Foundations
Artificial General Intelligence
Hypothetical AI that matches or surpasses human capabilities across virtually all cognitive tasks, in contrast to today's narrow, task-specific AI.
Artificial General Intelligence (AGI) is a hypothetical form of AI that matches or surpasses human capabilities across virtually all cognitive tasks. Unlike today’s “narrow” AI, whose competence is confined to well-defined tasks, an AGI would generalise knowledge, transfer skills between domains, and solve novel problems without task-specific reprogramming.
The term has no single, measurable definition. In 2023 Google DeepMind proposed a five-level ladder (emerging, competent, expert, virtuoso, superhuman), placing today’s frontier models only at the lowest rung. AGI is usually distinguished from the still broader “superintelligence,” which would exceed humans in every domain.
AGI sits at the centre of the 2025-2026 debate: leaders at OpenAI, Anthropic, and DeepMind cite timelines of just a few years, while sceptics argue that scaling laws and reasoning models do not necessarily lead to general intelligence. Because of the potential risks, alignment and AI safety have become central research priorities.