AI Architectures for Self-Reflective Never-Ending Learning

Abstract

The PI, Dr. Tom Mitchell, proposes to investigate how to build on AI progress in component parts of intelligence. Such parts as in computer vision, natural language, machine learning, etc. to design architectures for AI agents that incorporate multiple such components, that learn continuously through routine operation, and that can self-reflect on their goals, strategies and actions to reason about and to explain them, to themselves and to others; both agents and humans. If successful, this line of research could potentially lead to dedicated AI agents for each aircraft, each airport, each soldier, accompanying and assisting that aircraft or soldier, and learning continuously from experience and instruction how to best help them perform their intended role-mission.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 29, 2024
Source ID
FA95502310196

Entities

People

  • Tom M. Mitchell

Organizations

  • Air Force Office of Scientific Research
  • Carnegie Mellon University
  • United States Air Force

Tags

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Geospatial Intelligence and Artificial Intelligence Analytics
  • Military History of the United States in the 20th Century.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy