Wars are defensive in nature, that is, responding to an attack, or preemptive – often called, ‘wars of choice’. Two considerations drive governments to undertake preemptive wars: power and wealth. Two things stop ‘wars of choice’ in their tracks: high numbers of military casualties and increasing taxes. The latter can be defrayed by borrowing from the bankers, who enthusiastically profit from chaos and destruction. The bill will come due in time, but it will be a future administration that has to deal with the huge public debt, and will almost certainly get blamed for it by those who incurred that debt. In the matter of casualties, armament manufacturers – the bankers of death, if you will – have the answer: robotic weapons. I mean, no one ever turned up at a funeral service for a blown up battle tank. Did they?
In war there are ‘good kills’, ‘collateral damage’ (civilian deaths) and ‘friendly fire’ incidents (soldiers killed by soldiers from the same side). Whether the soldier is humanoid or robotic, the terminology used to describe a killing event remains the same. So does the likelihood of such ‘collateral damage’ and friendly fire’ incidents continuing to happen, despite any claims the armament manufacturers might make to the contrary.
When the Government of Nevia first initiated a conflict in the tribal badlands of a barren, mountainous region of the planet, known as, Arghistan, it had relied almost entirely on humanoid ground forces, augmented with remote controlled surveillance and attack drones. This was the initial step on the slippery path of dehumanizing warfare. The remote operator has no idea at whom he is launching his missiles. At best, the image on his screen is indistinct, as is the ‘context’ in which the imaged person is moving. Assumptions have to be made: should the target be there if he/she is a non-combatant? Is that gathering a group of peaceful tribal elders, a wedding, a funeral, or a group of ‘insurgents’ setting out on an attack? If the decision is wrong, who’s to know? Certainly not the drone operator. The remotely controlled hunter-killer drones the Nevian military brought in theater during the early months of the conflict inflicted a tremendous amount of collateral damage, equipped as they were with missiles and ground strafing energy weapons. The drone operators knew neither the language nor the customs of the people they were targeting; nor did they really care, having grown up with computerized war games in which, like cartoons, no one really died.
At this point, it might be useful to introduce a brief portion of the widely publicized operational log for a robotic weapon that was involved in a ‘collateral damage’ incident. The unit had the designation, ‘Zee-con-two-niner-seven-tee’. Its task was to identify, engage and destroy Arghistani ‘insurgents’ who were members of a group called, ‘Followers of the Illuminated Path’, loyal to the Warlord, Dester Zoachian. Under the rules of engagement issued by the Secretary of War, robotic weapons were free to eliminate a target immediately upon clear identification – autonomously – but authorization was required if a potential target was ‘non clear’. What was never made public was this restriction applied to soldiers of the Nevian Armed Forces, only, in an attempt to limit ‘friendly fire’ incidents. Collateral damage, in the form of civilian deaths, were not covered.
‘Zee-con-two-niner-seven-tee’ was a primitive, tracked, weapons platform, fitted with a ten-kilowatt energy weapon and an active denial microwave system, for crowd dispersal, designed to produce an extreme heating effect on humanoid skin. A similar model is seen, here (right), engaged in tunnel clearing operations during the Swut Valley Counter-Insurgency Operation, code-named: ‘Tusa Karzir’ [Rabbit Warren].
‘Zee-con-two-niner-seven-tee’ located a group of peasant farmers driving a herd of herbivores. Peasants carry weapons to protect themselves and their animals from predators. ‘Zee-con-two-niner-seven-tee’ did not contact its operational controller for confirmation it should open fire – it was not required to do so. Fifty innocent civilians were killed. Eventually, the incident made its way into the public arena, in the middle of an election year. ‘Zee-con-two-niner-seven-tee’ was declared dysfunctional and destroyed.
The use of fully autonomous robotic weaponry is an even more dangerous departure in war, though certainly an attractive one in terms of casualty reduction. To answer ethical concerns, should they arise in the minds of a somnambulant population, such weapons can, like their non-autonomous counterparts, be given a degree of discrimination in terms of which type of individual would represent a valid target. Even if every effort is made to limit the potential for ‘friendly fire’ incidents, or collateral deaths resulting from misidentification – and that is far from being the case on most worlds that employ such technology – these weapons must still be imbued with a sense of ‘them and us’ to prevent them from destroying other robotic weapons on the same side. This is usually accomplished by some form of ‘identification challenge’, based on tonal or coded communications. During Earth’s WWII, for instance, allied soldiers were sometimes issued with a child’s toy (also used in dog training), which they ‘clicked’ as they approached their own lines. If the click was responded to, with a ‘click’, they would assume all was well and move forward. Robotic weapons do not rely on such a simple technique. Rolling code sequences and encryption algorithms were the method by which the Nevian robotic weapons recognized one another. Nevia soldiers were equipped with an electronic code device before being dispatched to the front lines, but these were easily lost or damaged. In the absence of an electronic response to an I.D. challenge, the robotic weapons were programmed to contact their controller and request clearance to fire if they could not clearly identify the target as a ‘bad guy’. This procedure applied to humanoid targets only. If a robotic weapon was unable to respond to a challenge, for some reason, its destruction would be no more than a debit entry on a government balance sheet and a credit entry on the supplier’s one.
Since the day the war began, there had been three hundred and sixty-two deaths attributed to ‘friendly fire’ incidents among the Nevia forces. Of these, two hundred and eighty-three had occurred after the robot weapons were deployed in combat. ‘Friendly fire’ deaths were internally reported as arising from ‘confusion’ when, for example, Nevian soldiers had been engaged with the enemy and had called in robot reinforcements to assist, or ‘misidentification’, where the soldier had not been dressed in combat gear and had been deemed a combatant because he was still carrying his weapon. All of these deaths were dealt with by the posthumous award of a medal for exceptional bravery and some cock and bull story conjured up by a clerk in the Adjutant General’s office. Such stories were not always endorsed by the fallen soldier’s comrades. Returning veterans brought with them tales of robot weapons running amok, slaughtering all and sundry, without regard or accountability. ‘Make sure they’re between you and the enemy’, was the advice handed out to newly arrived recruits.
Clearly, something had to be done about these ‘accidents’ or the military and the government risked being discredited in the eyes of the public, as more and more of these ‘cock and bull’ stories were proved to be just that. This is one of the more bizarre aspects of the relationship between governments, the citizenry and the wars they support, or are asked to support.
In their day to day activities, such as running the economy, politicians are typically regarded as elitist, generally useless and totally corrupt. In matters of war, however, a population tends to display an almost child-like faith in the integrity of its rulers, since their actions can be said to reflect that society’s own moral sense of itself. Using lethal force against those without the means to defend themselves is not part of any warrior culture, it’s an atrocity. Most races get it, and claim to be above such brutality. Until, of course, they find themselves bogged down in guerrilla warfare. Then, rather than running the risk of turning their consciences into pretzels, they eagerly accept without question the findings of opaque and biased government investigations, namely: that ‘rogue elements’ within the junior ranks of the military, identified or not, were responsible.
If the people the troops are killing have been successfully labeled ‘insurgents’ and ‘terrorists’ by their government – as opposed to ‘freedom fighters’, which is often nearer the truth – most citizens aren’t that bothered and the war will go on, as long as it’s not on prime time news when families are sitting down to dinner. But, let one of their own die, and all bets are off!
Three years into the war, the Nevian humanoid controllers were replaced with A.I. systems – the final step on the path to fully automated conflict. Instead of disgraced officers appearing on the news, the public would, in future, be treated to images of machines being crushed and shredded. And, why not? Every day, the calm and efficient unfolding of people’s lives is undermined by the failure of technology. Technology breaks. Nothing to see. Move along. It was hardly surprising that reports of atrocities dried up once the war was fully automated. The public soon forgot about the conflict, the armaments manufacturers continued to make out like bandits and the politicians got re-elected.
Months passed. The war was in its fourth year, by now. Victory was within reach, according to the news media. Then, one rain sodden day, several Nevian ‘humanoid casualties’ were reported. All had died from energy weapon hits while, apparently, secure in their bases. ‘Nothing more than a blip’, the internal report said. Soon, however, the number had reached forty a month, rising over the next three months to twenty or more a week. Panicked by the unproven fact that the ‘terrorists’ had somehow managed to breach the security cordons around their forward bases – as well as having acquired energy weapons – the military leadership sent an urgent request to their political masters for a budget increase for the purchase of more machines. The politicians, every one heavily invested in the armament manufacturers, rushed the necessary legislation through in one hour. The machine producers ramped up production to three thousand units a month.
The number of casualties continued to climb. At its peak, fifty soldiers were killed in a two day period. The government imposed a black-out on the arrival of transports laden with the bodies of the ‘fallen’ heroes. The military commanders flooded the battlefield with their newly delivered fighting machines. A few enemy stragglers were found and eliminated, but the bulk of the insurgents (terrorists, or freedom fighters) had disappeared.
An attack on two Nevian bases occurred a month later, resulting in the deaths of four hundred soldiers, with more than a thousand wounded. No enemy forces were spotted before the attacks, though the bodies of a few ‘terrorists’ were found close to the base perimeter, afterwards. More than twenty robot units were disabled or destroyed in the attacks. Their memory logs, together with those of their A.I. controllers, were recovered and analyzed. Every log recorded that a large number of ‘legitimate’ targets had been liquidated with a combination of energy weapon fire and micro-nuclear mortar shells. Curiously, though the total number of kills recorded by the A.I. controllers and the robotic weapons tallied, it fell well short of the actual total of dead and wounded Nevian soldiers.
One particular unit, designated ‘Tee-con-seven-three-fiver-six-em’, had been involved in the deaths of twenty ‘targets’ before succumbing to a peer-to-peer communications malfunction which had forced its shutdown. This unit had recently been brought in-theater and, like the units that had perished, was running the latest version of the ‘Target Identification Protocol’. As confirmed by its log, it had identified twenty potential targets in total, before shutting itself down. Two were local Arghistanis, working on the base as contractors. ‘Tee-con-seven-three-fiver-six-em’ had shot them. The other eighteen targets should have been designated as ‘friendly’, since they were, in fact, Nevian soldiers headed for the dining hall for a lunchtime meal. The Nevian soldiers were likely startled by the attack and brought their weapons to bear on ‘Tee-con-seven-three-fiver-six-em’, assuming, one imagines, that they were under attack from ‘insurgents’. ‘Tee-con-seven-three-fiver-six-em’ killed them with a single ‘pulse’ grenade. In all probability, the blast from the grenade was responsible for the unit’s communication failure. When its A.I. controller was examined, its log was found to have recorded every dead soldier as a ‘legitimate’ target, as it had the two contractors. However, the time-stamp of the data recorded in the controller’s log was a full minute later than that of its assigned robot’s last transmission, and listed another A.I. controller as the source. By comparing the time-stamps of all the logs, the investigators were able to determine that the attack had been initiated when ‘Tee-con-seven-three-fiver-six-em’ killed the two contractors. The other robotic units had joined in once ‘Tee-con-seven-three-fiver-six-em’ came under fire from the soldiers. Tragic as this event was, there was a further shock for the investigators.
The logs in both the weapons and their controllers were held in a ‘molecular’ memory, which recorded data by modifying the energy state of individual molecules. When new, these energy states were at a level consistent with the degree of excitation normally present in the constituent material itself. When data was recorded, individual molecular energy levels were raised as each byte of data was stored. If, say, forty bytes of data had been recorded, the excitation level of forty molecules would be elevated, plus six more for the check sum. Data could be erased, if needs be, however a molecule which no longer held a byte of information did not revert to the excitation level at which it existed when the memory was first manufactured, it remained slightly elevated. This design feature was included in an attempt to win over the rank and file members of the military, who’d grown weary of the endless lies and cover-ups surrounding ‘friendly fire’ incidents. Any attempt to falsify the data content, they were told, could be easily detected because of this important modification. What they were not told was that, since certain individuals in the company that manufactured the memory had easy access to the technology which had been used to create it in the first place, the falsification of data remained an ‘official’ option. Forensic examination of the memory in the A.I. controller assigned to ‘Tee-con-seven-three-fiver-six-em’ revealed that the log had, in fact, been altered, some time after ‘Tee-con-seven-three-fiver-six-em’ had shut itself down. One minute after! A slightly elevated energy level was detected in a total of sixty molecules, indicating the new log entry was sixty bytes shorter than the original. In the Nevian language, tura means ‘legitimate’. The word for illegitimate is, nihtura. The conclusion was inescapable! A shaken Chief of Staff informed an equally shaken President that, not only had the sudden escalation in military casualties been the result of ‘friendly fire’ incidents alone, but that the robots, together with their A.I. controllers, appeared to be lying about their involvement in them!
It took months to determine the cause. It was traced to five thousand lines of code embedded in one of the sub-routines loaded into the robots during their initial activation. One might call it the mea culpa routine, since it instructed the robot to shut itself down in the event of an ‘illegitimate’ kill. The newly delivered robots had been hurriedly reprogrammed with a far more powerful survival imperative than previously, as well as continuing to be intensely ‘goal’ driven. The presence of this sub-routine now had the potential to cause a serious logic conflict. Without human controller intervention – as these had been replaced by A.I. units – and since every robot was now in constant communication with its peers, it was clear that they had, over time, and as a group, come to the conclusion that telling a lie was preferable to shutting themselves down. The A.I. controllers, programmed with the desire to protect their charges from unnecessary harm, had concurred!
The war, which was going so well, is now simmering on the back burner. The robot weapons have been stood down, the humanoid soldiers are languishing in their bases. The government is hoping the public won’t notice, while the armament manufacturers and several university research laboratories suck up even more of the nation’s taxes. For their part, the ‘Followers of the Illuminated Path’ have used this hiatus to regroup, re-arm, and are waiting for the spring thaw.