General AI describes a machine that could understand, learn and perform any intellectual task a human can. It would not need to be rebuilt for each new job. It could learn accounting on Monday, pick up a new language on Tuesday and apply ideas from one to the other.
This page explains general AI as a category in the standard three-level model of AI. For how the industry defines and measures it today, see What is AGI?
Where general AI sits
- Narrow AI: one task, very well. Exists today.
- General AI: any task, at roughly human level. Does not exist yet.
- Superintelligence: any task, far beyond human level. Hypothetical.
What general AI would need
Researchers usually point to a few abilities that current systems lack or only partly show:
- Transfer learning: using knowledge from one field to solve problems in another.
- Common sense: the background understanding people use without thinking.
- Reasoning and planning: working through new problems step by step, not just matching patterns.
- Continuous learning: improving from experience without being fully retrained.
Strong AI and general AI
You will often see “strong AI” used as a synonym. Philosopher John Searle introduced the strong/weak distinction in 1980. For him, strong AI meant a machine that truly has a mind and understanding, not only human-level behaviour. Today most writers use the terms loosely as the same thing, but the original meaning was about consciousness, not capability.
How close are we?
Nobody knows. Some leading AI labs say general AI could arrive within years. Many researchers think it is decades away, or that current methods will not get there at all. Because there is no agreed test for it, even the question “have we reached it?” is disputed.