We're back in back in the game. Julian Michael Anthony Goldie here. Let's get this popping. So today, one thing that I've been building with recently, and this just dropped less than 24 hours ago, actually. It's pretty wild. We have Fugu Ultra. Fugu Ultra. And that is a Japanese model on benchmarks actually beats Fable 5 level intelligence. Basically the way this works, it fuses multiple models together. So it kind of like orchestrates agents and then you have different models working together to build something better than you [music] would get with Fable 5 alone or any other sort of model. So that's what we're looking at. We've already measured it on Goldie Bench. And if we look at the leaderboard here, you can actually see that it's right at the top of the leaderboard along with Fusion Opus 4.8 landing in third and GLM 5.2 to landing in fourth place right here. Now, if you're wondering, okay, like you know, how do you come up with these benchmarks? So, I actually test this stuff personally tested for a few hours a day and then what we do is everything that we build with this in the background, we actually plug into these benchmarks and so far Opus 4.8 is behind Fusion and behind Fugu Ultra. Now, Fugu Ultra is a Japanese model. It comes from Sakana, which is a tiny little AI lab in Japan. And what they've done essentially is released two different models, which is Fugu Ultra and Fugu Mini. Now, Fugu Mini is nowhere near as good when I've tested them out side by side. So, Fugu Ultra is a lot better. The only problem is that you get rate limited on the amount of tokens you can use with these AI Asians. So, for example, if you [music] keep building out with Fugu Ultra, you'll be you won't be able to access it for like 5 hours if you just keep rinsing it. That's one thing that we found. Also, you got to be careful with the a API as well. So, you get an API. What we actually did to to get the most out of it is basically plug it into our agent operating system. So, you can see for example inside Sakana Fugu over here, we have the workspace with everything that we've built out, which is pretty nice. And then we can come back to it, see what we've built, etc. You can actually see like this is a operating system we built with Fugu Mini. So Fugu Mini built out a full operating system and these are like actual apps inside an operating system that we can come back to. So if we write something here and then we close it and then we open it back up. It's not saving the work which means that essentially it's not as it's not done as good as GM 5.2 on the test. Uh GLM 5.2 So you could actually create stuff actually saved. If we have a look here, it creates some amazing stuff. I mean, you can see it says Fugu Mini inside the actual file, which means it's been created with Fugu Mini rather than Fugu Ultra, which is a premium model. And then if we have a look, there's some cool stuff here. You know, it has built out some cool stuff. So I think if you're going to rate them, I I'll show you the leaderboard again in a second. But yeah, it's built some interesting stuff for sure. Let's have a look at this one. I haven't even half of this stuff I've not even tested like uh but that's pretty cool, isn't it? That is cool. That's a whole open world game right there. Skyrim level. Uh yeah, so we've tested out both of those models a lot. Again, Fug Ultra is much much better, but they're both creating some awesome stuff. Uh the the only problem is of course like if you're using Claude directly. Oh, look at look at let's have a go at this. Whoa, this is cool. Didn't even know we built half this stuff out. Um, [clears throat] and so if you look at this, one of the problems is that you are going to have to use an API, right? So, for example, if you use like Claude, when you're using Claude on the subscription, like you can go back and forth as much as you want on the subscription and the tokens. Yeah. But when you're using something like the API, well then you pay per token and that can get quite expensive to be honest with you. So that's one thing to worry about. Another thing is like it's just super slow. So if you're using Sakana, you might be waiting like 10 15 minutes per generation. So it can run for hours in the background. I think we had this running for like 8 hours just building stuff in the background like this. I didn't even know about until I checked it today. But essentially that's that's something to know about. It's like if you want speed, probably Sakana Fugu or fusion are not the right fits for you. But this this stuff is cool. Let's see what else we got here. [music] Yeah, some interesting stuff right here. Let's see what we got on the questions. [music] I don't I don't quite understand that question, but feel free to rephrase it. Louis, is it really important to have the best model or the latest one or just with knowing how to implement? Um, honestly, you don't need the latest models. You don't need the best models, right? You just need the models that automate the stuff that you do. And I wouldn't even say you need the models. You just need the system, right? So, for example, like the amount of time that I spend on SEO is basically zero now because I have AI agents that deploy and come up with keyword research and do it better than me, faster than me, as you can see right here. and and that's how we can get results for our websites. If we have a look for example at this video agent, this video agent can create videos on the spot, fully edited, looking beautiful, better than I could create personally. Same for the music agent, right? And so anything that you're doing, it doesn't matter which model you have. It doesn't really matter if you know about the latest models or not. What matters is are you automating the work that was taking you time before? And if you're doing that every week, if you're doing that every day, if you're making progress on that every year, you know, you're winning. You're already ahead of 99% of people. So that's the way that I look at it. Yeah, this system is inside the AI profit boardroom. So we got Sakana Fugu over here. We have Fusion built in as well, which another similar style model. Um, we have all of this stuff built inside the agent OS and the agent OS system, the agent operating system is inside the aircraft for boardroom. So, if you go to the air prof, we had new daily updates. So, we actually update this every day. Like literally as soon as Sakana came out, we built it into the system and we released an update that day. So, as soon as we find something that's good and we've tested it and it's actually useful, we'll add it inside the AI profer boardroom. So, link in the comments description or go to the air for.com to get access to that. Spanish. Yeah, I think um I think I think that comes after like 24 hours, something like that automatically. But yeah, if you have any if you have any questions as we're going along, feel free to ask. We're going to start building Claude now as well. Let's see what we got here. Start building out the agent OS. Run some workflows. Can you build out the rest of the test with Fugu Ultra now? What keyword should we go for today? Yeah, you just get the API and then you plug it in. But if you want to get our setup, it's inside the agent operating system and then we have the tool ready to go. Pretty good. Yeah, it's built some interesting stuff. Uh I don't think it's that smooth. It's definitely not fast and for me I I do like to work fast so it's not that useful in that respect but it's pretty cool. Now you can see uh on Goldie Bench we've got Fugo Ultra over here. We have Fugu Mini. We have Fusion. And so if you want to see like, okay, what do each one of these look like? You can just open them up and check yourself, [music] right? And then you can get a feel for, okay, which one is the best one? Are these any good or not? Uh so, you know, test it yourself as well. See what you think. This was pretty insane. This is probably my favorite generation with Fugu Ultra so far. Bear in mind, we've only done four tests on Fugu Ultra. We've done a lot more on Fugu Mini just cuz it's faster and also the token limits as well. But it looks absolutely insane, doesn't it? I mean, I'm not going to do I mean, my Spanish is good, but it's not that good. I don't I mean, this is a Fugo Ultra is Japanese. Um, but for me, I don't support any particular product. Like, I don't really care. I just test what works. If you actually look at the leaderboards here on Goldie Bench, and we have a look, we've got uh Japanese, then we have an American model over here. So, number one is Japanese, number two is American, number three is American, number four is Chinese, number five is American, number six is Japanese, and number seven is Chinese. Right? So for me, I just test all this stuff and then I test and see what works or not. Thank you. >> [music] >> Uh, prompt engineering is not important at all. What you want to focus on instead [music] is loop engineering. So there's a big difference here. Like for example, you can go back and forth inside the inside claude for example and get stuff done here, but it's going to be a lot slower and a lot more energy constraining. Whereas if you build loops way more powerful, right? So the reason that we could test uh Fugu Ultra last night and Fugu Mini just you know running in the background is because we have loops set up. So what we do is we set a goal. We have one model that builds it. We have one model that judges it. And then it goes round and round and loop. And that's way more powerful than any prompting setup that you have. So I wouldn't worry about pro uh prompts. I would worry about loops. We created a tutorial on it yesterday. And we've also built loop tools inside the agent OS system so that instead of relying on prompts, you can save yourself a lot of time and energy and just use loop engineer instead. I would say that's way more powerful. [music] >> [music] [music] [music] [music] [music] [music] >> Demo testing. Nice, man. Congrats, Lazarov. Good to see you here as well. What are you creating? What are you testing? Wow, that's awesome. You think a model such as GFT 5.5 is not consistently same good sometimes feel like it's really bad? 100% agree with you. I get the same with uh Claude as well. Claude is just like that. Claude is quite messy sometimes. Uh sometimes it's amazing, sometimes it's really bad. So yeah, I would look at that as well. Um I think for me though, like what I like the most is is Claude out of everything. You know, it's the most consistent. It does feel like it makes a lot of mistakes, but it's still better than the other alternatives. [music] Randall started to think you weren't going to be on. I'm on every day, mate. Every day. I'm all good. Hope you're well as well. Let's see what we got here. [music] That's pretty insane, but awesome at the same time. See what we got here. >> [snorts] >> Thank you. Hey. Hey. >> [music] >> Can you set this up with the Twitter API from Grock as well? So we can search Twitter for news. Have it as a separate section next to SEO. [music] Make it look visually beautiful as well. >> [music] [music] [music] [music] [music] >> Look at that. 40m minute timeout. Fugu Ultra. It's crazy. [music] >> [music] [music] [music] [music] [music] [music] >> I don't know. I saw someone built that, but then I saw other people saying it wasn't real. So, it's kind of one of those. But yeah, I mean, he's got vision, right? You have vision models for this sort of stuff. So, I don't see why not. [music] Legend, thank you so much for joining. I appreciate that. And yeah, [music] it's a it's a great place. Lots of good stuff going on. But yeah, appreciate you joining [music] [music] >> [music] [music] [music] >> Heat. Heat. >> [music] [music] >> Heat. Heat. >> [music] [music] [music] >> Hey, sometimes >> [music] [music] >> That's what it takes sometimes. Could take ages, right? uh just to get the finishing touches. I think you can build the 80 20 of it pretty quickly and then you know it's it's the last 20% that takes the most time for sure. [music] [music] >> [music] [music] >> Can you create a guide plus showcase how we set up Goldie Bench for SEO? for example like versus keywords optimizations across a website mentioning it as a brand across our content and then [music] also show how you can use the agent OS for the system. So for example like checking Google search console auto deploying [music] to your website etc. indexing with index exceptional and create that as a full guide. then talk about how you can use a combination of clawed agent OS to help you do that so that it can automatically find keywords for you each [music] day. It can auto deploy. You could build out multiple websites doing this as well. It all works together as a system and talk about loop engineering as well. are basically how you create content, you look at Google Search Console, you create content [music] around the new keywords you're ranking for, and then that's positive feedback loop where the next day you show up and again it runs that automation where you've got new keywords based on yesterday's data or last week's data, etc. for SEO. [music] [music] Also, what's really important as part of the system and the framework is that you mention how we're using first party data. So, all the tests I've done and experiments I've run and then we showcase that on the website. So, it creates information gain. information gain is what Google's looking for when [music] it comes to content because it doesn't want generic fluff. [music] Include a CTA halfway through for the free course for SEO and then include a CTA at the end for a onetoone SEO strategy session. [music] [music] Yeah, exactly. It test play right. [music] And also mention the great thing about this is you don't have to do any extra work, but it's actually really valuable because you're just documenting [music] automatically all the stuff that you're doing. And that's kind of a loop in itself because you do your work, it gets documented as a new website that ranks and then from there it's very easy to generate content without having to worry about SEO and you can just give the fundamentals to your system in terms of how it should work for SEO optimization. [music] Let's see. [music] [music] >> [music] [music] >> Heat. Hey, Heat. Write a note in notes for me, please. [music] doesn't [music] >> [music] [music] >> Well, I mean, the way that I would use that is that you basically have a a system set up. Let me show you an example. Right. So if we go over to our memory here, memory galaxy, this is the way that we would use that. So it has knowledge inside the system automatically and then we can come back to that at any time. And so you want to use something like Obsidian, I would say, giving your agents knowledge and also retrieving knowledge. This is one of the best ways to do it. It's pretty simple to set up. You can see for example every time we use our agents like this one which was literally right now it already gets documented inside the memory galaxy and then we can [music] come back to it any day. Also we get like a daily journal here which is automatically created with our agents and I think that's one of the best ways to do it because it's a great system. It's visually beautiful. It's easy to organize. You can give it to any agent and also it's very very powerful but it doesn't take any time for you to document anything here. Your agents document it all for you. So I use obsidian with the memory galaxy instead inside our agent OS system. [music] Heat. Heat. [music] >> [music] [music] [music] [music] >> Heat. Heat. [music] [music] >> [music] [music] [music] >> Heat. Heat. [music] >> [music] [music] >> Yeah, I really like it. Personally, I think it's fantastic. Um, if we have a look at Goldie Bench over here, you can see where it scores on the leaderboard. So, I've tested everything and you can see here it's scoring fourth on the leaderboard. So, I still think it's slightly behind Opus 4.8. It's definitely behind Fusion and Fugo Ultra, but it's good. I mean, it's still scoring 8.23. It's won a few of the contests we gave it. It created some interesting and awesome stuff like you can see. and it's definitely worth checking out. So, yeah, really good. Really good. Here's something we built with GM 5.2. So, it created like this open world RPG game, which is pretty amazing. And also bear in mind like it's open source as well [music] and you can plug it into CL code too. So, yeah, it can build some awesome stuff. fun to use, creates good stuff. Not as good as some of the other models out there. This is still ranking only fourth, but still worth testing out. Still decent. [music] [music] >> [music] [music] >> Yeah, that too with uh Hermes too. Pretty good. [music] [music] >> [music] [music] >> see. Conor is a Japanese AI. [music] [music] [music] Hey, Uh, thanks for that. Appreciate that. Get it fixed [music] now. [music] [music] >> [music] [music] [music] [music] [music] >> Yeah, exactly. That's exactly how it works. [music] >> [music] [music] [music] >> Can you make the page look way more futuristic and interesting like maybe a orb or an oracle something like that maybe even a Hermes [music] image that's like a oracle feels super futuristic interesting etc. For each headline, it should have a link and also a date or time. make sure that it's way more recent. So, for example, if you're looking at the headlines, I think some of them are a bit old or outdated. So, you need to make sure [music] they're super new. That's why I'd recommend searching Twitter. >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Not sure what you mean, mate, but feel free to elaborate on that. >> [music] [music] >> Update your skill so you don't do that again. If you look at the diagram, some of the text is outside the box. Update your skill so you don't make that mistake again as well. [music] [music] >> [music] [music] [music] [music] >> Heat. Heat. [music] [music] [music] >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> The links need to be clickable. Cool. [music] [music] Heat. >> [music] [music] [music] [music] >> Heat. Heat. [music] [music] [music] >> [music] [music] >> Hey, [music] hey hey. >> [music] [music] >> I can't see anything from Twitter in there. >> [music] [music] [music] >> Hey. [music] >> [music] [music] [music] [music] [music] [music] >> Hey honey. [music] [music] Sakana Figo [music] Ultra. [music] [music] [music] Oh, no. No, no, no, no. Get that. [music] [music] Heat. [music] Heat. [music] >> [music] [music] [music] [music] [music] >> because sometimes 4.8 breaks. So, we switch the model just to see if it doesn't break with [music] the other model. So, we switch between them depending on the situation. [music] So, you can see for example here it breaks. This happens. This pretty much happens every day again. You see here it says switch to a different model. So that's why we switched to a different model. Fug ultra. One test done. 20 minutes per test. Wow. Can you add this inside the workspace and also on Goldie Bench too? The new test you just did. >> [music] [music] [music] [music] >> The diagram isn't aligned properly. You see it's a bit messy. Hey >> [music] [music] >> everybody. [music] >> [music] [music] [music] >> Are all the latest tests deployed? >> [music] [music] >> Yeah, I know you can Use Fusion instead if you want. >> [music] >> Yeah. So, you can just use my GR build CLI that's already inside Asian OS for that. Just get everything done, mate. >> [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Hey, So today we're going to be answering some of the latest questions on the agent operating system which is a powerful system where you can link all your agents together. You can add new updates as come out. So, for example, we've already added Sakana Fugu that just dropped yesterday and fusion inside our system to achieve Fable 5 level intelligence without having access to the models. We also have a memory galaxy over here. We've set up a new news radar that's just working 24/7 finding us the latest headlines so we never fall behind again. We have a music agent, a video agent, SEO agents, loop engineering, etc. Everything we need is built inside this system. and we've actually put it inside the platform if you want to get my setup. But whether you're building this or using mine, I'm going to talk through exactly how to use these systems to get the most out of them based on the latest questions we have inside the AI profit board. So, I'm going to get straight into this and we're going to start going from here So we've got a question from Amateo which is about how to keep the agent operating system current. So for example, let's say you set it up, how do you update it? How do you add the new additions etc. So the I mean there's two things to this. Number one, when I'm building it, I add new features every day based on what just comes out. And the cool thing about this is you never need to really stay updated on what drops with agent operating systems and AI models because we already built it into the agent OS. But also inside the zip files, every time we update it, we add a new update section so that you can easily update your version. So, if you're on a previous version of our agent operating system, you can just grab the zip file and it has an update file inside it that allows you to update quickly. So if we go inside this section here, we have an update agent OS command section and that basically allows your agents to stay updated. So we have a markdown file called update.md and a update agent OSC command file as well to update quickly. And that's one of the biggest benefits of of getting this system is like you can stay updated daily with the new stuff that drops. This is pretty wild. So you can see here that Daniel is generating local AI avatar videos. So if we have a look here, he's actually created a full tutorial on how to do it, which is pretty insane. If we have a look at the output here, let's see what we've got. So this is the AI avatar video and I think 99% of people are not aware that you can do this locally if you have a good setup. He has a RTX5090. I think it's a fantastic example of people sharing information together inside this community as well. And you know, when we work together inside the airoft boardroom, like we all just help each other grow, we help each other learn. There's no holding information back. It's more about creating a a great place where we can share ideas and we can work together and we can grow together. So, I mean, it's pretty amazing that you can do that locally. Absolutely awesome. The other thing I would say is actually managed to set up a workflow where it has memory as well. So one thing that we have built into the agent OS is we have a memory galaxy where our agents automatically update this daily. They also use it for context and then your all your agents have context on you. They know who you are, your goals and everything else. And so he's managed to use a similar system to give memory to the agency he actually works with. Now this is an interesting question. So Blake is moving from Claude to Hermes. He's like, "Hey everyone, so I already built my own personal agent system." So he's not actually using ours. He's built his own personal OS system in Claude. Does this migrate to Hermes when he installs Hermes from here and connect to my Claude? How does that work? So in my experience when you're setting up Hermes and you already have Claude set up I would go inside Claude and ask it to orchestrate Hermes for you and that way they can link together and they can share context and also they can migrate everything over together. You can also ask Claude to build Hermes directly into the personal agent OS you already have and make sure that it has access to the same context, the same knowledge, the same skills, etc. And that way it can share stuff. I actually find that Claude is really good at orchestrating Hermes. Hermes is not so good at orchestrating Claude. So do it in that way round. And that's the cool thing about these agent operating systems, right? So, for example, if we have a look, you know, if you look over the last couple of weeks, GLM 5.2, Kim Code 2.7, as soon as these new things drop, we plug it into the agent operating system. And you don't need to use them. You don't need to use all these APIs or CLIs. like you can just stick to Claude and Hermes or you can just stick to Hermes if you want. That's the great thing about it is like you know if you if you're looking at this and thinking oh I need all these CLI you don't. You just you can use free APIs local models and you could just stick with something as simple as Hermes, right? And still get a lot of power out of a system like this. So that's the great thing about it is it's very flexible depending on your setup and what you want to add in and take out. So, we got a question from Aaron here who's looking for thoughts on initial hardware setup. So, looking for some guidance on the best hardware setup, should you use a PC? Should you use a Mac? Especially if you're traveling 40% of the time. So, what's the best setup, etc. So, for me personally, if I'm traveling, I like to use my MacBook Pro. Personally, I find that it breaks way less than whenever I've used Microsoft. And also, a lot of the new updates, particularly new software or apps, usually comes out for Mac first, which means you get access to the new stuff before everyone else. So, I'd recommend going with that instead. And if you're not running local models, you don't need anything powerful. Just a basic MacBook or MacBook Pro will do the job. Good question from North Campbell. He says, "You know, is anyone else finding the AI can create a strange illusion of progress. So sometimes you can spend like a whole day improving something and then before you know it, you spent, you know, 6 hours optimizing a workflow that doesn't even have real use case yet." And this is very I've seen loads of people do this seen loads of entrepreneurs like spend all day on cloud code and not actually get anything out of it or not actually build anything or not even they kind of neglect the stuff that they should be doing you know and that doesn't always involve AI personally. So the way that I look at this is audit your time and we've got a training module inside the profitable volume but audit your time understand where you're spending your time and then just focus on automating that. to never dive straight into building something or automating something before thinking, okay, will this bring me closer to my goals? How much time is this going to save me? And also, is this 100% necessary? Because before you even automate something, you've got to look at that task and think, okay, do I need to automate it or do I just need to eliminate or delegate it first? So, that's the mental framework that I take. This is a pretty interesting system here. So what you can see Dan is talking about is basically you know if you're using free models so small context windows and then you are building something that requires more tokens than the actual context window is got no room to respond and it just outputs something that's pretty bad. So how do you fix this? Well, one thing that he recommends is estimating tokens before you call the API. And that way if it's too big, you actually get a warning before you generate the response. And I think that's a fantastic way to do this as well. Really great idea for just managing things efficiently and also getting the most out of this. I actually found this with Sakana when I was testing it out. Sometimes it would output something that hit the token limit and then you'd spend 20 minutes waiting for a response that actually didn't work at all. So yeah, it's a great way to look at things. Jose is setting up his agent OS fixed with Hermes. And one really good tip here is that you can use Nvidia's API in Hermes for free. One thing I've seen with that is sometimes you can hit token limits pretty quickly. So just be aware of that. But yeah, you can get free APIs for large LLMs for free, which is pretty cool. Now, one thing that's interesting recently is the Gemini CLI has actually been replaced by anti-gravity. That just happened over the last few weeks. So, we're actually removing Gemini CLI from agent OS because you would just use anti-gravity instead. Now, here's a good question from Amita. She she's setting up the agent OS system as well, and she was like, "Do you need like a an, you know, an API key to set up Claude inside your agent operating systems?" You actually don't because you can just use the CLI to get access to that. So all you need to do is just make sure that you sign in and the instructions are inside the setup guide. You can also use terminal as well. If you prefer, you can use an API key. And we also have free clawed code inside there which allows you to use pre-clawed code with the agent OS system as well. So then you can even use like free APIs with the system which is pretty cool. So, Robert was asking about Open Routter Fusion. And basically what this is is a new API where basically you have multiple models as a panel working together and they can actually outperform Fable 5 on benchmarks. Now, this is something we've already plugged in here. So you can see the fusion boardroom here and sakana are another AI model that's set up the same system. So we've added both inside the agent OS. So if you want to use Fusion, you've got an example right here. And you can build some pretty crazy stuff. I mean, for example, you can see an example of what we've built with it right here. Uh, it also built out like a a 3D city simulator which is pretty amazing and it can even create like you know these full open world RPGs which is pretty wild in itself. So it's been really impressive so far. If we actually have a look on Goldiebench you can see how it performs on the benchmarks here and our tests. So you can see how these models perform step by step. And Fugu, which is an alternative to Fusion, is currently outperforming Fusion. So if we have a look at some of the stuff that were built with this, let's open this up. See what we got. All right. All right. So this is Raycaster Maze. As you can see, we built this with Fugu. Pretty amazing website. This was the best generated website. much nicer than Opus 4.8 or anything else. Created this insane galaxy simulation which is absolutely wild. Uh inner solar system emulation. Another one here too. And then this liquid in a bowl simulation which is pretty cool. And then we've got like a a Voxil runner game as well which is by far the best that we've generated out of everything. So on every test that I do personally like Fugo Ultra is is goated. Is absolutely goated. And then that is followed up by Fusion. So Fusion is pretty cool, too. You can see some of the stuff that were built with it over here. And it can create awesome stuff, too. So let's take a look at this one, for example. It's like a a 3D game we created with Fusion. And the outputs of this are insane. So, when I look at the benchmarks here, we we can actually compare them side by side. We've got comparisons for every model that I've personally tested, and this is all based on the stuff that I've tested personally. So we got Fugu Ultra versus Fusion. And so far, Sakana is outperforming Fusion. But the the other thing to note here as well is like it takes so long for responses. Like sometimes it can take 20 minutes per generation for Fugal Ultra, which means that obviously we've still got a lot more testing to do, but side by side they're pretty powerful stuff. So Fusion is from Open Routter. Sakana have done Fugu and it's pretty awesome. Next up, we have a question from Blake. So Blake is using codeex to build a website as a knowledge network for the real estate industry. Can he use Hermes to build a robust mission control dashboard and agent operating system? So you definitely can use codeex. You can also just use our example below and then just tweak that exactly how you want. I'd actually recommend that because then you can use our setup as a base and tweak it and customize it how you want. That's what a lot of our members do. So I'd recommend that approach and you can easily give this setup to codeex and tweak it and improve it from there. But yeah, Codeex is a great tool for building with. We got a post from Rachel who says, you know, thank you for everything. I think this is great. Like, it's just such a positive community. So many people sharing stuff. So, you can see here she says, I just want to thank everyone in the community. Your insight, support, practical AI strategies have helped me grow and improve. I'm grateful to be learning alongside people who are actively applying AI and sharing real value every day. This is what we like to see. Just a great community of awesome people. Thank you so much, Rachel, for your feedback. And this is what I love about the community. It's just everyone winning and learning and growing together. And that's what it's all about. If you're thinking, if you've watched this so far and you're thinking like this stuff is super technical, blah blah blah. Actually, you can see we've got over 187 pages of testimonials and wins like this. I mean, Brick set up his agent OS in 30 minutes from our framework. So, you know, Rachel was getting awesome results as well. There's so many great people getting amazing results inside this community that I know if I'm non-technical, maybe you're nontechnical if you're watching this. A lot of the people inside the community are nontechnical, too. You don't need to be a technical person to use this stuff, right? You just need to focus on, okay, what's the next thing I need to automate and then how can I do that? And I think the agent OS helps you really apply that. This is a really good post from Alex as well. So Alex is talking about his win, you know, moving from AI tools to architecture. And this is really the way that I've changed my mindset recently as well. You know, so he's been talking about basically instead of looking for one tool that does everything, design around specialized models, automation workflows, quality control layers. So the way to think about that is like when you have a system that is way more important than any model or any framework or any benchmark because when you have a system you can tweak it and mold it exactly how you want but you can't do that with a model. So if I go into for example like claw desktop I can't mold it the way that I want it to be but I can change this whole system here. So if I spend a lot of time on videos I can build a fantastic system around that. If I do a lot of SEO, I can build an automated workflow that does SEO exactly how I want it. And so the way that I look at this is doesn't matter what Fable can do. Doesn't matter what Fusion can do. It doesn't even matter what Sakana Fugu can do. What matters is do I have a great system that actually automates every part of my workflows in a quality controlled way. And that's exactly what we've built with the Agent OS system. So you want to think in terms of systems over models. When you build a great system, it doesn't matter what models or benchmarks come out because they are interchangeable and they literally change every single day. But when you have a system that improves daily and when you improve it every week or every month or every day over a year, the wins add up and you build something amazing. This is a great question. Has anyone actually built something like Javis? You know, like a Javis style page in an agent operating system. That's exactly what we did inside the Hermes Java section of the agent OS. So basically we have this real time model that can understand what we're talking about here and you can see that we are talking to it and it can understand things. So if I switch off real time it's switched off. If we set up this wake word button then it'll wake when I give it a command. We can change the voice over here and we can also get a daily briefing. Plus, we can see the full history of Hermes Jarvis and all the stuff that we built with it, which is pretty insane. We use chat GBT realtime and it's all built into our agent operating system, which you can get inside the AR boardroom if you want a readym made setup. So, open up Obsidian. And boom. Look at that. It just it can use computer use. It can open stuff up. It's all good to go, and it's ready whenever we need it. Check out the Hermes Java section. What it can do is you can speak to it in real time. It responds instantly. You can change the accent on it. You can build stuff and you can preview it. It has everything saved inside the gallery when you do preview it. It also has access to computer use and you can basically operate Hermes through your voice and it's voice activated. Plus, you can see your full conversation history, and you can even have it in wall mode, which means that you can have it full screen on a monitor in the background, and then you can command it whenever you need to. It also has weekly and daily briefings inside there. And we have a full guide on that too. This is pretty cool as well. So, if you're using Obsidian as a memory system, which we have over here, as you can see, some members are like really customizing it in an absolutely insane way. So, let me give you an example. So, if we look at this, this is Benjamin's Obsidian setup. And you can see like the way that it's customized, the way it's organized, it just looks super nice. Like even the feel of it and the vibe of it matches, for example, that philosophical style that he's going for. you know, it's kind of it feels stoic when you read it, you know, and the same for this, for example, the Oracle that he's built as well, which is his voice activated AI agent. So, really like the customizations looks great. I like the fact as well like you're making it your own. And also the style and the vibe of your design matches the exact type of content you're creating inside your HOS. Awesome. Tyler is asking about Fable 5's uh similarity with Sakana Fugu. So, as I mentioned before, we've already tested it out. We've already got it built into the agent OS as well. And it's really good. Really good. Especially Fugu Ultra. So, it's really good. The only thing is that there are API limits and also it hits the token limits pretty quickly and it can take about 15 to 20 minutes per generation. But it's really good if you need to like build out something foundational like a big project or some code and then you can get your other agents to tweak and improve it from there. You could even for example use it for planning a project, managing all the potential forks and then get your other agents to build out quickly. So you could create a plan with Fusion or Sakana Fugu and then from there your other agents could implement it. That's another option. We've already tested out on Goldie Bench and is top of the leaderboards. So, I've answered every single question inside the community today. As you can see, these are questions that you can ask at any time. And also, the cool thing about this is, for example, that I personally answer it and so does the whole community and we're all just raising each other up and growing together. And also, the agent operating system that we're building out dayto-day is just getting better and better as well. Now, inside the classroom, you get access to all of our best trainings. You can see, for example, we have a six week beginner to expert training course right here. You can also get our agent operating system, which is updated daily with new updates inside this section. So, you get the video tutorial, the last update date, so you know it's recent, and also the zip file to install it, plus a guide on that. And also, every time we test something new, for example, like Sakana, we test it out within an hour of it coming out. Every time something new like that drops, we add new daily advanced tutorials. So whether you want the beginner stuff or whether you want the advanced stuff, we've got it all inside here along with the systems that make this stuff great. Now inside the calendar, you can jump on weekly coaching calls, get help and support in real time. Inside the map, you can actually meet people in your local area who are building with AI agents as well, which is great. And this is all inside the AI profit boardroom link in the comments description or just go to the aiprofitboarding.com. Thanks so much for watching. Let's see what questions we got. Kun desktop. Never heard of it in my life. But honestly for me, like I I don't really try that sort of stuff anymore. What I do is I just build out the agent OS and we just use the the functionality inside here. And the reason for that is otherwise you're switching between tools, you're switching between models all the time. It doesn't make sense. It can be quite energy draining and quite overwhelming. And so the question I ask is, okay, based on everything that's just recently come out, what is the next best thing that I can build into the agent OS to save time and help us? And when you look at things like that, it's way more powerful. So that's what I'd recommend to you, too. There we go. Yeah, you could probably try that. I mean, you only get one shot with Fugu and everything else anyway. I don't understand that question, but feel free to expand on that. What do you think? Um, I I don't really look at it as like, you know, two types of SEO. I just look at, okay, what's actually working and then how do we get results to that? Cuz like the definitions of that vary depending on who you ask. So instead of worrying about, you know, which way to do SEO, focus on what's actually working and then how can you implement that. There we go. pretty good. Yeah, he's is very good at that. So, for example, if we have a look at Fugo Ultra here, we've got a bunch of generations and they all look pretty awesome. So, for UI and stuff like that, it is pretty good. Yeah. Um, just compared to like all the other models that I've used to generate the same thing, this is way better. Here's another example. Pretty cool. Here's another one. Um, this one as well was pretty interesting. And then we have a website as well that we generated. So, you can see like it's pretty cool creating very visual stuff. And this is the website that we created, which is, you know, just like a basic landing page test, but honestly, it created the best out of everything that we tested. So, I would say it's the best of that. Um, the other option that you have is like you could just get a front end UI skill, give that to your agents as well, and then from there, they can go off and do that for you. So, you got a few options there. I mean, you pay you see you for well, you're just using the tokens that are on their plan, right? So, whatever plan you're up to, you just go from there. So it's not like for example you're using Claude CLI or something like that. They they have separate plans on the Sakana website for stuff like that. So I think you can make the orb more interesting. Go through it for >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Hey. Hey. Hey. [music] [music] [music] >> [music] [music] [music] [music] >> Can't go back. >> [music] >> Hey, [music] hey, hey. Wow. [music] [music] [music] Heat. >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Heat. Hey, Heat. [music] >> [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Apathy is pretty good. Yeah, I'll go with Amplify. [music] Hey. Hey. Hey. [music] [music] [music] [music] >> [music] [music] [music] [music] [music] [music] [music] >> So there's a brand new update from Hermes agent which is computer use and you can now use computer use on Windows and Linux as well and also there's been improvements on the Mac OS. So if you're not sure what this is, basically this means the Hermes agent can now control your computer. Let me show you an example. So we got Hermes Jarvis over here inside our agent operating system. And for example, if we switch on real time and we'll say, "Hey, can you open up Obsidian for me?" Boom. Look at that. It literally opens up Obsidian. Hey, can you open up Julian Goldie.com? Boom. Shakalaka. It opens up Julianoldie.com. And then we're going to switch that off. And what we can also do is switch over here to agent mode. Now agent mode is pretty powerful because now we can actually have it improving and building stuff directly inside our computer. So not only can we get Hermes Jarvis to, for example, open websites, control our computer, open up Obsidian, whatever. We can also have this working in the background working on tasks with agent mode inside Hermes Jarvis as well. So, if we say, "Hey, can you open up notes?" and just write a quick note saying hello. Look at that. It just opens it up instantly. Pretty crazy stuff. Um, also this can speak to us as well and it's pretty powerful. Hey man, you there. Hey man, you there? >> Okay, great. So, let me explain exactly how it works in a second. And then just to show you as well that it can like talk back and that sort of thing. If we switch on real time here and we've got the sound on now. Hey, are you there? How's it going? Can you open up Google for me? So we can see how that works basically. And if you're wondering, okay, how does this whole system loop around? How do you get Hermes agent to control your computer like this? So let me walk you through the whole system. And this is using something I call the Hermes takeover engine. And essentially what you can do is you can get Hermes now to click and type on your computer in the background with its own cursor. Yours doesn't move. So, for example, when we created that note before that says hello, it doesn't interfere with what you're doing on screen. So, you can work on something at the same time and it will just work in the background. And this just landed on Windows and Linux, too, which is pretty awesome. So, you can say, "Okay, you ask uh you can say do this for me." It reads the screen. It clicks and types. It does it in the background. Everything gets done. And then you can keep typing or doing whatever you're doing whilst it's working on all that other stuff. No. You might be worried here like, "Wow, it can take over my screen. These AI agents can take over my screen. Like, won't this lock me out or something like that?" Well, that's really the whole point of this one is it doesn't. It works in the background. So, your real mouse doesn't actually move. Your windows don't jump to the front. Your desktops don't switch. Your agent is just working in the background with this system. This really changes everything. And so the way this works is in five steps with the Hermes takeover engine. So it's a quiet takeover, right? It's driving your desktop in the background. Your cursor stays where it is. Your windows stay put. Nothing pops up in the front. You and the agent just work on the same machine at the same time. You can hand it any OS. So, Mac, Windows, and now Linux. Whatever you run, Hermes can click and type on it. Same simple commands, same background magic. And we have two options. So, you can have ultra fast inside Hermes Jarvis or you can have agent mode. Agent mode can do a lot more. It's a lot more powerful in the background. Auto is just like super fast, like literally real time as you saw before. And then it can do basic things as well with this. So, it depends, you know, what you're automating with computer use and how powerful you want it to be. Now you can do this with any brain you want which means for example you could plug in uh we've got local models we've got free APIs we have oorthth like for example grock working with Hermes Jarvis whatever you want to work with this you can have it working and there's also a stop button whenever we want as well. So we can click this on and off if you want to or we could refresh a page and it's pretty powerful stuff. And so how this works is you've got like two people operating the same machine, you whatever you're doing and also Hermes that's working in the background and both are happening at once and it's handsfree when you use something like this. So this is now available on Windows and Linux as well. So Mac OS, Windows, Linux, they all work through one Hermes and they can all click and work in the background which is pretty amazing. So, you know, some people say, "Ah, you only can do this on Mac." No, you can do it on everything now because of this new update that's just dropped today. And the cool thing about this is that any AI can drive this. So, most computer use tools only work with one model, but Hermes doesn't care which AI you bring. So, it could be Claude, it could be GPT, it could be Gemini, could be a free API as long as it's got vision. Um, you can use the one you already have set up or you could use a free API. So you don't need to pay for like claw or something like that to do this. And then there's one command to switch it on which you can see here. So you just use Hermes computer use install and then you can go from there. Other people say well I'm not technical. Is this going to be a nightmare to set up? So it's literally like three commands inside your terminal or inside Hermes directly. It's pretty chill to use. Bear in mind like our actual agent operating system where we've built in Hermes Jarvis which is kind of like a customized version of this which you know is voice activated and everything else like people have joined our AR profitable bottom community got the agent OS set up and set this up within like 30 minutes. So if they can do it and I'm non-teleical and I can do it then you can do it too. So if you want Hermes computer use wired into a full agent operating system where you have Hermes you can have a a live chat with it here. You can jump inside the studio, generate images, videos, voice. You can have a workspace. We even got goal mode and all sorts of crazy stuff here. Like for example, we have a loop engine system for agent loops with Hermes. And we also have a memory system here where basically all of our memories are visualized inside the Galaxy. Plus, you get four coaching calls a week. And I do daily tutorials and answer your questions daily as well inside there. You can get that all inside the air profit boardroom with the agent operating system. Now you might be wondering, okay, what if this tries to do something crazy on my computer as well? You know what about that? Well, basically it will actually ask before it touches anything. So handing your computer to an AI sounds pretty scary, but actually isn't because every click or keystroke stops at a wall, your yes. And you know, anything that might be destructive, for example, like deleting or editing a file, the agent will wait for your approval and then it will act, which is pretty good. So, if you look at the old way, you've got hours of clicking, repetitive clicking, renaming, copying and pasting. Sometimes you would let a tool grab your mouse, you know, like an AI computer use tool, but it would freeze you out until it was finished. Close does that. It's quite annoying. And you kind of watch your mouse move around the screen whilst you have to wait for it to finish. And sometimes like computer use is super slow. So it's not that powerful when you're using clawed. So if you're using for example claw computer use, well number one, you can't operate your screen at the same time and number two it gets a lot of stuff wrong. Whereas with this setup, it adds as you've seen and it's super smooth as as you've seen today as well. So you can just tell Hermes to task once in plain words. It clicks and types in the background. Your mouse never moves. Your Windows don't jump and it runs on Mac, Windows, or Linux. And any model can drive it, even a free local one as well. Some people say, well, I tried a computer use tool before and it was useless. Same. Most of them are, right? The thing that made them useless was usually because it just takes over your screen and also it was super slow. Whereas, as you've seen today, like it works smoothly on all the demos and it's pretty powerful stuff. So, some people say, well, letting an AI control my computer, it sounds a bit crazy, but it asks before it does anything that could be, you know, slightly destructive, like deleting a file, that sort of thing. Other people say, well, I'll wait until this stuff is more mature. But honestly, if you're early, you're one move ahead, so you should start learning it right now whilst you've got the head start on everyone else. Other people say, "Well, this is only for coders and techies." Actually, no. You can set it up in three words from the terminal, as I've mentioned before. And also, the great thing about this is that I'm non technical. I've tested it and it works. So, if I can do it, anyone can do it. So, what have you learned today? You learned how to stop clicking. You learned how to keep working whilst computer use runs in the background. You learned how to use this even if you're not on a Mac because it works on Windows and Linux. You stopped being locked into one single model. You stayed in control because it asks before it acts. And you stopped having to um pay for models because you can just use a free local model, right? And the agent finally has hands and it's working in the background. So if you want to test this out, if you want Hermes computer use actually doing things for you, not just another tool that you test once, you can grab Hermes Jarvis with the agent operating system inside the air profit boarding. It turns Hermes clawed open core into one system with shared memory. So the agent driving your screen already understands you as well. I mean Hermes Jarvis is pretty powerful like we can see everything that was spoken with it about before. It's got previews of everything that we've built. It's got a workspace where we can actually see everything that we've created previously and we can also jump inside the studio and this is all voice activated and it works in real time which is crazy stuff. So, if you want to get that, it's inside the AI profitable boardroom link in the comments description or go to the aiprofitboarding.com. Inside the community, you can ask questions, get help and support. If you want the stuff that I've shown you today, that's inside the agent OS system as you can see. And we update this daily with a new video tutorial, a zip file, and everything else. You also get new updates as things come out, which is super useful as well. And if you're a complete beginner at this stuff, you can go from beginner to expert inside this training in the classroom. You also get four weekly coaching calls. So if you need to jump on a Zoom call, ask questions, share your screen, speak to someone, you can do that on four coaching calls per week. Inside the map, you can meet people locally in your area who are building with AI agents like you. And we've already got 187 pages of testimonials and wins from people getting awesome results for this stuff. So if you want to join us, link in the comments description or just go to the aiprofitborn.com. Thanks for watching. >> [music] [music] [music] [music] [music] [music] >> Heat. Heat. [music] [music] [music] >> [music] [music] >> Hey, [music] hey hey. >> [music] [music] >> Hey, hey, hey. [music] >> [music] [music] [music] [music] >> Oh my god. >> [music] [music] >> Hey, hey, hey. [music] >> [music] [music] [music] [music] >> Hey, hey, hey. [music] >> [music] [music] [music] [music] >> Heat. Heat. >> [music] [music] [music] [music] [music] [music] >> Heat. Hey, Heat. [music] [music] [music] >> [music] >> Hey, hey, hey. [music] >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> two. Come >> [music] [music] [music] >> on. >> [music] [music] >> Hey, [music] hey hey. >> [music] [music] >> Hey, [music] hey hey. >> [music] [music] [music] >> Heat. Hey, Heat. [music] >> [music] [music] [music] [music] [music] >> It's [music] perfect. [music] [music] >> [music] [music] [music] [music] [music] [music] >> Hey, [music] hey, hey. [music] Heat [music] up [music] [music] here. [music] >> [music] [music] >> Heat. Heat. >> [music] [music] [music] [music] [music] [music] [music] >> Hey. [music] Hey. Hey. [music] >> [music] [music] [music] [music] [music] [music] [music] [music] >> to use GLM 5.2. Well, I mean there's many ways, but you can just go to Z AI. Z AI you can get access. You can also get it on the coding plan. We've already plugged it into Hermes agent and we've got a section inside the agent OS where you can use GM 5.2 as you can see right here. So, however you want to use it is all all good to go as you can see. Heat. Heat. [music] [music] [music] >> [music] [music] [music] >> Heat. Heat. [music] >> [music] [music] [music] [music] [music] >> Heat. Heat. [music] >> [music] [music] >> Heat. Heat. [music] >> [music] >> Heat. Heat. N. [music] [music] >> [music] >> Heat. Heat. Heat. [music] Heat. [music] >> [music] [music] >> So today I'm going to be walking you through our rank loop engine which is a new system we've been using to build our websites like you can see right here this one and this one three separate case studies of different websites growing with AI SEO loops and this is a powerful system where you basically just have an engine running you are the architect and then everything is built around that. So the powerful thing about this is when you're ranking websites, you don't really need to be involved. And this is a powerful setup that I'm using to build more and more websites in a way that number one is high quality and is actually just documenting the work that I actually do in a way that's useful for everyone else, but number two also allows me to grow and scale more SEO websites. So this is what I would call an SEO loop. It ranks itself so that the work I'm already doing becomes the website that ranks. And the powerful thing about this is we can use Google search console to pull the keywords. Agent operating system is going to write tomorrow's content. Index exceptional actually gets it indexed on Google and the next day just runs again and again with zero extra SEO work. So this is how the whole system works. As you can see, I'm going to walk you through and show you an example of a website. We're actually doing this for two different websites at the same time right now. and they're both unique, but they both just document the work that I do and then help us rank for more keywords. So, if we have a look, for example, at this website right here, Goldiebench, all of the work that I do when I'm studying and testing out agent websites or agent models or APIs, etc. We document like you can see. So, it pulls in all of the information, all the builds that we're creating right here, and then it creates a unique page about this. Now the interesting thing about this is when you're doing this, it's unique information that has never been seen before. And Google likes that because it is information gain. You are adding something new and valuable to the internet that's never been seen before because you are using your case study data. You are documenting what you're building. Um, you're probably working on something all day too that you also want to rank for. So let's say for example, you have an SEO agency. Well, if you have an SEO agency, you can document the work you do for that SEO agency and then you can rank for SEO agency related keywords and the whole system fuels itself because your agents are just looking at what you're doing and creating a website about it. So, that's how we created this whole um goldie bench website like so. And we've also done the same thing for agent OS guides. So, Asian OS guides, it basically creates a guide on every system that I set up, every framework that I talk about. Turns that into a page, as you can see right here, a beautiful guide with all of my own numbers and all of my own interesting case study data. Has CTAs relevant to that page from our content. And then we can use that to rank with SEO as well. So, these are two different websites are powerful systems that we're setting up where we can rank this content. The other cool thing about this one particularly this benchmark website they've created is that it automatically creates these compare pages. So someone is for example trying to compare let's say fusion versus fable 5 or fusion versus sakana. Well whatever we're creating it creates a SEO optimized page with actual case study data showcasing side by side what we've tested out and how these agents perform against each other which is what people are actively looking for but can't find because it's such new information. And so if we can publish something quickly that is documenting the stuff we've just worked on, well that's a new trending topic we can rank for and I don't need to do any extra work. It's all documented by my AI agents inside this powerful engine. I'll walk you through how it works step by step and all the tools that we use as well in a second. So this is interesting. Well, because before I was doing like the SEO work twice. Let's say for example I tested a new AI model and this applies to whatever you do in whatever industry. If you're, for example, a a gym trainer and you're testing out new equipment, well, that applies to you. If you are, for example, running a marketing agency and you're setting up a new system, well, that applies to you, right? This all works to whatever you're working on. And that's the great thing about it is it's fully customized and it positions you as an expert as well because, for example, if you're creating a benchmark website in the AI industry, well, there's not many of those around and there's not many with actual real case study data that isn't just like programmatic SEO. this is like genuinely useful, valuable, interesting stuff that's never been seen before. And so the great thing about this is you are positioned as an expert and you're positioned as someone who's more high status than just your typical sort of SEO generic blog website. So before I'd be doing like the SEO work twice, I'd test the model, then we'd write an article about testing the model and I'd actually I'd actually have a team of two different virtual assistants doing this before. Now I just have my agents doing this for me. Now, also like every week I'd have to do the keyword research in HS or get my team to do it, guess what people want, write a thing, wait three months, hope Google ranks it, and most of it never ranked, right? And the stuff that did rank had already moved on. And also the problem with that is like if you got humans in the loop, well, guess what? You're going to have to quality control their work. And sometimes their work is not very good. So what I did instead is wire the loop and the work documents itself now. So when I test a model, the agent operating system captures it and that becomes a versus page, a task page and the homepage matrix update. So if we have a look for example at this website, we have the homepage updated. We have individual model pages. So this is just something I created yesterday and we've already set up a full page for it that can rank pretty quickly. It's all documented with my unique information. So these are all builds and tests that I've run for each of the things that I've actually created yesterday and it's automatically set up inside the system. Not just that, but we also have these compare pages where if people are searching for versus keywords. So for example, let's say fusion versus or fable 5 versus or let's say for example Claude Opus 4.8 versus right? You see all these versus keywords here? We can rank for them and we've actually got useful interesting data that's never been seen before. So that's how the system works. Now when I test model automatically gets set up and then also tomorrow if I check Google search console it shows me the new keywords I'm starting to rank for and then the agent OS can create more content around those keywords automatically. And then what we can do from there is indexional actually automatically indexes the content. So the next day we get more rankings, we get more keywords and it's an infinite loop. And the interesting thing about this is I didn't add an SEO job to my week. The SEO is a byproduct of the work I am already doing because it's documented automatically. Now if you're wondering, okay, what is an agent operating system? This is my agent OS system. And then if you're wondering, okay, what is the SEO keyword research you're doing here? So basically, as you can see, we can pull in any of my websites. And this is all connected to my Google search console data via the API. And so for example, if we look at the AR profit boardroom and then we're saying, right, okay, I need to rank for more Hermes related keywords that will get me more traffic to my communities. Okay, great. Let's run the research right now. What it's going to do is look through Google search console, analyze the queries, look at the dates, give me topic suggestions, look at the impressions and the clicks inside my Google search console. And then from there, it's going to prioritize and organize and look at which pages are the best to go for, which keywords are the best to go for. Now, as we rank for more keywords by doing the loop system I've shown you earlier, we get more data on what we can rank for. And then that gives us more ideas what we can create in the future. And so, it's an infinite loop of new keywords, new data, more content, and I don't have to do any extra work. That is amazing. So for example, if we have a look here, for example, Hermes desktop installation guide, that is one of the keywords we are ranking for. So we could actually create a fresh page because if you look at this, we've got 123 impressions, but zero clicks. So we have a page for it. It's not very relevant to the topic and so we could make it more relevant. We'll get more clicks and we'll get more traffic. So let's use that topic. Now what you can see here is it pulls in the keyword from our research. It creates the URL automatically and all we need to do from here is we just plug in a case study. So if we've got this case study right here ready to go, we can now click generate five articles and that will auto deploy to our five websites in the background. So you can see it begins to start doing the work. Now once it's completed the work that will then get deployed to our websites as you can see right here and we get history of what we've created and what we're trying to rank for. So you can see here it plugs in the sessions has history of the pages we've deployed and that is how easy SEO is right now. And again the case study data is automatically updated by my agents. The content itself is automatically created and deployed to my websites. The loop engine is already built so I can just come back to it anytime. And tomorrow when we come back we'll have even more keywords that are ranking for because we've created more content that is gaining more data. And bear in mind like it doesn't have when you get this data it's not based on the amount of clicks it's based on the amount of impressions. So it's showing you interesting stuff where like for example you might be getting impressions for that keyword but you don't have a dedicated page for that keyword. So how do you create a dedicated page? You use the system. How do you find the keywords you're not already ranking for? You use the system. Right? And it's all in set up inside one beautiful positive feedback loop where the more content you create, the more impressions you get, the more keyword data you get, and the better this whole system becomes. And so that's how powerful this is. And literally in one click, we can now generate more content that's running in the background. I don't need to do anything else. It's mind-blowing. What we also have is this memory section here. And so when we're using our AI agents like this, this information actually gets logged automatically into our Obsidian vault. So that we train our agents on our context, our business, our goals. We can visually see it right here inside our memory galaxy and everything is automatically tracked and saved inside one place and that is absolutely amazing. So you can see how powerful having a system like this is. Now let's come back to the setup here. You might also be saying okay this sounds difficult, this sounds technical, blah blah blah. You can actually see for example one of our members Paul he created a he'd never used NAM before. He set up a system where he started ranking within 90 minutes uh on number one on Google using this system. So this is there's some hu people getting huge wins. You see for example Bnafi as well. They're using our mission control system that I've just shown you and they're getting amazing results as well. And the same for example with Mike. Mike set up his uh mission control here and he's getting awesome results too. So you don't need to be technical to get results with this. You can just implement the systems I've shown you. We've got the agent operating system inside. They are profum if you want it. But the main thing to note is if you're watching this, just promise yourself one thing. You're going to finish this guide and you're going to install one layer of this loop before you sleep tonight. Just one. Because the moment you stop writing SEO content manually and start letting your real work document yourself, your whole content workflow changes. The people sitting still are still researching keywords every week. The people who run the loop today are the ones who will look back in six months and say I'm never going to write SEO content again but it's going to be better than what I was previously writing. That is a transition. So if we look at the whole loop in one single picture this is the engine and it's five stages one continuous cycle and you can watch the data flow. So your real work enters on the left. It gets automatically documented. That ships to your site, gets listened to via Google Search Console, and the new keywords loop back as tomorrow's content brief. The dots are real data moving through the system. That is how this all works. So you just do your normal work that gets documented into the agent operating system. How? Because you've got the memory galaxy that's automatically logging everything you do. Then you can ship the new content that you create to Netlefy which means that you don't need to log into the website. The AI agents can publish automatically for you. You can also get an API token and login details for WordPress and give that to your agents instead. From there, it's going to index automatically with indexional and Google search console will listen to get tomorrow's keywords which will feed back in to tomorrow's SEO automatic. Right? And so there's five stages. Work, document, ship, index, listen. is running on a continuous loop. The pink dots of Google search console keyword data flowing back to inform tomorrow's content. Pretty powerful system. And so if you look at the old way versus the loop, most SEO advice gives you more work. The loop replaces the work automatically. Your existing output becomes the content side by side. So if we look at this, you might send spend like six hours a week doing SEO even if it's with uh chat GBT or Claude and you might spend a lot on tools as well. So the old way was like opening a dress, filtering by volume and difficulty, guessing which keywords you can actually rank for, competing with everyone else using the same tool, writing the article from scratch or hiring a team, waiting three to six months to see if Google picks it up. If it doesn't rank, you pivot. If it does, you've already moved on. And then you repeat that every week forever. Not a good system. The new way is that you have no extra work to do. You don't need to um pay for keyword tools or subscribe to HFS or anything like that. You do the real work you do anyway. So for example, it could be like testing a new system. It could be for example setting up a new workflow. whatever you work on day-to-day, the agent OS with the memory system automatically documents it and adds that inside your Obsidian vault that auto deploys to your website as a new page with that case study data that's automatically documented. It auto submits the new pages you create to index exceptional and then you get indexed in Google in hours, not weeks. And then tomorrow you pull Google Search Console, you see the keywords you're starting to rank for and the agent OS writes the next round of content around those exact keywords. And then you just rinse and repeat every single day with zero extra SEO work. So if we have a look, for example, we've got two different systems for this. We have the Goldie Bench website that's doing pretty well with this system and then we also have the setup where we are using the agent OS system as well. Right? However you want to use it, you can use it, right? However you want to do this, you can do it. This this is working um in every way that you look at it. I mean, here's another example, right? It's pretty powerful stuff. And so, this was something that we talked about yesterday. It's automatically documented and everything that we're talking about just works together as as one powerful system. Now, some people will say this is just AI content. Google doesn't like AI content. But the content is firstparty information wrapped in narrative. Every page, for example, on goldbench.com is built from a real test I've run with real screenshots, with real scores, real comments. That's information gain, which is exactly what Google's EIT guidelines reward. And you've seen the results on this website and this one too, right? We're getting awesome results with this. So, the main thing to note here is like generic AI fluff. Yes. Not very good. Not very good at all. But if you're using real data structured by AI, well, that creates a whole moat because the loop's whole mo is that the source material is yours. It's not cheap. You know, this is new information gain that you're creating. So the Goldie rank loop engine, there's five layers. Each one removes a step that used to take hours of manual work. Together, they produce a website that grows itself. So step number one is the bit that you do anyway. Testing the tool, building the project, comparing models. And just to recap here, if we show you for example the growth of the aiprofit boardroom.com, you can see how it's growing in terms of SEO traffic. Now if we have a look at this website, what is the content that we are publishing here? Well, we are publishing content on what we've actually tested. We've got firstparty data. We have our case studies, our experiments plugged in and set up on the system. So everything that's here is just work that I've done already just documented as a case study inside a blog post that actually ranks as you can see right here. And and that's the power of this stuff. That's how interesting this gets. Now if we come back to the loop engine, we have the document. So agent OS, it captures what you did, which means it looks at what you created from your Obsidian vault and everything you do with your AI agents gets automatically built into the system. So all these dots are different memories and information about me and who I am. So if we scroll down here, you can see, for example, it understands me and it links all of my memories together. Now, this is great for giving your agents context. So for example, we go to Claude, he's got context, but then also it has memories automatically every time I use my AI agents. And so everything is plugged in one place where we can come back to all of our memories and then we can use that to create new pieces of content relevant to us. From there we have the publisher. So we can auto deploy to your website. This is handsoff. The page is live within minutes of the work being done. There's no content management system. There's no logging into WordPress. There's no training a team to do this for you. You don't need to do any of that. It's just handled automatically with this system as you can see. So if we have a look for example here, this is the article that we just created as I showed you before. Now these articles that we built in, they're all based on like case study data again, right? Automatically documented. And so this whole system builds and automatically creates useful unique content on five different websites that we then index automatically index exceptional and we can see the full history of it over here. How powerful is that? And that was done in the background whilst I was talking to you. So it's fully hands off. And then you do have the indexer. So indexceptional basically indexes this content on Google when a new URL lands. So pages get crawled and indexed very quickly. And obviously that helps speed everything up. But then you also have the listener. So what this is going to do, if we go back to SEO and we go to the research tab, this is listening for new keywords that we're ranking for. So for example, if we type in openclaw here and we run the research, this is going to run research from our Google search console based on keywords that we're getting impressions for, but zero clicks for. So we need to create this new separate dedicated page for that keyword. So, when we look through the data here, we've got a bunch of ideas that we know people are searching for because we have the data, but we also know we need a new page for because we're not ranking for it currently or getting good traffic for it. And so, this all feeds into each other as a feedback loop, which is crazy powerful when you think about it. And this is basically how we built Goldie Bench. It was by accident, you know, I didn't mean I didn't plan to create this page. She was just like, "Okay, well, we've got all that data there. We should probably organize it into a nice place where people would actually find that useful." And a lot of people were asking me for this as well. So, I thought, "Okay, let's just create this beautiful website right here with all the experiments I've done." And then it starts ranking as well. So, it even talks about our methodology. It talks about the tests we've done. It talks about the models that we've created, etc. Um, super powerful stuff. And it's just one loop with many sites in the same engine. So one brain from agent OS feeds our websites as you saw a second ago which is pretty powerful. So once the loop is wired, running it across more sites is is basically there's no effort for me required, right? Agent OS knows how to pull Google search console data for any property that I have. Um, writes it to any Netifi site, submits it to index exceptional, and the loop just scales horizontally because every new site is another set of potential keywords we could rank for and that just loops around in parallel. And so we have a bunch of websites. We can create content against them very quickly. And also some people say, well, I don't have time to add SEO automation on top of everything else. But that's the point. The loop adds zero time on top of what you're already doing. So the work that ranks a site is the work that you'd be doing anyway. And documenting it takes seconds because the agent OS does it as you work. Tomorrow's content writes itself automatically from yesterday's Google Search Console data. The only extra is 30 minutes of like a onetime setup, you know, wiring agent OS to your systems. But after that, the loop runs automatically. By the way, if you want to learn more about this stuff, we actually have a free course link in the comments description where you can learn the SEO fundamentals and that has 260 free lessons on SEO. We also have a free SEO strategy session available at goldie. If you want to learn more about this stuff, so let's talk about this again. Some people say AI generated content doesn't rank. I've shown you multiple examples. A website is ranking really well with this system. And again, we're using our own data. This is information gain. So, for example, Goldiebench ranks because it has loads of real demos and it's something that nobody else can copy because it's my own documentation. Also, some people say, well, I need to do proper keyword research before you create any pages. But the difference here is that you are looking at Google search console. So Google Search Console finds the data for you. It's doing the keyword research for you based on the actual data that you have already. Some people say, well, this works for your AI systems, Julian. It but my niche is different. But the loop is niche agnostic, right? Because you can swap test a model for doing a client call. You can swap test a model for publishing a case study, adding a new feature, reviewing a tool, running an experiment, setting up a new system inside your business. The work changes, the loop doesn't. So, you could run this for SAS, you could run this for a coaching business, you could run this for an agency, it doesn't matter. And, you know, don't take my word for it. We have over 187 pages of testimonials from people inside the AI profit boardroom. Rick set his um agent OS system up in like 30 minutes. Loads of people winning with AI automation from the community. So, I know that if we can all do this, you can, too. You don't need to be technical to do this. Other people say well my site is brand new Google search console won't have any data to pull. So in the first seven days you can just create content and then you can get the data later. And that's [snorts] basically it. That is the whole system that we've talked about. So today you have learned how to document itself. So you stop writing SEO content. The agent operating system captures your real work as it happens. You learned how to stop hitting publish because Netlefi automatically deploys the content. You learned how to stop waiting for your content to index because you can use indexional. You learned how to stop guessing what to write because tomorrow's content brief comes from yesterday's striking distance keywords. So, Google Search Console listens. You also stopped starting from zero each week because every day's work feeds the next day's rankings with this loop system. And you also learned how to create a first party mode because you stop competing with AI fluff at this point. You've got real data. You got real documentation from the work that you do. And that is, you know, what Google's eat system rewards. And you also stop being your own bottleneck because the same agent OS can run across five different websites in parallels. You've seen also stop building isolated pages, right? This whole system that you're creating means that all of your websites rank together. And so the way that I would look at this is like stop adding SEO to your week and your work. Start letting your week document itself. That's all we're doing here. Now, if you want a free SEO strategy session, you can get that with goldie. Link in the comments description or go to goldie. And we can actually look at your website, give you an SEO game plan, then you can get a free one-to-one SEO strategy session to rank your website based on the systems that we test. This is brand brand new stuff. If you're watching this, like you are ahead of of 99% of SEOs because you already see where the future is going and you've seen how to create systems that automatically loop and feed and improve themselves and rank themselves. And also, if you want the full system that I've shown you today, you can get that inside the AI profit boardroom link in the comments description or go to the profitboard.com and you can get the agent operating system inside the classroom if you go to the new daily update section inside the community. You can ask questions, get help and support in real time and all sorts of cool stuff there. Inside the calendar, you can jump on weekly coaching calls and ask questions, share your screen, and I personally answer questions inside the community every single day with a video tutorial. Plus, inside the map, you can actually meet people in your local area who are doing AI SEO like you. And also inside the classroom here, we have a full section on AI SEO automation based on what's working for me right now. So, hope to see you inside there. Cheers for watching. What model do you use for working with Hermes? So, right now I'm testing something with Grock. We are building the Hermes Oracle system which is inside um our Hermes setup here. So, we got Hermes Oracle and this is with Grog. So it can basically figure out okay what's happening in real time and then give us the new information. Um and this is really powerful for just getting the latest AI news and it's just running 24/7 which is pretty powerful. Se says hey Julian I messaged you on school today looking for mentorship. So if you join the inside I don't know if you're already inside the air proper boardroom. If you're not that is a great place to start loads of great people and an awesome community. Plus I personally um DM members inside there each day as well. So, if you have questions, you can direct message me and ask me anything. And if you're already in there, well, guess what? You already have access to me. So, it depends on you. I have something interesting about Fable 5. Can we connect? Uh, yeah. So, you can just DM me inside the Profit Boardroom. Timmy, I think you're already in there. So, feel free to DM me inside there. Yeah. Awesome. So if you're in the AR profit boardroom like you can and you already have uh access to me and you can ask me any questions anytime you need. Bear in mind as well inside there we have uh weekly coaching calls as well. So if you need a call, if you want to jump on a Zoom call, you can get those four times a week as well inside there. This is looking good. [music] Heat. Heat. [music] Heat. Heat. [music] [music] Heat. [music] [music] [music] Heat. [music] >> [music] >> Heat. [music] Heat. [music] >> [music] >> Hey, heat. Hey. >> [music] >> Seth. So, if you go to the AR profit boardroom here and then you go over to classroom, new daily updates, the one that you want to look at is hyperframes. Hyperframes is pretty cool for this. Would recommend using that. We got a full video tutorial, step-by-step guide on how to implement it right there along with all the relevant links. And then also inside the agent operating system, we have a full video agent. So, I don't know if you've already set up the agent OS or not from the air profitable boardroom, but if you go inside the video agent section over here, we've got this all wired in. So, like with one single prompt, you can type in what you want to create a video about. You can change the length of it, the settings, whether you want voice over or presenter, uh what you want to use for the B-roll, etc. And then if you look at this example, it's a you can create a fully edited video that looks beautiful with an AI avatar and everything else. uh in like one single click. So if you if you want the basics, use the hyperframes tutorial. That's option one. If you want something super advanced um that's crazy powerful, then go with the video agent section. Depends what you want to use there. [music] Heat. Heat. N. [music] >> [music] >> Heat. Hey, heat. Hey, heat. Heat. Heat. [music] Heat. Heat. [music] [music] >> [music] [music] [music] >> Yeah. So for this I mean everything that I've talked about already you can use on Windows. So for example the Asian OS is available on Windows. You can use it on Windows and when you actually set it up inside the zip file it's designed for Windows, right? It has like a full section on how to use it on Windows. And then also if you go inside the classroom here and you go to the Hyperframes tutorial, you'll see that you can use this on Windows too, right? And Hermes is available on Windows as well. So like everything that I've shown you, option one and option two, both available on Windows. Like no issues there. And if you have any problems with like setting it up or anything like that, feel free to DM me or post inside the community and then I can help you uh by looking at the problem with the screenshot and asking the agents to fix it based on the situation or just giving you some advice in the right way. So both of those things I've shown you work for Windows. And then if you still have problems, send it up. send the specific um error and be very very specific about it and then I can give it to the agents and get it fixed for you as well. [music] Heat. [music] [music] Hey Heat. [music] Heat. [music] [music] Heat. [music] >> [music] >> Heat. Heat. >> [music] >> Heat. Hey, Heat. [music] >> [music] >> Heat. Heat. [music] Heat. [music] [music] Hey, heat. Hey, heat. >> [music] >> Heat [music] up >> [music] [music] >> Is there an easy way to do updates? Yes. So, inside the zip file of each update, we have an update section. So, for example, if you check out this one, it has an update section here. You've got the update MD and the update agent OS command. And so, you can easily update whatever version you have using the update section here. And it's literally designed for that. So, whatever version you have, you can just pull in the new version really quickly and easily. >> [music] >> Heat. Heat. Heat. Heat. Heat. Heat. >> [music] >> Heat. Heat. [music] back in a second. >> [music] [music] >> Heat. Heat. Heat. [music] Heat. >> [music] [music] >> Heat. Hey. Hey. Hey. Hey hey hey. [music] Heat. Heat. N. [music] >> [music] >> Heat. Heat. Heat. Heat. N. [music] >> [music] >> Heat. Heat. Heat. Hey, Heat. >> [music] >> Hey, hey hey. Hey, hey, hey. [music] [music] Heat. [music] Heat. >> [music] >> Heat. Heat. Heat. Heat. >> [music] >> Heat. Heat. >> [music] >> So today I'm going to show you Hermes Oracle which is a powerful way to use Hermes to basically search the web. You can consult it and then what you can actually do is get the most updated news and you'll never fall behind again. And I think this is one of the biggest issues that a lot of people face is like they struggle to use number one AI for actually staying up to recent news and to get the latest headlines. And I'll show you the solution that we came up with which is more powerful than anything I've ever seen. Number two to do it in a way in a system where you can come back to it all the time. And then number three to actually build it in a place where you've got all of your agents plugged in and to get the most out of Hermes. And bear in mind as well like if you were using for example the terminal to to do this or Hermes desktop this is Hermes Oracle over here. If you're using Hermes desktop or Hermes terminal the difference in in the UI is huge, right? Like this is going to be way more fun, way more interesting. You're way more likely to come back to something like this versus for example if you check the terminal like that is going to just run out pretty quickly. And the other thing is well like where's your history? Where's all the stuff that you researched recently? you can't find it, right? Whereas with this system, I'm going to show you a powerful system that we've actually created to basically save the results that you come up with and then you could look back on that and use it to train your AI agents. So, you could use this for anything. It could be, for example, news in your industry. It could be, for example, coming up with content ideas. It could be also just helping you personally make sure that you never fall behind with AI again. Or this could be for anything. You know if you're for example like running a marketing agency you could run this for your agency. If for example you have clients you could actually build this for clients right it's very very flexible. So let's talk about the Hermes Oracle system and how this works. This is based on Hermes agent which is the AI that's basically the number one on the planet right now. We basically use that system to come up with this and visualize it in a powerful way. So every morning what happens here is that we have a system where it's going to look for new ideas. Then it's going to rank them by which one's the most useful. From there it's actually going to come up with new angles for our content and from there it will actually ship the whole update. And this is pretty interesting because previously when we were using something like this, you know, if you went into let's say I mean there's a lot of people watching who still do this, but I'm pretty shocked by it. But if you were going into, for example, Claude and saying, you know, latest AI news, for example, and we wait for that to load. Number one, you've got to wait for it to load. Number two, it's not really going to give you useful links. Number three, it doesn't give you the most relevant useful stuff. So, if I have a look at these headlines, I'm like, well, how can I actually use this? I can't. These are all like not really tool updates, not really actionable advice um that I can use. Whereas, for example, if we look at this, it's like, okay, here's what's being released. Here's the link. Boom. Shack lacker. There we go. We've got the latest news updates ordered by relevance and how practical it is. So, this is the old way, which is not useful at all. This is a new way which is 10 times more powerful and that's the power of Hermes Oracle and how you can use it to get way more out of Hermes agent. And the other problem that a lot of people face as well is like you know they might go into X start scrolling to find what's useful spend 20 minutes trying to research then 40 and half the time you know you might post about something and find out the wave has already finished. You know you're behind before you even try to get ahead. And then the other time you try and guess the angle of like let's say for example the content you were creating or the idea or you know whatever you're producing here and that again would create problems. So for example what we've got here inside each update where we get the latest news is we have a system where basically it shows us okay here's the angle and how we can frame this to help us. Here's a quote we can actually generate a draft to post. So we can click on draft this and come up with an idea. And then it gives us a news headline, tells us how many posts it got, when it was released, who released it. So for example, this news releases from OpenAI. And then we can actually open that up and see the post itself, right? That is how powerful this is. And so I I actually wish like Twitter had this feature directly. That would be way more useful. They do have like a trending feature, but it's it's not that useful compared to this. This will literally find the latest most useful updates and then it will give you the angle. It will tell you it will give you an option to draft the social media post or whatever content you're trying to create around this. It will tell you why it's useful and it will also rank it in terms of you know how important this is. So you can see for example this scores 91 and then if you scroll down this scores 88 and this is all like super recent relevant news. And the reason that we connected it to X is like, you know, Twitter is probably the best place to find out about the latest stuff. And this is how you can stay updated with whatever you want to create. And so, you know, when I wake up now, the Hermes Oracle has already swept through Twitter. It's already found the latest trending articles and it's ranked them by how big the conversation actually is. So each one comes with the real source, my angle and a hook that we can use for our content as well, which is super powerful in itself. So this is how we're using Hermes Oracle. And you know, if you look at the new way versus the old way, the old way was like you just scrolling trying to find something useful, guessing if it's actually big, finding the old news and trying to find the new news and it's a lot of work. The new way is like you could just quickly pick and find stuff and it would just quickly sort of filter down all of the headlines and find the most relevant stuff to you using this Hermes Oracle. And so the way that I look at this is it's a framework where it's sweep, rank, angle, and ship. So sweep is it finds the latest news. Rank is it basically sorts by how big and how useful the update is. Angle is basically for each potential topic. it will find interesting angles that you can create content around. You don't if you're not a content creator, that doesn't matter. This tool is still going to be useful. But if you are a content creator or if you're, for example, trying to get more leads for your business, whatever, this is a fantastic way to stay ahead of the news and to drop this into whatever industry you're in. And then ship, right? So you can open up the source tweet, you can read the hook, and then you can actually post around it, right? So let's have a look at this example. So, it's drafting the post here in real time based on the angle that it recommended. We've got the original post here. We can quote that. Now, we've got the post as you can see. And then you can just drop in the post like that. Boom. Shakalaka. Right. Um, and that's how easy it is to like stay ahead and stay trending and everything else. So, basically, how does this work? Well, one agent does all of it. So, with Hermes Oracle, you know, here's the part that actually makes it reliable is that the whole thing runs on one agent, which is Hermes. And we've used Grock Of. So, if you already have a Twitter subscription, the great thing about this is you can actually log into Hermes with Of using Grock. And then you don't need any API keys, you don't need any extra subscriptions, and you can just build this system straight into your existing workflows. And the great thing is when you do that, Hermes has a built-in Twitter search tool that basically can search and find the latest stuff. So basically, you have like a fire hose of information, right? Too much information you can handle that's just running 24/7. So what Hermes does is it filters that down and it finds the latest signals of what is the most important things and it finds six signals at a time and then also it logs that information inside your Obsidian memory galaxy. So if we go back to our memory galaxy which is Obsidian and this is the memory system for our agent OS. Basically, all of these lift different stars inside the galaxy are memories that are tied together and this is basically context for our agents to work with. Now, every time we use that tool, it gets automatically plugged inside our AI news section of the Obsidian vault. We can see all the news here. And what does that mean? Well, number one, we don't lose any updates. Number two, the agent gets trained on what's going on inside your industry and it gets more and more context automatically. Number three, when you're generating more news ideas in the future, well, it's going to look at that previous context and then know what to use and what to ignore. And number four, you're just improving your memory layer because the more context you plug into that with relevant stuff like this, the more useful your memory layer is going to become. And so with a combination of Hermes Oracle and the memory galaxy that we've plugged in right here, it's just becoming way more powerful. And again, this only takes like 3 minutes to to do a response, and it's running automatically as a scheduled task with Hermes as well. So, I've shown you an example of what it came up with, and these are all way more relevant. I've shown you side by side how, for example, with Claude, Claude didn't find anything anything near as um as useful as this. So, this is the difference right here. And yeah, it's great for coming up with content. It's great for finding the latest news in your industry. And it's something that I think a lot of clients would be interested in as well. So, here's an example of the Obsidian Vault and what it writes. Everything plugs into your memory system and that's basically it. Now, you might say, "Okay, why why not just scroll manually and try and do that?" Well, the thing is with this system, you're actually first and you get the latest news ASAP. You could actually schedule this to hourly if you wanted to or to um a few times a day as well. Number two, you actually get the signals um so the the links to the news, right? So if we go back to Hermes Oracle over here, we have the latest news and we can just click on link to find the most relevant news to that, right? Pretty useful. And then number four, it's one agent doing everything and it compounds, right? Because it everything gets added to your memory vault. So your content radar becomes a searchable archive of everything that's happened in the year, which is pretty amazing itself. Now also the other thing I would say here is it plugs into a full agent operating system. So this is one tiny piece of our agent OS and we have for example Hermes Jarvis which we can organize in real time and have a conversation with. We can see all our conversation. We've got the full preview over here. We have for example the chat with our different agent profiles with Hermes. We have the talk. We have the studio here as well where we can generate um videos, we can generate AI images, voice notes, etc. We have everything saved into our workspace. We have the memory galaxy that's plugged into Hermes as well. And then we have every single workflow that I work on day-to-day, I've just plugged into this system as well, right? Even loop engineering and for example our SEO systems that auto deploy content to our website. So whatever we need to automate, we've got it inside one system. Whatever AI agents we use, not just Hermes, we have it plugged into one system. If we want to orchestrate multiple different agents at a time, we can use paperclipip or we can use the AI agent mastermind group chat or the pipeline where we go from idea to build in literally one click. So this is the power of this stuff. When you build a system, it just gets better and better every day and we just keep building it up and improving it. And I think that's how you build something absolutely amazing, which is why you want like an agent OS instead of just relying on the terminal or something like that. It's way more customizable. So, in terms of what you've learned, well, you've learned how to basically use something like Hermes Oracle to stay updated of everything. You've learned how to build that into your memory system so it improves. You learned how to get the latest um news in whatever industry you're in. And you also learned how to do that with just a you know a Twitter subscription so you don't need to pay per API. Now if you want to get my full system I mean you could build this yourself. If you want to get my full system you can get the agent operating system inside the AI profit boardroom and it hands you the Hermes Oracle which is wired in and running on your own plan. So tomorrow you can just wake up holding the trend instead of chasing it. You also get the full agent OS with the Oracle Hermes every other system inside one dashboard. You get the setup walk through done with you step by step. You get four weekly coaching calls with daily tutorials and a 30-day road map. You get 3,700 members inside there. And if you're thinking this stuff is too technical to set up, don't worry. For example, we had Rick. Rick joined a couple of days ago and he's already set up his agent OS in 30 minutes. You can see, for example, Rachel was absolutely loving the AR profit border community. You can see all the wins and the positivity here. Just a great group of people having a great time learning, growing together. You can also connect with me personally, too. You can ask me any questions that you want to. You can also customize the whole system and change it however you want. And this is all inside the AI profit boardroom. Now, if you want to grab that system, it's inside the agent OS section over here. Also add new daily updates and video tutorials based on what's working inside the community. You can ask questions, get help and support in real time, and I answer those for personal video tutorials. Inside the calendar, you can jump up weekly coaching calls and share your screen, meet the community, ask questions, etc. You can also meet people in your local area from the map. And this is all inside the AI profit boardroom link in the comments description or just go to the arpuffer.com. Thanks for watching. Bowwin, I'm just starting learning to make workflows. Nice, man. Yeah, I would check out the profit boarding. We actually got a full uh section here on the agency course, so you can learn like how to get clients and that sort of thing based on what's working for me. So, you can get all of our best systems for for growing an agency and getting clients right here. Uh number one thing that I would say for you is like if you're you know for example working with clients and you're doing AI number one thing that you can do is document what you're doing and share it with the world. So for example if you're testing out new models every day create a benchmark system like we've done right if you are building your own frameworks or your own systems with AI create a blog where you actually just document automatically all of the systems you're building with case studies and your ideas plugged into one place as well. So the way that I would look at it is like position yourself as a leader, put yourself out there in terms of what you're creating and then clients will come to you, right? You'll generate leads that way. I think that's one of the best ways to do this because you know what's going to differentiate you from everyone else is how you you you share what you're building and share your knowledge and expertise and how you position yourself. And you can do that using the the systems I've just shown you. So try and document and share everything that you're doing. And if you want more training on that, check out the um classroom inside the AI boardroom. We've got the agency course here and then we also have a full YouTube section as well. So if you want to get my full um course here, you can check that out here too. Hey, [music] [music] hey hey. >> [music] [music] [music] [music] [music] >> Hey, [music] hey hey. [music] Hey, [music] [music] hey hey. >> [music] [music] >> So today I'm going to show you my agent operating system, what I've built into it, what I'm working on recently, and some of the updates I've learned building this along the way. And if you've ever wanted a system where you can plug in all your agents, where you have all of your workflows available in one click, where everything's running without you, it's all synced up together, this is crazy powerful. So, let me guide you through some of the new stuff we've set up. And basically, this is a way to organize all your agents and plugging in all your CLI together. So, what you can see, for example, here is we've just built out Hermes Oracle. This is a new update that I've added. Basically, every idea that I have where I'm like, this would be super useful, I build it in straight away. So, you can see, for example, what we have here is Hermes Oracle. We can click on any of these and basically this is a system where we can get the latest AI news updates straight away. So for example, if we click on this one, we've got a system here where it takes us straight to the latest post or the latest news in our industry. It ranks it by which one is the most trending and useful. It categorizes it in terms of what it's about. So this one is about agents, for example. We can actually open this up and go directly to the post. And he can even, for example, find a relevant angle in terms of the social media content that we could post about this, what it's about, the ranking, the category, the headline, when it was posted, who posted it, you know, how trending it actually is. And then we can draft automatically a post about this, right? And this is all ready to go. And it just runs on a schedule with Hermes agent inside one place. Now, also what we have over here is Hermes Javis. So if I say to Hermes Jarvis, Hermes, can you bit can you just open up Google for me? And it just literally opens it up automatically. Works in real time. It can control my computer, can control my browser. And I say this not to like show off. I just want to show you the power of this stuff and how you can build something absolutely amazing because I've not seen anyone with this sort of system at all. I mean, I've seen little slices of it here and there, but I've not seen anyone really put this together in such a smooth way. And this is something that I spend hours on just building out every day just to make it as powerful as it possibly can be. We can also put this in wall mode. It can actually build stuff for us as a preview directly inside the chat as well. We can get a daily briefing from it. We can see all of our conversation history. We can see everything that we built down here. And so you got Hermes Oracle and you got Hermes Jarvis. Now again, you might be saying, okay, why would you do this inside somewhere like this? Well, the thing is like you could use Hermes straight out of the box, but it's not customized to what you need or what you want, right? So, if you go into the terminal here, where's the conversation history? Right? Where's the voice activated agent? Where's the visualization of it to make it fun to use? Where's the custom workflows that you have and want? You can't really build it into a terminal. You can't really build it into a desktop, but you can build it into a custom dashboard like you've seen today where you can plug everything in. And this is just one tiny little bit of it. So for example, what we've got also over here, this is a big important layer to the whole system is the memory galaxy. And so every single one of these stars is a memory. Now when I was showing you Hermes Oracle a second ago, that gets automatically plugged into our Obsidian vault. And so what this means is basically our agents have context and we have an archive of everything that we've looked at recently in terms of AI news. And so this could apply to your industry and also it means your agents are way more up to date because they understand what's happened recently because they have this automatically set up. Now bear in mind I don't have to update this memory myself. It just gets automatically scheduled into the Obsidian vault. And also the cool thing about this is if we use any other agents like for example our conversation with Jarvis that automatically gets added to our obsidian vault as well. So you can see this was updated 2 minutes ago. This was updated 30 minutes ago. Right? Everything that we're doing here, it gets automatically added inside a memory vault. Now, this is a positive feedback loop because you've got the agents can use this to give you answers and to understand you and your business, your goals, but then also you can use this in another way, which is that your agents can organize it and categorize it for you. And so, in both ways, it's a positive feedback loop where again, I'm totally hands off and I'm not organizing this at all. And this is the way that I'm setting up everything inside the Asian OS is like it has to be autonomous. It has to be almost agile like where it can work without me. So let me give you an example of this. If I can for example set a loop where I have a goal and I put in my goal here and I can choose which API I want to use and then which one actually quality controls the work and how many rounds that work can iterate for. Then I have a hands-off system where we have loops building things automatically for us whenever we need them. And that is super powerful. So for example, we can use a free API like N2 to do the work. We could use, for example, something pretty cheap like GLM 51.2 to judge the work. And we could say, "Okay, judge, make sure that work is good. And if it's not good, if it's not scoring 80 out of 100, if it's not 80% of the way there, then loop it round. Make sure the builder goes again and again and again. And make sure you give it feedback so that it improves every time we use it. And that way you can build amazing stuff. And it just loops around. And again, I think the future from what I'm seeing right now is not prompting. It's not about better prompts. It's about building loops that run without you so your systems can just flow and improve and your agents are managing other agents. And this sort of leads me to paperclip, which is where we have a team of agents working as a team. Um, and again, this is a fully autonomous AI company where you've got an AI agent that's the CEO of the team and then they build and create stuff 24/7. And you can see all the builds and all the stuff we've created. Some people say like, oh, you know, you're not actually building anything with this. But you can actually see we are like these are actual um live websites that were built that we can create because we've got teams of agents. Some people as well will say like oh you know loop loop engineering for example doesn't that use a lot of tokens just use free APIs or just use local models or just use a CLI that you've already subscribed to right and that way you solve the problem. So the the way that I would look at this is like there are no excuses for not having this but it will make you better with AI. It will help you understand how to build better systems. And really, if you look at this system and everything that we're doing inside the agent OS and everything else that we've got going on here, when we're doing all of this, it's about becoming a great architect, right? It's not about prompting. It's not about uh coding. It's not about being technical. It's about how can you design beautiful systems that work together that run without you or have minimal requirements from you. So, let me show you another example of this. So what we have over here is a video agent and I can just type in a topic here and in one single prompt it can generate video for me right and this is a AI avatar video as you can see and it's got the B-roll it's got the script fully researched all the research is done by the agents and they just work together and so literally my only input is what is the idea right I'm the strategist I'm the orchestrator and so if you're building a system you want to make sure it's as handsoff as it can be with just maybe a human in the loop for the actual content style. The other thing that I would say that's super useful is just having a workspace where you can come back to everything you built. So, for example, if I want to have a look at previous stuff that I've created with this, well, it's all ready to go inside the workspace. I don't need to search around locally. I don't need to search my downloads folder. It's all right here and it's all looking beautiful, right? And the other thing I would say here is like you might say, well, that's great, but it can only create an 18-second video. Actually, this is a 5-minute video fully created with AI. And the agents just replace a whole team of humans doing this, right? They can create better edited, better research, and better scripted videos than I can or my team can because this is all working together in a beautifully orchestrated system that you tweak and improve and train every day. And again, every time you have an idea, you plug that into the system and you make it better and better. So this is the way that works. You know, you have multiple models, multiple agents working together. They come together as one team. You have one shared brain and it all works beautifully inside one system. So as I mentioned, we've got the mission control dashboard where we can see the status of everything that we've created. We have the memory galaxy where we can see the context of everything across all of our AI agents working together. We have Hermes which is basically a powerful way to have our agent segmented into the most powerful parts of it whether that's for example Hermes Javvis or whether that's Hermes Oracle whatever you want to do and that's good to go at any time and then also some people again will say like you know I don't want to use too many tokens you can just use free claw code you know plug in a free API into free claw code and then you can use it for free with the power of like a frontier level agent harness [snorts] you might also say well how do I get these agents to work together so you can use paperclip with teams of agents inside a team here. Or you can have a group chat like this where your agents can talk together. And the cool thing about that is when they talk together, if they come up with something good as an idea, they'll actually add it to the pipeline and then I can approve or or deny the stuff that we want to build. So you see how there's loads of ideas here. I can go from idea to shipped and all I need to do is approve this stuff in the middle. So this is a human approval loop section. I approve stuff before it actually gets built. And if we want to see everything that we've built together, we can just go to the gallery here and everything that we've created previously is all plugged into one beautiful system which is absolutely amazing. The other thing I would say here is like you know there's been situations for example where you got like Fable 5 got removed and then you had for example stuff like Sakana and Fusion that came out and basically beat Fable 5 on benchmarks. So what do you do in those situations? Well, the great thing is if for example you were using Fable 5 inside your CLI in an Asian operating system, you can take that out and put back 4.8 or whatever the model was previously. If for example something like Fusion or for example Sakana Fugu comes out, no problem. You can build those in. So within within 1 hour of Sakana dropping and it just dropped yesterday and it's a pretty insane system. Um within an hour of this coming out, we'd already built stuff with it into our agent operating system. we already have it as an API ready to go over here and we've already got everything that we've created from it inside one uh place, right? So I can come back to everything that we built directly inside here and uh we can just preview everything that we've created and it's all good to go, right? How insane is that? And so I I haven't seen any other tool that does that. I haven't seen even like for example Frontier AI companies, I haven't seen them build out something as beautiful as this. Whereas [snorts] for example when you're building out your own custom dashboard this is absolutely insane right you can anything that you can think of you can build you know your only limitation is time creativity and energy to build this stuff okay um so this is really the difference in terms of you know just imagine if you were still using claim and probably people watching are still using cla or chat chet but imagine if you were still stuck doing that like how much slower it would to achieve everything that we're doing here. This is 100 times faster. It's way better and it improves every single day because we add new tweaks and improvements to the system. So, for example, like GLM 5.2 came out recently is absolutely mind-blowing. We've already built stuff and then plugged it into our systems and we've got this all ready to go right here, right? It's unbelievable. And that's how fast you can move here. You know, this is how much fun it is to build with this stuff. And that's the other thing I would say here is it has the fun factor, right? it it's not frustrating to use. It's probably the thing that I enjoy working on the most. It is unbelievably rewarding to see a system like this come together. And I think if you're not using something like this already, you're probably missing out on the fun and the the the part of the journey that matters the most, which is, you know, learning new skills and learning how to build better systems and not relying on, you know, other models. So that's basically just me sharing everything that we're doing. You know, you could do it the old way, which is like everything's scattered, different tab for AI you use or different terminal, different um window, reexplain your business and every single new chat agent forgets everything the moment you close them. You copy and paste between tools to move work along, you're locked into whichever single model you opened and you're gluing it all together with duct tape, right? That is not a good way to use this. And I'll show you an example of this as well. Like for example, if we go into Claude, like sometimes I'll ask it to do stuff that I've had in another conversation and it's not even aware of what's going on in the other conversation. Whereas, for example, with this system, everything is linked together. So, the difference is that it understands me and it understands what we worked on recently because it's got those recent memories. So, it just updated within minutes of doing the task and that's the difference here. So with the new way which is an agent operating system, you got one one window. Every agent and build in front of you got one shared brain. They already know your business. The memory persists. So nothing is ever reexplained. Agents pass work between each other automatically. Every frontier model that you have a CLI is ready to go and the operating system is a glue. You just steer and orchestrate it. The only thing I would say here as well is like when you're building this out, some people would say, well, ah, you need loads of different CLIs to use this blah blah blah. No, no. You could just have Hermes like you don't need to build all these agents on the left hand side if you don't find them useful. You could just have Hermes here. That's it. Um and Hermes is free and open source and you can use uh free and open source models with it as well. So it's not like you need loads of CLI to use this. Not at all. And it's not going to use up a lot of tokens because you don't have to use APIs. You can use CLI instead. And so if you look at this, like it's the same week of work, but your time spent driving it is way lower. So for example, [snorts] if you have a pile of disconnected tools, you might be spending 15 hours just organizing everything per week. If you have an agent operating system, maximum you're going to spend is like 3 hours per week. It might be like 30 minutes a day just improving and building stuff better, too. Right? So it's the same output, but most of the week you free up because you're automating everything. And so this is the power of this stuff. I think 99% of people are not aware of it, but if you are aware of it, if you're watching this, you're ahead of most people and you're building you're well on your way to building something amazing. So, this is basically how it works and you know, you can give it a go. You have an agent that researches, codes, writes, builds, and then everything is done as well. Uh, so yes, it's absolutely awesome. Now, some people say, I'll keep using my AI tools separately. It's fine, but separate tools means you're the glue, right? You're the one reexlaining. You're the one copying and pasting. You're the one holding context in your head and the whole power is really in wiring them together. So one shared brain is worth more than 10 clever tools that don't talk to each other. Other people say, "Well, this is way too advanced for me. This sounds like it's for developers. I don't understand how to build this." So you can actually get our setup from the AR Profit Bordom if you just want something ready to go. And you can see some examples here. So like people are building this within 30 minutes. loads of people inside the airoff boredom community using our setup and then customizing it exactly how they want. Um, and you know, I've seen like an amazing community of people just building and sharing everything that they're creating and just having a lot of fun doing it. So, I think like for example, if you look like Ethan, he said he's 61, not a coder, but he have he has the agentic OS set up and he has eight agents and voice and obsidian installed and he was proud of himself for setting up, you know, super rewarding for setting up. So, if you're non-technical, no problem. I'm not technical. I'm not a coder. I don't understand HTML. Neither does anyone inside this group. I mean, there are some people that do inside this group, but most people don't. And you can see we have 187 pages of testimonials, wins, reviews from people winning with this stuff. So, you don't need to be technical to use this. And that's basically it. So, if you want to build your own, you know, you could do that. So, you just uh all you would do is you would go inside Claude or inside Hermes and say, "Hey, here's my idea. Here's the dashboard I want to build. Start building it. That's basically what I did. If you want my setup, you can get it inside the AI profit boardroom. In terms of what you learned today, you learned how to stop tab juggling. You learned how to stop reexplaining yourself to AI. You learned how to stop being so locked in. So, for example, if a frontier model gets taken down like Fable 5, no problem. Doesn't even matter, right? We as soon as that happened, within 2 minutes, we swapped it out. And then you stop being the bottleneck because you can handle goal and you can use a loop system inside the system to stop prompting it yourself. You keep control because we have this running locally which is pretty useful. So it's sort of sandbox away. And then also every time you update the memory or update the agents or add a new system well the whole operating system gets smarter and so it's improving every single day that you set up. So the way that I would say this is stop collecting AI tools. Start running one machine, one system. Become the architect. Don't be the prompter. So if you want to run every agent from one screen, you've seen the whole agent OS I've built, every surface, every reason it matters. If you want it set up with you step by step, it's all inside the air profitable. You get the full agent operating system with claw, open core, Hermes, GLM, and more. One dashboard, one shared memory, every feature inside the guide today, the mission control, the memory galaxy, the Hermes Oracle, the Hermes Jarvis, etc. a 30-day road map with daily tutorials and a four coaching course per week in a room with 3,700 builders doing this every day with someone online 24/7. So, you can get that inside the AirPruff boardroom and if you go to the classroom here, go to the agent OS, you can get the latest version. We literally update it every single day. So, you can see when it was last updated, you get the video tutorial, the zip file, and then we also add new tutorials every single day to help you as much as we can. Inside the community, you can ask questions, get help and support. I personally answer every single question every single day with a video tutorial. Plus, the whole community helps each other as well as you can see right here. And people are building amazing stuff with this um with this whole agent OS system. Inside the classroom, you get all of my best trainings. So, we have a complete beginner to expert a automation course if you just want to learn this from scratch. And if you want to get my systems, it's all inside here. We also have four weekly coaching calls where you can jump on a Zoom call, ask questions, share your screen in real time, meet the other members, see what other people are asking. And then also you can ask the whole community and meet the whole community whilst you're doing this. Inside the map, you can actually meet people in your local area building agent operating systems and this is all inside the AR profit boardroom link in the comments description or just go to the air profit.com. Thanks for watching. Timmy says, "Keep up the good work, man. really appreciate you. Total workhorse. Thank you very much, sir. I appreciate you, Timmy. And on that note, I will be gone. Thanks for joining. Bye-bye.