Artificial intelligence is a broad label for software that performs tasks we normally associate with human thinking: recognizing a face, translating a sentence, recommending a song, or predicting whether a loan will default. There is no single algorithm called "AI" — it's an umbrella term covering many different techniques, from simple rule-based systems to today's large neural networks.
Most of the AI you interact with day to day is what researchers call narrow AI: a system trained to do one thing well, like ranking search results or filtering spam. It has no general understanding of the world beyond its training data. The idea of a machine with human-level reasoning across every domain — often called artificial general intelligence — remains a research goal, not a shipped product.
What changed in the last decade wasn't a single breakthrough but a combination: more data, cheaper computing power, and better training methods for neural networks. That combination is why AI suddenly feels everywhere, from your phone's camera to your inbox's spam filter.