My Account Log in

1 option

Risk Assessment of Reinforcement Learning AI Systems Looking Beyond the Technology / Kyle Bunch.

Van Pelt Library UG479 .B863 2024
Loading location information...

Available This item is available for access.

Log in to request item
Format:
Book
Author/Creator:
Bunch, Kyle, author.
Hou, Alexander C., author.
Haberman, Ryan, author.
Herron, Marissa, author.
Jacques, Anthony, author.
Briggs, Gary J., author.
Contributor:
Acquisition and Technology Policy Center.
Rand Corporation.
United States. Office of the Under Secretary of Defense for Research and Engineering.
Series:
Research report (Rand Corporation) ; A1473-1.
Report ; A1473-1
Language:
English
Subjects (All):
United States. Department of Defense--Technological innovations.
United States.
Artificial intelligence--Military applications--United States.
Artificial intelligence.
Reinforcement learning.
Machine learning.
Risk assessment.
Autonomous Military Systems.
Machine Learning.
United States Department of Defense.
risk assessment.
Local Subjects:
Autonomous Military Systems.
Machine Learning.
United States Department of Defense.
Physical Description:
xi, 88 pages : illustrations ; 26 cm.
Other Title:
Risk Assessment of Reinforcement Learning Artificial Intelligence Systems
Place of Publication:
Santa Monica, CA : RAND, [2024]
Summary:
This report presents some of the challenges that the U.S. Department of Defense (DoD) may face in fielding an artificial intelligence (AI) technology called reinforcement learning (RL) in DoD applications. RL has been credited with expanding the decisionmaking ability of machines beyond that of humans in playing complex games of strategy. The fact that RL-enabled systems can beat world experts in these games raises the question of whether such systems could outperform humans in DoD applications. Especially relevant are "broad" applications having large, complex processes with multiple steps leading to few but consequential decisions for a military commander. Timely alternatives could lead to decisive advantages in such situations. What is not clear, however, is what risks such a system would introduce from a technical standpoint (i.e., technical failure leading to mission failure) or the risks to the force structure incurred in absorbing such technology. This report represents a first step toward understanding such risks associated with employing RL-enabled systems for operational-level command and control.
Contents:
CHAPTER 1: Narrow Versus Broad Artificial Intelligence and an Overview of Reinforcement Learning
CHAPTER 2: Further Technical Challenges of RL in Broader Problems
CHAPTER 3: Scenarios for Exploring Challenges
CHAPTER 4: Challenge Identification Framework and Workshop
CHAPTER 5: Conclusions and Recommendations
APPENDIX A: Additional Workshop Details
APPENDIX B: Challenges, Implications, and Mitigation Strategies Identified in the Workshop.
Notes:
Title from PDF document (title page; viewed July 3, 2024)
"Prepared for Office of the Under Secretary of Defense for Research and Engineering"
"RAND NATIONAL DEFENSE RESEARCH INSTITUTE"
Includes bibliographical references (pages 79-88)
ISBN:
1977413153
9781977413154
OCLC:
1444074996

The Penn Libraries is committed to describing library materials using current, accurate, and responsible language. If you discover outdated or inaccurate language, please fill out this feedback form to report it and suggest alternative language.

My Account

Shelf Request an item Bookmarks Fines and fees Settings

Guides

Using the Library Catalog Using Articles+ Library Account