Artificial General Intelligence

What is AGI?

Artificial General Intelligence (AGI) is the idea of an AI that can understand, learn and perform any thinking task a human mind can.

Today's AI is usually designed for particular kinds of tasks, such as translating, recognising images or playing chess. AGI would be flexible across many different tasks and areas.

It could apply something learnt in one area to a different problem, much as a person might use strategic thinking from chess when learning a card game.

AGI does not exist today. It is a goal that many AI labs are working towards, and experts do not agree on exactly how it should be defined or measured.

When will AGI happen?

Nobody knows.

• DeepMind CEO Demis Hassabis and OpenAI CEO Sam Altman have suggested it could arrive by 2030.
• Anthropic CEO Dario Amodei has suggested AGI-level systems could arrive around 2027 or 2028.

These are predictions, not confirmed timelines.