All right, let's get this party started. So, we're going to kick it off. We got some big news with OpenAI dropping GPT 5.6 Soul family. So, what's happened here, and I was consulting the Oracle before, as you can see, is that Open Eye just dropped GPT 5.6. However, it is not available to everyone. So, as you can see in this tweet from Philly, uh the US just locked it behind a velvet rope, which is pretty wild stuff. So, that means basically everyone who uses it right now has to be approved to use GPT 5.6. They are looking at a roll out over the next couple of weeks. I did talk about this yesterday, but that is not confirmed because they only said they believe in broad access and plan to make GPT 5.6 Soul, Terror, and Luna available in the coming weeks. But for now, at the request of the US government, they're starting with a limited preview among a small group of trusted partners in codeex and the API. So, what we're looking at here is basically a very slow roll out of the latest GPT 5.6. Pretty wild times, peeps. That that is the new reality. And then we've got three new models here. So, we have Soul, Terra, and uh Luna. And you can see the costs here. So, this is Soul would be the fl flagship model for aentic work. Then you got Terra which is kind of in the middle and Luna which is basically like the cheap fast version I think the high volume stuff. What else do we have today? We got Fable 5 comeback signals everywhere. So you can see for example here we're looking at a potential comeback from Fable 5 and apparently the US government has approved 100 uh companies to use this but it's in limited preview but we're making little steps towards it so that's good news. Then we've got the agent arena efficiency chart drops. So, token efficiency in agent arena and we've got some changes right there as you can see. But yeah, interesting times. So, today what we're going to look at is just building out some stuff for agent OS and I'll get straight into it. One sec. Come on. Good to see you. What have we got here? Guten Morgan. G Morgan to you too, sir. Ah, thanks so much. Yeah, we're live everywhere. I appreciate you joining, Sharon. Good morning to you too, sir. Yeah. So, what happened with 5.6 is there's three new models, but they're in limited preview only. So you can see here the next generation of GPT is now segmented by three models which I think is good because you know before it wasn't that easy to understand but you have basically Soul, you have Terra and you have Luna. Now if you look at these models these are a new frontier models. They're performing pretty well on the benchmarks but the main thing to note is that it's not available for everyone. So, they've just approved it for a small select number of people and then it will come out for everyone eventually. That's what they're hoping anyway. Uh, it's not been fully approved and I think for a full roll out. And then if you look at this, soul is like the best agentic frontier model for GPT 5.6. Terra is in the middle. It's kind of balanced. And then Luna is your sort of high volume chat GPT instance uh style response. Now also if we have a look on the benchmarks it's pretty interesting. So terminal bench 2.1 Claude Mythos is being outperformed by GPT 5.6 Soul and GPT 5.6 Soul Ultra. So Mythos is below these and Fable 5 is even more below not just Soul but also GPT 5.6 Terror. Um, this is just for Terminal Bench and they they they expect it to roll out after the limited preview within the next few weeks. So, we're probably looking like mid July for that. Morning, Max. Thank you so much. Oh, I appreciate that. And then, uh, Sharon says, "Yeah, we're we're live everywhere." Yeah. Yeah. Heat. Heat. Yeah, I saw that. I think it's just small steps forward. So, it looks like we're getting closer to a bigger release for everything else and for Fable 5. Like, that's definitely progress versus what happened a couple of weeks ago. But, it's going to take time. Might be like a few weeks before we get access. It might not even be able to access these models outside the US. You know, that's something else to bear in mind. There's a lot of people who watch this show that are not based in the US. So, it seems like we're getting closer to a launch, but at the same time, it's it's a very slow setup. Sharon says, "Thanks for the 5.6 explanation. Had the idea. That's where it's headed lately." Yeah, it's you look at these US, it's only in the US, but if you look at these US Frontier models, they're only coming out to a small preview of people. The problem with that is like you don't even know if they're number one good. You don't know when they're going to drop. And also, you know, if you look at these Chinese models that are coming out like GLM 5.2, they're getting better and better in the open. And so like it looks like we're going to h have a situation where you know China eventually overtakes us maybe this year with their open models if things keep going the way they're going which would be massive for companies like this you know bear in mind like you know OpenAI Anthropic they're trying to stay competitive they'll probably IPO soon it's not good for I don't think they want this. I actually think no. From what I've seen, for example, when Fable 5 came out was it was not very good at UI. It was not very good at user interfaces. So, I don't think the next update will improve it at all. We do have a full guide on how to improve UI with AI inside the AR profit board. So, if you go to the classroom here, new daily updates, you'll see a full Claude web design masterass with video tutorial and a full step-by-step guide, plus the guide and the skill that I use for web design. Um, but yeah, I don't think these new models are great. Yes, I've already got a tutorial on that. I tested out yesterday. Um, so I've got a full guide breaking that down. On is pretty good so far. Pretty good. It was the fastest model I've seen on Mac looking stuff. it it's nowhere near frontier level, you know, it's it's just basic. It's going to be sort of at the same level the AI models were at last year, if that makes sense. So, I tend to find local models are usually one year behind at least for the ones that actually run fast. Heat. Heat. Heat. Heat. Hey, hey, hey. Heat. Heat. Hey, hey, hey. Hey, hey hey. Hey, hey hey. Heat. Heat. Yeah, I think uh 5.6 will be a huge step up, you know, versus GPT 5.5. That's why they've been so careful in releasing it, you know, whether that's for building apps tools games websites whatever you want to automate. I think it'd be a lot better. And then Fable 5, I mean, you probably already tested it. If you didn't, it was really good. So, when it comes back out, I would expect it to still be pretty good. Yeah, I think that'll keep going, keep improving. back in a sec. Where's the link? So for the tweets you embedded, you need to talk through them and have like a a section that talks through each one, what's going on, etc. Update your skill based on that. Also remember with your guide skill, you always should send me the link directly. Update your skill to remember that. Yeah, if you go back to what I talked about in GPT 5.6, you you'll see why it's only available in preview for a select number of people. Today we're going to be running through the latest questions that I've had about the agent operating system. How to use it, how to get the most out of it. You know, we've got stuff like Hermes Oracle in there, which pulls in the latest news, and then we can automate SEO with it. We've got Hermes Jarvis, which is a fully automated voice agent for Hermes, and it can build and automate anything that we want. we have all of our agents in one place. And this is even more important now because if you look at the latest Frontier models, for example, like GPT 5.6, which is only available in preview, we haven't got access to it and there's no guarantee that we will get access to it. It's the same with, for example, something like Fable 5 that got put out there and now it's taken down again. My point here is that it doesn't matter about the model because we have to live in a world where we need great systems instead so that we can get more out of our models. And that's exactly what we're building with an agent operating system so you can have all your agents in one place. Now, I'm going to cover some of the top questions and some of the best ways to set this up based on the community questions I've had recently. So, let's get straight into it. This is the AR profit board and this where I answer questions every single day. And we're just going to jump straight into some of the latest questions we've had about the Agent OS. So let's get straight into this. The the first post we actually had was from Jean and Gene was saying you know Open Eye is going to limit the release of GP 5.6. I just heard that Open are facing the same security concerns about their new model GP 5.6 similar to the issues faced by anthropics Mythos Fable 5 Open AI will have to restrict it and also be limited to a few financial partners. Now, this is crazy and unfair. He says uh the USA is is slowly becoming less liberal with its releases of models. Now, here's the way that I look at it, right? It really I don't think it matters too much about models because for 99% of automations you don't need access to GPC 5.6. So you don't need access to Fable 5, but you do need a great system. And so one of the best things about the agent operating system is that you can easily be very anti-fragile and flexible to whatever comes out next. So, for example, if Claude Fable 5 drops again, we can plug it back into the system. When GPT 5.6 comes out in a few weeks, we can plug that into the system inside Codeex. When they're not inside there, no problem. We can use something like Fusion or Sakana Fugu to get Fable 5 level intelligence and build and automate anything they want. So, it's not ideal. I can understand why they're doing it. I will say as well like when you've got systems for example like GLM 5.2 GLM 5.2 2 is a Chinese model. It's open weight, is free to host locally, and it's completely open source. And these Chinese models, I think, are going to, you know, if this situation keeps going on for the rest of the year, Chinese AI models are probably going to overtake US because they're just releasing the models whilst the US is restricting the models. And that's a bit of a problem, right? If you look at Anthropic, if you look at OpenAI, they're trying to stay competitive, but they can't, if they can't even release the technology that create to the whole public. So, if their whole business relies on on releasing these models and they can't, that is a huge issue. And I don't know how long that's going to last. We got another question which is interesting. It's about my posting strategy with stuff like on on social media for example. And the question that Ryan asks is like you know how often do you post? Is it an SEO strategy? What are you trying to do? So, we actually have a training on this here. The reason that we post more often is just because if we get more attention, we tend to get more leads in. And so as long as you keep the quality high, more posts equals more attention equals more leads. It's the same for example, if you look on X as well. We've recently ramped up the posts that we create and that's gradually increasing over time. If you look over the last few years, as we've posted more and more, we tend to get more and more impressions. So, it's just one of those situations where as long as the quality stays high, more posts equals more attention. Now, at the same time, if you're just using AI without quality controlling everything, that would be a problem. So we make sure that we give our team a lot of feedback, a lot of coaching, a lot of support on how to make sure the posts are as good as possible. And we set up really good systems and prompts to make sure that the quality is high. But yeah, we tend to find like volume works pretty well. One of the best things as well about this community is like people are just getting awesome results. So you can see an example of Gil who posted about the weekly coaching call and how good it was. It's pretty insane. Pretty insane. There's some great people inside there and it's just a great place to network and meet other cool people building similar things. Giovani was posting about local models and whether he thinks, you know, whether it's something good that the clients would be interested in. So, for example, offering a local setup for clients, offering to set up local models for clients with AI and help them figure out how to build systems around this stuff and I'd say this is becoming more and more in demand right now. So, if you look at the rise of local models, I think it's becoming more popular over time. These local models are becoming better and better. And also if you look at a lot of clients, they don't want their data in the cloud. And so having a system where they can use free local models which keeps the cost down. And with local models, your data doesn't go to a server. Well, that's a huge benefit to them. So if you become an expert at setting up local models, that's a huge advantage that you could offer to clients too. So Lenny is asking about how to efficiently manage all of his Hermes agents. So, he said he recently learned how to use Hermes and wants to deploy an instance for each of his clients. He's looking for a way to control all of these Hermes instances as well as the sub aents from a single interface without having to constantly switch between Docker or VPS servers. So, one thing that we've set up and this might be great for your clients is inside our agent operating system, we have separate profiles for each API. Separate agent profiles inside Hermes agent. You could do the same but for your clients. So you can have separate agent profiles for each one of your clients and then you just manage it inside your agent OS. So if we go to the agent operating system here and we look inside the Hermes section, you can see we have each of these different profiles like Julian, Jarvis, default, local, north mini, GLM 5.2, SEO lead, etc. And these are all different profiles that you can set up with Hermes agent. And so what you could do is you could have each of these profiles for each one of your clients. And that way you could manage everything from one place. You don't need to switch between different VPS's and you can just have it in one setup. And if you want to get the full agent OS from us, we've got it inside the AI profit boarding link in the comments description or just go to the aiprofit.com. We got a video tutorial, the last update date and the zip file on how to use it. So air is asking about how to automate video and how to get started on this. So you know there's many options out there. You got like Hicksfield and stuff like that. For us, we actually have a video agent inside the agent OS as you can see right here. And it can fully edit videos in like one single prompt with the avatar, the B-roll, the research all put together. So that's one single option. Also based on my workflows, we actually have a setup here for AI avatar videos. So you can find this training right here and it's got a full step-by-step operating procedure. It basically shows you step by step our process for getting results with this and that's something we've tested relentlessly. So there's two options for that. Number one, you could use the training here and number two, you could actually use the video agent inside the agent OS. This is an interesting one. So Jay was talking about using a fact checker agent inside paperclip. And basically it job is just to fact check stuff before you publish it. I think that's an interesting way to use paperclipip. The way that I've done it in the past is like we have teams of agents and like if we're working on a project, we can just assign all of our agents to work on one single task and then we can see everything that they've done and they build together. But using it for a fact checker is quite interesting too cuz it's kind of a loop where you create the work with AI agents then you fact check it with paperclipip and then that goes round and round and self-improves and self iterates. So that's a pretty cool way to use it as well. I like the idea of that. Also, Ronald was asking about, can you use the agent OS on Windows? You absolutely can. So, we've actually set up the installation file so that you can use it on Windows or you could use it, for example, on Mac or Linux, whatever you prefer. It's all inside the zip file, so you can quickly install it for whatever platform you're building on. So Gil was asking about Hermes Javis. Now this is a voice activated version of Hermes. You can see it right here. So for example, if I say okay, open up Google, it's going to open it up, right? If we actually scroll down on the page, you can see the full conversation history. And we can also build things in real time as you can see here and and see what we've created. And then also we can put this in wall mode which means that basically we can speak to Hermes Jarvis in the background. We can also this is pretty interesting when we're using Hermes Jarvis we can actually switch between agent and auto mode. So if you using auto mode it will reply to you quickly in real time and it can still open stuff up like and still do browser use or computer use. If you switch to agent mode then it becomes even more agentic and even more powerful. And again, you can have this just on a monitor in the background or in a tab in the background and voice control it whenever you want. You get the full power of Hermes, but you get it with a voice activated version. So, if you need help setting that up inside the agent OS zip file. We have a full installation guide on how to use it and set it up. And you can just give that to Claude or to Hermes or to Codeex or whatever AI agent you prefer to use and it can set up and install it for you. Hobing low was saying that he wanted to give a shout out to paperclip. So he said this has revolutionized how I work with AI agents at the moment just using Claude through the agents OS but this paperclip is really helping him think through who he needs and what they need to do. Super helpful. Thank you Julian. And you can see the system that he set up here. So you set up a full social media team using the agent OS system that we've got with paperclip. So this is really cool. If you want like a zero human company and you just want AI agents running, building and organizing for you 24/7, it's a fantastic way to set up. Uh you can easily do it inside the agent OS. I also think this is an interesting way that you can use agents which is you can have the agent OS but you get Claude to operate it for you. So you don't prompt the agents. Claude prompts the agents. And I think that's a great way to use the agent OS is you get Claude to prompt it for you and organize it for you. And also just thanks so much for sharing. Like it's great to see you getting results on that. So congrats. By the way, you might be thinking, "Okay, this agent OS, it looks awesome, but it's it might be technical." I've seen so many people getting great results for it. Like, so many people raving about this setting up in just 30 minutes, saying the system is amazing. Um, people just getting awesome results. We've got over 191 pages of testimonials of wins from people getting great results with stuff like this. So, I know like if I can set it up, if they can set up, we can all set it up. And and it's just a fantastic system to get results with AI. It's the most powerful and interesting system I've ever seen for using AI, too. So, I don't know anyone else who's got a system like this. And that's why I built it really because it just sort of captivates everything I wanted AI to do, but just wasn't out there. Yeah, this is pretty wild as well. So, Jean shared this news about Alibaba and Anthropic. Who knows what's going on there, but I know that that was reported a couple of times this year as well. So, it's one of those, isn't it? Like China and and the US are really competing to have the best AI models. And obviously if you look at Quinn, you look at Alibaba, like they're releasing new models all the time. They're improving and accelerating at a rapid rate. And then you look at the US and they're gating models and they're um avoiding sharing the New Frontier stuff with people publicly. So it's one of those situations where who knows what's going on behind the scenes, but I can imagine it's is really crazy. So Ritz was asking should he uninstall Hermes desktop instead of using the agent OS. So for me personally, I actually don't use Hermes desktop anymore because I tend to find like they would overlap and it'd be pretty hard to sync them both up. So for me personally, I just focus on the agent OS. I don't use Hermes desktop as I find it's kind of like a limited simplified version that doesn't work with all my AI agents. So if you want like a simplified version, I would go with that instead. Just focus on the agent OS. And then if you're worried about breaking anything, you can always back up Hermes. So you can run Hermes backup before you change anything. Or you could have a separate instance where you experiment with it first before you have it fully set up on your live working main instance. This is a good question from Neil. So Neil was asking, "How do you use GLM 5.2 inside Claude's desktop? Is there a way to do it?" So as far as I know, I don't think you can use I don't think you can use GLM 5.2 in Cyclaw Desktop. Olama originally had an integration with these two and then they took it down. There might be a way to change your environment file, but I wouldn't like to mess around with it because it'd be pretty hard to restore. What you can do though is you can use GLM 5.2 inside C code. That's pretty simple and easy to set up and less volatile. You could do that with O Lama like this. And then inside GLM 5.2 you would make sure you have OAMA downloaded. And then from there you can run this command inside your terminal to set up GLM 5.2 to cloud with port code instead. We had another post about GPT 5.6 and Fable 5. This is one that I know like people feel strongly about. So with this whole situation, the way that I would look at it is build the system. Don't worry about the models. So if you look at the agent operating system, it's super flexible whatever comes out and you can achieve able five level intelligence if you're using sana figu or fusion. Anyway, we've tested it and it's you know fusion especially is top of our leaderboards when it comes to goldie bench. So I wouldn't worry too much about the models as frustrating as it might be. just focus on the systems and then the models become irrelevant because you got great systems to automate whatever you want. So for example like with automated SEO for automated videos using this whole system and so when you do that you don't tend to worry so much about the models because you know that you've automated the most important parts of your business and that's what really counts. So, I've got a question here from Justin, which is inside the agent OS, is the SEO section in mission control supposed to write a blog for you, or do you paste your already written blog in and then it ranks it? So the way this system works is that you plug in your keyword here and then you would plug in a case study that you work on. Make sure of course that you've wired in your websites with your Netlefi or WordPress access tokens. And then from here you can click on generate five articles and deploy that to your website directly. So that's the way that works is inside the generate tab you plug in your keyword and your case study. The case study has to be something unique because you want to create information gain and make sure your content is not just like chat GPT fluff. And then from there you deploy that content to your website so that you can rank it with a a fully completed blog post. just having a look at some of the questions in the comments as well. So, one question we got was like, can AI do like game development? And I would say yes. I mean like this is a really simplified version but we have a a video game studio inside the agent OS as you can see right here. But let me show you some more interesting stuff that we've worked on recently. So, for example, we actually gave a video game skill to Sakana, I think it was recently, to build out better games, and it can build some pretty cool stuff. Like, you can see an example here. But if you want to see something even more interesting, let me show you Goldie Bench and some of the stuff we've created around games. This is just for fun. Obviously, you know, for me, I'm very focused on on automating with AI for business. But if you want to see some examples of what we've done when it comes to video games and that sort of thing, let me open up a couple of examples right here. So, here's one. This is called Dragon Realm. Super nice. Kind of like open world RPG style. Looks pretty cool. Very visual. And, you know, we built out this whole world using this system. And let's open up some more here. So, here's another one. This kind of like a a Skyrim style game as you can see. 3D open world. Looks pretty cool. And again, like this this game is absolutely massive. Like I haven't found the limits of it yet. It is wild how big it is. And so that was using Sakana Fugu as I mentioned before. And the thing to bear in mind with these systems, sorry this is with Fusion. The thing to bear in mind with these systems is like we benchmark it. So, we'll test each of these prompts directly against everything else. And this compare section here, so you can see the outputs. But from what I've seen so far, Fusion seems to be the best at creating a massive open world game like you can see. So, yeah, you can do a lot of that with AI agents. Another question we got here was what's the difference between the existing Hermes cam board and yours? Do they still function the same? It's the same functionality but it's a better UI, right? Like if you have a look at the camb board that works normally with Hermes, it doesn't look as nice as this. And also it's not inside one place, right? Like if you want to access the camb you need to run terminal commands and then you also need to go inside your dashboard and mess around with it. And it's a lot of work. I know because I personally tested it and that's why we built this. So if you look at the camb board here, you can have separate camb boards depending on what you're building. So you know we're talking about game design here. This is for a dragon game that we built out. And then also we can have quality control on these systems as well. So we can have multiple different agent profiles, quality controlling the content whilst we build it out. And then also you can just drop in a new task here. It gets triaged. So it has the same functionality, but it just looks way nicer. It's way easier to organize. And also everything that you create inside the system goes inside your Hermes workspace. So when you create something usually with these sort of agents you lose track and it's easy to lose everything that you've built. Whereas for example with these systems that we have here we can easily come back to them anytime we want to. It's pretty simple and easy and that way we can save our progress you know and we can keep building on the stuff that we built before. So I think that's the best way to use it. Whereas, for example, if you were just using the terminal or Hermes desktop, you could easily lose all the stuff you've created. And then for Unreal Engine, you can get, you know, if you interested in game design, you can use uh MCP with Unreal Engine and Hermes agent or with any agent that you use like Claude and then that way it can control it for you. I'm not an expert at Unreal a uh Unreal Engine, but I know that you can get an MCP for it and then you can build games that way. So, that is it for the Q&A. If you want to be able to ask me questions like this and I answer them with a video tutorial, I do that every single day inside the AI profitable room. So, I personally answer all of the questions inside the community to help you as much as I can. Inside the classroom, you can get the agent aware system that we've covered today over here with video tutorial, the last update date and a zip file of how to use it. And then we add new daily tutorials like you can see right here. You can also jump on weekly coaching calls and you can direct message me as well if you need extra help. And on the coaching calls, the great thing is you can share your screen, you can ask questions, you can meet all the other community members and find out what they're building on too. And then you can also meet people in your local area. So the whole point of this is like we create a community where the world is changing, humanity is changing because of technology and we're creating a community where we can all raise each other up. We can learn and grow together. We can share our journey and also is focused on helping you save time, grow and scale with AI automation. So if you want to be part of that, feel free to join the AI profit boardroom link in the comments in description or just go to the AI profitum.com. Also, another question we just got here is so Mo said he spent eight hours struggling with setting up his own agent operating system. So, do we teach people how to do this without all the stress? Yes. So, the way that we've done this is we have this agent OS system and you can just get the zip file and we update it daily. So, you can get the zip file and then you can just quickly install that with your AI agents. So, we try and make it as seamless and as easy as possible to set up your system. And then, as I've shown you before, like I answer all the questions that you have daily on a video tutorial. So, I can help you as much as I can. And then, if you look, for example, like Rick, he said, he joined the school and he already built his agent OS within 30 minutes. You can see, for example, Hockey as well. Hockey just joined and he was like, I've installed the agent OS. Your system is amazing. Right? And there's so many reviews of like people getting amazing results using our agent operating system. The reason I built it and shared it is because I knew that the world needs something like this, but they don't know how to get it right. And so that's why we these systems just to help you as much as we can. You see all the testimonials around the agent OS and people sharing their dashboards and what they've created. So yeah, you can get that inside this section right here. So hope to see you inside there. Cheers. Happy to help me go. Thank you very much for the questions. Yeah. Hey hey hey. I noticed that you deployed the agent OS guides, but the CTA doesn't go to the school group and it should. We changed that yesterday. So, update your skills, you never forget that again. And then also I notice that on each of the guides that you've published the navigation at the top which should link back to other blogs and make it easy to navigate to other blogs is gone as well. Same at the bottom. Like there used to be a a linked resources section with images back to the other blogs. So can you just update your skills so you don't make these mistakes again for creating guides? Whatever guide you're using for creating guides, whether that's multiple as well, you need to update them all so you don't forget this stuff again. If so, what sort of models could we run that are better than what we currently get? I think the best that we've had so far in terms of outputs is orif, but it's still not great. Could we run something like, for example, Quen 3.6 with this or how does it work? Just don't want to slow us down. We got the same question three times from uh Arthur which is what is the best computer used feature or tool in the world that works with no mistakes to automate for example social media publishing. Um yeah I I would genuinely say like every single computer use model that I've tested makes mistakes. So, I don't think that exists yet. And I I still think it's very very early for computer use. And that's why 99% of people don't use it. However, what I would recommend is that if you're trying to post like that, probably you want to use API instead. Like that's going to be way more reliable. And then number two, if you do insist on using a computer model, then just go with Claude. But again, like none of them are good. None of them are that great. Uh they still make mistakes. They they're still very slow. But yeah, Claude Opus 4.8 is probably the best right now. Heat. Heat. Heat. Hey, Heat. out. Yeah. All right. What questions? Yeah, Lucas, I've I've been through this, but I'm gonna do a tutorial on it soon. Hello, Brett. Thanks for joining. Brett says, "Hi times, so must be happy to be here, mate. Welcome in. Already done. Already done. Today I'm going to show you my best systems for AI SEO based on what's working right now. So you can see, for example, this website, it's gone from like zero to 278 clicks a day. This website's following a similar trajectory. Here's another one that's come coming out of the sandbox gradually. You see was at zero clicks a day and now it's growing nicely. And the same for this one. This one has grown like crazy. It's going from from zero to 29 clicks a day. Here's another one that's doing pretty good. This went from 0 to 26 clicks a day in in like how long? From the end of May all the way up to the end of June, right? And then also this one as well went from like zero clicks a day to 68. So I'm just going to talk through what we're doing for each of these websites, how it's working. Let's start with the air profit.com. So basically what we're doing here is each day we are using our systems to create content in two different ways. So number one is that we use Hermes and Hermes Oracle to pull in the latest information, the latest trending topics and then create content around it. So you can see this is automatically running on a schedule every day. So every 24 hours it refreshes all the news. It pulls in the trending news based on what's working and then we know okay these are trending topics is organized by how trending that topic is. Obviously for example a lot of people are searching about cheap 5.6 right now. So what we can do from here is we can just click on publish to WordPress and that will start to write a blog post on our WordPress websites for three different websites and start publishing them directly. So you can see this is currently getting published and indexing. And if we have a look at the post here, if we open this up, this is published directly to our WordPress website, which is pretty cool right there. So this is basically a full blog post fully personalized to us with CTAs throughout the page on the current news topic and we can just quickly create content. Now also the great thing about this is automatically gets indexed with index optional so that when we're publishing this content we can get it directly on Google ASAP right so for example if we have a look at this it is already indexed and we only publish it 23 hours ago right so it gets indexed super fast how do we do that because we're using index section the API for indexing it then we focus on trended topics and then we automatically publish to WordPress the other cool thing about this is when we are doing this it saves a lot of time um because we don't need to go into WordPress, right? Like this post is already published whilst I was talking to you. It's embedded the news articles inside the post itself and then it runs through all the updates and we've got a fully SEO optimized blog for this particular keyword published ASAP, right? Super fast and easy. And the cool thing is like if you're if you're the first there or if you're the only there only person there talking about that topic, well then you're going to rank and get more traffic faster with SEO. Now, you might be thinking, okay, well, how does it personalize it to you? So, you it pulls in all my memories and all the context about me and all my goals and everything that I'm working on and uses that to write the article, right? So, we've got Obsidian as a memory folder directly plugged into all of our AI agents, for example, like Hermes Oracle, and then it can quickly pull in the news, create the content, pull in the contacts from me, and personalize it to me, my business, and and everything else. So, that saves a lot of time. And then what we've also got over here is a system where we can research what's ranking on our current websites that we're getting impressions for but not clicks. So I can easily identify okay what are the gaps and what should we be trying to rank for. So for example we type in open claw here we can look through our Google search console data and if we have a look for example this is a keyword that we are getting a lot of impressions for but no clicks for. So it probably makes sense that that is a content gap that we could create content around and we could quickly start ranking our content. Now if we have a page that is getting clicks for that particular piece, we'll already have a page here that is published and we can quickly see that as you can see right here. And so the great thing about this is we analyze. We see okay what are the content gaps and then what are the pages we need to create new content for and then which pages do we need to reoptimize as well. And we can quickly use this SEO keyword system to find the potential opportunities we should be ranking for. And it's so easy and fast. Then once we've done that, we can click on use topic. And you can see how quickly this is to navigate. From there, we can plug in a keyword, plug in a case study. And the reason we want a case study is it's all going to be personalized to me. And we just deploy to our websites and we publish that across our website. So for example, we go to this page here, Hermes desktop installation guide. We were getting a lot of impressions for this but not a lot of clicks to our main page about it. So we created a new dedicated page for this so that it's a lot more relevant to the search intent that people are searching for and that's how we're getting you know good traffic across all these websites because we can rank and find opportunities quickly and then we can start getting more traffic. So that's that's basically the system that we're using. The other one that we have finally is video agent. So, the video agent is pretty cool because you can just type in a topic here and generate content quickly. And you can see an example of a video that we generated in one click with the avatar, the B-roll, everything plugged into one system. We can see all of the videos that we've created over here. And videos ranked really well. So, if we type, for example, best AI community into Google, you can see that inside Google AI overviews, we are recommended as the number one community for the best AI community. Right? Perfect. How are we doing that? where we're using the systems that I just showed you for SEO and then the systems for videos too. So you can see us ranking directly inside the Google AI overviews over here and then if we scroll down we're also ranking here as well. And then if we scroll down on the page too we are ranking with our AI profit boarding website and we can create multiple pieces of content that rank multiple times for that particular keyword. So it's pretty powerful and easy to set up. And then for example if we actually have a look inside Google AI mode too, we can rank inside AI. So, if we have a look and we type this keyword here, guess who's ranking number one? Um, it it's us, right? It sends traffic directly to our page right here. So, this is the power of this stuff. And I think 99% of people think, you know, it's too hard to rank inside AI. No, like we're creating pages that are ranking really quickly using the systems. The only difference is now is that you can create more content. You can quality control it better. You can um get better results with it and it's easier. Right? So the beautiful thing about this is if you have great systems, you save a lot of time, you automate it, but you quality control it as well. That's what other people say as well. They'll be like, you know, is is this actually good content? What we don't need more of with AI is bad content. So if you have a look at the content that we're creating is actually pretty good. So if we have a look at this one for example, it pulls in a relevant case study. So this is my experiments. This is my content that's plugged into this particular article. And then you can see for example inside the introduction it talks about my experience. So this is not like for example generic stuff. This is not an article that's just chat GBT fluff. It's based on information gain. It's based on adding something new on that particular topic that no one else has talked about. And then it has relevant CTAs. The page itself looks better than 99% of blogs. It's easy to read. It's easy to scan through. It's all plugged in. And also it talks more about like relevant topics. And this is a a website that I personally cannot log into. It's only my agents can control it. And I I think the beautiful thing about that is I don't get distracted going into WordPress. I don't need to worry about what plugins we install. I don't need to organize the design. I don't need to have a team of people operating it. It is literally a zero person block, right? There's there's no human input inside this block. Is literally just AR agents updating it and creating it. So that's what we're doing for the AR profit boardroom in the websites you can see right here. Then if we have a look for example that's another website in the the five that we deploy to that is as well. Let's have a look at this one agent OS guide. So this website right here this basically documents everything that I do inside my tutorials. So when I'm talking people through a particular topic you know inside my latest tutorials we create a guide automatically for it. And the great thing about that is that number one it's beautifully designed. So, you know, it adds some If we have a look, for example, let's let's have a look at this one. Like the the hero image is absolutely insane, right? Like beautiful image. And then the design of this is great. It's got a CTA at the top that sends traffic directly to our funnels. And also, you can see like even the way the font is formatted looks beautiful. And if we keep scrolling down here, you'll see the progress bar at the top. You've got these beautiful diagrams inserted inside there. It's got relevant sources and external links based on the topic. And it it's it doesn't feel like your normal blog, right? It feels like something that's handcrafted. But the difference is that it's just my AI again with a great skill that's using it to generate this. And also formats the content really nicely as well. And you know, again, this doesn't look like generic AI fluff. This is all stuff that I've personally tested and I taught people through and then we turn it into a great SEO guide that actually ranks as well. So if we have a look for example at this website, this is agent OS guide. We created it at the end of May, Thursday, May the 21st. And by June the 24th, it's already getting 26 clicks a day, and the trajectory is pretty nice there, too. You might say, well, isn't that just people watching your tutorials and then searching for it? And that's a valid point. You know, part of that will be branded search. But another good thing is if you're getting branded search, that's going to help you with SEO as well because you look more like a legitimate brand to AI search engines. So, I think I'm one of the few people who actually documents what I'm doing and actually shows people what's actually working with AIO. I think if you watch a lot of other people's blogs, they just talk about other people's websites or they don't actually show the websites that they're creating or they don't even show the data behind it. But these are all systems that were built and personally work for me. So, if you want to get this stuff, it's all inside the if you want to get the agent OS systems, you can get that inside the AI profit boardroom. So the systems that were built for AI SEO as you can see here. So we've got the SEO section, we have the video agent as well and we have the Hermes Oracle. These are all custom automations that we've built inside our agent operating system and you can grab that inside the classroom if you go to this section here. We've also got a free SEO course link in the comments description if you just want to get the free stuff. We got like 260 lessons on SEO. And then if you want to get a oneto-one SEO strategy session that shows you how to rank number one with your website based on what's working for us and we can jump on a call together, you and my team. You can get free SEO domination plan, discover the secrets SEO link building. You can get your questions answered. You can learn the best link building strategy for your website and how to quickly outrank your competitors based on what's working for me. You can get that link in the comments description or just go to goldie.c to access that. So either way to see you inside the forum or on the strategy session. Thanks for watching. No, it's not. It's not. This is me right now. But I'm glad that I'm so smooth that you think it's uh it was created a while ago. That's pretty cool. Heat. Heat. Heat. Heat. So, there's three new models from GPT 5.6 today. There's a limited preview. There's a lot of controversy. And I'm going to walk you through everything that we know about GPT 5.6 today. what it means for you, how it works, how people can get access right now, who's got access, and also what the three new models are. So, let's kick it off right now. What does this all mean? So, GPT 5.6 has dropped, but it's in a limited preview, and it's only available for government approved people who have access. I think there's about 100 people have access right now, something like that. So, it's very, very limited, and it's not open for applications or anything like that. Now, if you're wondering, okay, when does it come out for you? I'll come on to that in a second. The second thing to note here is that there are three new models. So there's soul, there's terror, and there's Luna. Now, what do each of one of these do? So soul is the frontier model, and there's actually two versions of that, which I'll come on to in a second. And then you've got GPT 5.6 Terra, which is the balance model, right? So Soul is like the superpowered version. Terra is the balanced version that most people use day-to-day. And then you've got GPT 5.6 6 Luna which is a faster more affordable model for high volume work. So think for example like sub agents or chat cheapy instant style process right. So that is the setup. You got an announcement itself which is soul for the hard long horizon work terror for everyday production luna for cheap high volume jobs and three options which I think is a lot clearer than what they've currently got because for example right now they have GPT 5.5 then GP 5.5 instant and then they have pro and it's just like it's super messy. I think this is easier to understand. Now, here's the bad news, my friends. So, that is, as exciting as that is, you cannot actually use it yet. Read that one twice. EP 5.6 is a limited preview. So, I said about 100 before, no, it's actually about 20 partner organizations via codeex and the API only at the request of the US government. Now, apparently, and this is not confirmed because it sounds like they're just hoping, General Access is in the coming weeks. So, the most capable model on Earth right now, and it does beat Fable 5 on benchmarks, is one that you literally cannot touch. And this seems to be a trend, which is, you know, all these preview models come out, we hear about them, we see the benchmarks, but we actually can't use them. And I'll come on to how you can get around that in a second and what the good news is. But you can see for example open have announced we believe in broad access and plan to make GPT 5.6 soul terror and Luna generally available in the coming weeks. Now one thing to note here is it might be limited to country bycountry basis. So if you look at these situations, this is interesting because it might be a case where it's only available for people in the US. You know, we're seeing that with, for example, Fable 5, it was banned for non US residents and and non- US nationals. It might be the same with GD 5.6. Who knows? Who knows where this is going? Also, it tops Terminal Bench, and that's the headline number. So soul sets a new state-of-the-art and terminal bench 2.1 which is the test for real multi-step command line and agent work. So that's a genuine win. But at the same time you can't even access it. So you know it's all good. It's all good fun. But you know we can talk about this all day but it is all theory. We don't know until we've tested it. I always like to test stuff first before I actually see the you know before I even think about the benchmarks because the benchmarks are totally different. So let's have a look at this. Soul is a new flagship and a step function better than GPT 5.5. Terra delivers performance competitive to GPT 5.5 at 2x lower cost and Luna is the most costefficient model delivering strong capability at low cost. So together this is for developers and people. They say together the GPC 5.6 family gives people and developers that must be cut off. Yeah, more choice in how they balance intelligence, speed, and cost. So, if we look at the benchmarks right here, let's see where we're up to. So, Soul Ultra from GPT 5.6 is scoring 91.9 on Terminal Bench 2.1. GPT 5.6 Soul is scoring 88.8. Bear in mind the Soul Ultra and the Soul. So, there's two different versions right there. And then you have Mythos 5, which is 88%. Bear in mind the public never got access to Mythos 5. So something to bear in mind there. It was only available for project glasswing and now Fable 5 and Mythos 5 are both completely gone. G5.6 Terror is at 84.3%. And then we have Fable 5 which is on par. So you might be thinking okay Terra is in the middle not that good on S uh on terminal bench 2.1. These are par for par. Again, I would test it yourself because these are just benchmarks that are totally I mean, they're already theoretical, but now they're totally theoretical because you can't even access the model. And if we look who wins here, for example, Soul Ultra is crushing on benchmarks. But if you look at SW Bench verified, Fable 5 and Mythos are still beating GPT 5.6 Soul. Now, let's talk about the whole situation here. So over have shipped three new models, but the model was never the boat, right? And these are all strong models that were awesome. They may come out, they may not. Nobody knows. It's kind of like it sounds like they're just hoping that it will get approved to come out, right? And so you got GT 5.5, you got Fable 5, Myos, GT 5.6. The thing that I would say here is it doesn't matter what comes out next because the old way is like people would rebuild the whole setup on every launch, all their prompts and the treadmill never stops at that and it's a mess. I don't recommend that for you. The way that I would recommend using this is you have a system and you swap out the models depending on whatever happens. So, for example, if cheap 5.6 comes out tomorrow, no problem. We're going to plug it into Codeex. If Fable 5 comes out tomorrow, no problem. We're going to plug that into Claude. If when Fable 5 got taken away, we took it out within two seconds and we still had amazing systems. And the thing to note as well is like it doesn't really matter about the models because 99% of people don't need them, right? Unless you're you're trying to build something absolutely insane, which 99% of people are not. They're just trying to automate social media or or trying to automate lead generation or whatever it is. If you're just trying to do that, you don't need these these frontier models. You know, you can easily automate it with something like GLM 5.2, too for example. So my point here is that models come and go in weeks but the system is constant and that's the thing that you actually own and for me personally you might relate to this too. It's I used to chase every model and now I don't really care. So for example GPT 5.6 situation this morning interesting to read about but it doesn't phase me at all because the models can come and go. You know, I I was there testing out Chat GPT 3.5 when it came out, right? Or chat GBT3. And back then, you would chase the models and you'd switch your prompts and you'd be looking at all these different models in different tabs and messing around. Whereas now, it doesn't matter, right? Like for example, if Hermes has a new model, we'll plug it in like or if it just came out yesterday, we've already plugged that in as a local model into our AI agents. But it doesn't matter what models come out because we've got great systems. And that's the difference here. So now when a new model drops, I feel nothing but mild curiosity. I can add it in the system in one line and keep working. Last week for example, um I can just swap everything onto a local model or I can set up a new section for new models. And the main point here is like stop chasing the models, start chasing the system. Now, here's what I would recommend setting up your systems around. And this is what works for us inside the agent OS. So if you look at this system right here, this agent operating system, really we don't care about leaderboards because a model proof system doesn't care which model is on top this week. And here are the five things that actually make it work. So number one is building on a swap layer. So we can change one setting inside our system and then that decides which model runs. So, for example, if we're using Hermes, well, I actually have separate agent profiles for each API that comes out. North Mini Code came out recently, we plugged it in. GLM 5.2 came in recently. We plug that in, right? We can switch and swap and change between all these different Hermes agent profiles depending on what we want to use. So, that is the swap layer. Number two is keeping many models on tap. So, for example, you can have multiple different local models plugged into your system. We have a a local section down here where we can build and automate and then preview what we've created. And then we can have a look inside the workspace and see everything we've created. But the point here is like if you've got many models ready to go, it doesn't really matter about this. You could use local for the free private high volume work. You could use frontier for the hard 10%. If you've got soul or fable 5 or mythos or a local model all sitting behind the same agents, then you're good to go at any point. The other thing I would say is root each model to the right uh route each job to the right model. So for example, if I am trying to build something absolutely insane, then I might use something like Fusion, right? Fusion has a panel of agents all work together. If I'm just trying to do something super basic, then I can use my local model agent. And the other thing I would say is own the memory as well, right? You don't need to own the system, but own the memory. So you can see my memory model here. This is a memory galaxy powered by Obsidian. And every time I use my agents, all of the memories inside my models are updated automatically. And so I'm super flexible because no matter what model comes out, I can still plug the same context in and it's automatic because it's all inside the system. So if we plug in, if Fable 5 comes back out tomorrow, we can plug that into Claude and it's going to run off our memory galaxy, which literally doesn't care which model comes out next. And the main thing I would say here is like never chase a launch again. when the next model drops which will happen days from now might happen tomorrow might happen next week you can just add one line and that's it right everyone else spends a weekend migrating their systems and the workflows you just spend it building and that gap compounds so whatever happens next when GPT 5.6 comes out or when Fable 5 comes out, have the models, have the system ready and swap the models in and out. And that way you've got the swap layer and you can just do whatever you want, right? Whenever you want. Now, some people say, "I need to be on the newest model, the best model." The best model changes every few weeks. The best system doesn't. So, Soul is the new Frontier model that's beating Fable 5 on many benchmarks. Last month, it was a different name. Next month, it will be another name. anchor on the thing that stops moving, which is the system. And that way as well, you never feel like shiny object syndrome again. Other people say, well, switching models means rebuilding my setup. In a real system, switching is one single line of code because it's just a EMV file, right? If a new model launch costs you a weekend of migration, you don't have a system. You have a pile of tools wired together in one provider. So, you want to fix a layer, not the model, which is why you need an agent operating system. You want to be model proof. And other people say well whoever tops the leaderboard is the right choice. The right choice depends on the task. So you want to root you don't want to marry terminal work to today's um what I mean is like basically it doesn't matter who the leaderboard is because that changes all the time and also it's very subjective like you can see amazing benchmarks on something and then when you test it out yourself you're like actually that's not that great. So, let me give you an example of this. So, for example, Fugu Ultra has similar benchmarks to Fusion, but when we tested out, Fugu Ultra's um outputs weren't as good as Fusion's. And we actually have like a bunch of tests you can try on Goldiebench where I show you all 42 live demos and you can just test them out and see what you think. But my main point here is like not everything that's created is that good, right? Like it it might be great on benchmarks, but it's not great in reality. and what I focus on is reality, right? It's like how does this impact me dayto-day? So, those are the things and I think the people who start stop focusing on models and start building systems that the models plug into, they're the ones who are going to win. Uh, other people say as well like I'm not technical enough to build a system like this, but you don't build it, right? The agent OS is already the system. You can get it from us inside the AR profer link in the comments description or just go to the arpuffer.com. You could build yours from scratch as well. I I usually spend about three to four hours a day just improving and releasing new updates for this sort of stuff because I want to make it as good as it possibly can be and I want to help everyone inside our community. But either way, focus on the system, not the model. So, the next model will drop next week. You know it. Stop preparing to chase it. Build on a system that swaps any model in and keeps you working. And if you want my exact one, the model proof agent OS is set up and waiting inside the AI profitable boardroom link in the comments description or go to the AI profitable.com. And this is my AI automation community that's focused on helping you save time, grow, and scale with AI automation. You might be saying, "Okay, this stuff sounds technical to set up, Julian. Like, you know, I'm not techy." Well, for me personally, I'm not techy either. And also, I've seen a lot of people set up our agent OS system and absolutely love it. We've got 191 pages of testimonials and wins from community members. So, I know like if we can all do this, so can you. And it's just a fantastic place to share and learn and grow together on this journey through AI. Right? Inside the community, you can ask questions, get help and support in real time. Inside the classroom, you can get access to all my best trainings, learning, so you can grow from there. Inside the calendar, you can actually jump on weekly coaching calls, get help and support, share your screen, and inside the new daily tutorials here. We drop new stuff all the time based on what's actually useful and helpful for you with a video tutorial and a full step-by-step guide. And if you want the agent operating system, it's right here with video tutorial. You can see the last update date. So, we updated it today. You can get the zip file for that, too. So, hope to see you inside there. Link in the comments description or just go to the aiprofitwarm.com. Heat. Heat. Hey, hey hey. Hey, hey hey. Heat. Heat. Hey, thanks. Heat. Heat. Test this out but in more interesting variations. Try different colors etc as well. Heat up. Heat. Hey, Heat. Hey, hey hey. Heat. Heat. N. Hey hey hey. Hey. Hey. Hey. Wow. Wow. Here we go. Heat. Heat. N. Today we have a brand new update from Hermes agent and they've released something called Hermes mixture of agents presets. So this is a way to combine multiple models that you can access into one and then you can choose from a powerful variety of mixture of experts, a mixture of agents that work together in parallel. It's kind of similar to if you've come across Fusion, which achieved fable five level intelligence by having multiple models work together. We also saw this with Sakana Fugu that came out this week and that also achieves fable five level intelligence. And this is a way of basically getting multiple answers from different AI models working together. And the reason that they're doing this is because if you look for example at the release of GT5.6, six, which is a limited preview today. We also have the release of Claude Fable 5 that is gradually coming back, but only for 100 different approved partners. All of these models are getting gated, which means that for example, if you want to have Fable 5 level intelligence or Frontier level stuff, well, it's very difficult to do that with the new releases that are just in preview. So, how do you get around that? Well, what you can do as an alternative is have multiple agent models or multiple models working together to give you better answers. And this is essentially what Hermes agent are now releasing. And they're also working on a new benchmark which is Hermes Bench. And you can see on Hermes Bench, they've announced uh on Hermes Bench are upcoming agentic benchmark Opus 4.8 A and GPT 5.5 reference mixture of agents scores 8% higher than Opus and 11% higher than GPT. So basically you can improve the performance of your AI agents by using this system instead of relying on just one model at a time. Now obviously I can imagine that would use more tokens but at the same time you can see that mixture of agents will work better together than just having one model working in isolation. Now at this point you're probably thinking okay how do we get this working for me. So what you can do is if you go into your terminal you can type in Hermes model and this is a new setup so make sure you're updated. First thing you want to do is just run Hermes update inside your terminal. Or you can do that inside your dashboard. If you go to the Hermes manage section, scroll down and click on update Hermes. Once you've done that, then what you want to do is go to Hermes model. So you type that inside your terminal and then from here you'll see an option that says mixture of agents and these are named presets. So what that means is basically they have preset options for multiple different models working together with mixture of agents. So these are kind of like pre-made formulas where you can have these models working together. So if we click on enter here, we can then start switching those around. And then once you've done that, if you go to your model section, click on change inside your dashboard. You will see the model settings here and you can load mixtures of agents. So you can configure this directly. So this is coming soon. And they've also talked about fable five level stuff as well. Right. So they've said you might be wondering as we did if it is Fable 5 level. And obviously nobody really has access to Fable 5, but the Hermes benchmarks look promising according to Rome words. So it's a pretty cool way to work around this situation with the preview models and it's now available inside Hermes agent so you can get access to it. They're also going to be releasing a new benchmark for all this stuff too. And it's also provider agnostic. This is not limited to for example like newsportal or open router. It's provider agnostic which means you can plug in whatever you want. Now if you want the full documentation on that they've got that inside the Hermes agent news portal details. So you can select a preset through the normal provider services. So for example, you can type in model default provider MOA or model review provider MOA. and these are the terminal commands you can use to switch between them. You can also use these on agent loops as well. And if you want to configure the presets, you can configure that from your dashboard. You can also do this inside the desktop app and you can do Hermes MOA configure and then configure the name of your Hermes agent profile or you can go to the config.y ML. Now if we look at what's performing the best on these Hermes bench scores, you can see that Opus Aggregator, so Opus 4.8 plus GP 5.5 is scoring the best. And then you got Opus 4.8 A and GBC 5.15 below that. So if you're wondering, okay, does this actually work in reality? A panel of experts typically beats one genius. So if you picture for example like one brilliant person answering a hard question alone and then you have a panel of brilliant people each one writing their own take privately and then a sharp chair reads all of them and gives you the best combined answer well the panel would win every single time and that is mixture of agents the reference models are the panel the aggregator is the chair one question goes in then several models think and one clean answer comes out better than any single one of them could give and it's the same. For example, if you're using something like Sakana Fugu or Fusion, let's take a look at Fusion over here. So, all of this stuff that you can see here was created with Fusion directly. And when we test it out side by side versus stuff like Opus 4.8, you can check that out Goldie Bench, it outperformed pretty much every single model on the leadboard. If we actually open up Goldie Bench here, you can see that the mixture of experts, mixture of panel system, Fusion is outperforming pretty much everything else on the leaderboards here by a long way. So, it's a really powerful system that actually works and there's multiple ways. You don't just have to rely on Hermes to do this. You could use Fusion. You could use the kind of Fugu. Whatever you prefer. Now, it seems as well when I've checked this out that I I don't know if there's a limit on the number of agents and presets you can have, but it seems like it's just two models working together inside Hermes agent. But the main thing you want to focus on here is like stop chasing the model. Build the system instead. So, everyone else is like waiting on the next model, the next Opus, the next GPT, the thing that'll finally change everything. But if you look at what happened, a mix of today's models beats the best single model that's not available on the market anymore, right? And there's no new release of that. There's no gated access. It's just a smarter way to use what's already there. It's more efficient. And that's a lesson that mixture of experts or mixture of agents has you for free. I keep saying mixture of experts. I'm so used to saying that with local models. Mixture of agents has you. So the model is not the mo. The system around it is. And the model is a part you can swap. The system is the thing you own. And the timing on this is huge, right? You got Fable 5 in preview. You got GPT 5.6, which isn't confirmed to come out yet, but Open Eye are hoping they can release it to the public. So, the winning move isn't waiting. It's squeezing more out of the models that you already have by combining them. So, this is how you can get them working inside your system. I mean for example if you look at our setup here we have Hermes agent and we can use mixture of agents inside this section but then we could also switch over to sakana fugu and start chatting over here with it using the API and we can see everything that we built with it directly here as well right and it's the same for example with fusion we can chat with it anytime we want it's just one click away rather than a separate tab and we've got a system where everything that we built with this is all ready to go and preview inside one easy system and that's really the the method Now, you don't want to focus on the model. Focus on the system instead. It's really a pattern. I've been running this for weeks now. So, you know, mixture of agents. It isn't like a one-off trick. It's a pattern that I'm seeing recently. I've already built two systems working on the exact same idea, which is a panel of models fused into one answer. So, we've got fusion inside the agent OS as well. We've got Sakana, Fugu, and now we can use mixture of agents as well. Three different systems that were built into the agent OS to get the most out of this stuff. And you can see how they would work to get the best possible outputs you can from your system. And this really changes everything. If you look at the old way, it's like pick one model and hope it's the best. Hit that model ceiling and stop there. Wait months for the next release to save you. Beg for access to the gated Frontier models. And then when they finally do come out, you know, it it can be quite expensive using the APIs on those. Whereas if you look at the system, you move beyond the ceiling because you can mix several models into one virtual model. You beat the best single model today with mixture of agents. No waiting. You can squeeze more from what's already available to you. You go beyond the public frontier with no extra access needed and you hit frontier quality for a fraction of the cost because you got these agents working together. Bear in mind like you could have cheaper models working together and still get better outputs than if one of those models was working alone and it's just one command. It's just for/ MOA and then you can switch around it. So if you want Hermes agent with the panel built in with uh mixture of agents, you got Fusion, you got Sakana all running on the same idea, we have all of this inside the agent operating system which is a dashboard that I run pretty much my whole systems on. Right. So the full agent OS zip Hermes MOA. Yeah. A Fusion Sakana all wired in coaching calls where I build model panels with you live daily tutorials. So you're stacking models, not just reading about them. And a room with 3,800 operators running this exact stack plus the prompts, the presets, and a memo map for your city. You can get that inside the AI profitable boardroom link in the comments description or just go to the AI profitable.com. And this community focused on helping you save time and scale with AI automation. Inside the community, you can ask questions and I personally answer them with daily tutorials. Inside the classroom, you can actually get all of my best trainings including, for example, we've got trainings on Sakana versus Fable versus Fusion. We have a full system on loop engineering to achieve better outputs. And we also have a full training on Fusion versus Fable 5. And if you want the agent operating system with all of this built in and ready to go, you can see that right here. You get the last, you can see it's updated daily. You get a full zip file plus a 30-day road map to implement it inside the calendar. You can jump with coaching calls, get help and support in real time. Inside the map, you can meet people in your local area who are building with AI agents like you. And this is all available inside the AI profitaboom link in the comments description or just go to the Aprofit.com. So, we've got brand new updates on Fable 5 and Mythos today. So you can see here this was announced by the code letter and basically what they've said here is that the Trump administration has struck a deal with Anthropic which grants a company permission to release its Mythos 5 model to a group of 100 companies. So, if you're wondering, okay, when is Fable 5 come back out? Who's got access? When is it going to be released again? This sort of confirms it, which is that it's gradually moving towards access again. It's not completely removed. It's just that a smaller group of people have access to it. We're seeing this, for example, with GD 5.6 as well. Today, GT5.6 has been released, but only in limited preview. And I think what's going to happen now is that all of this stuff gets tested more relentlessly by a smaller group of people and then we could be looking at this coming back out next month. The other thing to note here is what's going to happen next in terms of is this only going to be US-based? Is everyone in the world going to get access? Is this the end of getting access to Frontier models? Do we have to allow mixture of agents or mixture of model panels instead to get better intelligence? Pretty interesting stuff. So if you're wondering, okay, what happened so far? And this is according to CNBC as well. So senior Anthropic uh staff apparently flew to Washington to meet Trump. Then Anthropic said earlier this month that it's been disabled and then the Trump and Anthropic have been like going back and forth on the latest models right here. And also if you're wondering, okay, how does this relate to OpenAI? Well, that's been gated too. So, if you haven't seen the news on that, like for both GT5.6 and Fable 5, they've both been gated in limited preview. So, the good thing is this is a partial lift. MIOS 5 goes to a vetted list of 100 institutions, not fully public, but I can imagine that Fable 5 will gradually start working towards being released uh to the public in the future as well. So, you can see this tweet by Chubby as well. said Anthropic is reportedly moving toward a deal with the US to lift restrictions on both Fable 5 and Mythos 5. Um, and apparently Bloomberg has been reporting on this as well. So let's have a look at both of these. So this is CNBC's setup as you can see right here. So Mythos 5 is available to a group of companies and federal agencies. The decision marks a step forward in the negotiations. So that is good progress in itself and that is available just for trusted partners right now. This is pretty interesting as well because there's so much going on with this, right? Like for example, Hermes agent, they just released a new mixture of agents model to get better outputs like Fable 5 level or better than Opus 4.8. So that's their solution to this. And we've seen, for example, Sakana Fu and Fusion as well drop, which are panels of agents that work together to get better outputs than, for example, Fable 5 on benchmarks. I've tested it myself. It was pretty impressive to be fair. And then for example, you've got China and China with GLM 5.2 just got released and that is pretty much opus 4.8 level but it's open weights. It's open to the public and so if you look at what's going on here it's like okay well GM 5.2 is open weight. China releasing open source models all the time. US is gating more models and less people being able to access to it. And then you've got for example all these solutions coming out like mixture of agents and panels of agents working together to get better output side by side. And then also you know this is a really good point is like by the time you get it will this be the frontier model or will there be something else? you know, so it might be like we're all publicly facing, we're all like a year behind or we're 6 months behind the latest models because we don't get access to the stuff until it's actually been fully tested by uh approved partners. This is pretty interesting as well. So German mentioned like you know he no longer believes that people outside the US will still have access to Frontier models or that their access will be limited. So we might be looking at the end of an era where we used to have access to frontier AI intelligence but now we might not be. That is pretty interesting. I mean the way that I'm seeing is like you know if China keeps releasing these frontier level models open source and then the US keeps gating them then I mean already more and more people are using Chinese models but that trend is only going to continue. And obviously that's not good for competitive companies like Anthropic or like OpenAI. It puts them at a massive disadvantage really unless they only focus on enterprise and even then I mean you can imagine all the all the gatekeeping with enterprise you have to be even more careful. So if we look at the timelines here you know on the 9th of June Mythos and Fable 5 came out on the 12th of June both those models were pulled worldwide on the 25th of June there's been a partial lift but just for 100 companies so that's pretty much where we're at right now. And bear in mind like the public didn't get the model back just a list of approved partners did. Now the thing that I would say is don't focus on the model focus on the system. So back in the day before I set this up which I'll show you in a second which is our agent operating system. You know, back in the day, I would like chase every model, be super excited about what's coming out, looking at the benchmarks, really focusing on that. Now, what I do is I just automate everything that I actually spend time on and work backwards from there. And I think that's a much smarter way. So, for example, we set up this system where, for example, Hermes mixture of agents update came out today where you can achieve potentially Fable 5 level intelligence. We've already plugged it into our agent OS, right? Fusion came out last week. Super powerful system. you can achieve Fable 5 level intelligence, no problem. We'll plug it into the system. And my point here is like don't chase the models, chase the systems. So when a new system comes out, plug it into your setup and then you're good to go. If you want our setup, it's inside the AR profitable link in the comments description or go to the profitboard.com. But the main point here is like a story like this, it doesn't really rattle me anymore because I stopped building on the model and I started building on the system. If my 5 gets pulled, if Fable 5 stays dark, fine. My agents run on whatever's reachable. My memory, my prompts, my workflows, they live in one system that can point to any model. It doesn't really matter. So when a model comes back, I plug it in. When one disappears, I will root around it. And you can build this way too because the gate stops mattering when the system is yours. And I think the thing that I I I I think about this is like, you know, if you looked at your setup right now with AI, ask yourself one question, which is if my main model vanished tomorrow, would my work stop? Because if the answer is yes, you need to fix that. And you can start fixing it today. I think the people who get caught out, for example, with Fable 5 were the ones who had everything in one place, who put all their eggs in one basket. The people who stay calm built a system that doesn't care which model is up and you want to be one of the calm ones. So that's why we set up the agent operating system. I mean this is more powerful than anything I've seen available publicly. Like for example, Hermes Oracle, it can pull in the latest news. We can create content to our website. We can create social media content and it organizes it by trends and automates automatically improves itself every single day. Right? If we look at Hermes Jarvis voice activated version of Hermes powerful AI agent, we can switch between auto and agent mode. We can build anything that we want with it. And we can see our full conversation right here. Hermes comes out and then orif comes out yesterday. Super powerful local model. No problem. We've already plugged that into our system along with GLM 5.2 and Quen 3.7 or whatever comes out tomorrow. That's my point. is like for example no matter what model comes out we've still got all of this context and this memory galaxy so that we can plug that into new systems and swap out the models I literally like I I really don't mind what models come in and what models come out right so if you look at the old way it's relying on one single model and the problem is like let's say for example a model changes their prices or for example a model get swapped out or you're on a wait list that's going to be a nightmare right you don't want to be in that situation but that's where 99% of people are. Whereas the new way is you have one system and any model. So you build your agents on a system, not a single model. You run on whatever model is reachable right now. You can swap models in seconds when access changes. If Mythos is pulled, you can route to Claude or Gemini or a local model. Your memories, your prompts, your workflow stay put. And the result is the work never stops. It only improves gate or no gate. So that's the way that I would look at it. So if you want a setup like this that is basically super anti-fragile, very flexible, improves every single day, no matter what happens, the agent operating system inside the AR puff boardroom connects Claude, Open Claw, and Hermes into one dashboard with shared memory. Your agents run on whatever model was reachable and you swap models without rebuilding everything. So if Mythos gets pulled, if Fable stays dark, your work keeps moving and improving on the next model in line. So you get the full agent operating system, every prompt, the memory setup, the dashboard. You get four live coaching calls a week with operators running this in production. You get daily tutorials as the model rules change, what to do plainly. A 30-day road map to set the whole system up step by step. A community of 3,800 founders now across 38 countries. There's always people online ready to help you and a member map to connect with builders near you. If you want to get all that, it's inside the AR prof description. This is my community that helps you stay focused and improve no matter what happens. So, for example, inside the community, you can ask questions, get help and support. I personally answer all the questions inside there daily. Inside the classroom, you can get all of my new trainings, including new daily updates. So, if you want the Agent OS, it's right here. You can see it was just updated today, and you get a zip file to install it quickly. And then inside the new daily tutorials, you can see everything set up step by step right there. So, for example, if we have a new update like Hermes learn, we set that up inside the classroom. And also if you're complete beginner, you can go from beginner to expert over here. You can also jump on weekly coaching calls, get help and support in real time, share your screen, meet other people, see what they're working on, and then inside the map, you can meet people in your local area who are building with AI agents like you. So, hope to see you inside there. Link in the comments description or just go to the aiprofitboard.com to get access. Thanks for watching. All right, my friends. I think we've pretty much answered every single question inside the comments, inside the classroom, inside the community. Boom shakalaka. Thanks everyone for joining today. That was a lot of fun. Enjoy as always. I'll see you tomorrow in the next one. And cheers for watching. Bye-bye.