Navigating the World of AI Coding: A User's Guide Paving the Path with AI Assistance

When (and When Not) to Use AI for Coding

If you’ve ever used AI to help write code or you’re thinking about trying it you’re not alone. AI-assisted coding tools can feel like hiring the world’s fastest digital bricklayer. And like any good worker, AI will build you something. The only question is: Will it be what you expected? Let’s look at how and when to use AI productively in your programming projects and where to draw the line.

Skills Exercised

Programming - Creating using AI

Programming vs. Coding: The Architect and the Bricklayer

Think of programming as being the architect. You’re the one designing the blueprint deciding whether it’s a house or a barn, one story or two, where the bathroom goes, and how all the rooms connect.

Architect = Programmer

Coding, on the other hand, is the trade work. It’s like being a bricklayer, plumber, or electrician. You follow the plan and execute with skill and precision.

Bricklayer = Coder

The larger and more complex your project gets, the more important it becomes to have an “architect” who understands the big picture.

The Issue

Without a blueprint, the bricklayer doesn’t know where to build the wall. If you ask a plumber to install a toilet without a plan, you might end up with it in the kitchen. Similarly, if you ask AI to “add a basement,” it might dig one but forget to rebuild the floor afterward. That’s the risk of coding without design. The bigger your project, the messier that risk becomes.

AI Can Make Unusual Mistakes Without A Blueprint

When AI Coding Works Best

Not every project needs an architect or a full set of blueprints. If you’re just painting a wall, paint the wall. If it turns out to be the wrong color, you can easily redo it. That’s the spirit behind what's called vibe coding the idea of pretending the code doesn't exist and just asking AI, in human language, what you want it to do.

AI Coding is Good for Simple Projects

Say you want to create a simple program to turn a lamp on and off. You could tell AI, “Write me code to toggle my lamp,” and if it doesn’t work, just have it try again. These small, low-stakes projects are like your weekend DIY jobs: you won’t burn the house down if something goes wrong. AI, in this context, is like a knowledgeable friend helping you out with the tools you have. And if you pay attention to how it builds the code, you’ll actually learn how to be a better “architect” for future, larger projects.

When to Take the Architect’s Seat

As soon as your project starts to grow such as multiple features, connections to other systems, or user interfaces you need to put on the architect’s hat. That means planning your foundation and thinking long-term. If you just keep saying “AI, add this” or “AI, change that,” the result will be a tangled, unreadable codebase that no human (including you) can maintain. A key danger here is AI might even modify or delete key sections of code without warning. Without structure, you’re building on quicksand.

Key Takeaways: The Pros and Cons of AI Coding

Where AI Shines:

• Speed: AI accelerates your workflow, letting you focus on creative or architectural aspects.

• Learning Aid: It’s like an apprenticeship. You can learn by watching how AI solves problems.

• Troubleshooting: Instead of waiting for help on forums, you can get immediate answers and fixes.

• Inspiration: AI often presents creative approaches you might not have considered.

Where AI Falls Short:

• Complex Projects: Not suited for enterprise or multi-layered systems where structure is critical.

• False Confidence: Just because AI says it works doesn’t mean it will handle all cases correctly.

• Poor Readability: Frequent edits lead to code that’s hard (or impossible) for humans to understand.

• Skill Decay: Overreliance on AI can weaken your ability to code and problem-solve from scratch. (I've experienced this personally)

  
  

Final Thoughts

AI is a powerful tool, but it’s still just that: a tool. Use it to learn, to prototype, to save time but not to replace understanding. The best results come when you, the architect, guide AI’s bricklaying.

See Our 3D Printed Smart Lamp Project

In our next post, we’ll put this into action by adding new features to our 3D-printed smart lamp using only AI coding. Thanks for reading, and stay tuned.