Artificial intelligence (AI) is the branch of computer science that builds systems able to do tasks we usually associate with human thinking. These tasks include understanding language, recognising images, spotting patterns, making predictions and choosing between options.
A simple definition
An AI system takes in data, finds patterns in it and uses those patterns to produce an output. The output might be an answer, a recommendation, a label on a photo or a decision. The system does not “think” the way a person does. It is very good at one kind of pattern matching, often at a scale no human could manage.
How AI works in practice
Most AI today is built with machine learning. Instead of writing every rule by hand, engineers show the system many examples and let it learn the rules itself. A spam filter learns from millions of emails marked “spam” or “not spam”. A language model learns from huge amounts of text how words tend to follow each other.
Older AI relied on rule-based systems, where experts wrote explicit “if this, then that” logic. These still exist in areas where rules must be exact and explainable, such as tax calculation or some medical checklists.
The main types of AI
- Narrow AI: good at one task or a small set of tasks. Every AI product in use today falls here.
- General AI (AGI): a system that could learn and perform almost any mental task a human can. It does not exist yet.
- Superintelligence: a hypothetical system that would far exceed human ability in nearly every field.
A short history
The term “artificial intelligence” comes from a 1955 research proposal by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon for the 1956 Dartmouth summer workshop. The field has gone through cycles of excitement and disappointment since then, often called “AI winters”. The current wave began around 2012, when deep learning made big jumps in image and speech recognition possible.
Why it matters
AI now sits inside search engines, bank fraud checks, maps, email, customer support and writing tools. For businesses, the question is rarely “should we use AI?” and more often “which tasks is it reliable enough for, and where does a human still need to check the work?”