A Retrospective in Engineering Large Language Models for National Security

Abstract

Large Language Models (LLMs) are powerful AI tools that can potentially save time and effort for users. However, LLMs alsodemonstrate behaviors that can lead to questions about their trustworthiness. These types of concerns may lead to significantconsequences, particularly in high-stakes contexts such as national security and defense.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2023
Accession Number
AD1214460

Entities

People

  • Andrew Mellinger
  • Angel Mcdowell
  • Bryan Brown
  • Eric Heim
  • Jasmine Ratchford
  • Nathan Vanhoudnos
  • Shannon Gallagher
  • Tyler Brooks
  • Will Nichols

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Autonomy
  • Biomedical
  • Cyber
  • Human Systems

DTIC Thesaurus Topics

  • Air Force
  • Altitude
  • Artificial Intelligence
  • Computer Science
  • Department Of Defense
  • Engineering
  • Engineers
  • Governments
  • Guarantees
  • High Altitude
  • Human-Machine Systems
  • Infrastructure
  • Intelligence Community
  • Intelligence Community (United States)
  • Language
  • Materials
  • National Security
  • Security
  • Software Development
  • Test And Evaluation
  • United States

Fields of Study

  • Computer science

Readers

  • Artificial Intelligence
  • Aviation Safety Risk Assessment.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.