How to learn AI (or anything technical) in 2025

How to learn AI without getting stuck in starter courses, certifications, degrees, or projects.

Every month, the Chocolate Milk Cult reaches over a million Builders, Startup Founders, Investors, Policy Makers, Leaders, and more. If you’d like to meet other members of our community, please fill out this contact form here (I will never sell your data nor will I make intros w/o your explicit permission)- https://forms.gle/Pi1pGLuS1FmzXoLr6

---

“Devansh, how can I learn AI”? is in the top-3 most common questions y’all ask me, so I’ve decided to put together a detailed guide for you. In this article, I will give you a detailed step-by-step plan that will help you learn AI w/o getting caught up in the mess of courses, certifications, paying for expensive college degrees, and more.

To start with, here is a mostly free plan. You’ll have to pay for the book, but everything else on here has very rich free tiers that you can use to keep up with AI.

1. Look up the core principles behind neural networks: backprop, learning rates, exploding gradients, error functions, activation functions, and non-linearity. YouTube is great for 101 videos, so is ChatGPT if you want explanations tailored to you.

2. Look up an intro to self-attention 101 to get the core idea behind MHA (or talk to ChatGPT about it until you understand all the core pieces).

3. Buy

   Sebastian Raschka, PhD
   ‘s “Machine Learning Q and AI,” and go through it For everything you don’t understand, use the following learning method we have for learning Math/AI (or you can just talk to ChatGPT about the book and what you don’t understand/next steps). Additionally, sign up to his Ahead of AI publication.

4. Then sign up to the following people—

   Sairam Sundaresan
   ,
   Logan Thorneloe
   ,
   Cameron R. Wolfe, Ph.D.
   ,
   Damien Benveniste
   ,
   Sahar Mor
   ,
   Interesting Engineering ++
   , and
   Ksenia Se
   . They’ll give you a nice round-up of cutting-edge research, ML Engineering, and more standard information.

5. Study for a while using this framework.

6. GG.

That plan will serve you well to get you to a high level. But it will take you a while, rely on you to keep self-studying, and you’ll have to be very proactive about engaging with sources and analyzing what you don’t understand. The rest of this article will provide you with playbooks, learning shortcuts, and frameworks that will help you speed up your AI process tremendously. In MMA, we have a saying that you will fight how you train. Therefore, the goal will be to build a learning plan (and more importantly, a learning philosophy) that mirrors our framework for effective AI/ML Engineering (if you’re not an engineer, focus on the principles and be high-level on the implementation; we’ll have some detailed pointers to you in the relevant sections of the article).

PS- This approach will require at least 4-5 hours weekly (don’t burn yourself trying to a1dd too many hours), done over 8-12 months to really click (this assumes you have no baseline knowledge- prior knowledge will speed your learning up a lot). You will start to see major improvements in the first 2 months, but the real benefits come later. In my experience, there’s no real way to build lasting knowledge without deep exposure, which comes with time. You can’t speed-run excellence. This guide is also best for people who can handle unstructured learning. If you’re someone who absolutely requires structure with graded assignments, learning paths, and clear instructions, you probably will have a hard time with this approach. But if you can handle some ambiguity, then the following approach is the best for getting to cutting-edge AI quickly and building long-term expertise in the space.

Assuming you meet the criteria, this approach will likely be as helpful to you as it will be for me. I know it’s helped a lot of other people (there was a time I used to work with people as a tutor/mentor and it was very helpful to them)-

To access the full article—and all premium breakdowns going forward/written prior—upgrade to a premium subscription below.

Each piece is rigorously researched, built from firsthand signals, and written to make you sharper than the noise. It takes a long time for me to compile, analyze, and verify research. If you believe deep insight deserves support, become a premium subscriber to allow me to keep doing the same.

Flexible pricing available—pay what matches your budget here.

Most companies offer learning or professional development budgets. You can expense this subscription using the email template linked here.