Stop Prompting, Start Designing | Issue #87


AI can write code now. But you already knew that :).

If you’ve tried using AI seriously, you’ve probably noticed something: the code works (kind of) and then slowly turns into a mess.

In this week’s video, I show why prompting alone isn’t enough and why software design is becoming more important, not less, in an AI-driven world. I walk through a real interaction with an AI coding assistant and show how thinking in terms of responsibilities, structure, and system boundaries completely changes the quality of the code you get back.

With this video, I want to show you why learning how to direct AI is so important, so it becomes a force multiplier instead of a liability. By using the right techniques and applying core ideas from software design, AI-generated code doesn’t have to turn into a mess. Instead, it stays readable, maintainable, and ultimately empowers you to build better software faster.

Have a great week,

Arjan


Do you enjoy my content on YouTube and would you like to dive in deeper?

🚀 Check out my online courses

My courses have helped thousands of developers take the next step in their careers. Check out these courses to help grow your skills and become a senior developer:

👕 Clean code and clean clothes

The ArjanCodes merch store features T-shirts, hoodies, hats, and more for the clean-code-obsessed. Careful though: you'll look dangerously professional while reviewing PRs.

👉 Check out the store here and grab something before your stand-up call starts.

Unsubscribe | Sent by ArjanCodes

Wolvenplein 25, Utrecht, UT 3512 CK

The Friday Loop

Every Friday, you'll get a recap of the most important and exciting Python and coding news. The Friday Loop also keeps everyone posted on new ArjanCodes courses and any limited offers coming up.

Read more from The Friday Loop

Have you ever hesitated between writing a method or a property, and then just picked one because “it felt right”? That small choice actually carries a lot of design weight. In this week’s video, I explore when properties make sense and when methods are the better choice. I talk about what properties promise to the reader of your code, why setters should be used carefully, how this choice affects abstractions and Protocols, and where async fits into all of this. This video is also a good...

Most systems don’t start out needing CQRS. They start simple: create, update, list, dashboard. Everything in one place. Everything reasonable. Until read requirements slowly begin to reshape the system, often in subtle ways 🧩. In this week’s video, I show how that happens using a FastAPI app backed by MongoDB, and how Command Query Responsibility Segregation (CQRS) provides a clean way out once reads and writes want very different things. Along the way, I cover: separating command intent from...

Code can be correct and still be unpleasant to read. A lot of software design isn’t about what the code does but about how clearly it communicates the sequence of ideas behind it. In this week’s video, I explore the Fluent Interface pattern by refactoring a small animation engine in Python. I start with a clunky but functional API, then gradually reshape it until defining an animation reads almost like a sentence: move, rotate, scale, fade. Enjoy this deep dive into API design, readability,...