A “bloodless” war between machines is impossible – the technological nature of LAWS guarantees alternatives that target non-military objectives
Wilson 20 (Neil Wilson, Master’s at Utrecht University, “Understanding the Battle for AI in Warfare through the Practices of Assemblage: A Case Study of Project Maven”, Utrecht University, 8/2/20, https://studenttheses.uu.nl/bitstream/handle/20.500.12932/37392/Neil%20Wilson%20MA%20Thesis%20%281%29.pdf?sequence=1&isAllowed=y) // EL
2.1. Lethal Autonomous Weapons Systems and AI in Warfare Debate relating to Lethal Autonomous Weapons Systems (LAWS) is a mile wide but an inch deep - vast but as-yet unsubstantiated. Ample discussion surrounds the potential political, legal and moral repercussions of deploying LAWS but there is little theoretically-informed analysis of their development. Given the revolutionary potential of such technologies, there is an urgent need to rectify this imbalance. Political challenges posed by the deployment of LAWS on battlefields are numerous. Unlike nuclear weapons, LAWS would require no hard-to-obtain materials and be relatively cheap to develop, leading some to fear a global robot arms race. Even those contending that “arms race” framing risks escalating rivalries nevertheless agree on the need to manage the proliferation of these technologies. In particular, the risk of LAWS proliferating among non-state actors seeking to “level the playing field” is severe. Even among democratic states, LAWS would overcome the “bodybag” problem and thus lower the threshold for going to war. In such a scenario, however, the obvious attraction of using LAWS - a ‘bloodless’ war between machines - is misguided. A change of footsoldiers does not alter the fundamental objective of war: “imposing and enduring costs to reveal relative capability or resolve.” In the absence of human soldiers to impose costs upon, belligerents may seek alternative means to demonstrate their resolve, e.g. by targeting non-military objectives and civilians. Indeed, as drones have enabled the removal of soldiers from battlefields, the inability for enemies to directly reciprocate violence has instigated fears of “blowback” by other means. Political concerns aside, the legal questions raised by LAWS are momentous. Formal and informal rules have guided conduct in warfare for as long as it has been carried out. Over time, International Humanitarian Law (IHL) has steadily evolved to keep up with the development of new methods of warfighting, as evidenced by bans on the use of Chemical and Biological Weapons, anti-personnel landmines and blinding laser weapons. Throughout history, however, IHL has been developed under the assumption that human intelligence is guiding conduct in combat, not artificial. As existing treaties make no reference to LAWS, it is unclear if they can be banned under the current legal framework. Nevertheless, because it is impossible to confidently predict the outcomes of commands on complex AI systems, there is a possibility of LAWS running afoul of IHL. Can a robot be entrusted with the ability to discriminate between legitimate targets and civilians? If it fails, who can be held accountable? Some fear LAWS subverting the Hague Convention if they can act autonomously but not assume responsibility for the consequences. The ethics of using LAWS are fraught. At first glance, there is obvious appeal in using robots to do the “dull, dirty and dangerous” work in war as in civilian life. A robot never gets tired, stressed or traumatised. Their dispassionate nature, immune to the intoxicating effects of adrenaline, may even reduce the risk of civilian harm in conflict. From this perspective, if LAWS lessens the danger to soldiers and civilians alike, some suggest utilising them is not only judicious but an ethical imperative. Even if LAWs could perfectly adhere to IHL and reduce harm to both combatants and civilians, many question whether they should. As a simple matter of dignity, some insist that taking a human life should involve a human decision. “Delegating the decision to kill to algorithms,” according to Rosert and Sauer, “is inhumane and unacceptable under any circumstances.” A compelling argument, but potentially relativised by the fact that there are many different interpretations of human dignity and many ways it can be violated - technological or otherwise. The political, ethical and legal concerns raised are troubling but overwhelmingly speculative. Are we jumping the (autonomous) gun? Such preoccupation with the future use of killer robots may obscure meaningful debate about their present day development. A few notable contributions from Lewis, Taylor and Verbruggen are to be commended for grappling with the development of these systems, but this remains an under-theorised field. Presumably owing to its novelty, Project Maven has thus far evaded serious academic scrutiny. One notable exception comes from Suchman, however it nevertheless engages primarily with the deployment rather than development of this technology. Common to these conjectural, dystopian perspectives on the future of warfare is the assumption that if LAWS technology exists, it will be used. Not a dramatic leap of faith but nevertheless it seems prudent to question what is driving the demand for these technologies in this domain; thereby drawing the line between where we stand and where some fear we are heading. LAWS will only be deployed, presumably, if there is some strategic, operational and/or tactical need for them. Understanding this will elucidate a fuller picture of why and how they are being developed. To satisfy this curiosity, we turn to the growing field of scholarship on remote warfare.