Consciousness in Artificial Intelligence: Insights From the Science of Consciousness

By Patrick Butlin et al

Published on Aug. 22, 2023

Read the original document by opening this link in a new tab.

Table of Contents

1 Introduction
1.1 Terminology
1.2 Methods and Assumptions
2 Scientific Theories of Consciousness
2.1 Recurrent Processing Theory
2.2 Global Workspace Theory
2.3 Higher-Order Theories
2.4 Other Theories and Conditions
3 Consciousness in AI
3.1 Implementing Indicator Properties in AI

Summary

The document explores the topic of whether current or near-term AI systems could be conscious through a rigorous and empirically grounded approach. It discusses various scientific theories of consciousness and indicator properties derived from these theories to assess AI systems. The authors argue that while no current AI systems are conscious, there are no obvious technical barriers to building AI systems that satisfy the indicators. They propose a rubric for assessing consciousness in AI and provide initial evidence that many indicator properties can be implemented in AI systems using current techniques. The report emphasizes the importance of further research on the science of consciousness and its application to AI, as well as the moral and social risks of building conscious AI systems.