Transcription
When I first started to learn Python in 2018, it was actually a computer science class and I struggled so, so much. I followed all the tutorials. I followed the code from class. I followed the textbook and just nothing clicked. I just fundamentally didn't understand it. People always said, "Oh, it's another language. Think of it as a language." But I just didn't understand when to use what, what I had to learn. It just seemed so much. I was just so overwhelmed.
And for years, I carried around this guilt thinking I wasn't technical enough. I wasn't good enough to be in tech. I wasn't ever going to do computer science. And it's ironic because a few years later, I have a master's in computer science and I've worked as a data professional in tech. So, a lot has changed since then.
I'm going to explain exactly how I would learn Python from scratch in 3 to 6 months if I was just starting over in 2025, 'cause trust me, it is not as complicated as you probably think it is right now, once you have a clear road map in front of you.
Now, before we begin, you might be wondering, is it even worth learning Python with the Age of AI in 2025? And that is a really fair question to ask. If your goal is just to generate a bunch of code snippets with ChatGPT and have code you don't really understand what's going on with a bunch of errors and bugs, sure, don't learn Python. But if your goal is to actually make code that's production-ready and actually build cool things that work, I would probably try to learn Python. 'Cause in the age of AI, you still need to build these AI systems. And to build these AI systems, you're going to have to use stuff like LangChain and LangGraph, which are pretty much in Python, where you're going to have to learn Python to use these. So, knowing Python will make you a builder and not just another consumer.
The analogy I like to give here is like it's like riding a car. You riding a car is like using ChatGPT to code. But if you want to drive the car yourself, you need to be able to learn Python. Use chat as your assistant while driving the car. Python is a 100% solo language to learn if your goal is to break into AI, data analytics, data science, data engineering. Trust me, you are going to want to learn the basics of Python.
Now, the first step you need to do is get set up. You need to actually find a place where you can actually code in Python. And the mistake I made was I tried to overcomplicate it. And by overcomplicated, I tried to install Anaconda, Jupyter Notebook, all these different IDEs when I had no idea what an IDE even was. I didn't understand why we needed this. And don't waste your time doing that. Go to Google Colab. Google Colab. It's completely free to use. It's pretty much Google Docs, similar to Microsoft Word. You can actually start coding immediately rather than installing all these softwares and dependencies, and it makes your life much, much easier.
Once you spend enough time in Google Colab and you learn the core fundamentals of Python, then you can move on to VS Code, Cursor, whatever IDE is of your choice. For those who don't know what IDE means, think of an IDE like Microsoft Word, where you have to download a software to actually write stuff. And Google Docs is, think of it as like Jupyter Notebook, where you can actually just pull stuff up on the cloud and just start using it immediately. So, that's pretty much the difference between the two. And you can see Python is still the exact same code whether it be on Google Colab or Jupyter Notebook. It just adds a level of friction that we don't want at this stage.
Now, step two, you need to learn the core fundamentals of programming and Python. And if you're coming from a programming background, you know, Java or something, a lot of this stuff relates. You just have to relearn syntax. If you're completely new, like I was, I'm going to tell you how to learn these new core skills for absolutely free. When I was starting out, I thought I had to learn everything. I was like, "What is a dictionary? What is a list? Why do I have to learn all these things?" And I just wanted to give up. And I think that's 90% the issue for most of you guys out there. And I got stuck enrolling in tons of Udemy courses that I didn't need and never finished.
So, if that sounds like you, I have good news for you. The truth is, you don't need to learn all of Python. You just need to know 20% to unlock 80% of its value. It's that simple. The best part is all these, the 20% of fundamentals you need are laid out in front of you, so you don't even have to guess. Python actually has this official training or documentation you can use. It's called docs.python.org/3/tutorial. You'll see it on the screen. And I want you to think of this as your dictionary. It's clear, reliable, and always up to date. You can just follow along with this.
Now, here's what I want you to do. I want you to work through the first couple sections: variables, data types, loops, and functions. Don't worry if you don't get this all on the first try, but just get started. As soon as you see a concept, I want you to open up Google Colab and just start coding and using that concept and applying it immediately. Even as something as simple as a budget calculator will take you a long way and further than most other people. And most importantly, don't try to memorize like I did. Do not memorize. Bookmark the doc and just get comfortable coding and just make sure you have no errors. That's all that matters up to this point. The skill is much more valuable than cramming a bunch of syntax. If you do this right, you'll cover the basics in a few weeks, and not a few months or years. And you'll be using Python the way it's meant to be used, which is solving problems.
Now, after this, you want to learn what we call libraries. And don't freak out if you don't know what libraries means. I'm going to teach you right now. When I first heard the word library, I had no clue what it meant. People were like, "Install pandas, install numpy." I was like, "What does that even mean? Why do I need to install these?" So, let me strip this down to first principles. Python by itself is like learning a language with its basic alphabet or grammar. You can write a bunch of sentences, but writing an entire novel with just Python will take you forever, right? A library, you can think of this as like a phrase book. It has pre-written lines already written for you, so you don't have to manually do it again. In code, that means someone else has already built the tools you're going to be using to code. And it makes it a lot easier. Someone's already built them, tested them, optimized them, and instead of spending weeks writing your own math functions or graphing code, you just import the library and it's ready to use right away.
And these libraries aren't random. They exist because programmers like you and I run into the same issues over and over again. And they wanted to solve these problems. People needed faster math, so they built NumPy, which is a Python library. People kept working with tables, so they built this tool called pandas. People wanted charts. They built Matplotlib. People wanted to customize the charts. They built Seaborn. You see how these libraries are built for specific purposes and problems that they solve. Over time, these just became the standard. And if you wanted to break into data, you had to learn these.
So now, this is where paths kind of split. There are different goals by learning Python. I know some of you may want to learn AI. Some of you may want to learn data analysis. Some of you data engineering. So I'm going to explain the packages you're going to need to learn based on what your goal is.
If your goal is data analysis and BI, like mine is, I want you to learn pandas, numpy, matplotlib, and that should be it. Pandas will help you clean and analyze tables. NumPy will help you with simple calculations, and Matplotlib will help you actually visualize data instead of using a BI tool.
If your goal is AI, you can learn NumPy again. You can learn scikit-learn, and you can learn this library called PyTorch. These are great for built-in machine learning libraries immediately.
If your goal is automation or software engineering, you can learn requests, Flask, or FastAPI.
And lastly, if your goal is data engineering, learn SQLAlchemy or PySpark. These are the libraries you're going to need. And obviously, there are a lot more libraries than these. These are just some of the core foundational ones. You're going to want to learn more on top of this, but just learn the basics for now. The key idea is libraries are shortcuts built by experts in this field. And once you learn the libraries, it just saves you a lot of time. And for me, it's like it's kind of like a low-code tool. I don't have to spend a ton of time like low-level languages building everything from scratch.
Now, once you learn the library, it's time for you to start building projects. So, before I took that computer science class back in 2018, I had an uncle who told me to build a calculator in Python in 2017. I voiced him. I was really interested in code, but I had no idea where to start. And he was like, "Just build a calculator." And I always, for the longest time, doubted him. I was like, "Why would I build a calculator if I don't know how to even code?" And to be honest, I had no clue where to start. For years, I just avoided projects and just procrastinated by taking a bunch of courses. But that's not the way we want to do things.
The truth is, the fastest way to learn Python is by doing projects immediately and not wasting any time. Projects for me are where all of the skills you just learned finally click: variables, the loops, the functions, and libraries. They force you to stop memorizing and start creating and start building these cool things with Python the way it's intended to be used. And I don't want you to build the next Uber or next Instagram. Just start really small.
Some beginner projects you can do: you can start cleaning up messy CSVs in pandas. Delete duplicates, fix dates. You can build a BMI calculator, a budget tracker. You can create a text-based to-do list app in Python. Want to go more intermediate? Maybe pull all your Spotify and YouTube history into a dataset and make a bar chart of your most listened-to songs or videos. If your goals be more advanced, start building machine learning projects. Create a chat with Python in an API. For me, one of the first projects I did with Python for work was I automated a weekly email. This is pre-ChatGPT. So, I'd actually spend my time on Stack Overflow and see what other people have done.
Each and every project is like a Lego block. Each one will build intuition for you to use for the rest of your career. You're going to go from knowing Python exists to knowing how to use it and knowing what's possible, which most people who are vi coding will never get. And here's the key: momentum matters more than everything. Just be consistent and have momentum while you're doing this intentionally. You're going to make mistakes. You're going to hit bugs. And that's honestly the point. Every project teaches you something a tutorial never can.
So, here's my challenge to you right now. Pause this video right now and think of one dataset you want to analyze and one problem you can solve with Python. The dataset could be simple. Could be your bank statement. Could be your Spotify most listened to. Could be your YouTube watch history. Doesn't matter. Write it down and start analyzing. And that's going to be your first project.
So, once you've built projects, the next step, the most important step, is showing your work, showing to other people you know this cool skill. When I finished my first project, I hid them on my laptop. I didn't show anyone. I thought I would share them when I'm perfect. And even to this day, I'm kind of afraid to post my projects onto GitHub in my portfolio 'cause I'm afraid of criticism. But here's the truth: no one really cares about perfection. People just care about proof. Proof beats promise every single time. If you go into a job interview and you say you know Python versus showing an actual portfolio of you using Python, you'll probably get the job if you have the portfolio. It'll look a lot better, right?
So, here's how you can start right now. I want you to go to github.com and create an account. I want you to upload one of the projects you just done. Add a small README file. This just in plain English explains what the project does and start updating it. Iterate as you go. It doesn't need to be one and done. Start adding to the project and you'll learn a lot of core foundational skills here. And then post about it on LinkedIn. Don't underestimate how powerful this is. So, don't just learn Python. I want you to document it and share it with the world. The world deserves to know how far you've come.
So that's it. That's how to learn Python in 2025 in months rather than years. So if you got this far, I want to thank you so much. Just leave a like, subscribe, and I'll see you in the next one.