Writing Code with ChatGPT

Whether you are a seasoned engineer or a complete beginner, ChatGPT is a force multiplier for coding speed. It excels at Boilerplate, Logic Implementation, and Translation.

1. Generating Boilerplate

Writing the basic structure of a project is tedious.

• "Generate a Python FastAPI project structure with three endpoints: /user, /data, and /health. Include a requirements.txt file."

2. Implementing Specific Logic

Don't ask for "the whole app." Ask for specific functions.

• "Write a recursive Python function to find the sum of all nested values in a deeply nested JSON object."

graph LR
    Goal[English Requirement] --> AI[ChatGPT Engine]
    AI --> Code[Source Code]
    Code --> Test[Compiler/Interpreter]
    Test -->|Fail| Dev[Human Fix/Refine]
    Test -->|Pass| Ship[Final Product]

3. Translation Between Languages

Knowing one language well allows you to write in others using AI.

• "Translate this Javascript map function into a Python List Comprehension."

4. Best Practices for Coding

[!IMPORTANT] Always verify AI code. It can generate Insecure Code (like hardcoded passwords) or use Outdated Libraries.

Use Markdown Snippets

When asking for code, ask for it in Markdown blocks for easy copying.

• "Provide the final code in a single Python markdown block."

---

Hands-on: The "FizzBuzz" Test

1. Prompt: "Write a Python script that prints the numbers from 1 to 100. For multiples of three, print 'Fizz' instead of the number. For multiples of five, print 'Buzz'. For numbers that are multiples of both three and five, print 'FizzBuzz'."
2. Review: Look at the logic. Is it efficient? Could it be shorter?

Key Takeaways

• Use AI for specific functions, not giant systems.
• Always include the Environment (e.g., "Python 3.10").
• Treat the AI as a Pair Programmer, not a replacement.