All right, we are live! So, welcome to the AI Agent Master Class. So, we're going to get started here in just a few minutes. We'll let people trickle in, we'll get the stream rolling, but let me know if you can hear me in the chat, okay? Let me know if you can hear my microphone, and let me know by telling me, uh, where you're calling in from tonight.

So, Kim can hear me, okay. Matt's got me. Got Florida, Philly, Dallas, Texas, Jacksonville, Quebec, South Bend, Stuart, Florida, Washington State, Colorado, Durham, California, Western Oregon, Maine, Dallas, Romania, Florida, New Hampshire, Arizona, California, Washington, Dallas. Welcome!

All right, everybody, let's waste no time. Let's jump right in here. So, I'm super excited to get to share this with you tonight. You know, welcome to the AI Agent Master Class. Uh, we're going to show you how to create a blockchain-based artificial intelligence agent tonight. So, you know, this is a really emerging industry. You know, artificial intelligence is one of the rapidly growing technologies of our time, and so is blockchain. And when you put the two together, uh, you're on the cusp of something really cutting edge that has a ton of possibilities that we're just scratching the surface for.

So, I'm super excited to give you an overview of what that process looks like tonight. You know, how AI is evolving, how it's intersecting with blockchain, and how we can put the two together and what you can create. And I'm even going to show you how to create your own AI agent tonight, uh, with some simple, step-by-step instructions, and show you how to customize that to make it do a lot more.

So, um, we're going to cover all the concepts that you need to understand, and also the actual coding setup. And we'll take a look at the code and how you can make changes to it, all that type of stuff.

So, you know, a couple quick house rules before we get started here. Um, you know, we're going to have some chat as part of our presentation tonight. This is a live presentation. Some people ask me like, "Hey, are these actually live?" Yeah, they're actually live. Um, you know, for example, Luro, and then podcast, intern news, and media, etc., etc. Yeah, we're live. Um, so we will have chat as a part of our presentation tonight. Um, stay inside the YouTube chat. I know we, we got people in the Blockchain Mastery Program, there's a chat room there, we got the mentorship, we got all this stuff, but, uh, any questions related to the presentation tonight, let's stay inside YouTube. It's too much for me to like, look at a bunch of different apps while we're chatting tonight.

Um, that being said, you know, we will have a time of questions at the end. So, it's a lot for me to present and like, you know, try to try to set everything up and demonstrate and screencast and, uh, watch chat. So, I will look at the chat periodically from time to time, but, uh, we will try to save the bulk of our questions, uh, towards the end of the presentation tonight.

Okay, so, um, yeah, that being said, we will have, uh, some coding as a part of our presentation tonight. It's not going to be extensive coding, it's not going to be hard stuff. So, if you're here just to kind of get a taste of what this is like, and you don't have to be an advanced coder to do this, um, but what I would say is, um, we will have some step-by-step instructions here. It may be hard for you to follow everything in real time live, just because there's some time in between different tasks. So, what I would recommend is definitely, uh, you know, getting an overview of the presentation tonight, and then watching it again to do everything step-by-step after you've gotten the high-level overview.

So, no matter how you join the class tonight, whether you join the Blockchain Master Program or you just join the one-time class, um, we will have a replay link sent out to you, so that you'll be able to, um, you know, watch this later and follow it again. You can rewatch it a million times if you want to, uh, it's yours for life. So, that's what I recommend is don't worry about, you know, following every step-by-step instruction on the live presentation tonight. Definitely give it a rewatch after.

So, that being said, you know, we're going to talk about how to set this AI agent up tonight, um, and, you know, there's a lot behind the scenes technically speaking here. We're going to show you how to do, uh, your own agent, but there's a million different things that you can do with AI agents. I mean, the world is your oyster. Uh, there's so many different possibilities that we'll talk about. And if you want to leverage the power of these actual AI agents to do cutting-edge things, then obviously, you know, having some advanced developer skills is going to give you the edge to do that. And so, the absolute best way to get those advanced developer skills, um, is through learning with hands-on mentorship, okay? And I do have a couple of slots open, uh, inside the Blockchain Developer Mentorship Program at DAP University, um, where we will work with you to get the skills to become a blockchain developer through a multi-month training program, work with you one-on-one, um, and even work with you to get hired in the industry if that's something that you're interested in.

So, I do have a couple of slots left open for this. I'll put a link down in the description below. Uh, you can submit your application there. And yeah, we've got a lot of people from the mentorship on the presentation tonight. I think that they will attest that it is a hugely valuable, uh, journey that they're on. So, um, yeah, that's pretty much it, guys. Um, I'm super excited to get jumping into this. Just wanted to cover some house rules, some ground rules before we get started. That being said, let's go ahead and jump in and get started.

All right, so, so Jamal says he's new to coding. That's totally okay. Um, no issues there. So, let's kind of start off with the basics, all right? Let's talk about what is an AI agent in the first place? That's what we're going to create tonight. That's what we're all hyped up about. You know, AI is one of the fastest-growing technologies of all time, so is blockchain. When you put the two together, the, the possibilities are are absolutely crazy.

So, let's talk about the current state of AI and the current technologies that people use the most, all right? So, you know, unless you've been living under a rock, you've probably seen ChatGPT or at least heard about it, okay? So, you know, ChatGPT is really just an AI chatbot, okay? Um, it's what's called a language model. So, you can talk to it, and it's trained on all this data, okay, that basically will just give you answers to pretty much any question that you have, all right? So, um, you know, for example, here, here's an ex, here's ChatGPT. I'm not even logged in. This is the free version. You can go test this out if you want to at chat.openai.com. And I can just say, you know, "Hey there, can you explain to me like, like I'm five years old, how a blockchain works?" Right? And it's going to give you a pretty good answer, um, to that question, all right?

So, it's like your buddy. You talk to it, it gives you answers back. It can also do things like say, "Okay, cool. Can you write me the Solidity code for a basic cryptocurrency token?" All right? It can do things like write code for you, okay? Pretty cool stuff. We've seen lots of developer tools pop up where people can create, you know, websites really quickly, um, and all that type of stuff. Now, that's really powerful, it's really awesome stuff, but there's a limitation with this type of thing where all it's doing is it's giving you answers back, okay? You ask it a question, it gives you an answer based on what it's trained to do, but it can't actually perform tasks for you, okay? It can just give you answers, and you can take that feedback and perform the task yourself.

But that's where AI agents come into play, okay? AI agents are actually artificial intelligence, um, apps that can do things for you, okay? So, what are, what are some examples of things that it can do for you? Well, you know, AI agents are really popular outside of even the cryptocurrency space. You could have an AI agent do things like any automated task you want to do. You can answer emails for you when they come into your email inbox automatically, all right? It can read what the person is asking for, and you could train it on certain types of emails that you want to answer. Let's say you run a, um, you know, customer service-related a business, or a business that has customer service. If somebody, you know, emails you and says, "Hey, I need a receipt for the thing I just bought," you know, you could train the AI to understand those messages, just like it's understanding your natural language right here, okay? And then you could train it to actually respond to the email and say, "Okay, sure, here's your receipt." And that saves the person time from having to check their email, all that type of stuff, right? You could have AI book travel for you. It could watch for airline tickets and see when the cheapest prices come up, and it could, you know, purchase the airline ticket for you, all right? They can do all types of things. Now, that's not blockchain-related, but that's just an example of what an AI agent can do. This is a language model that you talk to, it gives you the answers back, but an AI agent actually does things on your behalf.

So, with crypto and blockchain, you know, how, what can you do with an AI agent there? Well, the cool thing is that you can get, you can take an AI and essentially equip it with a crypto wallet, all right? So, just like a, a MetaMask wallet on the Ethereum blockchain, like a Phantom wallet, uh, to go to the Solana blockchain or something like that. You can essentially give an AI a key pair, like a, a wallet, a public key and a private key, and actually give it some cryptocurrency, and it can do stuff for you, okay?

So, what's an example of this, right? So, one pretty cool example that I've been following recently, you guys might have seen the video that I put about this on YouTube, is like, uh, Billy Bets, okay? So, what is Billy Bets? Well, Billy Bets is a crypto AI agent that essentially is, uh, doing sports betting, okay? So, what it's doing is it is taking a new approach to, you know, sort of making money on the blockchain that a lot of people aren't concentrating on. So, a lot of people are making AI agents to do things like, you know, listen for new cryptocurrencies that are, uh, put on the blockchain and trying to find new ones to buy that could potentially appreciate in price and then sell them. So, basically, like sophisticated trading bots with AI. This is a similar type of thing, but it's not betting on cryptocurrencies as much as it is sports betting, okay?

So, Billy, what it does, uh, is it plugs into Polymarket, all right? So, Polymarket is a really popular, um, predictions market. Basically, what it does is it lets anybody on the blockchain, um, list events that have verifiable outcomes. Like, so Polymarket got really popular during the US presidential election, um, trying to say like, "Hey, who was going to win the presidency? Was it going to be like Donald Trump? Was it going to be Kamala Harris?" And you could watch people bet on that in real time, and people sort of use that as a probability indicator. And again, not trying to get political here, but that's how Polymarket got really famous in the past year or so. And, U, you know, it was forecasting, you know, Trump victory before, you know, when a lot of the news outlets weren't. So, that's how it got really famous. But after that happened, after the election was over, you know, Polymarket was like, "Okay, well, we got to have another reason people keep using the application because the election's over, nobody cares anymore." So, they really leaned into sports betting, okay?

So, basically, you can see here, you got all these different types of outcomes on, uh, different, different sporting events that are happening. You know, "Who's going to win right here? Arkansas versus Kansas," etc., etc. You can place bets on who the outcome is. So, basically, you've got all these markets for this. And so, Billy is an AI that essentially is getting all the data from Polymarket on these different types of markets, uh, it's plugging into something called Sports Tensor, which is helping it determine the probabilities of different types of things, and it gives it sort of different options. And Billy actually picks the options that it thinks are the best, and it places the bets on the markets, and it actually takes profits, um, as it starts making money. So, it's an emotionless thing, you know, it's not like making really emotional decisions like a human would, and it's automating all this process. And you can sort of see on Billy's dashboard like what he's doing here, okay? You can see different summaries of some different types of bets, okay, with his ROIs. And what's also cool is Billy has his own token, okay? So, Billy also has a cryptocurrency associated with it that you can hold, because whenever Billy makes profits, he's actually burning some of his own token, okay? And that's allowing him to, uh, potentially third parties, uh, like you or me, hold that token and capture some of the value, uh, you know, if Billy is profitable. So, that's a really cool example of an early crypto AI agent. Again, there's lots of other things you can get an AI agent to do. You can get it to, again, evaluate new cryptocurrencies that are on the blockchain. You could train it to look for arbitrage opportunities, uh, you could train it to, you know, value things like NFTs. Other, those aren't quite as popular these days, but that's still something that you could get it to do, all right? You can get it to look for different DeFi strategies on like, "What has the best returns?" Okay, different pools. It can basically automate the yield farming process for you if you're trying to move in between different liquidity pools, all different types of things. You just have to train it on what it's looking for, and then equip it with a crypto wallet, and then teach it how to actually, uh, essentially implement this, the mechanics of what you're trying to do, whether that's like move money around or swap tokens or, you know, trade NFTs, whatever it is. I'll show you how to do all that, you know, type of, you know, adapters tonight, but, uh, or those types of strategies. But yeah, that's what it is. Basically, anything you can think of that a human might do and make intelligent decisions on the blockchain, you could potentially automate all that type of stuff with an AI agent, um, with these types of techniques.

So, you know, that, that's basically what an AI agent is as opposed to, um, you know, just a ChatGPT chatbot that you're talking to, it's giving you answers back. This actually does something for you. What makes a crypto AI agent different from other AI agents is that you're giving the AI agent a blockchain wallet, and it can actually perform tasks on your behalf, right, then just giving you suggestions on what to do.

So, uh, yeah, that's a really brief kind of overview of, of what these agents are, what they're capable of. We're going to go ahead and kind of jump into some of the implementation of this stuff. I want to give ourselves enough time to do that, get it set up, and demonstrate this. So, let's go ahead and take a look at this, uh, code right here, okay?

So, this is, um, um, you're going to have the code for this down below, okay? So, wherever you're watching this video, there should be a link, uh, to a zip file that you can download to get the code for this, okay? You're not actually going to have access to this right here, okay? But, um, this is a private GitHub repository, but everything that you're seeing on my screen right here, you will have access to with that zip file down below, all right? So, go ahead and download that.

Before we get into the coding steps, I'm going to talk about another critical piece of information that you want to understand is how do you actually create an AI agent? All right, well, uh, we are going to use a framework to do this. So, what is a framework? Okay, well, a framework is really just a set of tools that you install on your computer that makes development so much easier. So, you know, if you're not a developer right now, um, just don't, don't worry about the analogy that I'm about to use, but, uh, really, it's just a set of tools that makes your life a lot easier. So, if you are a developer, you know, you've probably used a framework on something, maybe like for front-end development, like React.js or Next.js. Basically, it's just a way where you can create user interfaces, if in JavaScript, uh, without having to do a bunch of stuff manually, okay? It makes your life a lot easier. Or if you've been developing smart contracts as a blockchain developer, maybe you've used a framework like Hardhat, okay, to develop smart contracts where you can, you know, save the smart contracts in, in a certain directory structure, you can deploy them, you can write scripts, you can write tests, all that type of stuff, right?

So, we're also going to use a framework for this, uh, called Eliza, okay? So, Eliza, let me just pull this up right here, is a powerful AI agent framework for autonomy and personality, okay? So, essentially, this is put out by an organization called AI 16z. They got pretty popular in the crypto space, um, and they saw, you know, what's going on with AI agents. They have a platform for launching their own AI agents, and they created a framework for developers to create their own, all right?

So, what can you do with it? Well, essentially, um, you're able to create an AI agent that you can talk to, sort of like your own personal ChatGPT or something like that, and then you can equip it with a crypto wallet, and then you can customize it to do all these different strategies with different plugins, okay? So, let's just take a look right here. So, we're not going to go through this, look at the, look at this page. So, let me look right here, one second, everybody. Yeah, here you go. So, like, here's an example of some of the stuff you can do, uh, with your AI agent, all right? So, I mean, you can, you can plug your AI agent into all types of stuff, okay? You can plug it into the blockchain. You can use all different types of blockchains like Aptos, Avalanche, Binance Smart Chain, you can use Ethereum, you can use Solana, that's what we're going to focus on tonight, okay? You can, um, use the Telegram ecosystem, you can use all different types of stuff. And you can also plug into centralized cryptocurrency exchanges like Binance, okay? If you're trying to look for, you know, meme coins and stuff, you can look at Dex Screener, you can look at Bird Eye. So, you can actually train it on all these different types of information. It actually has, sorry, let me see here. It actually has an arbitrage plugin. So, if you're trying to look for arbitrage opportunities on the blockchain, you can use that, okay? Uh, you can do things like, uh, Zapper for portfolio tracking on your DeFi. You can plug into all different types of DeFi applications like, uh, decentralized exchanges to swap coins. You can swap on Pump.fun, you can do all, all different types of stuff, 0x Protocol. I mean, you can browse a list like this to just see the infinite number of things that you can essentially do on the blockchain by plugging it in. So, that's going to be on the plugin side. So, plugins are what allow you to actually perform actions, and you have to configure the plugins inside the Eliza framework, um, in order to be able to do those things. But you can also, um, plug it into lots of other clients as well. So, you could essentially plug it into a Discord. Basically, you can create your own Discord bot where you can chat with your AI inside of Discord. If you have a Telegram account, uh, you can chat with your AI through Telegram as an interface and have it do things for you on-chain. It can, you know, tell you about coins that are coming out. If you're looking for arbitrage opportunities, you can get it to tell you that type of stuff too, okay? And you can also, um, configure it with a Twitter account or an X account. So, if you go back and look at Billy Bets, like I was talking about before, some of these AI agents essentially have social media presences. So, like, if you look at Billy Bets, of course, he's making, he's actually doing things on the blockchain with his crypto wallet, but you can see he also has a Twitter account or X. So, you know, the cool thing about these AI agents is they actually have personalities. So, you know, ChatGPT, like I was showing a second ago, sort of has a default personality associated with it, but you can actually train your AI agent to have a certain personality that'll talk to you with a certain tone of voice. And you can also get that AI agent to, like, you know, use that same personality and tone of voice and make social media posts, okay? So, that's like what Billy's doing here, essentially, is you're able to see, you know, Billy is sort of tweeting all the time about the bets that he's making. He kind of shows his wins and losses. And also, you know, you can talk to Billy yourself, all right, and try to influence Billy what, you know, he should or shouldn't bet on. And of course, Billy's not necessarily going to take your suggestions, but it actually can take those into account and, uh, use that part of it as a part of its data to make its decisions for its sports bets, okay?

So, um, let me go back to the Eliza page here. So, those are some of the plugins. So, that's what talking about plugins. Plugins are a huge part of the framework. Basically, these are things you can plug in to do all different types of things, whether it's swapping tokens, whether it's communicating on Discord or Telegram or Twitter, or, you know, doing different things on-chain, different blockchains. That's what plugins are for, okay? That's what's going to make your life so much easier. All right, so we're going to use Eliza. The code that you have down below, uh, we essentially have created a fork of Eliza and we've updated it with lots of different stuff to go ahead and get it ready so that you can, uh, start creating your own agents that have the capability to do things like send cryptocurrency around, uh, chat with it, all right, be able to swap tokens on-chain. So, all the things you might want to do to create some other type of bot or automate certain tasks, you've got all that stuff ready to go in this kit, um, that we have set up for you tonight, okay?

So, let's go ahead and get started, um, getting this set up. So, let's cover the basic prerequisites that you're going to need in order to get started, okay? So, if you are, um, if you're on Windows or Mac and you're using WSL on Windows, you'll probably have this first, um, first dependency installed. You have to have Python installed on your computer, okay? We're not actually going to write any Python code, but it needs Python installed in order to, uh, just run the actual process on your computer, okay? So, you can go here to python.org/downloads. If you don't have Python installed already on your computer, you can simply download it from here. If you're a developer and you're used to switching around different Python versions, you can use pyenv or something like that to manage your Python versions. But if you're just brand new and you need to get Python set up, you can just download it directly from this website, all right?

So, um, the next thing that you're going to need to install on your computer is also Node.js, all right? So, what is Node.js? Well, um, the majority of this Eliza framework and this AI agent framework is written in JavaScript, okay? So, um, JavaScript is a language that natively runs in your web browser, okay? It doesn't natively run on your computer. And so, in order to run, um, JavaScript on your computer outside your web browser, you need a special, uh, environment to do that in, and that's what Node.js is, okay? Um, we also use Node.js with npm, Node Package Manager, to install all the dependencies, uh, that we need for the project. So, that's going to get installed. All right, so you need to download Node.js if you haven't already installed. If you're new, you can download through this link, okay? Um, and if you're a developer already, see here, you're going to want to make sure you have a specific version of Node installed. Uh, I'm using, uh, Node Version Manager, okay, to manage my Node versions. And specifically for this project, uh, I'm using Node version 23.3.0, okay? Now, I highly recommend that you use that exact same version because we've had trouble with other Node.js versions, and 23.3.0 seems to be the one that, uh, is going to work the best for this. So, it says Node version 23+, but specifically, we're having results with 23.3.0, okay? So, if you are a developer and you haven't used a Node version manager before, you can just install nvm. It's a really great Node version manager. That's what I personally use, and you can swap versions pretty easily, just like that. All right?

So, the final, um, dependency that you're going to need is pnpm, all right? So, pnpm is just a package manager, uh, kind of like Node Package Manager or Yarn, if you use something like that before, uh, that's just what's there to, um, that's just what they use on the Eliza framework. So, unfortunately, we have to use that as well, okay? So, you can, uh, install pnpm if you haven't already, uh, with these instructions right here. You can, let me find it. If you have, um, npm installed or Node installed, you can do this command: `npm install -g corepack`, and then you can do `corepack enable pnpm`, and that will get pnpm working for you. And then you can use the latest version here if you want to, and that will let you install all your other dependencies, okay? So, those are some of the basic things, uh, to know.

All right, right, so what we're going to do is, you know, you do need a couple of basic things. If you're not a developer and you want to get into the code, you need some basic stuff like a terminal, all right, to run terminal commands. I know the terminal can be kind of scary if you're brand new, but as long as you follow these exact instructions that I'm telling you tonight, you shouldn't have any issues, all right? And you're also going to need a text editor of some kind, all right? I'm using VS Code. You can use whatever text editor you like, all right?

Sorry, let me just bump my font up here. All right, bump my font up here so you can see it better. All right, so go ahead and, um, open up your code in your text editor and go ahead and have your terminal handy here. Uh, if you want to use your terminal inside of your text editor, you can. All right, I've got VS Code with my terminal open here, all right? But the first thing that you want to do is go ahead and open up, uh, the code in the directory that you want to work in. So, I keep all of my coding projects inside of `code`. So, if you unzip that project folder down below, you're going to want to put it into whatever directory that you want to, and then you want to enter into that directory like this: `cd Eliza`. All right, now I'm in, uh, this folder right here, all right? And so, you're going to want to have that terminal open so that we can do more commands inside of here. Uh, I'll probably just do more of my commands inside of here, just so I don't have to keep swapping tabs. But the, we need to connect to a couple different services for our AI agent in order to get it to work.

So, what we're going to do as a first step is, um, we are going to open our, uh, `.env.example` file. All right, so what is this? Well, this is a starter template to create a new file on our system. All right, so right next to this `.env.example` file, what I want you to do is create a new file right next to it and call it `.env`, all right, and hit enter. All right, I've already done this, so I'm not going to do it. Um, I don't want to overwrite my other `.env` file. What you're going to do is create a new file called `.env`. That's the actual file name, okay? Sometimes people get confused about that. You know, it's like, uh, it's a dot file, so you might see something like, you might get confused because the file name starts with a period, but, you know, you might see an image like, you know, `my_image.jpg` (jpeg), um, that's got a file extension of the dot after it, right? But this actually starts with a dot. It's just `.env`. So, what you're going to do is create that file called `.env`. You're going to copy everything from the `.env.example` file and paste it into the `.env` file, all right?

So, what is this file and why do we need it? Well, um, basically, we need to store some sensitive information in our project, and that's what the `.env` is for. So, we need to do things like, uh, store a cryptocurrency private key from our wallet, okay? And we also need to store things like API keys for certain external services, okay? And we need a place to store those in our project that will never leave our computer and that no one else is going to have access to, okay? So, um, that's what the `.env` is for. So, I'm going to demonstrate everything instead of the `.env.example` file because I don't want to show my `.env` file on the screen, all right? But, um, you're going to want to do everything that I'm telling you, uh, in the `.env.example` file in the `.env` file, all right?

So, what are we going to do in here? Well, we're going to keep this open handy, and we're going to go walk through some setup steps. So, what we're going to get our AI agent capable of doing is a couple things. We need to be able to talk to an actual AI, okay? So, what you have to understand here is our agent is going to be talking to another AI like ChatGPT, for example, to sort of run the base language model technology, okay? And, um, we need to be able to talk to that. All right, we also need to connect our AI, we need to connect our, uh, AI agent to the blockchain so they can actually talk to the blockchain. So, we'll get access to that, all right, and we'll store that configuration in our project. Uh, we need to get our AI agent capable to talk to something like Telegram, so we'll get that, all right, and we'll put that back in our project. And then finally, we need to equip our AI agent with a crypto wallet. So, we'll, we'll get the wallet set up, all right, and we will, uh, essentially take that and then put it into, uh, the project as well. So, we're going to put all that information inside of this, uh, `.env` file. You can think about this like a configuration file, okay? But it's for secret information. You don't want to ever show somebody else this file. That's why I'm not going to show you my `.env` on screen, all right?

So, the first thing that we're going to do is we're going to connect our, um, AI agent to ChatGPT. We're going to use that as the base language model because you need a language model in order to run this. So, the agent's going to run on your computer, but the actual AI technology is going to run somewhere else, okay? That's the important part because language models are very heavy. It'd be very hard for you to have a sophisticated enough computer to host this type of thing. So, we're just going to talk to another AI, but we'll be able to train that AI, uh, with our agent. So, that's okay. So, we're going to use ChatGPT for this, all right?

So, um, basically, what you can do, all right, pull ChatGPT here, is, um, I showed you ChatGPT a minute ago, but you're going to want to go visit a certain URL in your browser. Let me pull it up here. You're going to want to go to platform.openai.com. Okay, so this is OpenAI, the creators of ChatGPT, and this is sort of their developer platform that's going to allow you to talk to ChatGPT programmatically, okay? So, uh, basically, what that means is you'll be able to talk to the actual technology behind ChatGPT without having to use the, the interface where you're chatting to it, okay? So, you do this over API calls, and you do that over essentially, uh, with an, you need an API key in order to do that.

So, um, what you have to do is you're going to sign up. All right, I'm not going to walk through the sign-up part, that's pretty straightforward. You just have to give it an email and password, verify your account, and log in, okay? But I will show you the screenshots of sort of the step-by-step, uh, screens that you'll see, uh, once you log in and show you how to actually create a new app and get the API key to put back into your project, okay?

Okay, so let me just pull that up right now. So, what you're going to do is, once you have, um, created your account, you've logged into the platform dashboard, you're going to see this dashboard that looks like this, okay? You're going to click on the billing tab, all right? You're going to see an overview. Now, this is going to cost a tiny bit of money. I'm talking $5. Okay, that there will be a small cost issue with this. You need $5 for API credits. It's not automatically billing, don't worry, it's not going to like recur on your account. It's $5. Should last you a long time to test this type of stuff out. Um, you're going to need maybe $5 just to test out the crypto side of things with your MetaMask wallet or Phantom wallet, okay? Um, so there's not a lot of, there it's a small cost, unfortunately, you can't do this completely for free, but it's, it's, so it's going to cost you less than a lunch, okay? So, click on billing, um, you click on payment methods, all right? What you're going to do is you're going to add a payment method right here, all right? Uh, whenever you've added the payment method and saved it, you'll, of course, you'll see it right there, all right? It'll be set as your default payment method, all right? You're going to click back on this overview tab, okay, once you set up the payment method, and you're going to click on the "Add credit balance," all right? That's going to bring up a popup that looks like this. I actually went through these setup steps earlier today, took all these screenshots. I'm doing these screenshots because I don't want to, you know, expose my admin area with sensitive credit card information and stuff like that, right? That's why I'm doing the screenshots. Um, I'm adding $5. That's all I did. That's going to be plenty of credits to test this out for a long time. Uh, click continue, all right? Whenever that's done, you should see your $5 balance, uh, pay as you go, right here, all right? Now, that's going to give you the funding that you need in order to actually talk to this thing programmatically, all right? And just for the record, you don't need a ChatGPT Pro account or advanced account or anything like this. You don't have to pay for ChatGPT extra. All you have to do is pay developer credits to, um, talk to the API. So, in order to talk to the API, you need what's called an API key. So, it's kind of like a, uh, proprietary, uh, sorry, sorry, it's, it's a, it's a way to talk to it. It's like a password, okay? So, what you're going to do is click on API Keys down here, all right? You're going to click on, uh, this "Create new secret key" right here, this button, all right? Whenever that pops up, you're going to fill in some basic details like the name. I just called mine, you know, "my AI agent." You can call it whatever you want to, all right? You want to make sure it's owned by you, that it's in your default project, and it has all permissions, okay? And you're going to finally click this button, "Create secret key," all right? Whenever you have, uh, whenever you've done that, you're going to see your secret key right here, okay? You're going to copy it to your clipboard. You're going to go back to your code in your actual `.env` file, not the `.env.example`, and you're going to open `OPENAI_API_KEY` and you're going to put it in right here. So, like, `abc123456 blah blah blah`, big long key, looks like this, right? You're going to paste it right there, okay? So, I'm not going to actually leave that there, but that's where you put the API key. You copy it from that screen, you just paste it in plain text right there, okay? And then you'll be set up to talk to the AI, okay?

So, the next thing that we need to do is get the AI to talk to the blockchain, okay? So, how do you talk to a blockchain in the first place? Well, you know, you, you talk to a blockchain by talking to a node, okay? So, a blockchain is a peer-to-peer network of nodes that all talk to one another. It's a distributed system, and in order to talk to a blockchain, you just have to talk to one of those nodes. That's just a computer, kind of like a server on the blockchain that talks to lots of other stuff. Now, just like language models, it's really hard to run a blockchain node on your own computer. It's very expensive, it's hard to do. So, instead of, um, instead of, uh, doing that ourselves, we're just going to use somebody else's blockchain node. So, uh, what you're going to do is head on over to quicknode.com. So, QuickNode is a developer platform for essentially spinning up blockchain nodes and talking to other blockchain nodes. You can read information from the blockchain, you can write transactions. So, you know, when we get our, when we get our AI agent to do things like send cryptocurrency around or trade tokens, um, we're going to use a node provided by QuickNode to do that, all right?

So, specifically, what we're going to use tonight is the Solana blockchain. So, uh, QuickNode has, uh, Solana nodes available for us, and we can just get one for free. It's not going to cost you anything, and we can, um, we can do that. So, um, you're going to want to create an account, okay? I'm not going to walk through the account creation steps, that's pretty straightforward. You just provide your information, it's totally free, you need an email, password, okay? And again, I'm not going to reveal all my account information, um, so, but I am going to show you the sort of screenshots that I walked through earlier today, just to make sure this worked properly. So, whenever you, uh, look at your account, you're going to see a dashboard that looks like this, all right? At your homepage, you're going to see this button over here on the left called "Endpoints," all right? You click on that, and endpoints just going to be a URL to a specific blockchain node. You're going to click "Create endpoint," all right? You're going to choose Solana, all right, and then click this "Continue" button. You're going to do "Mainnet Beta," all right, and then click "Continue." And then finally, you can skip all this stuff, you don't need to add these extra bonuses and stuff. Uh, you're just going to click "Create endpoint," and then, uh, when you're done, you'll see this HTTP provider link, all right? It's got some sensitive information after this part, that's why it's not blurred out on my screen. Uh, this part's safe to show you because that's pretty much what everybody uses, but it does have some sensitive information after that. You're going to copy that, okay? And you're going to go back to your project and, uh, you're going to put it down here in the `SOLANA_PRIVATE_KEY` portion. Okay, so if you go here, go back to your `.env` file, not the example file, but the actual `.env`, and then, uh, Solana, let's see, which one is it? No, sorry, I said `SOLANA_PRIVATE_KEY`, sorry everybody, `SOLANA_RPC_URL`. I had a, I misspoke there. Yeah, you can see `SOLANA_RPC_URL`, it's already populated. So, you want to back that out, and then whatever this, you know, whatever you saw here, you're going to copy that, you know, `blissful-bitter.solana`, right? Like `https://blissful-bitter.fire.blah blah blah`, right? So, whatever, whatever you copied there, you're going to paste it in plain text right here, okay? So, I'm going to backspace that out just so I don't screw up my example file. Okay, so that's the next step, okay?

So, now we've gotten the required information to talk to a language model, so our AI agent can work. We have the connection to the blockchain saved in our, uh, agent, okay? Now, we need a couple of other things. One is we're going to, uh, create the ability for our, uh, bot to talk to Telegram, okay? So, you don't really have to use Telegram, okay? But it's probably a good idea to go ahead and create a Telegram app, uh, because we've already added the Telegram functionality. We may not even have time to demo it tonight, that's okay, but let's just go ahead and set it up now. You'll have it in your app for later, and it'll ensure that your app's running properly because we already have a Telegram section in it.

So, what you're going to do is, uh, uh, open up something like Telegram. I'm not going to open Telegram right now, but what you're going to do is you're going to find, um, you're going to search for the contacts in Telegram, and there's one called BotFather. It's just called B-O-T-F-A-T-H-E-R. You can see his icon right here. All it does is it's a chat room where you can create new Telegram bots, okay? So, inside that chat, you're going to open it, you're going to type `/newbot`, okay? That's a command, and it'll say, "All right, a new bot, blah blah blah, blah blah. What's the name for the bot?" You give it a name, all right? You gotta find a unique name, so it might not look like something really meaningful, but you give it a name. So, it's got, uh, the name. Now you have to give it a username, okay? So, a username has to actually end in the word "bot." So, if it's like `my_ai_agent123456`, you have to do `_bot`. So, `my_ai_agent123456_bot`, okay? And whenever it's done, it'll say, "Okay, congratulations, here's your bot. You can find the link here. You can, you know, do all that type of stuff." And then finally, it's going to give you an HTTP token for it as well. So, that's how you can save the Telegram portion. Uh, you can also go back if you want to, you don't have to do this part, and save that in your `TELEGRAM_BOT_TOKEN`. So, if you go back to VS Code, `TELEGRAM_BOT_TOKEN`, you can put that here, right? So, whatever your token is, you can just copy that from the Telegram chat and put it in here.

So, real quick too, says, "Can we set up multiple RPC URLs if we want the agent to be multi-chain?" Yeah, you sure can. You can do, I was showing you before, you can do Ethereum, you can do Solana, you can do Aptos, you can do all this type of stuff. Like, if you search `RPC`, all right, here's all the different RPCs that's currently working for Eternal AI, Solana, Abstract, StarkNet, Conflux, Zog, Near, The Veil, Telegram Network. If you just search for `RPC` inside of your, uh, `.env` file, you'll see all the different, you know, chains that you can.

Can connect it to. And if you go back again to the Eliza framework homepage, and you look at the plugins, um, you can see like all the different uh chains like EVM, Solana, etc., etc., okay? You can see all the different chains you can connect to here, okay?

So, um, we have, um, gotten this thing to talk to an AI, talk to a blockchain, talk to Telegram, if you want to do that. Now, finally, what we're going to want to do is we also want to equip it with a crypto wallet, all right? So, uh, we're going to use Solana. So if you don't have a Solana wallet yet, uh, what you're gonna want to do is use something like Phantom, okay? So Phantom's pretty easy. Uh, let me just pull this up here. Let me create a new tab for this. All right, just go to phantom.in, or if you want to download a desktop app, that's fine too. I'm not going to do the wallet setup because you got to go through the process of creating a seed phrase, all that type of stuff.

Um, so let me, let me give a couple quick words of caution here, all right? So we're going to actually equip our AI agent with a crypto wallet. So if you already have a crypto wallet, do not use a crypto wallet that you've already used for something else, okay? What I mean by that is, don't use the private key from a crypto wallet that already has real money in it, all right? Just don't do it, all right? Don't do it. Why? Because what we're going to do is we're going to take the private key from a specific crypto wallet, all right, and we're going to put it into our AI agent. We're just going to copy and paste it in there. So if anybody found that file on your computer and you had some real money in it, you know, they could steal it, all right? So I mean, if you got five bucks in there, you may not care, but if you're holding any type of valuable cryptocurrency, do not use that wallet for this thing, all right? So set a different way, if you start using this thing and this thing actually has some substantial money in it, once start using this, uh, wallet, don't leave all the money in there, okay? Because anybody who has access to this wallet could clean you out, okay? So don't do that.

So, um, if you already have a Phantom wallet, you can just create a new account inside there. Sometimes the terminology gets confusing. People say wallet and they think like that's your address. Well, I'm talking about a wallet as in like an app that can handle multiple different addresses inside of it, okay? So, um, what I would recommend if you're setting up Phantom for the first time is just create a new account inside of there. Um, if you already have a Phantom wallet, you can just click the new account button, you'll see that at the top, okay? And then whenever you do that, you're gonna walk through these next steps, okay? Um, you're going to, I've created, I created a fresh AI agent, you know, wallet, or sorry, account inside of my wallet. So if I clicked on this, I'd see a bunch of other stuff, but you can just click on that. All right, you're going to want to go through these steps. Um, so you can see if you push on the side, you can click this plus button to create a new wallet, that'll give you that. All right, and then you're gonna click on settings. All right, you're gonna click on this account right here that you're going to want to use. You're going to want to click on show private key. All right, it might ask you for your password. You're gonna go put your password in and then it's going to reveal the private key on screen, okay? So you're gonna copy that private key. You're gonna go back to your, uh, .env file and you're going to fill in this Solana private key variable here. So like, here's your private key, you got to put that information in right here. There, copy and paste from the wallet, put it back here. All right.

Now, the next thing that you want to do once you've got the private key is you are going to go back to your wallet, the main page, and where it has your wallet name up here, you're click on this little button. It's going to show you like the Solana address, okay? You can copy that address. This is the actual address that you send funds to, and you're going to, uh, put that here in the public key. All right? So just copy and paste that from your wallet and put it here. So you to fill in your private key, you need to fill in your public key. All right? And then finally, if you want to actually use this to talk to the real blockchain and make transactions, like send cryptocurrency around, swap tokens, all right, then, um, you have to fund it with some money, okay? So I funded mine with like 10 bucks. You could do it with five bucks, you could do it $1 if you wanted to. I recommend probably five, so if you want to take a couple different actions, you can do it. But I mean, you can, you can do really minimal stuff just for testing this out. You can test this whole thing out for 10 bucks, five bucks worth of API credits, five bucks worth of, uh, you know, cryptocurrency, uh, to test the agent out, okay? But once you've done that, we're basically set up. We're basically done with the initial setup of talking to an AI model, talking to the blockchain, talking to Telegram, if we want to. I don't think we'll have time to demo that, but you can talk to Telegram if you want to, and then finally equipping it with a crypto wallet so you can actually make transactions on chain, right? So that's the important, uh, heavy lifting aspect out of the way. All right?

So, um, now that we have all these variables filled out, so optionally, you know, later, if you want to, you can also add a Birdeye API key, okay? So Birdeye is like a trading platform. So if you want to look for like, if you want to get your AI to like learn about the crypto markets, uh, it can actually pull data in from Birdeye, all right, and actually actually learn from that information. So you can, as a, as a homework, you can go through the Birdeye, uh, steps to essentially, you know, get a developer key, and then based upon everything that I just showed you from these other steps, you can put that here in your Birdeye API key, and you can train your agent on what's going on with the crypto markets, all right?

Okay, so, um, let's look at a couple things here. Let's look at some, we're about to boot our agent up, but let's look at some diagrams about how it's going to work, all right? So, you know, we are building this agent here in the middle, uh, with this Eliza framework, all right? So that's really what this is. This is the framework that's, we're going to run our AI agent with a process on our computer. It's talking to the blockchain, it's talking to OpenAI, it's talking to our crypto wallet. It could talk Telegram if you wanted to, Dr. Birdeye if you wanted to, lots of other stuff. And here's sort of our agent in the middle, all right? So the plugins are what I talked about before, like Telegram, like Solana, like Ethereum. And also our AI has a character. All right? It has a, a background and a personality. And you can actually define the background and the personality if you want to. I'll show you how to do that in a minute. We've already pre-programmed it with a certain personality. I'll show you how you could update that as well, all right?

Some of the other, uh, things I understand about how a plugin works, we'll kind of go over this when we look at the code here more in a minute, but, uh, essentially plugins have providers, um, where you can do things like read information inside your AI, like, uh, a data provider, basically. So if you're talking to blockchain with the, with the Solana RPC from QuickNode, you could do things like get token information. You can get the symbols, the decimal places, you can fetch prices from exchanges or get liquidity information from exchanges, okay? That's what a data provider is inside of a plugin. It also has specific actions, okay? So that's when you're actually getting it to do things for you, okay? So provider just gives you information to act on. Actions allow you to actually perform actions. That's what makes your agent valuable. It can do stuff for you, like transfer tokens, like swap tokens, okay? And then finally, it has another aspect to it, like an evaluator, which is it can like track transaction history, it can learn about tokens, it can learn favorite tokens, it can, you could teach it how to look for trending tokens, stuff like that, okay?

So here's sort of the high-level flow that we're going to see with our agent, all right? So basically, um, like step one is, you know, we're gonna be able to talk to our agent with our own sort of personalized ChatGPT, which you're going to see here in a second. But here's the flow: is we'll talk to the agent. You could do it through Telegram, you could do it through Discord, you could do it from Twitter, but we're going to do it from a specific web interface here in a minute. All right, the agent's going to sort of think because it's actually a thinking thing. It's going to fetch information from the providers like I just talked about here, all right? It's going to receive the information from the providers to sort of formulate what it's going to decide to do. Um, if it needs to do an action, it will do an action that's in this actions part of the plugin, all right? And then if they need to evaluate anything, it will evaluate it from this evaluator part. And then finally, it will reply to you with the results. So if it took an action, it will tell you, all right? If it's, you know, if you're just asking it to give you some type of answer or prompt you to take an action, it can do that type of thing as well. So for example, if you wanted to look for certain tokens and then give you some suggestions to trade, you could tell it what to do based on the suggestions rather than it just automatically doing it for you, or you could get it to do it automatically for you. Um, so here's an example of an action: like you send it a message, it thinks, all right, does it send tokens? Does it swap? Does it provide liquidity? Does it buy? And then it analyzes the keyword from what you're doing. That's what natural language processing is about. And then it takes the appropriate action based on the instructions that you give it, you know, give you a balance, swap, transfers, etc., okay? So for example, if you tell your agent, hey, I want to swap 0.1 SOL for USDC, all right, it's going to understand your request. It's going to think about it. It says, oh, you're trying to swap because it learns the exact keywords. And then it's going to determine the appropriate actions to take on the blockchain, uh, with code that you sort of, uh, prepared it to be able to execute. Can it do an action and evaluate both? Yes, it can. Can you link multiple agents to work together? Yes, you can. All right?

So, uh, that's kind of what we're gonna do. So now let's get everything set up, all right? So you're gonna want to go to your terminal, all right? And, uh, I'm going to do VS Code here. You can use your external terminal if you want to. You do `pnpm install`, okay? I've already done this step, so it, you're going to see it says already up to date, okay? That took like two seconds, but for you, it's probably going to take a little bit. All right, it'll take maybe up to a minute for all your dependencies to install. You'll see a bunch of stuff spit out on your terminal. Basically, `pnpm install` is just installing all the packages that are included inside of this, um, `package.json` file, which you can see right here. This is what this is where all the external dependencies are installed, all right? So once that's done, uh, you're gonna run this `pnpm build` command, all right? Mine's probably gonna run kind of fast, but yours might take a second. All right, yeah, mine ran pretty fast, but yours will might will take several seconds, maybe even up to a minute, depending on your computer speed, okay?

Now, the next thing that you're going to want to do is, uh, you're going to run this `pnpm start`, okay? So I'm list my terminal here. You're gonna see it kind of like going through a bunch of stuff, booting up. It's gonna say I'm doing this, I'm doing this, and doing this, okay? And whenever that's going, it's going to sort of just run in your terminal there, okay? And what you're going to want to do is create a separate terminal tab here, and you're gonna run the following command: `pnpm start`. You're going to run, um, `pnpm start clients`, all right? And that's going to start a new, um, a new web server for you here, all right? You can paste into your browser, all right? And you'll see, uh, this popup for you. Now, I'm going to demo this actually on a remote box here. I had a little issue with my system earlier today, so I'm gonna have to demo this remotely, unfortunately. But we verified this is working on other systems, so shouldn't be an issue for you. But, uh, whenever you get that set up, what you're going to see is basically this right here, okay? So on localhost port, uh, 5173, you're gonna get your own personalized agent interface. So this is pretty awesome because this is going to be like your sort of personalized ChatGPT, all right? It's like your own private, uh, ChatGPT running locally for you. Now, again, you can interface this agent different ways. We set up Telegram, you could talk to it directly inside Telegram, uh, you could hook it up to a Discord, it talks from Discord, you know, you do all different types of adapters and plugins, which is pretty cool. But we're just going to make it easy and run it from our own personalized sort of ChatGPT-like interface, uh, on our computer here, okay?

So what you can do is you can already see, uh, you've got your agents listed here. So you can create multiple agents, but we just created one here. You click on that, go on chat, and now you're in a chat interface where you can talk to it, okay? You'll also notice, um, I got the Phantom wallet here, okay? So I got my Phantom wallet that I funded with some cryptocurrency. We're actually going to monitor the Phantom wallet while we're talking to the agent because it's going to do stuff for you. Yeah, Marco's saying Telegram bot tokens required, giving error. Yeah, that's what I was thinking could happen. So yeah, just make sure you walk through the Telegram setup steps like I was talking about a minute ago. It's pretty straightforward, um, and just make sure you throw in the Telegram bot ID and it should get it to work.

So before we start demoing this, I want to point you towards an important part of the code about the agent itself, okay? So I'm going to kill the terminal here, okay? And we're going to go into our main folder here. We're gonna go to agent, all right? We're going to go to source, and we're going to go default, okay? So we've already set up an agent, all right? And this is going to be the, uh, Eliza agent, okay? That's what you're seeing here whenever I boot this up on my screen, all right? We clicked on Eliza, now we're talking to Eliza. I can just chat and say, "Hey, hello world." All right? And you say, "Ah, hello world, the classic greeting of programmers everywhere. All right, whether you're in coding for the first time or just testing the waters, let's always have a good start. What's on your mind today? Need some guidance on your next blockchain venture, or just looking to chat about tech and beyond?" So you can see, you know, it already responded to us with a personality, okay? And the personality is different than just like the super dry, calculated responses from like ChatGPT, all right? That's because if you go into, uh, this configuration here, we've already sort of set up the character to have some type of, uh, personality, okay? So basically, we have a system configuration, roleplay and generate interesting dialogue on behalf of Eliza, okay? And then I'll just close this down. So you can actually create a bio and you can customize all this stuff yourself. You want to rename your agent, if you want to name it Billy Bob, you can change that here. You know, Billy Bob, okay, username, you go Billy Bob, whatever, right? Um, I've got the plugins installed here for Telegram and Solana, which are importing here, okay? You can see that we're using the OpenAI model provider. Again, you can use this with different, uh, different AI. You can use with Claude, you could use it with Grok, you could use with all this other stuff. Uh, we're just using OpenAI because ChatGPT is pretty popular, pretty much everybody knows what that is. Uh, but you could change that out to different language models. I also, people who are using DeepSeek in the chat, you could probably get this to DeepSeek, even though I haven't tried that. Um, so yeah, um, but yeah, going back to the bio, you know, basically right here, you know, someone who can help with performing different transactions and tasks, a seasoned Web3 developer with deep expertise in Ethereum and Solana, okay? You know, we're here, uh, you know, dApp development is all about blockchain development, so we're creating our AI, you know, with blockchain development knowledge, a security-minded, always emphasizing the best practices in smart contracts, a natural educator who believes in Web3, firm believer in decentralization but pragmatic about challenges and trade-offs, encourages people to experiment but always with a safety net, code like an optimist, test like a pessimist, and, you know, has an knack for making Solidity, Rust, and smart contracts feel less intimidating, okay?

Now, again, you can change this to whatever you want to, but you have to follow the structure. It's this array of strings. You need to make sure you've got these things in strings, you've got commas at the end, okay? Um, and also if you want to, um, if you want to get some ideas on what you could change in here, you could just go to ChatGPT and say, "Hey, I'm trying to write a bio for my agent. I want to give it a certain personality. Can you help me essentially write the prompt for that?" Okay? You can even copy and paste this format of what's inside of here. Be like, you know, go here. And you could even go back to ChatGPT. Let me see if I can do it. Um, sorry, let me just, I'll just show you. Me create, create a, a new AI personality bio for a sports better like this. Okay? So here's an example. I just said, "Help me create a new AI personality for a sports better like this," and I just pasted in the bio from our configuration. And it's like, "Okay, here's one for a sports better, right?" And so you, you could do that. Like if you're like, you want to really deeply describe what your personality should be like, then you could do that and it'll give you sort of a good answer. And you could, um, you could do that.

So, does the LLM you use affect how smart your agent is or becomes? Well, it can for sure. I mean, um, you know, it's like the, it's sort of like, you know, everybody's going to argue about whose AI is better, whose models better, right? But there are some measurable, uh, benchmarks between the different models out there. The course is being updated and changed all the time, okay? But yeah, um, there could be some difference. I don't think it's gonna make a massive difference at this point in time, but yeah, there could be some differences on which LLM you use and how smart the agent could become, okay? So there's that.

Um, let's look at the lore again. Lore's like the background. You started in Web2 development, for getting hooked, Web3 has contributed to major Ethereum, Solana projects, host online workshops, etc., etc., all right? You can see some major examples of like, how did you get into Web3 development? Okay? And then it tells you like how it would answer that question. Ethereum or Solana? You know, it depends on the use case. You're teaching it like how it would respond to certain types of questions. And he can learn from those responses about how it should actually respond to the end user. I'm not going to go through all these, but you can read through that and see kind of an idea because there's a lot there, and we've done a lot of heavy lifting to get it to actually train like that. Uh, here's some other post examples of like, if you're going to post on social media, you know, what it could do. And then topics that it's interested in, like Web2, Web3, Solana, Ethereum, cybersecurity, DeFi projects, smart contracts, technology, meme coins, blockchain architecture. Uh, style: keep responses concise and sharp, blend tech knowledge with street smarts, use clever wordplay and cultural references, be confidently quirky, stuff like that, respond with quick wit, use playful banter. All right? And then adjectives like brilliant, enigmatic, technical, witty, sharp, cunning, elegant, insightful, chaotic, sophisticated. This is where it gets really fun. You can play with this stuff.

Now, I will say, if you change anything inside of here, um, you need to rerun the `pnpm build` command, all right? So if you do that, you need to do `pnpm` uh, `build`. All right? You to stop, stop your process, do `pnpm build`, and then `pnpm start`, okay? You got to restart your app, basically, if you make code changes. So if you're going in here and you're, um, changing your models' personality or your agent's personality, the name, all the type of stuff, you got to rebuild your app, okay? All right, so, um, that's sort of the gist of, of how you would sort of tweak your character here.

Now, let's kind of chat to the character, all right? So, kind of did a special thing here. Uh, let's go ahead and kind of interact with the blockchain. I'm gonna get my Solana wallet here. I'm gonna go ahead and, uh, copy my Solana address. Go and copy that here. All right, so please check my Solana wallet balance for my wallet here. All right, send message. This is gonna be a quick smoke test. Um, oh, that's not the right one. That's the wrong, that's the wrong address. Hold on a second here. I wish I had that much SOL in there. Sorry, let me copy it again. All right, let's see here. Me paste it in there. There we go. Huh? I'm not sure why I was doing that. Hold on a second. Let me make sure I got the right, because I, I was on Solscan a minute ago and I, me just copied this. Yeah, let me paste it back in there. Make sure it's the right one. Yeah, okay, that's the right one. Sorry, I, I copied the wrong address. I was messing around this earlier. I was on Solscan and I think I copied the wrong one. Okay, so that, that's the actual one from my Phantom wallet. Yeah, GLYR. Okay. Yeah, there we go. One second here. Oh, I'm on the wrong one still. A second. My thing is messed up here. I'm not sure why it's telling me that. Hold on. Oh, my bad. I forgot it's doing it in lamports. That's right. Ah, that's right. I forgot. Okay, we're good. No, it is actually working. I, I was testing on a different Solana wallet out earlier. Copy from Solscan. Yeah, I forgot. So this actually, you got to convert it back to, uh, a decimal places. So it's 0.09 because it's doing it like in the, uh, at. So basically, at the blockchain level, um, you've got like, if you're coming from Ethereum or some other blockchain ecosystem, like at the blockchain level, um, you, you don't use decimal places. You do everything in lamport written strings, okay? So basically, like, um, you know, SOL tokens have nine zeros after the decimal place. So that number is divisible by nine. So really is like, it's this much, 0.953 blah, blah, blah, blah, blah, blah, blah. So, and that's exactly what I've got right here. And this is telling us in US dollars. All right, yeah, 0.09537, 0.953, that rounds up to seven, so it's good. Okay, awesome. Cool.

Well, let's, uh, so that's our smoke test. It's actually talking to blockchain. It's talking to our Solana wallet. It can fetch the balance of our wallet. I know it got a little, I got myself a little confused there because it was showing it without the decimals, um, and I was looking at a bigger Solana wallet before, and that's why I thought it was the same one. So what we're going to do is we're going to, um, essentially just send cryptocurrency around, okay? We're going to get our agent to actually make a transaction for us. Um, so let's go here. Let's click on, just, I'm just gonna get a random Solana wallet. Let's look at this signer here. Let's go ahead and copy this to the clipboard, okay? And, uh, ZER CFY. All right, and let's say, "Please send 0.01 SOL to," let's do that address. Paste it there. All right. All right, so let's go and copy this transaction hash. And then let's go over here to, um, Solscan. We'll paste it in here, click transaction hash. And there we go. It sent 0.001 SOL on our behalf with our agent to a different crypto wallet. So that's our smoke test. It's able to, uh, read our balance from our wallet. So it's actually accurately pulling information from the blockchain. And now it's actually sent, uh, cryptocurrency to a different address, okay? So that's, that's the basics. All right, so we know it can use our wallet. It can actually perform tasks for us, okay?

So now what I'm gonna do is, let's go back to our wallet. All right, this is our, uh, signer, okay? And, um, we've actually done that, okay? So now let's get it to do something fun, okay? So let's say that you wanted to get your API, sorry, your AI agent to like, you know, get into coin trading, maybe wants to do meme coins or new coin launches, and you're trying to get that functionality up and running. Well, the good news is, we already have the adapters in place to start actually swapping tokens. We can do it through the Jupiter aggregator so that we can swap things on things like Raydium. So if you want to create a trading bot, you've already got the infrastructure for doing that. If you want to trade tokens on pump.fun, we can do that as well, okay?

So I'm gonna do, I'm gonna pull this off screen. I'm gonna pull pump.fun off screen because sometimes pump.fun doesn't always have the best stuff that I can show on screen, if you know what I mean. Um, so I'm just gonna pull up a random token on pump.fun. So I have no affiliation with any of these coins. I don't know what the heck they are. I'm just going to pick a random one. Sorry, let me find one. Uh, I'm trying to find one that's like not terrible. Hold on. Okay, so I'm literally just picking a random coin here. One second. What happened? Okay, yeah, I'm just picking a random coin here. Uh, let me see if the chart will load. Okay, here we. Oh, no, it's already dumped. Hold on. That's a problem with pump.fun demoing this live. Too many rug pulls for. Okay, here's one. We're gonna try to do this really quickly. I just picked a random coin here. Again, not trying to, there's not an endorsement of this coin or anything like that. I'm just going to pick it. So let's go and copy the coin address here, okay? And yeah, we'll go ahead and get it ready to paste. All right, so let's say, "Please swap, you have to say please, just tell it in direct command, swap 0.1 SOL for," let's place in that token address on pump.fun. We'll see if that's a big enough amount. We may need to do a bigger amount. We'll see, okay? We got it. Let's go ahead and copy this transaction. And we should be able to see it in our wallet here. Yep, we got it. We got some Teddy Bear. So it's like, uh, 10,000 Teddy Bear, okay? So let's look at the transaction on Solscan. If I just refresh, I should be able to just see it. Yep, there we go. So that's entering into a trade. Let's make sure Teddy hasn't rugged yet. All right, Teddy hasn't rugged yet. So let's go back to our Eliza and let's say, "Swap, uh, let's actually, let's copy the Teddy Bear token again. Let's copy that. And let's see how much we've got. Let's just do, we won't sell all of it. We'll sell 10,000. 10,000, uh, let's paste the address. Teddy Bear token for, uh, SOL on pump.fun." All right, let's see if it worked.

So somebody says, um, "If we tell it to arbitrage between a certain pair from DEX X to Y, would it just do it?" So, uh, the answer is yes, but you need to be able to actually have the adapter for it, okay? So there, you know, to start C, start implementing advanced strategies, we've got to get into doing some of the adapters, um, and and the plugins, excuse me. So I, I'll get to that here in a second. Uh, the short answer is yes, but it's not going to be able to do it exactly like this just right now, okay? Um, so that's how you can get in and out of a trade on something like pump.fun, okay? So, uh, yeah, so you can trade on pump.fun, you can send tokens around. Let's get it to also do something like just trade another token like USDC, for example, all right? And let's use a different, uh, exchange so we don't have to just do coin launches. If you want to do like, you wanted to create a trading bot that just traded cryptocurrency price movements, let's say you want Solana coin was going up and down, you wanted to go in and out of USDC coin, you know, you can swap for USDC on something like, uh, Jupiter aggregator, which would plug into Raydium and all the other DEXs on Solana, okay? So, um, basically what you can do is, uh, let's look up the USDC address real quick. Sorry, keyboard's kind of slow today. My S key is not working for some reason. That's okay. Let's just go here. Yeah, let's just go [Music] here. Here we go. Let's copy that real quick. And then so we can do, I'm just getting the USDC address from here, okay? And then let's say, "Swap, uh, all right, let me restart this real quick. Everybody hold on."

So while that's restarting, um, I want to answer a question. So people are talking about arbitrage. So if you go here, um, and you go to the, you look at awesome, um, you look at this GitHub repo for the plugins, you can browse the plugins on Eliza's site here, okay? But you can also at this GitHub repo that I talked about before, okay? You've got a couple things here. You've got CCXT, cryptocurrency trading and arbitrage across multiple exchanges, and also EX identifies and executes cryptocurrency arbitrage opportunities. So this is an actual, um, plugin for, uh, Eliza right here, okay? So you can see you can install the plugin like this, all right? This is for, uh, uh, EVM private key flash bots relay, okay? And it can actually, uh, user find arbitrage opportunities and the trader is like scanning for arbitrage opportunities, okay? So the answer is yes, it can do that. You're gonna have to either develop your own plugin to do it or use one of the existing plugins to do it and kind of, uh, set it up.

So, um, while that's cooking, let's go ahead and take a look at some of the plugins that are already inside, um, already inside of the project. So it's a good question. How could smart contracts be integrated either to mitigate hallucinations or unauthorized transactions? How do you know if any adapters or plugins for smart do you know of any adapters or plugins for smart contract interactions? Okay, so yeah, let's just, let's just go ahead and dig into the adapter for a couple minutes, and I think that'll help sort of explain some of these questions. So let's take a look here, um, let me pull up the VS Code. Let's go and start talking about adapters, uh, and the plugins. So basically, you can see the plugins here, you have Telegram and Solana, all right? And if you go down here, all right, go to packages, all right? You can see the plugins that I have, all right? And look at the specific actions, okay? So you can install those plugins like I was talking about, but even if you have plugins like the ability to talk to Solana, you can also create specific actions that you can take. So let's say that you wanted to focus on arbitrage, and let's say you had your own specific arbitrage contract out there on the blockchain, um, to answer Lou's question, how would you get it to like not lose you money? Well, let's say that you had a smart contract that was capable of performing arbitrages, and you wanted to define the rules about the arbitrage inside that smart contract so that it won't lose money, well, you could put that contract on chain, and then essentially get your AI bot to talk to that contract, and that contract becomes the rule. So if the AI hallucinates and tries to capitalize an opportunity, then the smart contract actually be a backup set of rules to say, like, all right, I know this thing told me to try to trade, but like the trade's not going to be profitable, it's going to lose money, so I'm not going to let the transaction go through because the smart contract actually orchestrates the multiple steps of doing the transaction on chain, and if it's not profitable when it comes back, then of course it will just fail a transaction. Now, of course, that would lose a small amount of money in the gas fees, but that's how it works. So inside the project, if you go to, uh, the actions, Solana plugin, okay, go to source, go to actions, okay? You can see all the different types of things that it can do. So if you want to get your app to do things beyond trading, like, like sending tokens around, seeing trading, see, it knows how to do swaps, it knows how to trade on pump.fun, it knows how to get balances, all right? So you have to teach AI to be able to do the things that you want it to do. So if you want to do an arbitrage, maybe you have your own smart contract that could actually, if you want to make sure that it never loses money, you could essentially get a, um, you can essentially get a smart contract to enforce those rules, and you can write your own action inside of here to talk to that contract, okay? So for example, you can see here, here is the, uh, let's look at the get balances one. This is, so this is like whenever we got Eliza to simply check our cryptocurrency balance, okay? Uh, this is the action that it used, okay? So essentially, we have a template for these. So if you want to create your own actions to get to do things, you want to do arbitrage, you can do that, all right? Um, then you can, uh, essentially create one with this template, all right? And here's the template for that. Let's kind of go over the get balances one. So it does all the imports here at the top, all right? You get connection, public key from Solana web3.js. So behind the scenes, it actually knows how to write code with JavaScript, all right? Or execute code, I should say. So essentially, what it's doing is, you, you set up a name for it, so this is get SOL balance. It's got similar ones like get account SOL, get account balance, get account lamports, okay? And then get native SOL balance from an agent's wallet for a specified address, all right? It knows how it's going to get some JSON back, and it knows how to find the specific data out of that JSON, like the account ID, right here. So extract the following information from the requested SOL balance account address, all right? And then right here, this export default, where you give it the name, similar ones, description, all right? But the handler is the most important part. This is where it actually executes the code, okay? So essentially, this is the, uh, this is the, this is where you're executing the code, and you're creating an action that the AI knows how to do on-chain. So essentially, it's like connection equals new connection. This is to the Solana blockchain with the RPC URL. Um, this is the public key, all right, from the account. This is it gets the account balance, wait connection, get balance, recipient public key. And then essentially, uh, it calls back with the information that the language model needs. So like, um, content, success is true, the amount is the account balance, and the account is the content that account. So that, that's a simple example of creating an action that simply fetches the account balance from off-chain. Um, now for a template, you can say the same for swap. Um, so for swap, it's the same type of situation, but, um, you're looking at like this swap token down here, all right? Essentially, it's doing all the code for, uh, getting a quote from Jupiter, all right? And then actually doing the swap and getting the response, okay? And then returning the the information and a lot of other kind of complex code inside of there, but it gets, it gets handled inside this handler right here, okay? So for your own, if you're gonna code your own actions on-chain to get to do other things that it's not currently supported, you're gonna create a new action inside the Solana, um, plugin here. You can use this template as a starting point. You fill out the basic details here, and then, uh, for the actual code that's going to be written to the Solana blockchain, you put it inside of this handler right here, okay?

So, can Eliza schedule tasks or do polling? Is there a plugin for that? That's a great question because at the end of the day, um, you might want Eliza to do things for you in the background. So yes, it can. Let's just go, I'll show you that. Let's, let's answer that question right after we demo the, uh, USDC here. So let's say, "Swap 0.01 SOL for USDC," and let's paste in that address here, okay? All right, let's go back to to our, uh, thing here. We can see that we got some USDC. All right, there we go. There's our transaction right here. We go, and let's swap, um, let's see how much. Let's just do 0.11 USDC. Let's get out of the trade. Let's paste that address, uh, for SOL, swap the opposite direction. Awesome. There we go. And we got other trade as well. So yeah, you can swap on pump.fun, you can swap on, uh, Jupiter aggregator to do any, you know, token that you want to that's supported on the Solana blockchain, um, hook up your crypto wallet, get to do all that type of stuff.

So let's kind of talk about some of these more advanced topics. So, uh, I'm going to try to answer some of the questions have been popping up in the chat. Everybody go ahead and throw your questions out there that you want answered. I'm gonna start answering some of the ones that were, uh, put up right here. So, uh, yeah, Landon says he's not going to sleep tonight. So, can Eliza schedule tasks, um, or do polling? Yes, it's a great question. So I'll pull these docs right here. You click on the documentation tab and Eliza. This is going to be, uh, your help. So let me just open it over here. Got documentation. Um, the tasks are going to be your friend. So we talked about a lot of different nouns here. We talked about plugins, we talked about adapters, uh, we've talked about, um, these actions, okay? So the actions are actually things you can do on-chain. But tasks, um, are basically a way to manage deferred, scheduled, and interactive operations, okay? So the task system allows agents to queue work for later execution, repeat actions at defined intervals, await user input, and implement complex, sorry, complex workflows across multiple interactions, okay?

So, VT's asking about that. How can we get access to the code? The code is in the description down below. You can just download that code. You're good to go. Can you get Eliza to trade for me? Yes, you can. So we just demonstrated the actual trades happening on-chain, but if you want to do automated trading or automated anything, um, so Sans says, "Can Eliza trade actively trade on crypto market on the blockchain profitably?" Yes, so that's kind of what I'm getting into. We demonstrated how to trade with the, uh, command prompting, but here's, um, here's kind of like an example. So, um, I'll give you a really great example. So I talked about the mentorship earlier. We've got one of our students who's working on a project inside the mentorship, which is pretty cool, where he's actually using AI to essentially kind of turn it into an AI like meme coin trader, okay? So one of our students is doing. And so basically, I don't remember if he's using Eliza or not, but what he's doing is he's essentially, uh, using something kind of like this, and he's got the Eliza setup into his Telegram, okay? Is either Telegram or Discord, I can't remember. But so instead of instead of like always looking at it like this, he's got it set up to his Telegram. What it's doing is it's actually, uh, using stuff like these tasks, okay? Where it's like routinely fetching new information from the blockchain, like new coins that are coming out, new projects that are even trending, okay? And it's listening to all that, and it's sort of surfacing some options, uh, to him through his, you know, user interface, right? So basically, like if you had your chatbot open, it could be listening for this stuff on a regular interval. It could be like, "Hey, I see this thing." Okay, it could pop up and like, "Hey, do you want to trade this?" And you could actually say, "Yeah, that looks good. Let's do it." All right, you can hit enter and it would do it for you. Now, that's a really great step to test things out because what you can see here is it can queue up work for later execution, repeat actions, define intervals, or await user input, okay? So sometimes when you're trading, you may want to just say, "Okay, yes, that's what I want you to do." Sometimes you want to happen automatically in the background. So if you're working on something like this, you may want to have it await for user input while you're testing it out, okay, before you actually get it to do it completely automatically. All right, but you can basically do these later execution, repeat actions, and define intervals, await user input, and also implement complex workflows across multiple interactions. So you can see you can define tasks just like this, okay? You can create task workers. So this is, uh, the actual logic that executes whenever the task runs. So each task worker is registered and the runtime is defined by name. You can see how that works right here. And then you can also, uh, uh, register the worker to handle the execution. You create a one-off task, create a recurring task. Someone's asking about doing things in regular intervals, like polling. That's exactly what this is. Create a recurring task, and then creating a task await user choice. So if it's doing something, it says, "Hey, here's this cool thing I found. Do you want to do it or not?" Here's how you would do this. It's confirm action, okay? And then you can manage all your tasks, um, like this, okay? And you can see how they process your tasks. So tasks are going to be your friend if you're trying to do like scheduled or recurrent things.

um what if the interface is closed? Well, you see the messages the AI sent to us when you open later. Not necessarily. Uh, that would be a good thing to have some like Telegram for. So if you want persistent message history for the AI, um, that's probably what you want to have your Telegram account set up for because Telegram will always keep the chat history. And if you're like walking around and you want to have on your phone and get notifications in real time, then yeah, you could have Telegram with alerts set up and, uh, yeah, you could just have it pingy that way.

Is there anything that will affect if we get front-run on arbitrage opportunities, like speed-wise? Um, I mean, if if you're doing it like if you're doing it with a smart contract that enforces a certain profit, uh, not losing money, a smart contract, then the only thing you'd lose at all is the gas fee. So automated trading, how to set it up? So unfortunately, we're not going to have time to do a complete automated trading demonstration right now. Let's see here. Actively trade crypto market. Yeah, same one. Um, let's see here. Somebody else also asked about Billy Betts. So I mean, the Billy Betts thing, I mean, so Billy Betts is not going to be a simple prompt that you simply add to the block, add to the agent, and then suddenly it turns into Billy Betts. I mean, that's going to take quite a bit of, um, you know, programming to like teach it how to do that, to train it to actually plug it into like sports tensor to get all that type of stuff. Um, you know, you, if you want to take a first immediate step or something like, um, you know, crypto markets again, we do have the Bird Eye, um, implementation inside of here. So if you want to hook it up to Bird Eye, start getting crypto market data, um, then yeah, you could do that.

So that's a great question. So, um, somebody says, would ARB code from the previous courses be compatible with the agent? If so, are edits required? So, um, and then, uh, Khim says, how will we set up to add a smart contract to your brief in to it? Yeah, I'll kind of answer both those questions at the same time. So, you know, obviously there's a million different things that you can do with this. Um, you know, we're not, we, we could, we could probably even do another class if you guys wanted to on how to do some more advanced topics on this type of stuff. If you're some interested in, just shoot me a message and we can try to make that happen, maybe do a follow-up class. But, um, you know, whatever you're going to decide to do there, there is going to be some, you know, kind of advanced coding, uh, to get it to work. Um, like I said before, we got even some students at the mentorship program who are working on this type of thing. We've had some students creating AI agents and some hackathons and some of our recent alumni work on this. Again, if you guys really want to dive in this further, uh, we do have people working on this this inside the mentorship program. It's the absolute best way to learn the skills that you need with hands-on help, with feedback, you know, with guidance in real time. Again, I'm gonna put an application for that down the description below. So if you're interested in building your AI agents in crypto, uh, you know, getting the industry, seeing what's possible, uh, we do got have a couple slots left and this is the absolute best way to do it. But, um, so yeah, it's a question, how would the ARB code from the previous master classes work with this? So that AR code is really broken out into two things. It's like the Searcher bot and the smart contract that actually does the execution. Okay. So, um, what you would do is your agent's going to be more like the Searcher bot that's watching for things. Okay. So that's, you could definitely plug it into, you could, you could get it to watch for the opportunities. But I'm gonna answer, uh, Khim's question in conjunction with this, which is the real benefit from that pre previous class will be the actual, uh, trading contract. So the trading contract is what takes out the flash loans, you know, does the arbitrage opportunities and enforces that the trades are actually profitable so that you don't lose money. Okay. So, uh, that's going to be a smart contract. So to answer Colleen's question at the same deal, uh, let's go back to VS Code and let's look at, uh, the source actions. Now, that previous arbitrage bot was written in Solidity for Ethereum, so it wouldn't work directly with this plugin. You'd want to have an EVM plugin, which they do have. Okay. But you basically would create your own action for a a smart contract function call. Okay. So basically this swap really is a smart contract or Solana program function call. Okay. So what you do is inside this template, um, you would import what you need. You need the connection to the blockchain, you need the smart contract ABI, you need the smart contract address, and you need the, um, library, which is like Ether's JS or for Solana, it's Web3.js. Okay. Um, to actually talk to the blockchain. So, um, again, I'm kind of having to translate this between Solana and Ethereum, but you would essentially, um, inside of this action, all right, you can see, um, you can see the stuff right here. So perform the action here. Example, set up a connection with Solana. If you're going to do on Ethereum, you can set up a connection with Ethereum. Set up a transaction. So in order to set up a transaction, you need to create a JavaScript version of the smart contract. So you'd use like Ether's JS on Ethereum. You'd put in the smart contract ABI, you put in the smart contract address, and then, uh, then you have a JavaScript version of the smart contract. And let's say that the smart contract call was, you know, the function call was called trade. Then you just do like trading contract.trade. All right. Obviously, you need to connect the wallet. So like trading contract.connect, pass in the wallet, and then do trade, pass in the variables, and then do wait for the transaction to finish, get the transaction hash, and then you just return the results here. So like, let me just pull up some code here. Let me just pull this up here. Hold on. So here's like the smart contract for the one, the last trading bots we did. Okay. Um, so I'll just show you like, hold a second. So like, you know, here's, I mean, literally I'm talking about something like this. Basically, it's almost like this thing right here. If you're on Ethereum, you'd essentially just put that right here. Okay. So hopefully Colleen, this kind of helps answer your question and also, um, I'm sorry, who Al, who answered it before? Basically, like you're creating a signer. Obviously, you don't have this if statement in there. All right. It is basically that part that actually does the trade. That's where you put in your own action, like this. It's a trading action. It's an arbitrage action, and you get your AI to learn how to do that. So anyways, I just want to show that.

So, Floronic was asking, is there a group for students in this class? Yes, definitely. So, um, if you joined the, uh, one-time class tonight, or so, if you joined the Blockchain Mastery, uh, program at checkout, there is a Slack channel for that. Uh, if you didn't get access to it, you can just shoot to Gregory at university.com. If you join the Blockchain Master program, we can get you access to that. Uh, if you're talking about the mentorship program, um, yeah, we definitely have a really tight-knit community inside of there of students who are really digging into this stuff. We're meeting every single meeting, three times per week, um, kind of going over a lot of this type of stuff. Um, if you want to, you know, get access to that where we're, you know, meeting all the time, again, we got an application for the mentorship down below. Let's see here. Yeah, so Kim said he had that thought too. These actions, smart contracts? So the actions aren't smart contracts, but the actions are things that you're going to do on-chain. Okay. And those can include smart contract calls. So we, the best much watch master classes in BL versus un-teachable. Um, so I would say like, definitely the trading bot master class is very popular, the flash loan master class, and I think this will probably be another must-have class in there for sure. But those are the two very popular ones. We've got lots of other ones, but I would check out those two first. And we are getting close to our time limit here, unfortunately. I wish we had all day to do this. Maybe we will try to have a follow-up class, um, if you guys are interested in that, let me know in the chat section below. Go ahead and drop your questions in there. We are going to run out of time here shortly. Um, but again, I'll leave, I'll leave a link down below for the mentorship program. If you guys want to get in there, again, we've had students get into the mentorship, go on to create their own businesses together, launch their own apps, go on to win hackathons, or I should say place highly in hackathons, collaborate on all sorts of types of stuff. We do have a couple slots open for the mentorship. I'll leave a link to the description below.

So, need a check sheet, step-by-step process for the whole setup. So the good news is the step-by-step process setup is in the GitHub. But definitely go back and watch the setup steps. Re-watch this video again. It's a lot to take in when you watch the first time, but if you go back and watch it again, follow each step, um, you should be able to get this thing up and running. So what's the difference between actions and smart contracts? That's a great question. So smart contracts are the, uh, programs that exist on the blockchain. So you create a smart contract, you put it on the blockchain, and then the action could talk to that smart contract. So, yeah, somebody said, what if the interface is closed? How do you get the persistent messages? So again, that's what a Telegram bot would be great for if you want to have persistent messages over the long period. Yeah, and it's all happening on-chain. Yep. How do we get avoid? How do we avoid getting crushed by people who are more nerd than us in this space? It's robot wars right now. That's a good question. So I mean, the good news is there's a lot of room to develop skills and create things of value, even if you're not like super duper duper duper duper advanced because you can create small things of value that other people actually want to use who are too scared to even touch it in the first place. Okay. Um, the other thing is, I mean, it's a big world out there. The blockchain is a huge space. The AI space is a huge space, and there's a lot of opportunity to go around. And there's, you know, I mean, yeah, there's a lot of people who would look at like what we're doing tonight and say, that's too scary for me to touch, but you know, you're here doing it. So, uh, yeah, you should have gotten Slack access on the email when you subscribe for monthly. Uh, if you didn't, just shot to email Gregory at university.com and we'll get you square away. So are all actions one-step processes? No, they don't have to be. They could be multi-step. All right.

Everybody, we'll go ahead and wrap it up there for today. This has been a great session. I've really enjoyed digging into the AI agent stuff. Uh, again, you've got links down below for the code. Um, you've got the setup instructions inside there, inside the README file. Again, I know there's a lot to take in on this class. Um, definitely rewatch this. You're going to get a replay link afterwards, sent out to you. If you're in the Blockchain Master program, you can go access out the video vault. That's going to be the time to slow down and do everything step by step and really ingest this stuff. Spend some time with the Eliza documentation. That's going to really show, that's open your world up to what you can do. Again, um, you know, there's a lot that you can do with this stuff. Obviously, you're gonna get your have to have your hands dirty with some code. Um, you know, you could try banging your head against the wall and try to get that figured out, but if you want to, you know, level up your skills and actually get the coding skills that you need to build advanced AI bots, arbitrage trading bots, you know, flash loan applications, all that type of stuff, or if you just want to work in the industry, you know, uh, the absolute best way to get those technical skills is going to be with hands-on mentorship. That's what we're doing, the Blockchain Mentorship Program. I'll put a link to the application down below. We got some really great students on the call tonight, um, who are part of that program, people working on AI agents inside of there, lots of amazing stuff. So, uh, that's all I got for today. Great class today, and I'll catch you guys on the next one. For.

Get Our Mobile App

The AI Agent Masterclass - March 20, 2025

Transcription