Blog Details

Welcome to the Lab 🚀

Today, we’re going to use AI as your coding partner. You’ll see how a Large Language Model can turn your problem description into working Python code — instantly. By the end, you’ll be able to generate functions confidently, tweak them, and learn from AI suggestions.

Step 0: Set the Scene

Imagine: You’re building a small math toolkit for students. One core function: factorial.

• Input: integer n
• Output: factorial of n (n!)
• Special case: if n is negative, raise an error.

📝 Mini-challenge: Write down, in plain English, what your function should do. Be as explicit as possible — this is the blueprint for your AI prompt.

Step 1: Crafting Your Prompt

There are many ways to ask the AI. Here are two:

Option A — Zero-shot prompt:

Write a Python function factorial(n) that returns the factorial of n. 
Raise a ValueError for negative inputs.

Option B — Few-shot prompt (with examples):

Example:
5 → 120
0 → 1

Now write a Python function factorial(n) that behaves like this.

💡 Pro Tip: Few-shot prompts give the model context and usually produce cleaner, more reliable code.

Step 2: See the AI Work

AI might generate:

def factorial(n):
    if n < 0:
        raise ValueError("Input must be non-negative")
    result = 1
    for i in range(2, n+1):
        result *= i
    return result

✅ Already looks good! But let’s test it.

Step 3: Test Your Function

print(factorial(5))  # 120
print(factorial(0))  # 1
print(factorial(-1)) # ValueError

• Works perfectly!
• If it didn’t, think: “What could I clarify in the prompt?”
• Maybe add type hints, docstrings, or specify loop vs recursion.

Step 4: Iterative Refinement — Make It Better

Refine prompt to include style or efficiency:

Write a Python function factorial(n: int) -> int using recursion. 
Include a docstring and handle negative inputs with ValueError.

AI Output Example:

def factorial(n: int) -> int:
    """Return the factorial of n. Raise ValueError if n is negative."""
    if n < 0:
        raise ValueError("Input must be non-negative")
    if n == 0:
        return 1
    return n * factorial(n-1)

🔍 Notice the difference? The AI adapted to recursive style and added type hints + docstring.

Step 5: Mini Lab Challenges

Try generating these using AI:

1. is_prime(n) — check if n is prime, raise errors for invalid inputs.
2. reverse_string(s) — reverse a string safely; handle empty/invalid input.
3. Compare zero-shot vs few-shot prompt outputs: which one is cleaner? Why?
4. Try adding efficiency constraints (O(n) vs O(n^2)) to the prompt — see how the AI adapts.

Step 6: Pro Tips for AI Code Generation

• Always specify input/output clearly.
• Few-shot examples = better code.
• Iteration is normal: AI might not get it right first time.
• Use AI-generated code to learn coding patterns, not just copy-paste.
• Include constraints, styles, or documentation in prompts for production-ready code.

Lab Summary 🏁

• AI is your coding assistant, not a replacement for thinking.
• Prompt clarity + examples + iteration = key to reliable code.
• Experiment with style, recursion vs loops, error handling, and documentation.
• Each small prompt tweak teaches you how the LLM “thinks” and how to guide it effectively.