Machine Learning
1 min read · Reviewed 17 Oct 2026
In short
Software that learns patterns from examples instead of following written rules.
In plain English
Traditional software follows rules a person wrote. Machine learning works the other way round: you show it thousands of examples and it works out the pattern itself. Nobody writes the rule; the rule is inferred.
A real example
A spam filter is not a list of banned words written by hand. It has seen millions of emails labelled spam or not spam, and learned what spam tends to look like — including patterns no human would have thought to write down.
Why it matters
It explains why AI is good at fuzzy jobs and bad at precise ones. If a task has clear rules, ordinary software is more reliable and cheaper. Machine learning earns its place where the pattern is real but too messy to write out.
