How to make artificial agents more like natural agents.

Journal: Trends in cognitive sciences
Published Date:

Abstract

The quest to make artificial intelligence models more human-like could profit from a study of biological agents and their evolution. Recent research suggests that animal species on the evolutionary line to humans employed a series of qualitatively distinct agentive architectures of ever-increasing complexity, with humans then forming shared agencies coordinated via linguistic communication.

Authors

  • Michael Tomasello
    Duke University, Durham, NC, USA; Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany. Electronic address: michael.tomasello@duke.edu.

Keywords

No keywords available for this article.