Learning to make video games in C++ was always slow and frustrating. With an AI assistant as your co-pilot, the path gets shorter without you losing track of what you're actually writing.
I'm Carlos Montiel, Enterprise AI Solutions Architect, and this is the first episode in a 13-lesson series where we'll build real video games in C and C++ using an AI assistant (Claude Code or Cursor) as a programming companion. This isn't a course about "magic prompts" that generate a complete game out of thin air. It's a real game-programming course — pointers, memory, game loops, collisions, networking — where AI is used for what it actually is: a tool that multiplies your productivity if you know how to direct it.
By the end of the 13 lessons you'll have built a complete Pong, a sprite animation system, collision detection, a save system, a state machine for menus, enemy AI with pathfinding, and a networked multiplayer game. All in C/C++, all compilable, all explained.
C++ has a real learning curve. Unlike Python or JavaScript, here you manage memory manually, deal with pointers, understand the difference between stack and heap, and a single-character error can produce a segmentation fault that tells you nothing useful about its cause.
Traditionally this meant spending hours searching forums, reading SDL2 documentation line by line, or asking on Stack Overflow and waiting for a reply. The result: a lot of people give up before they ever see something move on screen.
A beginner might spend an entire afternoon figuring out why that code "sometimes works and sometimes doesn't." A well-used AI assistant explains in seconds that the problem is the lifetime of `local_value`, and shows you alternatives: return by value, use `new`/`delete` responsibly, or better yet, a `std::unique_ptr`.
The core idea of this course is that AI doesn't replace your judgment, it accelerates it. An assistant like Claude Code can:
- Explain, line by line, what a piece of code you don't understand is doing. - Generate repetitive "boilerplate" (initializing SDL2, setting up a `Makefile`, writing a `struct` with getters) so you can focus on game logic. - Catch memory bugs before you compile, flagging dangling pointers or memory leaks. - Suggest design improvements as your code grows and becomes harder to maintain.
What doesn't change is that you're still responsible for understanding the code. If you ask AI to generate a function and paste it without reading it, you're not learning to program, you're delegating without supervision — and in C++ that will blow up in your face sooner or later.
This course is designed as a practical progression, not a collection of loose topics. Each lesson builds on the previous one:
First we set up the environment (compiler, SDL2, your AI assistant). Then we build Pong from scratch to understand the game loop, rendering, and input. From there we add layers: sprites and animation, collisions, save systems, state machines for menus, enemy AI with pathfinding, and finally networking — first the fundamentals of sockets, then a real multiplayer game. We wrap up with performance optimization, migrating from C to C++ when it makes sense, and how to package and publish your game.
This isn't a course for absolute beginners to programming. You should be comfortable with:
- Basic C or C++ syntax (variables, functions, `if`/`for`/`while`). - The concept of compiling a program (even if you don't fully master the process). - A willingness to write code yourself, not just read what the AI generates.
If you've never compiled a C++ program in your life, don't worry: the next lesson covers exactly that, from installing the compiler to getting your AI assistant integrated into your editor.
One practical tip before we start: in each lesson you'll find complete code blocks. I recommend not copying and pasting them blindly. Type them yourself, line by line, and when something isn't clear, ask your AI assistant "why does this line do this?" before moving on. That habit — writing and asking, instead of copying and moving forward — is the difference between finishing this course knowing how to program games and finishing it with a pile of code you couldn't reproduce without help.
In the next lesson we'll install everything you need: a C++ compiler, the SDL2 library for graphics and input, and we'll get your AI assistant (Claude Code or Cursor) configured and ready to work side by side with you for the rest of the course.
Carlos Montiel is an enterprise AI solutions architect. He implements LLMs, Agents, RAG, and orchestrators for companies across Guatemala and Latin America. Reach out for a consultation.
Contact Carlos Montiel