How to start Learn AI | AI roadmap | start learning artificial intelligence

 How to start Learn AI | AI roadmap | start learning artificial intelligence

So you want to learn artificial intelligence, then this video is for you. I'm going to provide you with a complete roadmap that I would follow if I had to start over today on my artificial intelligence journey. And now for context, I started studying artificial intelligence back in 2013, ten years ago, and over the past years, I've been working as a freelance data scientist, helping my clients with various end to end data science and artificial intelligence solutions and applications. 

I also share all of this knowledge and my journey on this YouTube channel, which as of today has over 25,000 subscribers. And at the end of this video, I will also provide you with a resource completely for free, where you can follow all of these steps to complete the roadmap, even with training videos and instructions. So make sure to stick around for that. And now, before we dive into the seven steps that I would take today to go from beginner all the way to monetizing my data and AI skills, it's important to provide some context on what is currently going on with the AI hype, because I see a lot of new people entering the field, and for a good reason. Because the AI market size is expected to grow up to 20 fold by the year 2030, bringing it all the way to nearly 2 trillion USD. So it's really one of the best opportunities, I would say, right now to get into, because we're still early, we're still at the beginning of this AI revolution. And also with the release of these pre-trained models from OpenAI, it's now also easier than ever to enter the field. But that said, that is also where a lot of the misunderstanding and just wrong expectations arise from, because I see a lot of people online as well as on YouTube explaining like how you can quickly start, for example, your own AI automation agency. And while there are great tools already online out there like bot Press and stack and flow wise, which are also made a video on where you can quickly spin up prototypes and simple bots and even can get a little bit more advanced.

 Don't get me wrong, you can definitely build some great solutions with that, but if you really want to learn artificial intelligence and build applications that companies can count on and build upon, then you really have to understand the coding part, the technical part, really of it. So that's really where our starting point should be for you and for your learning path. Figuring out, hey, do I want to just learn how to use these no code, low code tools already available? Or do I really want to learn artificial intelligence? And with that said, there is also just a general misunderstanding, I believe, of of what really AI is, because AI is such a large umbrella term and it's also nothing new. It's been around since the 1950s, but right now, with the ChatGPT hype and the OpenAI models, people think AI is that really, if we look at what artificial intelligence really is, it's like I've said, a real big umbrella term with various subfields. So for example, within artificial intelligence, which is here explained as programs with the ability to learn and reason like humans, machine learning, then we have deep learning, which is another subset focusing on neural networks. 

And then we have the field of data science. But in my work as a data scientist, I use artificial intelligence, I use machine learning, and I also use deep learning. It's a lot more than what people think. The first real question that you got to ask yourself is, do you want to be a coder? And now there's no right or wrong answer here. There are plenty of opportunities right now and also in the future for both pathways, for both low code, no code tools and building custom applications. But you just got to be aware of the pros and cons to both of the sides. And now to be totally clear, this roadmap is for people that really want to learn AI with a depth of understanding, really learn the technical side of things. And now if you've decided that that is not for you, that's of course totally fine. Like I've said, there's no right or wrong. But then if you want to still want to do things with AI, then I recommend starting out by checking out Bot Press like I've said, or stack AI, which are excellent resources. Or you can check out my video on flow wise here on YouTube, where I show you how you can get started with a low code, no code tool as well, completely for free. But if you do decide that you want to join the dark side and become a coder, then let's proceed with the next steps. My approach is quite different from anything else you will find online. And now why is that? And what I typically see online is you have two ends of the spectrum basically, where on the one hand you have the people talking about these low code and no code tools, not really getting into the specifics, the theoretical part. And then on the other hand, you have the more classical approaches towards artificial intelligence and machine learning, where people really get into the mathematics and the statistics, giving you roadmaps where you really have to get theoretical first. I'm a firm believer of learning by doing reverse engineering things that people have already done. Putting in practice and then trying to fill in the gaps. Now, the technical roadmap that I'm going to provide to you will really focus on the fundamentals that you need in order to get started in either artificial intelligence, data science or anything in between. Like I've said, I've worked in all of these fields over the past ten years, and I've really identified the core techniques, workflows, and tools that you need in order to get started, regardless of what you want to do. 

So this will work for you if you just want to build applications with large language models and Lange chain, for example. But it will also work if you aspire to become a data scientist or a machine learning engineer. Now, the actual first step that I would focus on on my journey would be to set up my work environment. Now what does this mean? So Python is the go to language that we have to learn. If you want to get started in AI or in data science. But the thing is, if you learn Python, if you start to follow these tutorials, online videos, training videos, courses, even you can quite quickly understand Python and how it works because it's one of the easiest languages to get started with. But I found in my personal journey that there's this initial bump where you see things online and you see people run some code, but then you are missing some information on okay, but how do I now actually do this on my laptop or my computer? And I would really focus on this first, setting up an environment on your laptop, on your computer where you have an application, a program, and a Python installation that you are confident with. And now I have a specific approach that I take over here with in fiasco. And a lot of people seem to like that, so make sure to check that out in the resources. But this really is step one. Then getting accustomed with that. And that brings us then to step two, which is actually getting started with Python. It's like I said, the most important language. 

This is going to be your tool that you're going to build these applications in. Now, if you're new to programing at all, I would first focus on the fundamentals of programing, which I will have resources to, but then quickly transition into learning the basics of Python and then specifically some libraries that are very useful for AI and data science in particular. So these would be for example, the numpy library, the pandas library, and the matplotlib library. Now these are all libraries that you can use to do data manipulation, data cleaning, creating visualizations. This is really your starting point for starting to work with data, because in the end, all AI applications, all AI tools are created from data with data. So being able to work with data and turn raw and unstructured data into information, into valuable insights that you can actually do something with is really at the core of of artificial intelligence. And now step three would be to learn the very basics of git and GitHub. 

Now why is that? Some would argue that that would be a little bit more advanced. And it's not required in the beginning. But what I've found, especially with artificial intelligence and also the video tutorials that I make, is that a lot of examples online, people will make that code available via GitHub. But you have to understand kind of at the very basic, how these tools work, because that allows you to easily copy and clone is what they call it tutorials. That brings us to step four, which is working on projects and building a portfolio. And for this, it's convenient if you already know how to use git so you can download some projects, download some code from other people, and then try to reverse engineer it. To me, that really is the best way to to learn Python, to get good, to actually understand holistically what a project looks like, how people are structuring their code and trying to run it, and then you don't understand what's going on, but then trying to reverse engineer. So it's really like beginning with the end in mind and then trying to change things and see how that affects the different outcomes. And this also provides you with an opportunity to explore what it is specifically that you like about artificial intelligence. 

All the areas we've discussed computer vision, natural language processing, machine learning. Here. You really find out, okay, these are all the kinds of things that I can do. And this is really what I like to do. And then as you're working on these projects, selecting them, picking them, you there will be a lot of gaps and things you don't understand. And that would be a good point if you interested in that, to find specific pieces of information or courses to help you with just that. And now when it comes to projects, probably the best place to start. If you want to learn more about data science and machine learning is Kaggle. So Kaggle is an. Excellent resource that you can go through. And they host machine learning competitions here. So you can see all kinds of requests and you can even win prizes. So this is one from Google. And the cool thing here is if you click on the actual competition, you can also actually have a look at submissions that people have made. So here you can see an entire notebook from someone that is trying to solve this problem for Google, all with documentation and and even the code. So this is such an excellent learning resource source that you can go through. Like I said, there are plenty of resources available on here. But if that's not for you, machine learning data science if you want to just explore large language models and OpenAI for example, right now, then I recommend to check out my GitHub repository on Lang chain experiments. So I also have videos on my YouTube channel for that, but here on the repository. That's why it's good that you at least understand the basics of git and GitHub. So you can take this code, know how to work with it. So here are some cool examples of how you can create a YouTube bot that can summarize a video, or even a slack bot, or a pandas agent that can ask questions and answer questions about large data tables. And now, if you're really serious about learning artificial intelligence and data science, and another great resource that you can check out is Project Pro, which I've recently discovered. 

So Project Pro is a curated library of verified and solved end to end project solutions in data science, machine learning, and big data. So overall, this is just an excellent resource with with so much information and all the projects on here that you can pick from, all from the various fields are all created by top industry experts from leading tech companies. So what I really like about this is, first of all, you have about 3000 free recipes that like anyone can check out. But if you get to the subscription and that is where it really gets interesting, you have access to 250 plus end to end projects. So you can really like go in here and see, okay, what is it that you're working on? So maybe it's data science and you want to specialize in machine learning. And you go in here, you literally have all kinds of projects. And this is not only a great resource for you to learn from because you will have complete video walkthroughs 24 over seven support. And you can ask questions and you can even download all of the code. So literally the entire project will be made available to you. So it's an excellent learning resource. But also for me personally working as a freelance data scientist, this can also like really help me in my professional work, the projects that I take on. So for you, it could either be in your job or in future jobs, freelancing, whatever you really have. A library that you can pick from that can really give you that extra kind of confidence you need, for example, to take on a project. Now, like I said, really, you see video instructions, you can go through everything and then also download the code. So this really is a great resource that you can check out. And if you want to learn more about this, I will leave a link down in the description. And project Pro also has a YouTube channel which you can subscribe to if you want to stay in the loop. Learn more on that. And that brings us to step five, which is picking your specialization and sharing your knowledge. 

So right now you understand the fundamentals of Python. You have a work environment and some some efficient workflows that you can follow. You also have some project experience. So now you get a little bit more clarity of what it is that you want to do within the world of AI or data science or machine learning. So this would be the point where you pick a focus area, you specialize, you try to learn more. And also what I really would recommend, and what I would do is to start sharing your knowledge. So you could do this through a personal blog. You could do this through writing articles on medium or towards data science, or you could even potentially, like I'm doing, share your knowledge on YouTube. And by doing so, you're not only contributing to the collective knowledge on AI and data science, but it's also an essential method for you to strengthen your own learning. Because in doing so, in explaining concepts that you're working on, that you're learning to to someone else, you really start to identify the gaps within your understanding. And this again allows you to fill in those gaps accordingly and really focus on some specialized learning versus just going through course after course after course. And then step six would be continue to learn and upskill, because now that you have clarity on your specialization and kind of the direction that you want to go, and you also start to identify these gaps within your own understanding, it might be time for you to, for example, focus on math, focus on statistics. If you want to become a better machine learning engineer or data scientist. But if you've decided to go with the large language model and generative AI.

You can also have a look video demo

https://www.youtube.com/watch?v=h2FDq3agImI