Artificial general intelligence describes broad, transferable capability rather than excellence on one benchmark or task. A qualifying system would adapt knowledge across domains, reason about unfamiliar problems and perform many kinds of useful work.
There is no universally accepted operational threshold for the term. Claims about an internal or future system therefore need clear evidence, defined tests and separation between demonstrated capability, forecasts and rumors.
Acronyms and aliases
AGI acronymgeneral AI variant
General terms
Related terms
Frequently asked questions
How is artificial general intelligence different from narrow artificial intelligence?
Narrow systems are optimized for limited tasks, while artificial general intelligence would transfer competence across many unfamiliar domains.
Has artificial general intelligence been achieved?
There is no broad consensus that it has been achieved, and public claims depend strongly on the definition and evidence used.
