Rules vs learning, explained simply
Not every problem needs AI. If the answer follows a clear, simple rule, such as "a school day is Monday to Friday", a programmer can just write that rule. It is faster, cheaper and easier to explain.
Machine learning helps when the pattern is too messy to write down. Try writing exact rules for recognising a friend’s face or telling whether a message sounds cheerful. There are too many exceptions.
Good builders choose the right tool. Sometimes that is a rule, sometimes a learned model, and sometimes a mix of both.
Example: Two problems, two tools
- Is this number even? A rule works perfectly: divide by 2 and check the remainder.
- Is this drawing a cat? No simple rule works for every cat, so learning from many examples is the better choice.
Try this
- Write rules for spotting a triangle. Then try writing rules for spotting a happy face. Which was easier?
- List five everyday decisions and sort them into "rule" or "learned from experience".
- Find a situation where a rule would be fairer or clearer than an AI model.