Regular Programs vs AI: What's the Difference? (Kid-Friendly)

When people say "AI," they often mean it as if all software is the same. It's not. Understanding the difference between a regular computer program and AI is the foundation of AI literacy — and it's simpler than it sounds.

Regular programs: following rules

A regular computer program follows instructions written by a human programmer. Every possible situation is anticipated and coded:

• If the temperature is above 22°C → turn on the fan
• If the user clicks "submit" → save the form data
• If the score reaches 100 → show "You Win!"

Regular programs are predictable and consistent. They do exactly what they were told — nothing more, nothing less. A calculator is a perfect example: it follows precise mathematical rules and always gives the same answer to the same input.

AI programs: learning from examples

AI programs work differently. Instead of being given rules, they're given thousands of examples and asked to figure out the pattern themselves.

To build a face recognition system using regular programming, you'd have to write rules like: "A face has two eyes, a nose, a mouth, roughly oval shape..." — but faces look so different (different lighting, angles, expressions, ages) that you'd never be able to write enough rules. It's an impossible rulebook.

Instead, AI is shown 1 million labelled photos ("this is a face / this is not a face") and learns the pattern itself. The resulting model handles new faces it's never seen, at angles the training data didn't include, and in lighting conditions the programmer never anticipated.

Side by side

Why both still exist

AI is NOT replacing regular programs. They solve different problems. Your phone uses a regular program to set an alarm at exactly 7:00am every day — that doesn't need AI. But it uses AI to unlock your face, because that's too complex for rules. The best technology combines both: precise logic where it's needed, AI where patterns are too complex to code manually.

The activity: rule writer vs AI trainer

This is Lesson 1.2 in the AI Adventures course. Try this at home: pick a task (sorting toys by colour, or identifying animals from photos) and try two approaches:

1. Regular program approach: Write out every rule you'd need to cover all cases. How long is your rulebook? How many exceptions do you find?
2. AI approach: Instead of rules, gather 20 examples. Would an AI trained on those examples do better than your rulebook?

Almost always, the AI approach handles edge cases that the rulebook missed. That's the insight that makes AI powerful.