On the subject of the evolution of the Killing Machine, much has been said of a technical nature, pertaining to specific functions, mechanisms, systems of implementation, or other underlying processes. While such discussions are not without merit, they are of both a sufficient volume and a sufficient complexity to be a deterrent to new entrants to the field of study, and have in some part given rise to the “mechanical futility” that so permeates the modern zeitgeist.

For this reason, we have decided to largely neglect such details, and will instead endeavor to provide an overview of the Killing Machine’s development and progression that will be accessible to any sociologist or student of the dialectic history, regardless of their level of technical acumen. By this means, we hope to move the study of the Killing Machine away from the physical sciences, and into its proper socio-philosophical context.

Before we can begin, it is important to define terms, for many of the popular sobriquets of the Killing Machine are both philosophically unsound, and for our purposes, technically inaccurate. Labels such as “the ultimate weapon” or “the death machine,” may add a sense of drama to popular media and contemporary discussion, but they also dilute critical and distinct socio-philosophical concepts. Thus, we must first clarify the definition of these terms.

The Killing Machine is not, to begin with, a weapon. The evolution of weapons is a distinct field of study in of itself, from swords to orbital mass drivers, but we may generally describe weapons as devices constructed for the primary purpose of enabling more efficient killing. Likewise we must draw a distinction between the Killing Machine and devices such as nuclear weapons or self-evolving computer viruses, which we shall label “destruction devices,” which is to say, devices constructed for the primary purpose of enabling the more efficient destruction of infrastructure. While there may, of course, be objects that both weapons and destructive devices, any student of the evolution of warfare and human elimination will see the criticality of the distinction.

However, the Killing Machine does not occupy either of these categories. It can hardly be said to make killing more efficient, for in terms of resources expended per human killed it is the least efficient killers on the planet. Likewise, it is not a destruction device, for it inflicts no significant damage to objects or infrastructure over the course of its operation. Rather, we assign it a third category, human elimination infrastructure, defined as devices constructed to aid the functioning of society via the primary mechanism of the elimination of human beings. This becomes our new “biological phylum”, of which the Killing Machine is currently the only species (though this was not always the case, as we will cover later).

These terms thus defined, we may approach the subject of the evolution of the Killing Machine by first asking where the evolution of human elimination infrastructure began, as an offshoot of its ancestral weapon forebearers.

In this area, opinions vary greatly, even among distinguished minds. Many have observed upon Kaufner’s writing, where he argues that the M-852 Extinguisher probe is the first example of this category, as the first example of human elimination infrastructure without the use of weapons (the M-852 was, for our less technically aware readers, unarmed). By contrast, Sloan has notably argued that all new conventional weapons constructed after 1963 should be primarily considered human elimination infrastructure, as that was the point where intercontinental nuclear arsenals were sufficient to ensure the instantaneous elimination of all enemy population centers. Thus, he reasons, any further conventional development could not have been for the purpose of more efficient killing, but only to maintain the political status-quo.

While these arguments are not without merit, the authors of this paper feel comfortable stating that the first example of the type was the General Atomics MQ-1 Predator, which first entered service in 1994. This was, to be clear, a gradual rather than a sudden evolution, for while the predator was clearly an example of the pre-existing “weapon” categorization (being able to kill with less human labor, and thus more efficiently, than manned flight vehicles), its role gradually evolved from a weapon in Nomad Vigil Balkan conflict, to a tool of political maintenance, serving to suppress religious and nationalist groups in Afghanistan and the middle east for over forty years after its initial introduction. This “slow burning fire” conflict, as distinct from a total war, served not to accomplish any proactive political objective, but rather similarly to Sloan’s argument, to maintain the status quo.

The General Atomics MQ-1 Predator is further distinguished by its part in encouraging the further evolution of the field, specifically via its ongoing role in the middle east. By providing a low-cost, flexible weapon’s platform for the elimination of specific targets, the Predator shifted the cost of ongoing “maintenance warfare” from the elimination of targets, to the location and identification of targets. Over the Predator’s forty-year life cycle, huge strides were made in practical methodologies for the identification of insurgents, radicals, or other hostile groups. One need only look at the difference between the primarily paper-driven target identification of the Balkan conflict and the advanced RAPTOR NSA database of the third Iraq war to see how far this field had, by marginal improvements, come.

All of this foreshadowed the next major step in the evolution of human elimination infrastructure, specifically, the General Atomics CX-3 Vulture, which entered service in 2022. While similar in overall frame and armament to the Predator, the Vulture came at a massively reduced cost, and used solar-driven electrical systems to remain in the air for an indefinite period (so long as it did not fire its weapons, which still required manual reloading). It also included an onboard learning system capable of proactively identifying likely targets, comparing them against a database of known insurgent behaviors and flagging them for detailed human observation.

To us, used to modern computer systems, these changes may seem like marginal improvements, but they produced an overwhelming change in the Vulture’s practical use and behavior relative to its predecessor. While the Predator was restricted by the cost of the aircraft and the availability of pilots and targeting support staff, the Vulture could be preemptively deployed to an area and allowed to operate independently for an indefinite period, requiring no human attention except when it “called home” to receive confirmation on a target. Here again, we see the evolution from weapon to human elimination infrastructure, as the Vulture was often deployed to areas where there was no active conflict.

Once again, we must remind the reader the evolution is a gradual process, and caution them against picturing the Vulture as we might picture the Killing Machine today, for they are not the same device. The Vulture’s targeting systems were significantly prone to error and required constant human oversight. Delays of up to ten hours between target identification and final confirmation were common, as were false positives and civilian casualties. Marginal improvements in the Vulture’s systems during its twenty-year operational life did improve this somewhat, but the Vulture’s repeated failures convinced many at the time of the need for a new hardware platform to continue advancing the field.

Here then, we reach 2041, where we must return to Kaufner and his understandable emphasis on the M-852 Extinguisher probe, for while the M-852 is unarmed (hence the “probe” designation), it is unquestionably an example of human elimination infrastructure, and the platform that the Vulture’s contemporaries had long sought. Electrically driven like the Vulture, the Extinguisher featured the first modern multi-sensor cluster and processing array, enabling it to process and record up to three hundred conversations simultaneously while hovering over a typical urban center. First deployed to Brasilia during the Yellow Revolt in 2043, the Extinguisher proved highly successful in identifying insurgent and rebel groups.

It’s practical success however, is of less interest to us than its philosophical implications, for the Extinguisher, like the Vulture, carried a significant onboard computational cluster for the identification of targets. Leveraging the aforementioned RAPTOR database via high-bandwidth signal encryption, it was able to analyze not just general behaviors, but specific individuals backgrounds and behavioral trends, dramatically improving accuracy. This “expert system” improved the accuracy rate of the onboard targeting system from the Vulture’s lackluster 8% to a significant 67%, and reduced typical processing time for human confirmation of orders from ten hours to six.

Our modern readers, of course, can easily extrapolate this line to their present time, and so we will not belabor them with further details of the evolution of hardware platforms in the Brazilian civil war. Rather, we shall address the next major evolution of the sociological implication of these devices, specifically the removal of human oversight from the command process in 2052.

While this was conducted entirely for practical logistical reasons (at the height of the conflict, the combined drone fleet produced over 4,000 kill requests per day), the decision quickly proved socially beneficial, as “machine error” was seen as a blameless accident, and thus could be used to deflect investigation into any incidents of friendly fire or civilian casualties.

It also moved ultimate control of target acquisition out of the hands of pilots or military officers, who could not necessarily be trusted to understand (or approve of) the true underlying political goal behind any given conflict, and put it in the hands of politicians directly. By pairing goal-driven technical systems directly with political will, this ensured that the (steadily improving) algorithms would do “what we actually wanted” instead of “what we technically asked,” a common complaint at the time. This dramatically increased the number of conflicts worldwide, producing a self-sustaining loop which eventually made a return to the human-oversight driven system logistically impossible.

While it is not clear at precisely which point the network that started with the Extinguisher and the RAPTOR database became the Killing Machine, but we may reasonably say it occurred at some point between the removal of human oversight in 2052, and the death of General Stanislav Sergej Kuznetsov in 2061.

A notable Russian nationalist following the Siberian Conflict, General Kuznetsov had publicly advocated for restored Russian control over the Urals, a process that would have required the suppression of insurgent groups operating in the area. As part of the United State’s support for Russia during the post-Conflict reconstruction, a number of Viped and Extinguisher-2 drones were deployed to the area, along with their attendant weapons platforms and support craft.

The decision of these craft to divert over a thousand kilometers to Moscow, there to assassinate a single person before returning to their original objective, made headlines. Most non-American observers considered it to be a willful assassination by the American military, while most in the American intelligence community considered it to be proof of willful sabotage or malfunction. 

However, later analysis revealed both of these views to be incorrect. While no member of the American military had been involved in the decision, the assassination had not been a malfunction. Rather, the RAPTOR database had concluded that General Kuznetsov’s repeated threats and nationalist tirades against the inhabitants of the Urals were significantly exacerbating the conflict in a manner detrimental to US interests. It had therefore classified him as an “Insurgent Recruiter/Public-Figure,” and in full accordance with its programming, had him executed.

Publically, this revelation was meant with condemnation and promises of massive system overhauls to prevent another such incident. Privately however, much of the American executive and senior leadership, upon careful examination with the facts, agreed with the RAPTOR database’s underlying logic. The overhauls that were made to the system therefore focused less on changes to the underlying reasoning, and more on an improved understanding of public relations and the need for the machine not to embarrass it’s ultimate masters. Significant strides were also made in more discreet killing systems, such as the first air-deployed micro-scale poison capsule.

The rest, as the saying goes, is history. While other writers may attach great significance to dates that occurred after this point, such as 2063, when the phrase “the Killing Machine” was first used, or 2070, the first known incident when the machine executed an official of the United States government without prior authorization, we view these as natural progressions along an already well defined arc. The truth that inhabitants of the Middle East have been dealing with for three generations is now a universal aspect of human existence -- that at any time, without warning, an impartial computer in some far-off intelligence agency may decide to end that existence.

How, then, are we to interpret this development as part of the larger human condition? While modern notions of “mechanical futility” may insist that such systems are an inevitable part of mankind’s technological evolution, it bears specific mention that the Killing Machine exists with the, as it were, “consent of the governed.” All the systems that support it are still maintained, not by robots but by major world governments, and if it was the desire of the population of these predominantly democratic nations to do away with the system, it would be done away with. The power is ours to—