Briefing to the Air Force Scientific Advisory Board: T and E Contributions to Avoiding Unintended Behaviors in Autonomous Systems

Abstract

Testers rely on making valid inferences about a systems performance in untested situations in order to evaluate and certify the system. However, when systems are black boxesas is frequently the case for AI-enabled technologiesthese inferences are not possible. Avoiding unintended behaviors in these systems, however, requires us to make valid predictions about behavior. This means we must have models of system decision makingan understanding of what causally drives systems to make one decision over another.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 2020
Accession Number
AD1122120

Entities

People

  • Daniel J. Porter
  • Heather M. Wojtan

Organizations

  • Institute for Defense Analyses

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Abstracts
  • Actuators
  • Air Force
  • Artificial Intelligence
  • Autonomous Systems
  • Autonomy
  • Bayesian Networks
  • Contracts
  • Delphi Method
  • Department Of Defense
  • Detectors
  • Distance Learning
  • Environment
  • Governments
  • Human-Machine Systems
  • Instructions
  • Instrumentation
  • Machine Learning
  • Organizational Structure
  • Perception
  • Procurement
  • Security
  • Simulations
  • Supervisory Control
  • Test And Evaluation
  • Training
  • Validation
  • Verification
  • Virginia

Fields of Study

  • Computer science

Readers

  • Aerospace Test and Evaluation
  • Artificial Intelligence
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - DoD AI Strategy
  • Autonomy