Skip to content
SI.info

SI Glossary · Core concepts

Artificial General Intelligence (AGI)

Published 1 min read
On this page
  1. What AGI means
  2. AGI vs SI vs ASI
  3. Why it matters

Artificial general intelligence (AGI) is the long-standing goal of building a machine that can learn and perform essentially any intellectual task a human can: write a novel, run a business, do original science, learn a new language from scratch. The key word is general: AGI would not be limited to one domain.

What AGI means

There is no single agreed definition. Common versions include:

  • Human-level across the board: matching a typical skilled adult on most cognitive tasks.
  • Economic: a system that can do most economically valuable work. OpenAI’s charter defines AGI as “highly autonomous systems that outperform humans at most economically valuable work”.
  • Levels: Google DeepMind researchers proposed a ladder from “emerging” to “superhuman” AGI, measured by performance and generality.

Because definitions differ, claims that AGI has or hasn’t arrived often say more about the definition than the technology.

AGI vs SI vs ASI

TermMeaning
Narrow SIBuilt for specific tasks (most systems today)
AGIMatches humans across most tasks
ASIFar exceeds humans in nearly every domain

Under Executive Order 14434, all of these fall under “Super Intelligence” when they exist. But the order did not rename AGI: the term “artificial general intelligence” is unchanged.

Why it matters

AGI is the stated goal of several major labs, including OpenAI, Google DeepMind and Anthropic. Many researchers believe a system with general ability could accelerate its own improvement, which could lead quickly to superintelligence. See intelligence explosion. That possibility drives much of today’s safety and policy debate. For timelines, see When will superintelligence arrive?

← Back to the SI Glossary

Frequently asked questions

Has AGI been achieved?

There is no consensus. Some lab leaders argue today's frontier models are close to or already at AGI for many tasks; most researchers say systems still fall short on reliability, long-term planning and learning on the job.

Is AGI the same as SI?

No. SI (Super Intelligence) is the U.S. government's name for all AI, including narrow systems. AGI is a much higher bar: human-level ability across most tasks.

Written by

· Editor

Editor of SI.info. Writes about Super Intelligence, technology policy and the people building frontier models.

How we research and fact-check

Free newsletter

Get The SI Brief

One short email a week: what changed in Super Intelligence, policy and models — and why it matters.

Free. One email a week. Sent via beehiiv, which counts opens and clicks. Unsubscribe anytime. Privacy policy

Keep reading

Core concepts

Artificial Superintelligence (ASI)

Artificial superintelligence (ASI) is a hypothetical intelligence that greatly exceeds the best human minds in virtually every domain, from science to strategy to social skills.

Core concepts

Narrow SI

Narrow SI (formerly narrow AI) is any system built to do one specific task or a small set of tasks, like translating text, recognising faces or recommending videos.

Futures

Intelligence Explosion

An intelligence explosion is a hypothetical runaway process in which an SI system improves its own intelligence, each improvement making the next one faster, rapidly leading to superintelligence.

Core concepts

Frontier Model

A frontier model is one of the most capable SI models at a given time, typically trained with the most compute and the target of the strictest safety rules and commitments.