I'm a civil engineer who works mainly with physical engineering, including quantity and structural tasks, but I also need to handle some automation. I learned Python a few years ago and can generally understand code, but I never built many complete projects. Now that AI can generate code, I'm worried about accepting low-quality code without being able to debug, improve, or rewrite it myself. When I start a new project, I often go completely blank and don't know what the first step should be. I already use AI for research and code reviews, but I'd like to strengthen my fundamentals and become more confident building useful tools for my work. If the comment about a building falling down is literal rather than figurative, I'd contact a qualified structural engineer or emergency service immediately rather than rely on code or AI.
2 Answers
Start with a few small, practical projects instead of trying to learn every topic first. Build something like a calculator, a spreadsheet processor, or a tool that works with engineering drawings or quantity data. Break each project into tiny steps: describe the input, define the expected output, write pseudocode, then implement and test one function at a time. Building projects is what turns isolated Python knowledge into usable fundamentals.
Use AI as a tutor, research assistant, and code reviewer rather than as the person driving the whole project. Before asking it to write code, explain the problem yourself, sketch the solution, and decide how you’ll test it. Then compare its suggestions with your own plan. Read every generated line, ask why it works, and deliberately introduce small changes or debugging exercises. If your main goal is shipping useful tools, AI can speed things up; if your goal is stronger fundamentals, make sure you still write and troubleshoot some projects without it.
That makes sense. I learned Python topic by topic when I was younger, but I never really built anything substantial. I’m returning to it as a working professional, so starting with small tools related to my field seems much less overwhelming.