The simple timeline
Researchers began asking whether a machine could do things that looked intelligent. In 1956, a summer workshop helped give the field its name: artificial intelligence.
People tried telling computers exactly what to do: if this happens, do that. It worked for narrow problems. Real life was too messy for a rule about everything.
Instead of writing every rule, people gave computers many examples. The computer found patterns inside them. This is called machine learning.
More data, faster computers, and better methods made it possible to recognize speech, faces, objects, and complicated patterns much more accurately.
Tools such as ChatGPT can create new words, pictures, sound, and video by using patterns learned from enormous collections of examples.
Why does it feel sudden?
The work is old. Easy access is new. A few years ago, most people needed special knowledge or expensive systems to use advanced AI. Now you can open an app and ask a question in ordinary language.
That change—from hidden technology to an everyday tool—is why it feels as if AI arrived all at once.
What you need to remember
You do not need to memorize the history. Keep one idea: AI got better because computers could study more examples, find more complicated patterns, and use those patterns to make something new.
That explains both the impressive part and the problem. It can make something that works. It can also make a very believable mistake.