Master AI In 2025 The Only Roadmap From Beginner To Builder (No-code, n8n)

You've seen what AI can do, but there is a huge gap between what these shiny, nice, cool things online that you see and what can actually improve your quality of life, improve the revenue that you generate as a business and ultimately make you more money. Well, this video is designed to bridge that gap. Okay? Forget the hype, forget the noise, forget the 90% of the information which doesn't exist. I Today I'm going to give you the full playbook of what is necessary in the artificial intelligence arena for you to become a success. We're going to break down the entire AI landscape together. I'm going to give you some of the key and core elements of life and business that will never leave. You'll learn the fundamentals that are necessary in order to understand AI as a brand new beginner with the core tools that you can't live without in the next five years. Moving forward, this isn't just another top 10 top 5 AI tools. This is a foundation for everything else in terms of AI. Stick with me. This is going to be part of a four or five part series and we're going to go all the way from this introduction to me exploring different AI solution tools to actively building with you. So let's go. So your roadmap to using the learning AI begins right here. This is a step by step guide to mastering the concepts, tools and skills you need to get ahead and make sure that I do it in a way or we do it in a way which is not overwhelming for you. So if you're new to the AI arena or you feel lost in it, this video is 100 for you. Now the first question is why learn AI? Okay, why? Why is it something that should take my attention? Well, AI has become non negotiable now, right? It's no longer an option. It's a core skill that is required for working smarter, sparking new ideas, new opportunities for businesses, automating your manual repetitive tasks and ultimately improving your quality of life as a business or of a business's quality of life by improving its functionality. Now this is your roadmap and you will learn key concepts to understand how AI works, the right tools to focus on so you're not bogged down by all the latest new flashy trends that this, this new launch of software. This version navigate the landscape with confidence that you know exactly what is important and where and a clear action plan. So at the end of this video and presentation, we're going to go into a step two, which is me explaining some more in depth information and Then ultimately going through moving for a few video sort of structure playlist, where we actually start building stuff together that can improve your quality of life without any of it feeling overwhelming. So the goal is to get ahead of 99% of the people. You're not late to the party right now whatsoever. Now is the time to learn. So that's why I'm producing this video at the day that I am, because I think that it's important for people who are struggling and feel that they're drowning to understand the AI space now. One of the most key and core things which is harming people's growth and development at the moment is the barrier syndrome. So everybody sits there and thinks, well, I'm not technically savvy. I'm not. I don't. I don't understand AI. Like, I'm too old to learn AI even is something that my dad has told me, which is completely untrue. We have to overcome common barriers, right? The AI landscape can feel overwhelming. It's overwhelming for me every day, and I do it every day for a living. So it's going to be overwhelming to somebody that doesn't understand. But the thing is, is that I have now ensured that I have streamlined the information that I digest so that it's only for specific things that can benefit quality of life, business or is indeed the future, even if I can't use it. The first myth is I'm not technical. I referenced it a moment ago. In reality, most AI tools are no code anyway. Curiosity and a willingness to experiment is exactly what you need if you have that. There are over 30,000 templates available across things like Gumroad, across NAN itself, make.com. there are 3,000 templates available for you to go and play with right now that take minimal to no setup. They're built for you already. I have videos where I've given them away. So that barrier doesn't exist. The technical barrier is not a thing. Obviously, when you. When it gets really detailed and deep and you're on about building really custom flows, nobody goes from being a child having its first swimming lesson to being an Olympic swimmer in the same day. You have to obviously build the foundations first. So right now you don't need to be technical. But you'll be surprised what happens with a little bit of effort and a little bit of willpower to be able to go through and how quickly you pick these things up. The second myth is it's changing too fast. Everything changes too fast. Everything in life has changes too fast. It's never the right time to buy a house, it's never the right time to have a baby. It's never the right time to start AI, Right? That is the same sort of pattern that every, everything in life follows just on steroids, just juiced up a little bit. So in reality, the fundamentals and core skills for AI do not change. Doesn't matter whether you use NAN or make.com for just as one example, the core fundamentals don't change. Focus on the core fundamentals and not every single new model that comes out. Because generally speaking, the ones that are important are actually just evolutions and better versions of the previous ones that existed. So they're more powerful. Yes, but you don't have to go and relearn everything from start to finish. The myth. There are too many tools. You only need three to five core tools to accomplish 90% of what you need, right? And only you know that while you're watching this video. But thinking of every scenario where AI can help or maybe AI can produce from a business perspective is not what you need. You need to take, let's say you own a business, you take your tech stack and you need to deploy that. And what is the most efficient way of using AI? Go from here. Take this. To overcome this barrier, go and find templates that use your tech stack, then put them into action inside of an end. You've got fully functioning flows within the set. Within the space of half an hour, 45 minutes, you've built out three or four flows that save you a little bit of time and you can expand your horizon from there. There aren't too many tools, there's just a lack of understanding about how you should use them and the focus and the direction of having the blinkers on to look at exactly what you need to for your individual circumst. So most people fall into one of three paths, right? And you can easily move between them as you go. There's no, there's no blocker, there's no barrier. You can easily move between each one as you go. The first one, which everybody on the planet should be, my 14 year old daughter is the everyday explorer, right? Somebody who wants to make life easier, buy back time. She doesn't necessarily want to do that, but she just understands that AI is a thing. When we're talking about adults like of age, you are watching this video. The goal is to make life easier and buy back time, summarize documents, write clearer emails and notes, transcripts from, from calls, you know, presentations, organizing notes. All of those things are everyday Explorer things, those are things that everyday explorers can do. The templates exist for them already, right? And they're there. You can find them, I give them away. You can watch content creators like myself produce them at scale. Then you have the power user got their goals to do more, faster, and they stack tools on top of one another for a multitude of different things. Content creation, in depth, research, analysis, quotation, production, complex problem solving and solution. And then finally you have the builder and the builder. From a perspective of the builder, from a perspective of, of what we're trying to do, they are the ones which create a significant amount of systems. From a point of view of like Nan, they will have this system, but then it will call this sub workflow and it will go into this and then this flow will have a sub workflow from it and it ends up being, you know, treacherously long, not the right word, but a really long fly. Well, these are the builders, right? These are the guys and girls that automate tasks, build custom agents and scale parts of business, parts of your business. That's where they fall into. It can get really, really complex depending on who the business is and, and what you're trying to solve. But the, the, the point is that you fall into one of these three paths. None of them are wrong. But pick whatever fits right for you. Everything starts with the first step, and that is the everyday explorer. Then you move to power user, then you move to builder. I did the same thing. Every single person you see on YouTube or is posting on LinkedIn does the same thing. And that's the way. Now this is where I want to simplify core concepts, because core concepts are extremely important and can be broken down into three sections. AI in the general umbrella as to what it is here. So the broad umbrella of intelligent systems, there are millions, right? There are so many that fall into this category that it's impossible for anybody to ever master them all. Nobody could could know, but they all follow. They all tend to follow the same core principles. And that's why there is, you shouldn't be scared. Google Calendar and Gmail, for example, example, they are under the same, the same category, of course, of suite, but they have different functions. You have different abilities with Google Calendar as you do with Gmail. So the AI umbrella of intelligent systems, it's so broad that it falls into vehicles, it falls into everything, right? AI is everywhere. Then we have machine learning. So AI subset that learns from data and patterns. So this is like calling individual systems. Machine learning can allow history to be used in order to continue to create new and improved responses or prompts, etc. That the machine learning element is a significant factor. I'm jumping ahead a little bit there, but machine learning is an AI subset. It learns from data and patterns, right? The process of how AI systems learn by finding patterns in data. Things you'll guarantee to have to know already, like Chat GPT for example. The more information that gets piled into it, the smarter it becomes. That's, that's machine learning. Now, when it comes down to generative AI, this is a machine learning subset that's focused on creating new content. So, and that's the focus of this guide, it's AI that creates new content. Text, images, videos, code flows. That's what you want to end up being. But it all goes back to, it all goes back to the previous step of where you want to be. You'll find your path by the end of this video and either you'll follow along and you'll, you'll continue to go through, or you will sit in the position of, of just being the everyday user. And that is absolutely fine. The AI tool landscape. There are thousands of tools, but they fall into a few main categories, right? And these are ones that are important for us to be able to use. There are many more. I could have listed hundreds, but I've broken down into the critical and core ones. Right? Number one, the first one on the list is obviously LLM. That's why it's in a big castle, right? Large language. Large language models such as Chat GPT are absolutely critical for functions for every everyday life. But there's a difference when using LLMs between being, you know, between being an emperor or an operator. An operator uses an LLM when it has a question, you know, what's the S P 500 at today? Whatever. Whereas the emperor, which is happens to be the 1% of people. 99 of people do that, right? 99 of people search for. 99 of people search for support on their day to day tasks to speed them up, to give them a boost in productivity. Whereas the 1% use large language models to build those systems that we referenced before. But ultimately the AI tool landscape is made up of six areas. Large language models, research tools, whether that be weather, whether that be like perplexity to do some deep research. Image generation. So creation of images, whether that be for your company or selling ad creatives and stuff, nano banana, image, imagen4dalay, whatever. From, you know, from a video perspective you, you have that, you. The, the VO3 image, video generation, audio tools such as 11 Labs VAPI there are so many that exist on the market and specialized wrappers that allow you to be able to to combine individual tools and piece them together. They all exist they the AI landscape is full of them. We need to stay focused on the on the actual goal. So when we look at this the tool deep dive so large language models what are they? The LLMs the large language models are the most important tool in your toolkit Neutral networks that they are neutral networks that are trained on massive amounts of data and tech of text to understand and generate human language to speak backwards and forwards. The key players in the game GROK Gemini Claude Chat GPT everyone knows Chat GPT Gemini is my personal favorite is the one that I use the most core functions. They're incredibly versatile. Use them for content creation, research, coding, translation and more. Key turns key terms to know when you look through this really really important terms prompts your instructions to the AI okay but can also deviate into a system prompt which which is what what you give what you deploy inside of a system to give it a brain like the chip that the the LLM users can be is a system prompt. But prompts generally speaking your instruction to the AI when you when. When you give it that information. Don't get confused with your message to the to the LLM here with that being a prompt. You are prompting an LLM by asking it a question but you are not giving it the the system prompt which is what it should do and be. You are a detailed research analyst for major medicinal companies looking at the laws on production of X right? That would be a system prompt. Very very vague. System prompt token a small chunk of text how AI's process language and calculate cost Use of tokens the larger your request, the more tokens you'll spend. The lower your request, the less tokens you'll spend. There's nothing to be scared of in terms of tokens. Hallucination is when the model confidently makes something up. Always verify important facts. You cannot trust LLMs 100%. A little tip for this. Turn the temperature from 1 to 0. Whenever you're using an LLM make sure you turn it down. I operate on zero temperature. You want to use LLMs all of the time. And if temperature confuses you, I'll quickly show you just a studio here just to showcase it to you because it's clear in AI studio this is the temperature gauge. Temperature should be considered as clarity. If you set it at 2 temperature it would give you something crazy. If you asked it a story it would Be wild about fairies and floating mushrooms and stuff. Whereas if you kept it at zero it would keep it more realistic and grounded. That's the way that you should look at, look at temperature hallucinations which is the core element of this is when. When AI decides to make something up, it decides to do something that it shouldn't and it leaves it in a position where it goes on its own tangent. Always Fact check your LLMs always with critical data. It's really important. RAG RETRIEVAL Augmented Generation RAG is important for grounding the AI's answer in real time data to improve accuracy and later on down the road when you get into our builds and we will cover this in one of the later videos in this series. RAG retrieval is something that is critical and allows you to pull from say like a knowledge base where it could only pull information the model could only pull information from a specific area where we know it's concrete. Those are four key terms to know. Don't be scared by them. So we're pretty much halfway through the understanding of this documentation. Maybe slightly underneath and I wanted to step in here. We've broken down the barriers to entry. We understand now that you shouldn't be scared of AI. You should. It's very easy to lean into. Just take your time to go at your own rate. Understand some key and core important points and follow along to make sure that you understand this this overview because this is going to set us up for all of the future NAN workflows and things that we build together as part of this series. Without the understanding to this of this then you're not going to be in a position where you're. You're going to be able to continue. So the next step is research an image. So research tools Research tools effectively combine LLMs with real time information. So weather summarizing specific events that happen in real time. Perplexity AI is probably the most Perplexity is probably the most common one. It's an AI powered search engine effectively gives answers grounded in real sources. So it's grounded by the search engines results page pages to make sure that it's put out properly. Notebook LMS like a second brain it lets you upload your own documents so you can provide your your own documents notes sources information whatever it be that's queryable. So that's you can see where the power would that come in would come in for an individual business Image tools create hyper realistic scenes graphic illustrations from text mid Journey Chat GPT Dal A3 Ideogram those are three examples of different elements where they are stronger. Nano Banana is the latest and one of my favorites at the moment for producing content. Extremely, extremely strong. So when it comes down to some of the tools, other tools and things that we can use. Video tools so video tools are one of the fastest moving areas in AI. Video and speech are two of the fastest moving areas. They're capable of generating full scenes from a single prompt, right? From a single, a single line of code that you don't see. That's what the wrapper is, which is what's up next. That's what the wrapper is. But we can create, you know, text to video, image to video using things such as AI studios and other LLMs to generate images. We either write a prompt and it generates a full video. We upload an image and ask it to provide a video based off of the image. Some of the key players, outside of the big ones that you'll be aware of here are Sora, Kling, Runway and Pika. Those are four that are actually really powerful. But Google's AI Nano Banana, Google AR Studios, Nano Banana and Google VO3 are extremely strong text to speech. So generating hyper realistic voiceovers of information, music generation and voice input. So have a real time conversational back and Forward with an AI chat GPT voice. Suno and 11Labs cover all of those. We will touch some of these during the build that we've got coming up. So don't worry. Now understanding all of those products that I just explained, ElevenLabs, Suno, VO3, well, what are they? Well, what they are is they are specialized wrappers because you give one simple prompt or one detailed prompt if you, as you progress and you, you understand how you need to manipulate prompts to get the best output possible for what you're looking for. But you need to understand that they're just, they're just all that like Chat GPT is or all that 11 labs is, is just a wrapper. It's a wrapper that's around. Think of it as the engine that's inside of a car and the front end that we see like from what you guys will see from Chat GPT, this is a wrapper, makes it extremely clear, makes it so that you can be, you know, you can use it, you can function. But behind it is where the magic happens. The that's what they are. Most AI tools you see online now are wrappers. They've wrapped up individual, individual tools and wrapped up individual coding techniques and their own flows even to be able to produce effectively what is a SaaS for you that you can use to put in. So a few examples. Support bot a data analyst to summarize and visualize data sets. Generate and fix code snippets. Marketing wrapper create targeted marketing copy Foundational LLM core model power in all wrappers. So when you start with the LLMs that you use now, as per my analogy of being an operator, the wrappers are produced by the emperors. They take the foundation from the LLM and they build all of these beautiful things out of it that we get to use to then create even further and more powerful flows. But you shouldn't be, don't be worried about like third party tools and how they operate and the way that they perform because effectively a wrapper is literally just that. It's the Christmas wrapper around, it's the, you know, it's the paper around the, the individual, the individual LLM and its capabilities. That's why everything is possible these days. That's why you see every day a new SaaS product pop up that move something from notion to here and It's a new SaaS product that's cheap. It's because they build it using the LLMs functionality and then from there they can proceed. Now the four core skills that will not change no matter what you're doing. Inside of AI, there's four tools that, that will, sorry, the tools will evolve, but there are four skills that will never change. These four skills will always be valuable no matter what tool you're using, what, what wrapper you've decided to unwrap today. Prompting the art of communicating clearly with AI to get the results you want. Landscape literacy, understanding the main categories of tools and what's possible and what's best in class for you. Workflow thinking, breaking big tasks into smaller tasks, AI assisted steps to help you along the way and creative remixing, combining different tools in new and unexpected ways to solve problems, which is how new SaaS, new products, new solutions for businesses are invented in the first place. That's it. Those four skills. If you have those four skills and, or grow and evolve into having those four skills, then you're in a very good position from an AI perspective because you'll never fail no matter what's released, you have the fundamental basics there. And when you look at it like that, when you look at AI learning like that and you see that actually all you need is four core skills. Prompting landscape literacy, workflow thinking and creative remixing in order to be able to evolve the old AI. Suddenly it becomes a lot less scary because learning four skills is A lot less complicated than what it looks like from the outside when you're looking at AI tools and usage. So that's important skill Deep dive Prompting a simple structure can dramatically improve Results. The formula aim context rules aim. What do you want the AI to do? For example, write a 500 word blog post. Context provide relevant background. Who is this for? What's the target demographic? What is it about? Example for busy entrepreneurs planning a productive Monday. Rules add constraints, sometimes strict constraints. Tell them strictly this at the LLM I strictly want this output. You know, for example, make it casual. Include three actionable tips. Strictly include three actionable tips and end with a quote. There's a difference between saying strictly and asking it to put output. Actionable tips. Do you see the difference there in the prompting? That's how you end up in a position where you get differentiating pieces of information. The pro tip from a role prompting perspective is tell the AI who it is. If you I've used this analogy before. Like Big Hero 6 in the film, Baymax has a chip in him that they change between a green and red chip, whether he's good or he's bad. If you put the green chip in and the red chip in, the system will do different things. Well, make sure that every time you use an LLM to system prompt, you put the correct chip in it. And you can do that using many tools. Inside of AI Studio Inside Inside of AI Studio. It's literally at the top of the page here of the chats that you had a system instruction. You know, like you say you're a travel vlogger. If I just told it you are a travel vlogger, it would instantly have significantly more power than if I just didn't put that in and wrote the question instead. So that's, that's important. You go a lot deeper than that as you go forward. But I'm not going to stick on it for too long. Level up from power user to builder. Go beyond using individual tools by connecting them into automated systems using NAN or make.com so step fixed. There's two, there's all. There's two types of automations here. And, and they do different things. So you have automations that are fixed step by step workflows. You go from point A to point B to point C and they follow a predetermined path, every single path, every single time. They don't change agents. On the other hand, they take on dynamic workflows. They can call on an abundance of individual tools. They can have Google Calendar, Google Gmail, you know, perplexity, tavily, whatever. They can have them all connected at once. And that's where the. The change sort of comes in. One of the key functions, something you'll have. You may have heard me speak about AI agents. I'm not a particular fan. I like them for specific use cases. But an AI agent needs a brain, right? It needs. It needs a brain, which is the LLM. It needs a memory, which is a context, which is usually simple memory in NAN and tools and actions it can perform. So if we were to go inside of NAN and look at this individual workflow here, you see, if we add an AI agent, this is the difference. We can have a basic LLM chain. And I know we're getting ahead of ourselves here, but I'm just showing you, this is a basic LLM chain inside of here. We can add prompts, system prompts. You are a travel vlogger, for example, right? You. You can add. You can add the individual messages there. We have AI and user. User is information that we put into it that could be the message from previous, from before. It can be anything. We're not going technically into an end today in this video, but I'm just going to show you the differences. That is how you would use an LLM. So this is that we're creating a wrapper inside of here. We're creating a new flow that's a basic LLM chain. But what it can't do is it can't do what AI agents can do. And what that is is that an AI agent can have a memory. So use simple memory so it can remember things. It can remember what you said to it a moment ago. If we look at it like this. Let's see. My name is Michael. The chat model, that would help. Let's choose Gemini. Clear the logs. My name is Michael. What's my name? Let's say. Okay, see what it says now. What's my name? See, it remembered my name. Anyway, your name is Michael. Now, if I delete this simple memory node. Hi. Clear the chat logs. Hi, I'm Michael. What's my name? As an AI, I don't know who you are. So that's the difference between them. Okay, memory and tools. Now, you could see from the other node, but I don't want to geek out too much because we're going to get into building later. But tools, you can select from a whole array of individual tools. Okay, you can see here. Get rid of that. Not you. So we can add Supabase. We could add Anthropic. Asana, we could add, calculate a calculator. We could add Cloudflare, we can add a crypto tool, Discord, Dropbox, everything, Facebook, Graph, Fresh Desk, GitHub, Gmail, Google Business Profile, Gemini Sheets, slides, anything. You can add anything to an AI agent and you can just keep adding. I just deleted the the connection, but you can just keep adding Air table. You can just keep adding and add in and add in. And that's the difference between an AI agent and an AI and an AI, sorry, an LLM system that's standard. So let's get out of here because that's not what this video is about. But that's the difference between and how you level up from being a user, which is chat, GPT, typing things in to actually building things. And the difference between automations is as the example of what I've just given you, the the example of what I've just given you, what a basic LLM flow might be would be this. So sorry, what a basic flow might be might be this. Google Sheets, append a row and then send an email. Gmail, send an email, send a message. So you can see here how this is fixed. This is automation section. This is going from steps A to B to C. The one that I showed you before, that was of a flexible AI agent that can do numerous different things depending on the prompts. Okay, so I appreciate that one took a bit longer than the others, but it's needed now. The future of building vibe Coding. A new intuitive way to create software. You describe what you want, you describe what app you want in plain English. The AI generates the code in a basic app structure. You test and vibe. You test the app, describe the changes that you want. You refine the AI by making updates to prompts and keep pushing. This makes software creation more accessible than ever it ever has done previously. And we can go and build that. Maybe that can be part of our video. We can build a software piece of software together. Your five step action plan should be to identify a pain point. The first step is identify a pain point. What is a task that is slow, stressful and repetitive for you or your business? Then you find the right tool. You start with an LLM, see if a more specialized tool is needed from there. But you then find the right tool. You iterate and refine, use good prompting, prompt the system, break the task into smaller steps. Then you combine tools and then you automate. You build your first automated workflow. Those are the five steps and it and this is exactly what we're going to go through Together we're going to go through identifying a pain point, finding the right tools, iterating and refining combining tools and automate. That's going to be the first set of action plan for us. The final takeaway is don't just learn about AI, use it to solve a real problem. Start with one friction point in your life or work. The tools will keep changing but the core skills always remain the same. By applying even a small part of this roadmap to your day you are already ahead of 99% of the of the industry. My name's Michael like subscribe comment Let me know that you're following along with this this video for motivation because I'm going to give you I'm going to continue now into NAN to give you a 1001 step of how to build and we're going to build our own for workflows as we go. This intro is a really important sector to knock down your barrier of thinking and barrier of entry to AI. Anybody can do it. Follow along. I'll catch you in the next.