AI Agent OS Just Got 5 MASSIVE Upgrades!
AI News Today | Julian Goldie Podcast
October 18, 2026
AI Agent OS Just Got 5 MASSIVE Upgrades!
Speakers Julian Goldie
Julian Goldie (0:00)
Today, I want to show you five brand new updates we've just dropped inside our agentic operating system to show you number one was possible, number two to inspire you, and to show you how powerful it is when you build a system like this together. So the first thing that we've actually built in is something called OpenCode, which works with three new models, right? And OpenCode are dropping three new free models every single day, it's pretty crazy. So for example, yesterday they dropped GLM 5.3 Flash for free. Now you can get Nemetron 3.5 Super VL for free. And there's also DeepSeek V41 Flash for free as well. And so what's pretty cool about this is you can basically mold OpenCode into your own agentic harness, and you can change any features that you like. So for example, I can switch the models over here really quickly. I can tell it exactly what I want to build. And then also inside the workspace here, we can preview everything that we've built directly, which is absolutely awesome because it's cool to see what you've built, and to see it in one place. The thing that I would say about this is normally when you're building out with AI agents, it's very difficult to see everything that you've built the previous week, what you've done before, etc. With this whole system, you can easily take a look and see and go back to what you've created. Now, what we've also done over here is we've improved the UI massively. So if you have a look at this whole setup, one of the best things about Opus 5.5 that only just dropped over the last few weeks is that it's way better at design. Now, what does that mean? That means you can make your agentic operating systems way more beautiful.
And if you're wondering what an agentic operating system is, basically it's a system where you can have all your agents plugged in one place. We've got OpenCode, we've got Hermes, we've got Claude, we have custom workflows and AI automations like the SEO OS, and we also have the video agent over here. We even have a game studio, for example. We've got a memory galaxy. And so you can have all your agents plugged into one place. We even, for example, added OpenDesign, which I'll come on to in a second. That is a brand new way to design, and it's really, really powerful. Now, also, what we've got over here is Hermes Agent. Now, not only have we improved the design and the UI, so for example, we ask it to do something. We'll literally ask it to do something. You can see here that it like speeds up and it looks really cool when you're using it. But also, if we stop that over here, we can use these new free models from OpenCode directly as the brain of our AI agents. So for example, we can use CATCHAMP, which is a new free model inside OpenCode. Step five, which comes from Hermes Cloud. You can switch them around. You can have DeepSeek V41 Flash. You could have, for example, Nemetron 3.5 SuperVL, and you can easily switch out the models and the brains and everything else. But the cool thing is, when you set it up like this, you can basically have all your agents working together in one place, which is absolutely amazing. And then if we scroll down to the memory section over here, this all gets updated across our agents. So, for example, if we do something with Hermes Nemetron 3.5 Super VL, that's actually going to plug it into our memory galaxy. Like you see right here, we get the new memories. And then we can basically have that plugged into Claude, into any other model we use with Hermes, into open code. It just saves a lot of time and it keeps everything way more organized, especially when there's so many models, so many genetic harnesses. The last thing you need is another app to open up or another tab to open up.
And the thing is as well, all these models are always overtaking each other. So, for example, sometimes you've got Claude, and that's absolutely amazing. But then sometimes Chat GPT releases like a, you know, let's say they bring out Chat GPT 7 tomorrow. Well, that means you have to switch everything over normally. But if you have a system like this, it's just like, okay, Claude has a great model. Well, we click over there, right? Codex is over here. We'll just use that instead, right? If, for example, Chat GPT bring out a new model. And also we can swap out the brain of our AI agents as well. So, for example, let's say you still want to use Codex, but you've run out of tokens. Well, the great thing about that is that you can actually use GPT's Astra, but then if you run out of tokens, it's like, no problem. Let's just switch to HY3. And then if you run out of tokens on HY3, no problem. Let's change the brain to OmniRoot. And we can still continue to build the same projects inside Codex. We can view everything inside our workspace. It's just super easy to have a look at everything that we've created over here. Now, one other thing that we've created, because Opus 5.5 is just so good at creating content, what we've got over here in this system is a way of basically analyzing our content. And then we have a three-step system for actually going from idea to posting content on our channels. And we can do it all from this one workspace, which is absolutely amazing. So what this means, we save a lot of time. We don't need the team anymore. I used to have a team who was writing the scripts, coming up with the content ideas, posting the content, etc. And that was a lot of work. With this whole system, we can just quickly go inside the write section. We write out a script, then we go to the edit and post section, and we just edit the content that we've created. You can see the style of content that Claude can do with Claude Motion, which is unbelievable. If you look at the way that it can design, the systems that it can build, the way it can basically design better than a video editor. But the other difference is that when it designs, it doesn't take Saturdays off. It doesn't forget something on a Sunday night. It's just doing it 24-7. And then the other great thing about this is that when we're posting, we can post directly to our platforms. So you can see we've got our platforms plugged in here with BlowTato. We can edit the captions before it goes out. And then boom shakalaka, we can see everything that we've created and all our views over here as well. So what this means essentially is that if we have a look here, we can see the number of views, the number of likes, the number of comments. We've got our next milestone to aim for. We can see which content has performed the best recently, and then we can see our recent top hits as well, which is pretty cool. So you see here it actually breaks down which ones have performed the best and how they performed across each channel plus which channel was performing the best. It's super easy to just manage this in one place. So if you want this system, by the way, you could build it yourself, but if you want to get my system ready to go, you can get that inside the AI ProfitBomb. Link in the comments description or go to the aiprofitbomb.com. If you go to the classroom here and then go to the new daily updates section, you can find our Agent OS system, and we show you exactly how to use it. You can see when it was last updated. You actually get the zip file to install it. And then we also add new daily updates, that you can see here, when I test new things out. I test everything out personally. And then if you have any questions about the Agent OS, you can post them inside the community. And inside the calendar, you can actually jump on weekly coaching calls, ask questions, and get help and support in real time. Link in the comment description, or go to the aiprofitbomb.com. So, so far, we've talked about OpenCode and the new free models for that. We talked about how we can take the brains from OpenCode, which are free, and then plug them into Hermes Agent. We've talked about the cloud system that can basically automate creating content, editing the content, and then posting the content, and analyzing the content's performance. So, three amazing workflows right there that have really just dropped, plus the improvements on the UI, which look absolutely awesome. It's just so much more fun. I mean, for example, we can see how our content is performing over here. We can see the latest news headlines and everything else, right? So, what is the final system? Well, that is Open Design, right? Open Design, this is a really powerful way to basically build and automate anything that you want. And it is an open designed system, which means that basically it's open source. We can edit it. We can see our workspace over here of stuff we've created. And then also, for example, if we want to build a new website out from scratch, it doesn't look like AI Slop. We can just click on a template here or we can describe exactly what we want over here. So if we click on this, for example, and then we're just going to hit enter, we can now start building an amazing website in like one single click. So this is Open Design. It's basically like an AI design agent workspace. It's a free open source project. You can plug any brain that you want into it, and you can see like in just build beautiful designs. Now, you can use this, for example, for slide decks, you can use this for image generation, you can use it for motion design, and you can also use it for building websites as well. And then it has all these powerful design systems. You can have plugins. It's got all your projects in one place, and it's basically creating like the ultimate design system, but it's open source. You're not limited to one model. So you could do some of this stuff with Claude, but then what happens when ChatGPT brings something else out, right? All your design systems are in one place, unless you use something like open design, and then you can switch out the model and the name whenever you want. And you can see here, it builds like super nice designs in like one single click. We literally didn't have to try very hard. And if we want to change anything, we can just go in the chat over here, and we can polish the design until we're really happy with it. Now, obviously, you can see you can use this for prototypes, for like web, desktop, mobile apps. So a lot of people watching this will be like, I want to create a website. I want to create an app. Here's how you can do it. Also, other people say, well, I want to create like a dashboard or some sort of live artifact that updates in real time. You can do that using this system. You might say, okay, well, I want to create a slide deck. You can do that with this or images, whatever you want, right? Even motion design videos. And it's all available inside this system that we've actually added to the Agent OS. So this is the whole system. It saves us a lot of time. It's really powerful. It's fun to use. And you can get it all inside the Agent OS system that we've got in the AI Profit Boardroom. Link in the comment description, or go to the AI Profit boardroom.com. Inside the community, you can post questions, get help and support in real time, and we'll answer your questions directly. Also, inside the classroom, you can get access to all of our best trainings. If you want the Agent OS system, you can go to the New Daily Updates and get the Agent OS system over here. If you want to jump on weekly coaching calls, you can do that over here. So you can ask questions, get help and support in real time, meet other members, share your screen, etc. And also, when you ask questions inside the AI Profit Boardroom, often what I'll do is actually answer them with a video tutorial to help you as much as we can as well. So link in the comments description or go to the AI Profit boardroom.com to get access. Thanks for watching.
Thousands of transcripts fetched by people building searchable podcast archives
Fetch the whole transcript
The demo key returns a sample episode in full, no card needed:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000796928640