What is AI?
Quick answer
AI (artificial intelligence) is technology that lets computers perform tasks that normally need human intelligence, such as understanding language, recognizing images, making predictions and solving problems. Most modern AI works through machine learning: instead of following hand-written rules, it learns patterns from large amounts of data.
You’ve probably used AI today without noticing: in the spam filter on your email, your phone’s photo search, map directions that avoid traffic, or the recommendations on a streaming service.
How does AI work?
Traditional software follows instructions a programmer writes step by step. Machine learning works differently: a system is shown huge numbers of examples, like millions of photos labeled “cat” or “not cat”, and adjusts itself until it can make good predictions on new examples it has never seen.
Deep learning is a type of machine learning that uses neural networks, layers of simple mathematical units loosely inspired by brain cells. It’s behind most of the big advances of the past decade, from speech recognition to self-driving research.
Key terms
| Term | Meaning |
|---|---|
| Artificial intelligence | The broad field of making machines do “intelligent” tasks |
| Machine learning | AI that learns from data instead of fixed rules |
| Deep learning | Machine learning with many-layered neural networks |
| Generative AI | AI that creates new text, images, audio, video or code |
| Large language model (LLM) | A generative AI trained on vast amounts of text, able to answer questions and write |
What is generative AI?
Generative AI produces new content rather than just sorting or predicting. Chat assistants such as ChatGPT, Claude, Gemini and Microsoft Copilot are built on large language models, while tools like image generators create pictures from a text description. They work by predicting what comes next, one piece at a time, based on patterns learned during training.
What AI can’t do (yet)
- It can be confidently wrong. Language models sometimes produce false information that sounds convincing, often called hallucinations, so important facts should be checked.
- It can reflect bias in the data it learned from.
- It doesn’t understand the world like people do. Today’s AI is narrow: very good at specific tasks, but without the broad, flexible understanding researchers call artificial general intelligence (AGI).
A short history
The term “artificial intelligence” was coined for a 1956 workshop at Dartmouth College. Progress came in waves, with “AI winters” when funding dried up. The modern boom took off in the 2010s, as faster chips and huge datasets made deep learning practical, and again in the 2020s with generative AI.
Many AI tools connect to other software through APIs. Here’s what an API is, explained simply.
Still curious?
Jump to a random question from the archive.