Beginner

Artificial intelligence: what are we actually talking about?

Artificial intelligence (AI) refers to software that can handle tasks we associate with human intelligence: understanding a sentence, recognizing an image, making a prediction, writing a text. If you use ChatGPT, a navigation app or your inbox's spam filter, you already use AI.

Data as of October 22, 2026

Key figures

LevelBeginner
Origin1955
Introduced byJohn McCarthy, Marvin Minsky, Nathaniel Rochester, Claude Shannon (proposition du projet de Dartmouth, 31 août 1955)

A plain-English definition

Artificial intelligence is a set of computing techniques that let a machine do things that used to require human judgment. It is not one single product but a family of approaches. Most of them today rely on machine learning: instead of writing every rule by hand, engineers show the program thousands or millions of examples and let it work out the patterns on its own.

A concrete example: your email app moves a message to the spam folder. Nobody wrote a rule for that exact message; the system learned, from a vast number of emails already sorted, what spam tends to look like.

Where the term comes from

The phrase appears in a research proposal dated August 31, 1955, written by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon. It announced a summer workshop at Dartmouth College in 1956, built on a bold conjecture: every aspect of learning could, in principle, be described precisely enough for a machine to simulate it. The text was republished by AI Magazine.

Why it matters to you

Since ChatGPT launched in November 2022, generative AI has moved into offices and homes alike: drafting emails, summarizing reports, planning trips. Knowing what AI is helps you decide what you can hand over to it and what you still need to check. At work, it also helps you follow conversations about approved tools, data privacy, or the obligations set out in the EU AI Act.

Common myths

  • “AI knows everything.” A large language model produces plausible answers, not necessarily accurate ones. It can be wrong with total confidence.
  • “AI is brand new.” The field is more than seventy years old. What is new is the quality of conversational models and how widely the public uses them.
  • “AI is conscious.” Nothing suggests it is. Artificial general intelligence, a system able to do everything as well as a person, remains a debated goal, not a reality.
  • “AI means robots.” Most AI is invisible: software running on a server or inside your phone.

To see where your own knowledge stands, the AI level test gives you a starting point that fits your level.

Related terms

Try it

Full glossary

Frequently asked questions

Does AI think like a human?

No. Today's systems spot patterns in huge amounts of data and produce the most likely answer. They have no intentions or awareness, even when their replies sound remarkably natural.

Who coined the term artificial intelligence?

It appears in a research proposal dated August 31, 1955, signed by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon. That document launched the Dartmouth workshop in the summer of 1956, widely seen as the birth of the field.

Is AI the same thing as ChatGPT?

ChatGPT is just one example of AI, from the generative AI family. AI also covers face recognition, video recommendations, credit card fraud detection and machine translation.

Is there an official definition?

The EU AI Act (Regulation (EU) 2024/1689) defines an AI system as a machine-based system that infers, from the input it receives, how to generate outputs such as predictions, content, recommendations or decisions. It is the main legal definition in the European Union.

Data sources

Data as of October 22, 2026

Report an error

Does a figure, a date or a fact look wrong? Tell us: every report is checked.