The militaries of developed states are on the cusp of witnessing the second Revolution in Military Affairs (RMA), with the wide-scale introduction of automated and autonomous systems and weapons. These will be based on a wide range of technologies, including IT, biotechnology, robotics, AI and nanotechnologies. The development of these kinds of weapons was pioneered by the Nazis and Japanese during WWII, although their autonomous systems were not effective; they remained part of the totalitarian dreamscape. During the following decades, this area of military research experienced gradual improvements, but it has still largely been limited by computer speeds. At present, with the ever-evolving computer technologies, the second RMA is about to commence. Autonomous weapons are militarily desirable and it is a matter of time before they become fully available. On the other hand, few can adequately anticipate the manner in which autonomous weapons will evolve. Enter Krishnan who assesses the upcoming revolution in Killer Robots: Legality and Ethicality of Autonomous Weapons. While the work provides a very detailed description of many future systems, the author does not focus on the technological aspect alone, instead, Krishnan analyses the entire problematic as a phenomena which affects the international community as a whole.
The book is divided into six chapters, excluding the introduction, in which Krishnan provides the reader with a general description of what autonomous weapons (AWs) are, and what enables their development. The following part, called ‘The Rise of Military Robotics,’ provides a historical context for the rise AWs, including the aforementioned Nazi and Japanese experiments. The next chapter, ‘Weapons Autonomy and AI,’ focuses on the question of artificial intelligence and possible problems that armies deploying highly intelligent AWs could face. Further questions regarding the use of AWs, in combination with lowered levels of human command, are discussed in the third chapter, ‘The Robotics Revolution of Warfare.’ These concerns lead Krishnan to think about international regulations for AWs which are described in a chapter called ‘The Legality of Autonomous Weapons.’ The introduction of artificial intelligence represents several social dangers, which are further developed in the chapters ‘Ethical Considerations’ and ‘Dangerous Futures and Arms Control.’ The book provides a complete analysis of current and future situations regarding AWs, and poses serious questions which require urgent addressing.
The standpoint – from which Krishnan commences – is based on the perspective of postmodern societies’ attitudes towards the sacrifice of lives during wars. As respect for human freedom and dignity rose, it seemed natural that the willingness to allow high soldier casualties declines. The armies are facing a situation, which pushes them to figure out how to perform desired operations and to wage war with a lower number of people involved. This led to the introduction of new weapons, such as nuclear weapons and new types of military strategies and tactics, which work with smaller and more efficient troops of soldiers. This need to make armies more efficient also sprung the development of AWs. Although Krishnan talks about nuclear weapons as the main subject of the first RMA, he makes a sharp distinction between them and AWs, because, as he claims, AWs can lower the level of human sacrifice in actual battle, contrary to the huge and fatal destruction abilities of nuclear weapons. AWs can be truly revolutionary weapons because, if used properly, they could save human lives. Also, there is a significant role of military expenses that Krishnan points out—autonomous weapon does not need salaries; they could fight with a precision unattainable for human soldiers, and they feel no pain and experience no fear. As a consequence, when AWs would be deployed on a large scale instead of human soldiers, the whole defense budgets of states could be substantially lowered. Therefore, AWs are militarily desirable and as mentioned above, it is only a matter of time, when they become available.
The author then focuses on technological aspects of the topic; pointing out that the development of AI started a few decades before. In 1980, several companies in the US launched government-supported AI research, but those projects were not as successful as the involved scientists hoped. The main obstacle was insufficient computer speeds. However, throughout the three intervening decades, computers reached speeds that cannot be compared to the levels available in the 1980s. Krishnan argues that according to scientists’ assumptions, robot on human on higher level of intelligence could be available no later than by 2030. Once armies buy (or will be offered) such technology, they will face serious dilemmas. In these terms, Krishnan focuses on the clash between traditionalists and revolutionaries in US Army headquarters. Accordingly, there is a significant ambivalence of army commanders to let a machine make decisions and act independently. However, Krishnan assumes that the voice of society is going to be stronger and will out-shift the resistance of traditionalists.
Krishnan then moves his thoughts towards a more philosophical field as he attempts to outline the possible need to regulate the use of AWs by international law and describes probable ethical impacts of AI proliferation across the whole society (this means also commercial and home-serving autonomous devices). This includes pressing social issues, such as robots taking jobs from people, or robot rights. When speaking in terms of law regulation of AWs, Krishnan justifies this need by serious threats that AWs represent. He expands these theories in the last chapter, where he poses scenarios of possible dangerous futures, including full-scale nano-wars and swarms of killer-machines that could turn against their creators. Krishnan warns that it is probable, that political leaders will not fully understand the impact and power of AWs, such as they did not in the fifties, when nuclear weapons were introduced.
The core argument of the book is represented by the uncertain attitude of the author himself. Krishnan talks about tremendous scientific progress, he describes the unbelievable effectiveness of future robots, but he warns of a future that will bring the world new dangers, dilemmas and questions, and that could possibly destroy life on Earth. According to Krishnan, the whole phenomena of AWs needs to be implemented into international law norms and the technology scene has to be supervised; otherwise machines could get out of control and perform in undesirable ways. Despite these facts, the author acknowledges that limited use of AWs could lead to more effective and less lethal warfare and in some constellations, it could bring worldwide peace.
Killer Robots by Krishnan is one of the first publications dedicated to the topic of AWs and therefore, almost naturally, presents some critical arguments and theories. The author uses a very wide range of sources, which vary from scientific and academic studies to military reports. The book does not lack historical evidence, as Krishnan spends a substantial amount of time explaining the roots of AW technologies. One of the most significant characteristics of the work is that the author understands the topic of AWs and AI as a phenomena that is, in the future, likely to change the whole society as we know it.