Army of None
Autonomous Weapons and the Future of War
Paul Scharre
W. W. Norton, New York, 2018, 448 pages
Book Review published on: February 17, 2023
Paul Scharre, the author of Army of None: Autonomous Weapons and the Future of War, conducted extensive research into the development, use, and policy surrounding automated and autonomous systems. Scharre’s book outlines the various methods of employment for automated systems and discusses several policies governing their current use and potential use. While looking at the future of warfare, particularly within the cyber domain, understanding the capabilities, vulnerabilities, and dilemmas autonomous systems bring to the fight is paramount for military and political professionals. Scharre outlines three salient concerns: lack of global consensus, autonomy versus automated, and convergence of the preceding elements within the cyber domain.
Early on within Army of None, Scharre discusses the lack of consensus on what autonomous systems were and the ethical use of those systems. The concerning lack of consensus may inhibit development of autonomous systems within some nations, while allowing for explosive growth among other nations. The speed at which autonomous systems operate creates a significant advantage for nations that employ them. However, these systems potentially create ethical dilemmas through their employment. As a result, nations willing to accept ethical risks, such as Russia and China, are more likely to develop and employ autonomous systems than more risk adverse nations such as the United States of America and the United Kingdom. This risk aversion leads to either a human-on-the-loop or human-in-the-loop approach when developing autonomous systems.
The human-on-the-loop/human-in-the-loop approach leads to an automated approach versus an autonomous approach. Scharre provides several examples of automated weapon systems that have humans within the process, to include the Aegis and Patriot systems. Despite human involvement, there have been accidents resulting in the loss of life. The question is whether removing the human or adding additional personnel would reduce the loss of life. While these machines can exponentially decrease the decision cycle, the human(s) within the cycle slows it down. However, the ability to look at ethical situations and trust the gut instinct is still highly valued within certain societies. As a result, the decision to pursue automated weapons as opposed to autonomous weapons is largely based within a nation’s culture and world views.
Lastly, autonomous weapons within the cyber domain present both a unique solution and a unique problem. Scharre explores an automated system that conducts vulnerability scans on networks and develops patches for the vulnerabilities it discovers. However, if this system was deployed as a cyber weapon, it could not only attack networks through identified vulnerabilities, but it could also defend itself. The primary risk with this type of weapon is if it spreads beyond the intended target similar to Stuxnet in 2010. As a result, a highly sophisticated autonomous cyber weapon could virtually cripple the digital infrastructure that our world relies on. The speed at which this weapon would evolve could only be combated with similar weapons.
In conclusion, Scharre’s book is enlightening on the benefits, vulnerabilities, and ethical issues surrounding the development of autonomous systems. However, like the development of the nuclear bomb, it is difficult to say where the development of autonomous systems will lead. The potential they have is revolutionary and could exponentially enhance human civilization, or they could destroy it. In either scenario, the development of these systems will likely lead to an arms race closely mirroring that of the nuclear arms race.
Book Review written by: Maj. Travis S. Johnson, U.S. Army, Wiesbaden, Germany