Welcome to the first lesson of Master AI — the course that takes you from "what even is AI?" to confidently building and creating with it.
Before We Start
I originally wanted to call this course Build with AI, but "Master AI" just sounded better — so that's what stuck.
This course exists to answer three simple questions:
- What is AI, really?
- How do you actually use it?
- How do you stay ahead as it keeps changing?
By the end, you'll be comfortable working with today's AI tools — how far you go depends on how much you put in.
What You'll Need
- A computer — minimum: 16GB RAM, 512GB SSD, Intel Core i5 (10th Gen or newer) or equivalent
- Accounts with Claude, Gemini, and ChatGPT
- Obsidian for taking notes
- A reliable internet connection
- Anything else we'll mention as the course goes on
Quick Intro
If you're new here — I'm Moses Mbadi, a software engineer and tech content creator. I also built Soma Stories, an app that lets people earn directly from the stories, books, and podcasts they create.
Now, let's get into it.
What Is a Piece of Software, Really?
Here's a simple test. Open Microsoft Word and type a heading — it becomes a heading. Every single time, no exceptions. Open a video file in VLC, and it plays. Try to open a random image file in VLC, though, and it might just fail.
That predictability is the whole definition of traditional software. It either works exactly as designed, or it doesn't work at all.
AI breaks that rule completely. And to really understand why, we're going to do something most explainers skip — we'll walk through, step by step, how a company like OpenAI or Anthropic actually builds an AI model like GPT or Claude from scratch.
The Formula That Explains Everything
Traditional software follows one simple formula:
Data + Rules = Answers
Here's what that means in plain terms: the software stores data, a human programmer writes the rules, and when you give it an input, it follows those rules exactly and hands you an answer. Facebook, WhatsApp, your banking app — all of them work this way. Predictable in, predictable out.
AI flips that formula on its head:
Data + Answers = Rules
Instead of a person writing the rules by hand, you show the AI a huge amount of data along with the correct answers — and the AI figures out the rules on its own.
Nobody sat down and coded "a cat has pointy ears and whiskers, so flag it as a cat." The AI learned what a cat looks like just by studying thousands of examples. This is what people mean when they talk about AI "learning," "reasoning," or "understanding language" — it's not following fixed instructions, it's recognizing patterns.
A Real Example: Spam Filtering