Skip to content
Longterm Wiki

Center for Human-Compatible AI

Academic

Also known as: CHAI

The Center for Human-Compatible AI (CHAI) is an academic research center at UC Berkeley focused on ensuring AI systems are beneficial to humans. Founded by Stuart Russell, author of the leading AI textbook, CHAI brings academic rigor to AI safety research.

Facts

1
General
Websitehttps://humancompatible.ai

Divisions

1

Related Wiki Pages

Top Related Pages

Approaches

AI AlignmentCooperative IRL (CIRL)Constitutional AIAI EvaluationAI Safety Training Programs

Analysis

AI Compute Scaling MetricsAI Safety Intervention Effectiveness MatrixAI Safety Research Allocation ModelAI Risk Interaction MatrixAI Safety Researcher Gap ModelGoal Misgeneralization Probability Model

Organizations

Center for AI SafetyAnthropicCenter for Applied RationalityOpenAI

Other

Vipul Naik

Risks

Corrigibility Failure

Concepts

Safety Orgs Overview

Historical

Deep Learning Revolution Era