"Artificial intelligence" is a category, not a product. Today it usually refers to systems that learn patterns from massive amounts of data and use those patterns to generate text, images, code, or decisions. The version that's everywhere right now — writing emails, summarizing documents, powering chat assistants — is built on large language models (LLMs).
The term "AI" went mainstream in late 2022 when ChatGPT made these models publicly accessible, but the underlying research is decades old. What changed is scale and accessibility, not the core concept.