AGI (Artificial General Intelligence)
AI that can perform any intellectual task a human can the long-stated goal of frontier AI labs.
AGI describes hypothetical AI capable of generalizing across any cognitive task a person can do, rather than being narrow specialists at chess, image recognition, or chat. Definitions vary widely: some emphasize economic benchmarks (e.g., 'autonomously completes most knowledge work'), others emphasize learning ability or self-directed goal pursuit.
There's no consensus on whether we're close to AGI or far from it. The capability of frontier 2026 models clearly exceeds 2020 expectations of what AI would be capable of by now, while still falling short of any reasonable AGI definition. Whether scaling existing transformers gets us there or whether new approaches are required is unsettled.
OpenAI, Anthropic, Google DeepMind, and others all state AGI as a goal. The discussion has become as much about safety and governance as about capability research.