Rules vs Learning: Two Ways to Program

There are two fundamentally different ways to get a computer to do something useful. In traditional programming, a human thinks through the problem and writes down every rule the computer must follow — "if this happens, do that." In machine learning, the human does not write the rules at all. Instead, the human provides many examples together with the desired outputs, and an algorithm figures out the rules on its own.

Almost every AI system you use — voice assistants, photo tagging, spam filters, translation apps — exists because engineers realized that for some problems, writing the rules by hand is practically impossible, while learning the rules from examples works remarkably well. Understanding this difference is the key that unlocks the rest of this course.

In this lesson you will learn to:

• Explain how traditional (rules-based) programming works and where it shines.
• Explain how machine learning flips the process: examples in, rules out.
• Walk through a real comparison — building a spam filter both ways.
• Decide, for a given problem, whether rules or learning is the better tool.
• Recognize hybrid systems that combine both approaches in real products.

In traditional programming, a person analyzes a problem, breaks it into precise steps, and writes those steps as explicit instructions — usually chains of if–then rules. The computer then follows the rules exactly, every single time.