody that is part of running Arweave India. I'm a big fan of Ropas, he's another Developer DAO member as well, he runs Arweave India as well and they've put together some really really cool stuff.

There's one project I have to mention that Ropas was part of was they put together a physical camera. This is so cool. They put together a physical camera that when you take a picture with this physical camera, it stores the image immediately on Arweave. And that's such a I mean to be honest like it's such a it's a concept thing. Let's be real. They're not looking to mass produce this product or anything, you know, but it's such an interesting thing of like what is truly capable inside of Arweave and AO now. It's so cool.

So then for some reason I can't see the questions. Are there any questions you want to read any of those for me? Basically everyone's giving you fire and rocket emojis and amazing, which I agree with. Cool, cool, cool. All right. So why is it easy? All right. So yes, you could build with several different languages and stuff like that, but you really only need to know JavaScript, TypeScript and Lua. And in this case TypeScript isn't even like a hard requirement, you can do anything with JavaScript.

Lua is a ridiculously easy language to learn. As a matter of fact if you type in Lua tutorials, you're mostly going to see tutorials for like Roblox and stuff like that because it's a language that is often used for like doing game mods and stuff like that for like Roblox and like Minecraft and different things like that. And they're targeting anywhere from as low as like eight-year-olds to as high as you know like maybe 15-year-olds or something like that. And so I mean Lua is an extremely easy language to learn. So if you're unfamiliar with Lua, instead of writing out these, you know, async functions or sync functions or anything like that, then you know, what you end up actually writing is these handlers. And so these handlers do what's called message passing and this message passing is basically what handles all the functionality within either the middleware or the back-end functionality of any application that or autonomous agent or anything like that.

And so basically these handlers are either waiting to receive a specific message or a tag of a message. It takes in the data from that message and then you know based on the results of you know that tag or the message data, it will execute a specific function for you. It honestly it is very very easy to learn. As a matter of fact if you do want to learn it, we'll I'll show you guys a link to the cookbook for AO Cookbook pretty soon. We have some we have two really cool tutorials. One's called the begin tutorial. It's a lot of fun to play with because I built out a couple autonomous agents in there based off of the movie The Matrix, so you got Morpheus, you got Trinity in there, and they basically set up a like a as you're going through the tutorial, you're within AOS, the operating system, and you basically get to interact with these characters and they give you specific challenges and those challenges basically teach you how to build within AOS and AO and everything. So yeah, an amazing and fun learning experience if you ever want to jump into the cookbook learn a little bit about Lua, learn a bit a little bit about programming with AOS.

So and then a lot of people kind of say, well okay, it this seems too easy. You know, there's no Rust, there's no Solidity. Why you know, are you just foregoing consensus altogether? No, not not at all. And so what's really interesting too is like consensus very much in many blockchains, especially Ethereum, is the main reason why Ethereum can run slow because of how consensus is managed. And this happens in many different blockchains. Well how we do consensus, specifically on AO, is despite its significant computation capacities, AO is able to maintain traditional smart contract execution capabilities while using holographic state mechanism.

And what that means is instead of coming to a consensus about a state of computation itself, AO focuses on ensuring that the logs of interactions are written to and available on Arweave. So what does that mean with like logs? So if I have a process which is kind of the core component of like what you're going to be building with in AO and AOS, I spin up this process and I'm executing specific handlers or actions or things like that with this process. Well, consensus is reached by basically checking the logs of that process. And it's significantly faster to do that rather than trying to check the computation itself. It's checking logs for that process and seeing what, you know, the history of the message passing and message receiving are like on on that specific process.

And then it will do the same thing on the process that it's interacting with as well. So this allows TPS to be extremely high you know with within AOS, sorry in AO, and allows for a lot of the interactions within let's say the games and stuff like that to happen very, very, very quickly. Just a question on that real quick. What is exactly being verified in the logs? Is it basically a verification of the chain of computation that went down and like a quick verification of that or how maybe if you want to dive a little bit on that?

Yeah, in a nutshell, yes. It will check the logs as far as like what kind of data is being passed, what kind of actions are being executed, the size of data that's being the passed and everything as well. Because you have to remember everything is being ran in parallel, which allows this kind of like horizontal scalability which you know, in theory you're based on how it's built with the horizontal scaling, it's infinitely scalable in theory. And so that being said, it is important to be able to verify, you know, computation size and everything as well during the holographic state mechanism. Does that make sense? Yeah, totally. And I guess that kind of lends itself to AO being able to support non-deterministic outputs like AI inference, right? Absolutely. Right, right.

So instead of having another set of nodes re-compute the same thing and maybe get a different set of results, you're just verifying that the computation basically occurred in in the transactional logs. Yes, 100%. Yeah. And of course there is kind of like a layer of trust here where like obviously if a process has been running computation for X amount of time and it has, you know, a long series of basically trusted transactions, it allows for the trust and the computation size and everything allows for that transaction to happen faster. And so that's also an important mechanism of the holographic state mechanism as well. So yeah. To get started, for those that want to jump in and play with this or anything like that, the barrier to entry is super low.

As a matter of fact, I highly recommend even if you're not an engineer or you're very new to engineering or anything like that, I highly recommend anybody that is even curious to see how easy, you know, jumping into something like this is, head over to Cookbook at cookbook_ao.arweave.dev and go through those tutorials. Because the tutorials are very well laid out and it's very, very easy to get into and once you're done with those tutorials, jump over to Quests. We have a quest page where in those quests there's some opportunity to earn some wrapped AR and things like that as well.

Really, really cool stuff. And so I highly recommend if anybody wants to, you know, learn and jump into this. I also say that Arweave in my opinion is probably if not one of the easiest if not the easiest place to go from like a Web2 engineer to a Web3 engineer because you're not required to learn you know, really, really complicated languages like Solidity or Rust or anything like that because I mean those languages are obviously other than Rust really, but you know, Solidity is very, very, you know, exclusive to contract writing and everything inside of Web3, whereas, you know, with Lua, JavaScript, and TypeScript, I mean you pretty much probably already have those skills already coming from Web2 and going into Web3.

So this I think anybody that is let's say they're Web2 engineers and they're Web3 curious, this is a great place to start because of how easy it is to get in and just start building things. We have some really cool tutorials, video tutorials and all kinds of stuff and so I highly recommend, you know, jumping into the cookbook and then the next next part, which I highly recommend as well, is making sure that you kind of follow these these accounts on X. And it's not just like a oh hey there's a call to action so I get more followers or anything. No, like we do really regular ecosystem updates.

We do online hackathons, the Weavers' Org is a is an online community that does some really great bounties, which they call raids, and it's this time time period where people can kind of, you know, execute specific actions to be able to complete these raid quests kind of thing and you know, you can earn some wrapped AR and things like that or or just AR token in general. And so I definitely recommend following Weavers' Org. AO the computer I haven't done one in a little bit because I was traveling for three weeks in Asia doing workshops and everything like that. But we do a show called The Testnet Show and The Testnet Show is usually like one or two videos a week where we just talk about ecosystem updates.

We talk about opportunities, we talk about live events, online events, offline events, we talk about new bounties, we talk about all kinds of testnet updates. Different projects that are either building new stuff or we even talk about like who's hiring and all kinds of stuff. So I mean highly recommend following the AO the computer if you're really, really interested in like getting into the Arweave ecosystem. AO the computer is a great place to start because a lot of people follow there, a lot of people get their ecosystem updates and it's a really easy way to find relevant information, most up-to-date information, so you can you know, be able to make sure that you're joining those live events or those online-offline events and things like that.

The next part, you know, kind of a bit of a plug for some of the ecosystem communities that I've been really proud to work with is like Arweave Oasis. I spent three weeks doing workshops in seven different cities and with with Jerry from Arweave Oasis, he put together some really amazing offline events all through China, Japan, Taiwan and and pretty soon I'll actually be you know with him in Singapore and then immediately after Singapore, I mean little teaser, don't know if I'm supposed to say this or not but we might be in Malaysia immediately after Singapore.

So you know, yeah, so a little bit of alpha there. But you know, highly recommend if you're if you're anywhere in Asia, highly recommend following Arweave Oasis. Arweave India, you know, I already shilled them pretty hard earlier, but if you want to be in a community of some really amazing developers, go head over to Arweave India, great community. Arweave Japan and Arweave Philippines are very, very new, but they're growing very quickly. And I highly recommend if you're in Japan or if you're in the Philippines, go follow them. They've had some really great offline events. I think Arweave Japan has an offline event and online event this weekend. Starting this weekend I think.

So head over there you can see some of their announcements. Arweave Philippines just had a hackathon not too long ago. Arweave Africa is constantly doing things. So you know, highly recommend if you're in those geographic locations, head over to those accounts and definitely give them a follow. Aside from that, that's pretty much the end of my presentation here. So let me go ahead and just end the share. I know we only have like maybe what 15 minutes left, but a couple things I do want to hit on is if you are going to Token2049, we do have an event called Arweave Asia. It's on the 19th. I think it starts at like 1:30 something like that.

So um, I don't have an active link for the Luma here, but you know, I'll make sure to post a link and maybe Developer DAO account and the DevNTell account could probably share it. It's for there's a Luma link for it. It's going to be a massive event. I mean we're I think we're estimating like 80 to 100 people, something like that are supposed to be there. Got a lot of sign-ups stuff like that. Some really really great speakers, I don't want to give away who the speakers are yet, but I'm super excited. A couple things I'll shill really quick, Tate from Community Labs, he just kind of hinted at the very first mobile wallet for Arweave.

And I don't know if we're going to hear about that at the event or if we might hear about some other things because I mean they man, Community Labs has been pumping out some really great developer tooling, some user tools and user platform stuff like that. Huge fan of what Community Labs has been doing. Tate has been doing some crazy stuff. Not too long ago they just held together held a incubator for 30 like $30 million incubator, which is insane. It's and it was all dedicated to people building with AO. And some amazing projects came out of that incubator.

So back to the Singapore part, if you are going to Token2049, hit me up on Twitter, hit me up on you know or sign up for the Luma link anything like that. I'd be more than happy to get to see you there. I'll actually be running as like the MC for the event so you'll see me just probably just dropping F-bombs on the mic and stuff. I know Sam's going to be super upset about it, but whatever, you know, that's what, you know, when you hired me as the lead DevRel, that's what you're going to get, man. And so but no, it's going to be it's going to be a lot of fun. We're going to have a ton of fun at the event and super excited to get to see so many people, so many Developer DAO members I haven't even met yet at Token2049.

Really excited to you know, take some selfies and stuff like that with those guys and because I feel like I've known you all for like three years, maybe three year plus and I've never seen you guys in person so it's going to be super exciting. But yeah, man, I feel like that's it. Do we got questions, comments, concerns, complaints, whines, gripes, cries, anxiety, anything? We we have questions. One one question here from ZQ. Not a dumb question. Is AO ZK capable? Any project on that? Yeah. So we got like two or three projects on this. This is super cool.

Pado Labs is a great project that is doing ZK on AO and that's I mean they've got some really cool stuff and then there I mean honestly there's like a line of projects that are just like waiting for like these two or three projects to like finish their ZK tooling so they could start using the ZK features, which I mean I believe some of them are like right around the corner almost done. So actually as a matter of fact it's beyond ZK. Pado is like is trying to implement ZK FHE. So fully homogeneous encryption and everything as well including the zero knowledge. So not a dumb question, there's yeah we got all kinds of cool stuff that is coming out inside of the AO ecosystem for that. So yeah, ZK is definitely a capability.

Awesome, awesome. And great question ZQ. Question from me. In regards to how the computation is performed on the network, obviously some tasks are not as computationally intensive as others, specifically if we're talking about like Llama 3 like you mentioned before that is quite a computationally heavy thing to run with all the memory and hardware requirements. What actually supports the underlying infrastructure that does the computation? Is it people providing their own hardware to do it or are you guys utilizing another infrastructure layer to do that?

Okay, so this is a great question and I'm going to basically say that like it's to be determined in some cases because right now in testnet, obviously we're handling all the computation and everything. But when we go to mainnet, there there are going to be and I know Sam and you know, a lot of the node developers and everything, they are actively I think they even had a big meeting yesterday where they're actively working some of this out as to exactly how this will function and getting ready to test it and everything you know, within the near future. So I don't have a great answer for you at this point.

I do believe that there will be a an opportunity to make some of your own computation available. I do know that there'll be some opportunities to be just basically a node provider or even own or sorry join basically kind of like a network of node providers. Kind of similar to like you know the staking pools almost kind of the same functionality as that as well but for computation essentially. So you're kind of looking at this as like almost like a computation marketplace kind of idea. Nothing is genuinely set in stone just yet on on that, but there are a lot of different people working on that that part. So it's a really great question though.

Awesome, yeah, yeah. For sure it's a very very very bleeding edge sector of of this cross between Web3 and like heavy AI required computation. But yeah, very fascinating to see how things develop there. We have a question here from Eason. How do Arweave and AO compete with other L1 blockchain ecosystems? What are the advantages? I guess the question comes down to like why choose Arweave and AO over a competitor? What what would compel a person to come use this stack?

Well yeah, I mean there there are a few options and so like your immediate thought of like okay an L1 you're thinking like obviously like Ethereum and but then you think of like Solana and then you start kind of getting closer to things that like what else do we do with decentralized computation? So you're thinking of like ICP, Filecoin and things like that as well because I know Filecoin has like a virtual machine system as well and and so the biggest thing that I would recommend, number one, number one is obviously the big difference is consensus and how consensus is run.

The vertical scalability sorry, horizontal scalability, horizontal scalability is a massive unlock for the parallel computation element. Barrier to entry, I mean honestly like I don't know if you have ever gone through the various onboarding experiences for like different layer ones that might do some sort of decentralized computation. But man, I mean I can get through all of the tutorial basically tutorials on the cookbook on AO in a matter of like a couple hours and start building right away and I should have you know, if I spent an entire day just hard coding stuff like that, I can have an application done in a day without having any knowledge of like what AO was in the beginning, which is really freaking cool. So barrier of entry is a big big thing.

Cost efficiency for storage and things like that. So like a lot of layer ones have both the computation side as well as the storage side. And I would say the cost efficiency. You have to look at this in a kind of a mathematical way as like you know let's just exclude the Arfleet side, that temporary storage element. With the excluding the temporary storage with Arfleet, which by the way we should have a live discussion about what Arfleet is and everything because we just kind of shilled it without giving too much information but we're going to do a deep dive discussion either Monday or Tuesday. Sam is going to go live with Sergey the the developer of Arfleet and then a few other people and so you know, I don't even think we've even announced this event, so you know, sorry David if you're watching this and I just totally blew the lid on this one.

But Monday-Tuesday, more alpha, yeah, but you know, we should have event very very soon discussing the temporary storage side. But anyways, going back to the permanent storage side, people say man storing on Arweave is so expensive. Man, if you do the math of like the because you know, let's say Filecoin or anything like that they they do it like contract basis and so it's a set amount of time, okay? That set amount of time, that cost analysis, the set amount of time or you know let's compare AWS or any of those things. Trying to map that out to a 200-year storage, dude you're getting a fraction of price in comparison to anything else.

So you have that use case of like okay, permanent storage, that's what it's for. So like now that target market really shrinks down as like why would I want that data to be available for 200 years? Well, okay, a lot of that data is very very important depending on what you know project you're building. Now you have that temporary storage side of Arfleet. Well, I don't want to go too deep into exactly how that works because obviously we're going to do a deep dive discussion exactly how that is, but the cost efficiency of the temporary storage side and the modularity capabilities and every modular capabilities and everything with Arfleet are I mean they're massive. It's going to blow everything out of the water as far as like how scalable and cost-efficient data is going to be inside of Arfleet. So sorry inside of Arweave ecosystem altogether.

So yeah, as far as like choosing layer ones and everything, we're talking TPS, you know, so super fast transactions inside of AO. We're talking temporary and and permanent storage options. I mean we're talking about you can you can officially have a fully decentralized application from the static page to the domain to the middleware to the back end everything every single thing could be fully decentralized and cost-efficient now, which is like what we weren't really able to do before AO. But now everything is very cost-efficient. I mean it's a no-brainer at this point as to why you would choose Arfleet or Arweave AO. And on that, I guess I think we've reached a point in the Web3 ecosystem where we have enough infra and all the necessary boilerplate to be able to build really interesting things.

So it's a very very awesome time to be alive and to see all the innovation that kind of comes out after this point, which I'm really looking forward to see. And there's a tangent question here, I think you basically answered it as you were giving the answers to the previous question. Great question by the way Eason. Is basically AO is being compared with ICP and why is that? Yeah, well okay, so I mean you know ICP they they came out as a layer one with the full intention of decentralized computation. And their underlying technology is honestly I've you know I've spoken to some of their team members, I've spoken to some people building within the ecosystem.

We have people inside of the AO ecosystem that build you know in AO and ICP at the same time. Their technology is really freaking cool. And some really really intelligent people building ICP. The biggest difference is I would say like they're two different tools, kind of I would say very different approaches to solving similar similar problems. And so obviously I do have to admit that if you were to say the immediate difference when you get into their documentation versus our documentation right away is barrier of entry. Like it's and it takes a little bit to really kind of get to understand what you're doing and and how to do things inside of ICP. Biggest thing is like governance as well.

They have a large governance system and to be able to do anything in the governance or anything I believe there is a series of like staking and things like that that's required. And so the governance model is really robust and and it takes a little while to kind of understand and get into. Everything on AO is extremely light. We are very very light in consensus, we're very very light in the governance. The only real governance that we're going to have is through the gateway providers, which you know if you want to dive into like how gateways and everything are run, highly recommend heading over to ar.io and you know you can get a full tutorial on how to spin up your own gateway and be a gateway provider yourself.

But they are the only ones that I mean they can come to a consensus or a vote really as to if something really egregious needs to be removed off the network, those are the only people that can really kind of do that. But that's the only series of like governance I would really say. We really believe in permanence and permissionless. That's a big difference between like ICP and and in Arweave and AO. I would say ICP is very permissioned based which you know obviously has its benefits in itself depending on what you're trying to build and then you know Arweave and AO is very permissionless. It's very easy to get into, very easy to start building, very easy to start deploying things. So yeah, I think that's a major difference other than just the various languages and stuff like that that you that are required to build and do decentralized computation.

Awesome, awesome. Great answer and great question ZQ. Unfortunately that brings us to time. Actually we've gone double time here on DevNTell and turned this into a special special edition which is really awesome. I'm so sorry guys. No no no. It makes up it makes up for the time we had to wait from yeah the first time around. True, true. So I hope everybody enjoyed this impromptu special edition of DevNTell with Patrick and AO. Yeah, it was great to see everybody, great to answer questions and stuff like that. And you know if you if you have any more questions or anything you want to really dive into it more, you know head over to the the AO Discord.

You can hit me up on X, my X account is PSkinnerTech, pretty easy to find. And then obviously if you're Developer DAO member you can find me there as well. So you know super excited to always be here Narb. DevNTell is a great show and just lucky to have you, man. You're the you're the you're the most consistent person I think I've ever met in a Web3. And this show I could I could set a clock to the DevNTell show. Thanks man, really appreciate it. And yeah just trying to keep up with everyone else.

But yeah always a pleasure to have you on to speak with you Patrick. You're always welcome to come back on to tell us the latest and greatest of AO or anything else you've been cooking up behind the scenes. But yeah man, thank you so much for taking the time to come and speak to us today on AO. Really appreciate it. Absolutely, looking forward to it. And thanks man. Take care everybody. All right guys, happy Friday, happy weekend wherever you may be and we'll catch you back here next week for another great DevNTell. All right y'all, see ya.