AGI (Artificial General Intelligence) is a hypothetical form of AI with human-level, general-purpose intelligence — able to understand, learn, and apply knowledge across any task, as a person can. It does not yet exist.
What it means in plain English
Today’s AI is narrow: each system is good at specific tasks but lacks broad, flexible understanding. AGI would be different — a single system that could reason, learn, and adapt across essentially any domain, transferring knowledge from one area to another the way humans do. Whether, when, and how AGI might be achieved is a major topic of debate among researchers.
It is distinct from the narrow AI behind every real product today.
A simple example
A narrow AI can play chess brilliantly but cannot then use that skill to plan a holiday. A hypothetical AGI could move fluidly between chess, holiday planning, writing, and science — applying general intelligence to whatever it faced.
Why it matters
AGI is the long-term aspiration (and concern) that frames much discussion about AI’s future. Understanding that it does not yet exist — and that all current AI is narrow — is key to separating realistic capabilities from science-fiction expectations.
Related terms
- Artificial Intelligence — the broader field AGI would be the pinnacle of.
- Narrow AI — the task-specific AI that exists today.
- Alignment — a central concern for advanced AI like AGI.
Frequently asked questions
Does AGI exist today?
No. Current AI systems are narrow — highly capable at specific tasks — whereas AGI refers to hypothetical AI with human-level general intelligence across virtually any task. AGI does not exist and its timeline is heavily debated.
How is AGI different from today’s AI?
Today’s models, even very capable ones, are trained for particular kinds of tasks. AGI would flexibly learn and reason across any domain the way a human can, without task-specific training.
