Artificial intelligence (AI) is software that does tasks we normally think need human judgement: recognising a face, translating a sentence, recommending a job, or writing a paragraph. It is not one technology. It is a broad label for many techniques.
AI, machine learning and deep learning
These three terms are nested. AI is the big circle. Machine learning (ML) is the part of AI where a program learns patterns from examples instead of following rules a person wrote. Deep learning is a kind of machine learning that uses many-layered neural networks, and it powers most modern language and image systems.
- Rule-based systems: a developer writes the rules ("if the CV has no email, flag it"). Predictable, but they break on cases nobody thought of.
- Learning systems: the program is shown many examples and works out the pattern itself. Flexible, but the result can be wrong in ways that are hard to predict.
Narrow AI and general AI
Every AI product you can use today is narrow AI: very good at a limited set of tasks. A resume checker does not drive a car. "General" AI, a system that can do any intellectual task a person can, is a research goal and not a product you can buy.
Where you already meet AI
- Spam filters, search ranking and recommendations.
- Voice assistants and speech-to-text.
- Resume parsing and applicant tracking systems (ATS) that sort applications.
- Chat assistants such as Claude that write, summarise and explain.
Limits you should know about
- AI learns from data, so biased or missing data gives biased or poor results.
- It can be confidently wrong.
- It does not "understand" the way people do, and it has no goals unless a person builds them in.
- It needs checking by a human when the stakes are high, such as hiring, health or money.