Fast.ai Chapter 0: Unofficial - Why Most People Quit Fast.ai at Chapter 4 (and How You Won’t)
First, a little bit about me so you know where I’m coming from (context matters, right?). I’m a Computer Science Engineer with more than 10 years of experience. Over the years, I’ve dabbled in multiple roles—.NET, Python, Django, React, Vue, —you name it. I’ve built systems and apps from scratch, and helped companies launch products successfully.
In July, I left my role as Head of Engineering at a construction-tech firm. That’s when the million-dollar question hit me:
👉 Should I take up another job or dive deep into AI/ML?
Spoiler alert: I chose the latter.
Like any good engineer, I went to my old friends—ChatGPT and Gemini—and asked, “Where do I even start?” Their response? “Math. Lots of math. Matrix multiplication, partial differential equations, probability, statistics…” Now, I've always been a "learn-as-you-go" kind of person. My practical approach has served me well, allowing me to pick up what I need to get the job done. But with AI, I was a complete newbie, staring at a map with no "You Are Here" sticker.
To be clear, the official fast.ai course (which is free, by the way!) doesn't require any of this pre-learning. But as I started the course, I felt like I was trying to catch smoke. The information was good, but it was slipping through my fingers. When I finished a chapter, I didn't feel confident. As a long-time teacher, I have a personal benchmark for knowledge: "How easily can I explain this to someone else?" The answer, in this case, was a resounding, "Not very well."
So, I hit the pause button on fast.ai and went back to the drawing board. I took a few foundational courses, and when I returned, it was like a new world. I was more confident, and the concepts clicked into place. Here’s the detour I took after 3 lectures on the course:
Math Refresher a.k.a. Don’t Let Matrices Scare You
Linear Algebra
Coursera: Mathematics for Machine Learning – Linear Algebra
Or, the absolutely superb 3Blue1Brown series: Essence of Linear Algebra
My take: Coursera gives you structure, exercises, and accountability. 3Blue1Brown makes you fall in love with vectors. Win-win if you do both.
Calculus - Yes, the scary one
Coursera: Mathematics for Machine Learning – Multivariate Calculus
3Blue1Brown’s free courses:
Why both? Coursera drills concepts into your brain through assignments. 3Blue1Brown makes you go, “Ohhh, so that’s what a derivative actually means!”
Neural Networks
Honestly, I did this 3Blue1Brown series just because I love Grant Sanderson’s visuals. Also, the intro tune reminded me of the Game of Thrones theme song. (hey, you take your motivation where you can get it!).
Numpy
When I hit Chapter 4 of the Fast.ai book (fun fact: Jeremy Howard says this is where most learners drop out 👀), I realized I needed a crash course in Numpy.
I found this tutorial super helpful. A couple of hours in, and suddenly lesson 4 felt like watching a movie with subtitles—I could actually follow along.
🎯 Final Thoughts
To be clear, you don’t have to do all of this before starting Fast.ai. The course is designed to be practical and beginner-friendly. But if you’re like me—someone who likes to really understand what’s going on under the hood—then these courses will make your journey much smoother.
Remember: not understanding everything the first time is completely fine. The trick is to keep going, fill in the gaps when you can, and most importantly—don’t quit at Chapter 4.
I will keep updating this list as I go through the Whole course.
That’s all from my side, folks. Happy training your models! 🚀
💌 P.S. If you enjoyed this post (or at least chuckled once), consider subscribing to my mailing list. I’ll be writing one blog for each chapter of the Fast.ai course, along with my experiments, failures, and occasional “aha!” moments.
Don’t worry—I won’t spam you with cat memes… unless you’re into that 🐱. Just practical insights, resources, and maybe a joke or two to keep things fun.