OpenClaw Agents

Agentic AI is changing the world at lightning speed. In this episode, we discuss why it's so important to get started now. You do not have time to wait. Each passing day is another day wasted. You must start now. Listen to this episode to understand why it's so imperative to learn these tools because time is of the essence. You don't want to be the one in 3 years looking back wishing you would have done different. The technology is here to stay, forever. Harness the change and embrace it. Use it to your advantage. For it's all we can do. 

Transcript

SPEAKER_01 0:00

Open claw on the Zanben podcast today. You've probably heard about it. You've probably seen the little lobster with the claws, but you have no clue what anybody's talking about. Or maybe you're the person that does have an idea that they're what they're talking about, but you don't know how you can install your own agent and get it running. Today that's what we're going to talk about on the Zanben podcast and more. So you've heard about OpenClaw, the AI agent revolution that is completely transforming not just technology, but the world. No, no, no. It's a very complicated program, but it is simple for us to use. All we have to do is get it onto our local machine and kick it off. Once we do that, we're set up and running. But you need to connect your agent to an API key that will allow you access to some provider's model, whether that be OpenAI and Chat GPT or Anthropic in Claude or Claude Code. What you do in OpenClaw is you go into the when you're when you're setting up your your model, or I'm sorry, when you're setting up your agent, it will ask you to input the whatever model you want to use. And I personally use anthropic um claude opus 4.6 is a latest update for um for the opus, which is a deep reasoning uh coding, agentic coding um or model. And so it works fantastic. There's a lot of great models out there. You know, it's kind of the the Ford versus Chevy argument in a lot of ways, right? It's just like a preference. So so use whatever model that you want to use, but get it installed on your open claw agent and get it running. And if you don't know how to do it, just run to Google and and Google uh how to set up OpenClaw on my machine and it will tell you. Or go to Grok or go to Chat GPT and it's going to tell you all the information that you need because this technology is a must have. It's not a must-have, it is a must use. And you know, we've talked about um uh thoughts versus action, and the only thing that is gonna change right now for you is if you take action. The thoughts don't even matter. You must put those thoughts into action. We've all had great thoughts, and um some of them have turned into um turned into action, and some of them we have not turned into action. So I think that um it's very important to take those thoughts, turn them into action, and don't even look back, right? We're gonna have failures, we're gonna have mistakes, but those are the things we learn from, and that's how we how we move forward and we take the next step and the next step and the next step. And we don't even look back and and see the the things that may have not worked out in our favor. We just learn from them and move on. So the great part about these agents is that that you really don't have to learn anything. All you need to know is something that you want to do. So my brother recently, um, I was telling him about this stuff a week or two ago. And um he so the next day he contacts me. He had just set up his agent, or he was trying to set up his open claw thing. Didn't understand really what was going on, but but he got it set up. And so I hadn't heard from him in a week or so, and then he sends me a message yesterday, and he's like, he's like, hey man, I've been working all week, but I finally got a chance to come sit back down with my agent, and um, I'm I'm messing around with it. This thing's really cool. A few minutes later, he sends me a picture that he took of his computer screen that showed his Telegram open, and he was he was having a conversation with his with his agent there in Telegram, and it had a um a chart that had some real estate data in it that I guess that he was putting together or he had his agent putting together. And he's like, Man, this thing's really cool. I said, Yeah, that is cool, but check this out. Ask it to create an inter interactive dashboard for this program. A few minutes later, he sends back this this uh this picture of his computer screen that had this beautiful interactive dashboard open that had variables in place that he could swap out and change. And um, and he was simply blown away with it. And this is just very, very, very rudimentary stuff that we're doing with real estate data, right? Like this is data that's out there. And he just had his agent go in and fetching the data from wherever it was and putting it together into something that made sense into a program that you could then input variables and and get an output on the other side of whether it was a good deal, bad deal, you know, um rate of return on investments and all sorts of different stuff. And so it was really, really cool. But more so than that, it's the ability, like I was telling my brother, I said, find your pain points and build solutions to them. And that is something that that I've really been able to capitalize on, just just that idea is that we all have pain points in our life and no one knows the next person's pain point, right? We only know the pain point in our lives and the friction in our lives. And this doesn't matter whether it's like like dealing with finances, emails, work, uh, your your yard. I mean, anything that you have going on where you have friction, think about that. And think about um the friction and then what it looks like on the other side of that. What what what would you like it to be? What what process is there that could make your friction a little bit easier on the other side? And um, and then what you do is you just go tell your agent, hey, I'm having these issues. This is what I do in my profession and my daily life and my in my fitness, whatever it is. And here's the here's the problems that I'm having. Go research this field, body of knowledge, whatever you want to call it, and tell me what and come back with ways that that I can improve whatever it is, or build technology to support it, or a program, or a dashboard, or or something to give me outputs on the other side, whatever, or ideas, whatever it is. And your agent will go go research it, then come back with some with some ideas for you. And then you can ask it, it will ask you questions about about what you want. And it will ask you questions that that you didn't even know that you needed to be asked, especially because we don't have the answers, right? If we had the answers, we wouldn't be asking these agents to do it for us. So if we don't have the answers, we don't know what we don't know, and that's the most terrifying thing in all of this stuff. We don't know what we don't know. So our agents don't know what they don't know either. But the good part is is they know a heck of a lot more than we know about generally everything. They just may not know the little tiny details about what we want and what we're actually trying to convey, which is why the prompt, what you tell it, is so, so, so important to get that right. And so, but our agents are able to go do all this research and come back and ask us clarifying questions. And you can ask it in the prompt at the very end, ask me clarifying questions, and it will come back with questions. Hey, what about this? Is this how you want this dashboard set up? Well, you have three options here, here, and here. Which one would you like? And then you can tell it what you want and it will go through and piece by piece get it built out so it understands what you want, and then it will start building a program for you to do whatever it is. We've built everything from iOS apps to um law, um a program that that ingests um data from four attorneys. So the idea is this program ingests a law firm's entire uh active work history, work cases going on, active cases going on. And what it does is of course, every case is different, client A through Z, right? Everybody has a different case, everybody's at different parts in their case, and no case is any is the same. And so this program that I built, my agent built, is that it takes all that information and it comes up with a a summary for the attorney so they can see where they stand. Um, it comes up with a jury jury report, a mediation report, um, and it has uh an AI layer built into it where the lawyer can go in and ask it questions about specific cases to see the strengths and weaknesses and just do all this stuff. And it's amazing. And so from building guitar tuners to health apps to uh uh programs for lawyers to run their law firm to 3D printing programs to literally anything else you can possibly think of. It is possible. And every single day that passes, it becomes more possible. And I mean, this is the most exciting time for anybody who who loves loves something new because every single hour there's something new coming out with this stuff, and it's and it's new and improved. And um, I mean, there was just a claw, there was just a claw, there was just something that came out today I was reading about this morning that was uh Chrome had updated to uh Chrome 146, which allows agents to operate the browser, right? I mean, like if you understand the gravity of that, we're talking about before agents, you know, we're running through programs like Playwright and Puppeteer to try to run through these browsers, and everything is a pain in the butt when you're trying to deal with that. And uh you must be technical to understand how all that stuff works. And so even with the agents, with the agent revolution that's recently happened, the agents have um leveraged the previous programs like Puppeteer and Playwright so they could uh browse web pages and take screenshots and so forth. But with the latest update of Chrome um 146, this new update allows agents to operate it essentially in a native manner without having to use any acts, any outside um third-party library or program to do that. So the the middle is getting squeezed out, right? Like we've discussed before. And so, I mean, think about it. If you were one of these, if I mean, there's been companies, companies, browser use is one of them. Uh, you know, browser use is a company, a for-profit company that has offered this as a service. And now with this latest Chrome update, what's the need for browser use anymore? So uh not that I wish that company bad. I mean, uh, they're good, they need to, they need to adapt with the times just like everybody else is adapting with the times and come up with something different, something unique that they can put out and and keep keep working with, um, or else they'll go under. And so, but the exciting part for me and you is now that that that technology that they were just, I was just on browser uses thing today. I mean, they charge a lot of money for you to use it. It's like $400 a month for to use browser use in the medium plan. And um, so now with just this Google Chrome update, heck, you can just go out and do it yourself for free, just with your agent. And everybody sees all these companies are trying to position themselves to being the agentic workflow solutions, and uh, that's a really good thing for for us. And so I really encourage everybody to start start messing around with these agents because okay, they're not perfect, right? There's security risks involved, just like there's security risks involved with everything. But um, but now's the time to really start learning because now you can um now you can get your feet wet, and so it'll be easier for you in the future when you see something that you're like, oh my goodness, I want to go do this, and all I need is to be able to run a program to do it, but you've never done any of it, and now you have to go through the learning curve of doing it at that time, which sets you behind when you could be doing it now and putting yourself ahead of everybody else because it's a level level playing field right now. That's what we have to understand. Nobody is ahead of anybody else. It doesn't matter if you're Carpathy or Zayn Benton, nobody is ahead of anybody else. I feel like it is, and please understand, I'm not comparing myself to Carpathy. What I'm saying is that he's up here and I'm way down here, but the playing field is now level because of these agents and the technology and everything going essentially um really strong open source. So so long as we have this technology open source, which it seems like um there is still a really strong push to keep it available and free, generally speaking, then uh we're all in good shape. And you know, right now the AIs aren't free using the AI models, but here soon, but you can you can run your own local AI models and um bypass the cost that you incur with OpenAI or Anthropic making those API model uh calls to external uh APIs. If you have a computer that has enough hard drive space and enough RAM, uh then you can run your own local AI models. And uh they're they're really, really good. I'm running a a couple. I have uh I've ran Quen. Uh Quen has a uh 3.5, I think is what it is, uh Coder, Quinn, Coder 3.5 or whatever that is, and um one of the deep seek models. And so they're really good, but they're not up to the standard of an open AI or an anthropic model, uh Claude. Um, but they're not lagging that far behind. I mean, they're only like I gave an estimate, I think, on the last podcast, it's saying like three to six months behind. And I saw something today about someone made a post on X saying that they they think that they're about nine months behind. Whether it's three to six months or nine months behind, what that means is that this free technology and the AI models that are completely open source that you can download today are only a few months behind, maybe up to a year, these other frontier models that are the best. And so with each passing day, uh it's it's those models get better. So all of the models get better. And so here in six months, we're going to have free models that are equivalent to the paid models today. And if I can have a if I could if I can have a free model of Opus 4.6 in the future, I mean, that's all you need. Like I can't imagine what comes after uh Claude Opus 4.6. I mean, I think that that model is just absolutely amazing. So what comes next? Well, Claude hopefully is going to come up with a with a uh Opus 4.7 that hopefully blows us out of the water once again. And then the open source models will creep right back up to the 4.6 standard, and then they you know prevail from there. So it is a very, very exciting time to be alive. And and if if you're scared right now, that's understandable too, because some of us um aren't as accepting of change in some areas of life than in others or at all. Um, and maybe that this is one area that you're not comfortable with change, and that's understandable. But um, I mean, you really have one of two options. You can either embrace it and capture it and harness it and use it to propel your life, or you can sit on your hands and get railroaded by everything going on around you and just sit and watch, I guess. Um, and it's your choice. I mean, it's not saying that that's really bad either. I mean, it's just up to whatever whatever anybody wants to do, I'm fine with. But but I hope that I hope that you're curious like me, because that curiosity will drive you to figuring out new things and just not settling with being okay with whatever it is. And what you the main thing, if I could get across one point with all of this stuff, is you have to be curious and you have to be willing to just persevere, just overcome those errors because in every little step there's going to be errors. There's gonna be, I don't know how to do this, I've never done this before. Well, guess what? Google it, go to go to go to chat gbt, go to go to grok and copy the error and paste it into there, and it will, you know, tell you how to fix it and just keep going back and forth, and you'll get it. You will get it. So I really hope you enjoyed listening to this podcast. I think that I think that we are all off to to bigger and better things, and um there's no better time to start than now. It's much better than starting tomorrow. Okay, maybe you didn't start yesterday like you said you're going to do, but that doesn't matter. All we have is now and this moment. So let's kick it off and let's get it rolling. See you next time on the Zambin Podcast.

SPEAKER_00 20:30

So I'm just gonna call you some little bit of a little bit of a little bit of