If I Started AI Automation in 2026, I’d Do This

        Do you ever feel like every new AI tool is the next big thing, but you feel stuck on exactly where to start? I run a 7figure AI automation business, I sold my last startup with 60,000 customers, and after doing this every single day for multiple years, if I had to restart with AI automation, this is the exact road map that I would follow to save time, stop overwhelm, and give you the systems that work without the need to relearn everything new every 3 months. so you know exactly what to do and what stuff to completely ignore. But the thing is, no road map is ever going to help you unless you understand the key mindset shift because what worked in 2025 and 2024 is not going to work in 2026. And that's because everything is changing which can feel super overwhelming especially if you're trying to grow your business and the new tactile changes. So, look, I know, I get that it can feel confusing, but in this video, I'm going to take you from this guy on the left who's confused to this guy on the right. And you're going to say, "Jack, did he just deage 10 years?" Well, look, these kind of great things happen in the Jack Roberts YouTube channel. But I want to go back in time for a second to 2021. And let's think about Make.com, a no code builder, as an example. Life was simple. We were just talking about connecting Gmail modes to, you know, AI and it was very, very cool. It was beautiful and things were fantastic. And now we're building full-blown SAS. We're seeing examples of businesses that are canceling their providers to build their own SAS tools. And essentially, we can automate and build anything that we can possibly imagine. And everything you see is crazy, insane, insane, wild, blowing your mind. And I don't blame any AI creators for this necessarily. It's just that topics and videos drive this in the AI space. Anything that has the word insane in it, anything that's big, that is new, will get the clicks. It's the stuff that you click on. And because of that the incentives are such that people will create more content that is more outliers. But the problem with this is that if everything is insane, everything is gamechanging. Nothing is. And it becomes very difficult for the business who wants to scale and uses technology to understand where you actually need to allocate your time, which is what we're going to address in this video. So what on earth do we actually do? We stop focusing on tools and start learning systems. Most people find a tool a technology. They go down a huge rabbit hole and learn everything about it. when in reality, you're not going to use 80% of that knowledge. We want to bring it back down to fundamental systems. And AI is a tool that we use to achieve an outcome. And if you have that mindset, you're going to save yourself hours and hours of time. And when we're thinking of systems, we're thinking of inputs, outputs, bottlenecks, constraints, and data flow. And the skill of workflow building is one of the highest leverage skills that you can learn, whether it be in NA10 or meg.com. But this is fundamentally changing for 2026. It is becoming commoditized. Blueprints used to be super rare. Now they're a dime a dozen. And the key differentiator is building in systems. And I just want to show you how far this technology has come. And I'm going to show you this blueprint. So for example, if you come into any temp for instance, you can now just speak to it to create workflows. Like if I come over here to create workflow and if I want to build something really simple, I can click on build with AI and I can say something like, "Hey, I would like you to build for me a Reddit scraper that will scrape the top 10 posts from the NA10 subreddit and then I want you to create a Google sheet and put them in there with a hook for short form content on Instagram." You can give it posts like that. Of course, we're using Glider Beautiful G for our text to speech. But the point here is that you can actually do this stuff now just through text. Now, understanding system design is really important, but we're seeing the technology evolve significantly. And so, the principle here, and as you can see, it's now working in the background, is that you want to focus your time on the 20% of tools that are going to get you 80% of the results. And as you can see, guys, it will literally go ahead and create this this entire flow for us. Now, we've got all these different posts from your any subreddit. It's got the text. It will now go ahead and generate some hooks. And the key here is that workflow building is incredibly useful. This technology is evolving very quickly. So these are the exact five things that I would do in 2026 learning AI automations. The first thing I would do is start with foundations. So if you're a complete beginner, you need to understand the basics of AI. Understanding stuff like chat GBT, Gemini, some of the really basic tools that's going to get you really big quick ROI and some of the specifics. I'm just going to put my classroom for this cuz I organized it in this way specifically. If you look at foundations, you want to think about skills that basically aren't going to become irrelevant. So for example, the typewriter was great, but now it's just an antique that gathers dust in an office somewhere because technology rendered that thing irrelevant. So there are certain number of skills that we can learn that regardless of the technological changes are always going to deliver you a really high return on investment. But before you can understand where we apply AI, how we learn it, how it can accelerate your business or your agency, you've got to understand the foundations. That's things like prompt engineering, context engineering, understanding things about memory, AI fundamentals, AI safety. And when you have this kind of core fundamental series of understanding, you have the context in which you can understand everything else. So the very first thing that I would do is learn the foundations and the basics. And I'd have fun with models like this. I'd learn how to create images. I would go ahead and I'd learn tools like Notebook LM that lets you basically add any conversations, talk to them, and kick some ass with it. For example, you can feed this as many YouTube videos that you want to. It really will accelerate your learning. There's so many things you can do on level one on the personal side. That is the most foundational thing that you can do. It basically means you learn information faster than everybody else and get started. So, don't skip the foundations. Learn the basics. Get the blueprint sorted out. Second thing that I would do is build my very first AI automation. Now, I'm not trying to become an expert in make.com architecture. I'm not trying to become super nerd in NA10. I'm just trying to build one thing that works. Like I'm trying to connect a chatbot to an agent that can then send emails. That's it. I need to get that magic moment of oh my gosh it works. So you can see it for yourself. So this is not about going deep and you there's no shortage of tutorials online that will spend 20 hours, some of them actually 20 hours that want to go down every single crevice. Honestly guys, it's it's the classic. There's like 20% of things that will give you almost all of the results. You don't need to know so much crazy specificity and everything. I would build my very first automation. Make.com is great if you're just getting started and you're learning. You want visual you can see things pulse and things just to get an understanding and then you can have nm which is another awesome tool to learn a little bit slight more of a super learning curve but really really simple build that very first automation. Step three will be then to graduate to learning more about NA10 and getting into intermediate to advanced level stuff. Here you'll be building things like AI agents and systems like understanding retrieval augmented generation and understanding how you rerank slight more complexity. Once you reach this level, you probably be able to build like 95% of stuff within an hour or so maximum. It really does. It's like a little bit of learning goes an incredibly long way, especially if you focus on the things that actually make a difference, which is what we like to get on this YouTube channel. At level three, you want to be looking at learning more about going deep into NA and its configurability, but also then foring into things like these, which are your vibe coding platforms where all we do really is just give it text. Um, give bolt text. You know, bolt new is one, Google AI studio is one. The best of these right now in terms of creating gorgeous, beautiful interfaces is Google's AI studio. So, for example, if I just even show you this one on the left hand side I've got here that I built a couple of days ago. The strength and the power and the design of this is exceptional. And all you really do is you just talk to it and it builds and it's crazy. And if you don't like what it builds, you can talk to it some more and it builds even more. So it is really crazy and having fun and just getting used to the concept of building this stuff and understanding it will really make a significant difference. So you can see and this takes you through to a full dashboard. I did a full video breaking this down just so you understand the differences very well. Lovable can connect to NA10 really easily with something called MCP which is what we call model context protocol and you can connect it and it means that you don't need to do a lot of plumbing but they all connect very well to different platforms and databases and here you learning about things like superbase which is kind of like if you think about Microsoft Excel on steroids you're learning about databases you're learning about different things I would understand the basics of how to cover all of these things off as a priority and step four would be learning how to use things like clawed code and also gem Mi 3.0. This takes you from this to Interstellar. Now, if you've got a business and you're scaling it, you don't necessarily have to press the buttons, but you should understand the basics of how incredible this technology is. I talk a lot about my last business. My CTO said to me, Jack, it's going to be a few months until we can build an admin dashboard that can show you all the stuff. And I built it in about an hour two days ago. So, this is truly insane. And what I want to give you here are the tools and systems so you can leverage this in your own business. It's going to save you insane amount of cost. It's stuff that you can sell to clients and your agency. It really, really is crazy. And in reality, your interface looks a little bit something like this. Now, you there's many different tools for this. In reality, guys, it can come down to preference. Okay? So, all it ever really is if you think about it, now I just shut this down and show you a quick instance. You can understand it. And all this really is is just taking your projects as folders on your computer, which you can access by coming up here, clicking file, opening folder, and using it. Now, this would be where I would build everything. On the right hand side, we have our agent. It could be called code in what we call the terminal. Don't worry if that's a little bit confusing. I'll I'll put a video basically at the end of the video so you can learn all about this stuff. And effectively, we just build and create here. And we can have agents that work for us constantly. We can have multiple agents working on the same things. And the level so we can build here is truly exceptional. And so with all of this tool confusion, what do you use when? Now, this is the key difference that is going to save you a lot of time. It's not about do I stop using N810, do I stop using make.com, do I move everything over to claude code. Think of them as tools for the job. Like a plumber walks into a house with a big beautiful box. At least I believe they do. And they have different tools for certain things. Sometimes it's the Allen key. Sometimes it's a screwdriver. Sometimes it's that big sucker off Mario when he's killing mushrooms. The point is that you want to use and think more in terms of system architecture and just tools rather than going really down into anyone specific thing. So the idea here is that based on the thing that you want to accomplish, you'll know exactly what to leverage and where. Which brings me on to the fifth step and that really is about your business and your agency. And this is the driver. The biggest mistake that I see people make is that they it's the tail that is wagging the dog in the sense of they are pursuing tools in the they're following tools in the pursuit of an outcome rather than having a desired outcome and then using the right tools. This is just like for example saying well there's a destination over there let's build a bridge to get there rather than the way that most people do it which is let's follow all these cool windy roads and hopefully maybe we'll get there. It becomes the tail wagging the dog. So you want to think in terms of the business the bottleneck the outcome I want to achieve and when we know what that is the tools actually make perfect sense. There is no shortage of shiny AI tools that will crush it. And that leads me very nicely onto the sixth bonus thing which is invest your time in assets and skills that will never ever ever go out of fashion and that even a grain of understanding will give you insane results. That is stuff like retrieval augmented generation which is a really fancy way of saying memory. We know that AI has a memory problem and one of the reasons we don't see it fully adopted in a lot of businesses sometimes is because it has a tendency to forget. It is confidently incorrect. Sometimes it can forget every like 2 to 10% of the time. Now, the idea is that memory is in just one example of an area of AI, but regardless of the tool changes, it's going to be so valuable in the future. Context engineering is everything. It's like why for example, it's what Siri was supposed to be, right? If Siri knew everything about your business, your personal life, your workout routines, you know, your favorite caps for example, whatever the thing is, it would give you way better tailored advice. And the way that and that's what we call context engineering. That's like one of the futures of AI and the big trends that we're seeing. The way that you achieve that is with good memory and retrieval augmented generation. So you want to invest in areas like that that are never going to get old and that the skills are going to compound over time. And understanding what to do in 2026 is one thing, but that means nothing unless you actually know how to implement that stuff. And you can build your first AI system with the latest technology by following this step-by-step guide right