Levels of AGI: Operationalizing Progress on the Path to AGI

\\correspondingauthor

merrie@google.com

# Levels of AGI: Operationalizing Progress on the Path to AGI

Meredith Ringel Morris
Google DeepMind
Jascha Sohl-dickstein
Google DeepMind
Noah Fiedel
Google DeepMind
Tris Warkentin
Google DeepMind
Allan Dafoe
Google DeepMind

Aleksandra Faust
Google DeepMind
Clement Farabet
Google DeepMind
Shane Legg
Google DeepMind

###### Abstract

We propose a framework for classifying the capabilities and behavior of Artificial General Intelligence (AGI) models and their precursors. This framework introduces levels of AGI performance, generality, and autonomy. It is our hope that this framework will be useful in an analogous way to the levels of autonomous driving, by providing a common language to compare models, assess risks, and measure progress along the path to AGI. To develop our framework, we analyze existing definitions of AGI, and distill six principles that a useful ontology for AGI should satisfy.
These principles include focusing on capabilities rather than mechanisms; separately evaluating generality and performance; and defining stages along the path toward AGI, rather than focusing on the endpoint. With these principles in mind, we propose “Levels of AGI” based on depth (performance) and breadth (generality) of capabilities, and reflect on how current systems fit into this ontology. We discuss the challenging requirements for future benchmarks that quantify the behavior and capabilities of AGI models against these levels. Finally, we discuss how these levels of AGI interact with deployment considerations such as autonomy and risk, and emphasize the importance of carefully selecting Human-AI Interaction paradigms for responsible and safe deployment of highly capable AI systems.

###### keywords:

AI, AGI, Artificial General Intelligence, General AI, Human-Level AI, HLAI, ASI, frontier models, benchmarking, metrics, AI safety, AI risk, autonomous systems, Human-AI Interaction

## 1 Introduction

Artificial General Intelligence (AGI)111There is controversy over use of the term “AGI.”
Some communities favor “General AI” or “Human-Level AI” (Gruetzemacher and Paradice, [2019](https://ar5iv.labs.arxiv.org/html/2311.02462#bib.bib30 "")) as alternatives, or even simply “AI” as a term that now effectively encompasses AGI (or soon will, under optimistic predictions). However, AGI is a term of art used by both technologists and the general public, and is thus useful for clear communication.
Similarly, for clarity we use commonly understood terms such as “Artificial Intelligence” and “Machine Learning,” although we are sympathetic to critiques (Bigham, [2019](https://ar5iv.labs.arxiv.org/html/2311.02462#bib.bib10 "")) that these terms anthropomorphize computing systems. is an important and sometimes controversial concept in computing research, used to describe an AI system that is at least as capable as a human at most tasks. Given the rapid advancement of Machine Learning (ML) models, the concept of AGI has pass

... (truncated, 86 KB total)