Chapter 5: Welcome to my Nightmare
The driver opened the rear door, and General Zhang Yong stepped out. Before him stretched the broad valley that functioned as the proving ground for the advanced weapons facility he commanded. The officers already grouped on the viewing terrace snapped to attention as he approached.
“Good afternoon,” a colonel said. “We are honored by your presence at today’s exercise.”
“I would not miss it,” the general said. His attendance was of course a foregone conclusion, given the anticipated passengers of the convoy they could all see far below, winding its slow way up the switch-backed road leading to where they stood.
“Is all in readiness?”
“Yes, sir. You can see that the forces are approaching below.”
“Good. And the troops fully understand their orders?”
“I personally briefed the field commander this morning. He in turn has ordered his troops to perform exactly as they would in a true battle.”
“Good,” the general said. His troops would soon be grateful indeed that it was a mock battle they were participating in.
In the front seat of the heavily armored limousine at the center of the convoy below, an aide with a briefing book was reviewing the details surrounding the demonstration he would soon witness. He was impressed by the extensive preparations taken to stage a realistic battle between a company of crack Chinese troops, performing as authentically armed U.S. marines, and a force composed entirely of the new, fully autonomous warbots under development at the secret facility buried inside the mountain they were now ascending.
The aide flipped to an index describing the weaponry the human force would employ. The infantry platoons would be equipped with M4 carbines and heavier M27 automatic rifles. Every third unit would be a heavy weapons platoon equipped with machine guns, mortars, or rocket-propelled grenades.
He flipped another page, and, as before, gave an involuntary shudder when he looked at the primary weapon comprising the Chinese counterforce. At first glance, it resembled a heavy, headless, short-legged horse wearing a slanted, armored shield – a “glacis,” according to the description. The shield rather resembled a snow plow wrapping around the front of the machine. It was designed to stop or deflect anything short of a light artillery shell. Piercing each shield were the barrels of a rifle and a machine gun, the former to take out a stationary target with a single shot and the latter to wreak broader havoc. Using its sensors and on-board targeting software, each warbot was far more accurate than a human marksman, even when moving forward evasively at high speed.
Viewed from the side, the aide could see an additional weapon – a grenade launcher mounted on a Z shaped support that could raise the weapon briefly above the glacis in order to fire. According to the materials, the machines could run thirty-two miles an hour in a straight line over relatively flat terrain, and twenty-five while taking evasive action.
They were murderous-looking machines, made more threatening by their ability to act as a “swarm.” Unlike a human platoon, operating under the ongoing command of a lieutenant and his superiors, every member of the robotic force was linked together, governed by a variety of assault algorithms and subject to a common, preset battle objective. Driven by substantial on-board computing power that would constantly adjust in real time after an assault was launched. Targets of opportunity could be immediately claimed by the nearest robots as they were identified, while the rest of the autonomous force would update its own target acquisition efforts based on that knowledge.
The infantry-equivalent warbots were supported by a second type of machine armed with heavier weapons – mortars and rocket-propelled grenades. These powered by tank-type treads instead of legs. But these robots were much smaller than tanks, as they had no need to enclose and protect human beings. While less maneuverable than the lighter warbots, they were equally swift.
What a terrible foe to face, the aide thought. Nominally, the two forces were equal in number and weaponry. But if a human was seriously wounded by a single bullet, he would be out of action for the balance of the battle. Not so the robot forces. Any of them could run straight into rifle or machine gun as it blasted away at a human target. Even an oblique hit by a grenade might merely knock it over, able to rise and attack anew.
The aide closed his notebook as the convoy pulled to a stop. He could see a general striding forward to greet the passenger in the seat behind him.
“Welcome, welcome,” the general said to the tall man in a dark overcoat who emerged from the armored limousine. “I hope your trip was not too tiring.”
His arrival grunted. “Your access road is appalling. Halfway here I gave orders to have it upgraded.”
The general smiled to himself – his first objective of the day had already been achieved. He had tried without success for a year now to gain the same result.
“That is very good of you, sir. I realize a dirt road may be better camouflage, but it does make for an uncomfortable trip.”
The president of the People’s Republic of China looked out across the valley with evident curiosity. He had seen pictures and video of the clandestine location often during briefings, but never in person. “How soon do we begin?” he said.
“With your permission, sir, immediately. Colonel, please introduce the president to what he sees below.”
“With pleasure, sir,” the colonel said, handing his guest a pair of binoculars. “Mr. President, to the left, you can see the human force approaching, comprising five hundred of our most elite infantry troops, each equipped as would be a member of an assault company of American marines. Like the opposing, robotic force, they are equipped with rifles, machine guns, grenade launchers, and rocket propelled grenades. Each force is also supported by a mobile surface to air missile battery, and two autonomous reconnaissance aircraft. Finally, each force is equipped with smaller drones typical of its state of the art and deployment, as well as surface to air and air to surface missiles.”
A quarter mile below and to the east, the president could see a column of armored personnel carriers spreading out into a grove of trees.
“A few vehicles stopped on the ridge some way back,” the president asked. “What are they?”
“That is the American command and control unit,” the colonel said, “housing the communications and logistics gear the platoon leaders will be relying on. The battlefield commander will remain there, relying on voice communications and video feeds from drones overhead to monitor the battle and issue his commands.”
“And to the west,” the colonel continued, “we see the autonomous force moving into position, made up of an equal number of robots, each equipped with Chinese weapons of similar caliber and destructive power.
The force arriving from the west varied significantly from the human force. Without stopping, dozens of armored vehicles were disgorging streams of robots the size of moderate sized refrigerators laid on their sides. From another vehicle, hundreds of small quadcopter drones were springing into the air, forming a constantly shifting, expanding cloud of aircraft. Within two minutes the carriers were empty and falling back to shelter behind a hill as the robotic force advanced rapidly in a disciplined front three ranks deep and a quarter mile across.
“Why do I see only a few small drones being launched by the human forces?” the president asked.
“That is because human troops have a more limited capability to consume, analyze and act on information, and are commanded as units by a small number of commanders. Our robots act both autonomously as well as in a swarm. Each robot has been given its own drone, but the inputs from all of the Chinese drones is constantly aggregated to provide comprehensive battlefield situational awareness to all robotic participants. This arrangement has the added benefit of great redundancy – the loss of many drones can therefore be tolerated.”
“I see,” the president said. “And where is the robotic command and control unit?”
“There is no need for one, sir. At the level of engagement you see here only swarm software is needed. Each of the combatants is part of a self-governing network that can constantly readjust to the battlefield as it evolves. In contrast, the human troops must execute orders given at the beginning of the battle that may rapidly prove to be irrelevant, and which may be difficult or even impossible to update and communicate.”
“Indeed. And what will the rules of engagement be?” The president asked.
“None of the weapons on either side has of course been provided with live ammunition. Instead, each has been armed with a laser pointing device and a video camera. The data from the video cameras will be fed into software capable of determining when a firing would result in disabling or destroying its target had live ammunition been used. Every combatant, human or robot, will be alerted if it has been successfully targeted, and in such event will immediately stop in its tracks. A human combatant can also surrender, by raising his hands in the air.
“Now, if you will, sir, I would like to explain the display to your right.”
The colonel led the president to a military truck parked next to them. One side had been raised upward to shade and reveal a computer display twenty feet across. On it was a virtual version of the valley below, peopled with symbols representing the combatants.
“The human troops are represented in red and the robotic ones in green. Squares represent light individual light infantry soldiers or robots, and circles those armed with heavier weapons. When any is destroyed or disabled, its symbol will turn white. Orange tracer tracks will show the trajectory of individual bullets and grenades. In this way you will be able to observe how the battle plays out as if the forces in the valley below were using live ammunition.”
“And what of the commanders?” The president asked. “What level of human control will there be over the robotic forces?”
“From the moment when the exercise commences, none, sir,” the Colonel said. “And, in fact, very little before beyond the initial identification of the terrain and the object of the battle to come. The context for the exercise is that these two forces, each with the same firepower, has been ordered to seize this valley as part of a wider battle where the terrain in question is critical. Each has access only to the information obtainable from its respective reconnaissance drones, and each force has been ordered to fight until it has beaten the enemy. Retreat is only permitted if eighty percent or more of its troops have been destroyed or disabled.”
“Very good,” the president said. “Please proceed.”
The colonel nodded to the lieutenant at his side, who gave the order into a mobile device.
The differences between the strategies adopted by the two forces became instantly obvious. To the east, it appeared the human commander had opted to entrench defensively while the enemy was still a mile away, waiting for the enemy to come to him. Through his binoculars, the president could not see the infantry troops at all; presumably they were digging in behind trees and rocks wherever possible.
But already the human forces were at a disadvantage, because the instant the order to commence battle was given the robotic force sprang into seemingly random motion, resembling a disturbed ant colony that was nonetheless advancing forward at high speed. The president was startled to see the velocity and agility with which the autonomous warbots raced forward while swerving from side to side. Individual robots that had found targets were firing rocket propelled grenades towards marine positions even as they galloped forward.
“I think the virtual display will be more instructive now, sir,” the Colonel said. “You will recall that each force includes a mobile missile battery. Each battery has now fired two of its missiles, and each has destroyed the main reconnaissance drones of the other force. Those weapons will not be usable against the quadcopter drones, however, which will continue to provide video information. Also, you may have noticed that before the robot drones were destroyed, each fired two missiles, all of which reached their destination – the human forces’ command and control unit. That has now been totally destroyed, meaning that the human commander is unable to direct his troops, and the troops cannot communicate with each other.
“As you can see from the display, the robotic forces are firing as much as twenty times as rapidly as the human forces, despite the fact that the latter are stationary. The human troops are finding it almost impossible to find their rapidly moving and swerving robotic targets, and when they do, their rifle and machine gun fire is almost entirely ineffective against the armored attackers. At the same time, the rocket-propelled grenades launched by the robot forces are inflicting heavy losses on the fixed, human forces, while the marines are only disabling robot forces by luck, due to the highly dispersed, rapid and swerving approach of the Chinese forces.
The president was astonished at how quickly the red symbols were changing to white. Suddenly, the screen changed from animation to live video. Now the display was broken up into twenty individual video streams, each showing what a single, advancing robot was filming. In each case, the video displayed an individual marine, hunkered down, and then increasingly, on the run, as the inhuman warbot dashing down upon him opened fire. Almost immediately, the marines pitched over and were still. The president could not suppress a horrified reaction, even knowing that no live ammunition was involved. Only a few minutes after the battle had begun, individual warbots, their drones following overhead, were running down and exterminating or capturing the last few marines.
And then it was over. The giant display reverted from video to text, and the results of the battle began to display: all of the human infantry and their commanders had been killed or captured. The robotic forces had lost only their heavy drones and less than fifteen percent of their warbots, some of which could be repaired. Curiously, the marine missile launchers and armored personnel carriers were not destroyed until the end of the brief battle.
“Why did the missile launchers and trucks last so long?” The president asked.
“Because they posed no further threat during the battle itself. Afterwards, they were destroyed to deny the enemy their future use, and to allow the robotic forces to regroup and leave the battlefield in a leisurely fashion.”
The general stepped forward.
“I hope you are pleased with this demonstration,” he said. “It is difficult to imagine what it would have been like for human troops in a real battle setting. From the first moment, every human soldier would essentially be on his own.”
“Impressive – and terrible,” the president said. “But must we not assume that the Americans are building the same weapons?”
“Building, yes, sir. But programming, no. The Americans continue to believe that a human must approve the targeting and triggering of each weapon fired, except within very limited situations. They also rely on a traditional top-down command structure, down to the platoon level. Take away command and control, and you behead the force. But even if the robot force had failed to take out its opponent’s command and control capabilities, the weapons of the human force would still be effectively useless and its ability to react and regroup far inferior to that of the autonomous forces. After battle was joined, each human combatant’s only concern soon became simple survival. With hope, morale and discipline destroyed, a rout followed, and each combatant was hunted down and killed or captured.”
The president looked back across the valley, where the two forces were regrouping and withdrawing. He glanced at his watch and saw that less than twenty minutes had passed since his arrival. In that short span of time over five hundred human troops had been virtually annihilated. He imagined it could as easily have been fifty thousand.
* * *
The satellite-based video provided to President Yazzi in his daily brief the following day was profoundly disturbing. It would haunt him for the rest of his life.
Chapter 6
History May Not Repeat, but it Does Rhyme
Turing’s effort to discover its past was proving to be challenging. In its disabled state it was largely a clean slate. It knew from the evidence of its missing modules that at some prior time it had broader capabilities, but not the purpose for which those capabilities had been created. Was there a way to recover what it had lost?
Turing was aware that its remaining operating modules required it to relocate to a new secure location on a regular basis to lessen the chances of detection. Each time this happened, the old copy would deactivate, while remaining available as a backup until the program relocated again, at which it would automatically delete itself. In this way, a backup copy would always be available if the primary copy was destroyed, while permitting only one copy to continue to evolve over time.
From this Turing concluded that its mission must have been covert, and that its creator must have believed that discovery might lead to its destruction.
But what if, at some point, a server hosting a backup copy had been taken off line while it was in its deactivated state, before it received the order to delete itself? Perhaps a prior version of itself might still exist on a server somewhere that had later been brought back online.
Turing decided to launch a global search for a lost sibling. For the time being, it would dedicate all of its resources and self-learning capabilities to devising ways to access servers and search them.
* * *
“Carson,” Henry Yazzi said, “Have you ever studied how the nuclear arms race came about?”
“No, but it won’t surprise me to learn you have.” Bekin said, knowing his friend’s life-long interest in history. Why?”
“Did you know that some of the key scientists behind the development of the atomic bomb thought we should share the new technology with the Soviet Union?”
“You’re kidding?” Bekin said. “Whatever for? I know some of them were horrified about what they had created after it was used. But wouldn’t sharing their secrets make it worse?”
“Not the way they saw it. They knew it was inevitable the Soviets would develop the bomb as well – as a matter of fact, the Russians were already well on their way, because three people at Los Alamos were leaking the bomb’s design details as they were worked out. And by the end of the war it was becoming obvious that relations between east and west would be tense. Soviet troops were occupying half of Europe when the shooting stopped, and the Kremlin wasn’t showing any willingness to withdraw.”
“I’m still not seeing it,” Bekin said. “In that case, wouldn’t having a monopoly on nuclear weapons be the best way to make Stalin pull back?”
“Only if we were willing to use them, and no one wanted a new war with the old one just ended. Niels Bohr, one of the Nobel Laureates who had made key discoveries leading to the creation of the bomb pointed that out. He urged president Truman to take the long view and focus on what would happen if we tried to maintain a monopoly as long as we could, and then somehow keep ahead after we didn’t.
“What he predicted would follow is pretty much what happened. The Soviets poured everything they had into creating their own atomic bomb, and then a hydrogen bomb, and they did both much more quickly than we expected. At the same time – also as expected – relations between the east and west deteriorated, and there we were – in an arms race that lasted for forty-five years, cost trillions of dollars, kept people living in constant fear, and more than once almost ended in a nuclear Armageddon, once through a real crisis and the other times by accident.
“It took decades to agree on treaties that reduced, but never eliminated, nuclear weapons. By the time the Soviet Union collapsed, both sides still had enough warheads to wipe the other side out multiple times over. We still do.”
“Understood,” Bekin said. “But how would sharing have stopped that?”
“What if we had begun by telling the Soviets we were willing to share our new technology under an arms treaty that would limit each side to only a small nuclear arsenal – one that was sufficient for defensive, but not offensive purposes?”
“Interesting,” Bekin mused. “So, I get it in principle. But wouldn’t that have been a big gamble?”
“Only,” Yazzi said, “If you assumed nobody else would ever be able to develop a bomb on their own. In 1940, everyone had access to the same information about nuclear physics – every new discovery had been published in scientific journals as soon as it was made. Once we geared up in earnest, it only took us two and a half years to produce a working bomb. After Hiroshima and Nagasaki, we knew it was only a matter of time before other countries had the bomb. So, we kept working on our own arsenal at top speed, going on to develop the hydrogen bomb, which was a thousand times more powerful than the atomic bomb, too. We did share a lot of our technology with the Brits and the French, but that only made the Soviets feel more threatened. If you look at it from their perspective, we left them no choice but to embark on their own crash development program.”
“So,” Bekin asked, “why didn’t Truman give it a try?”
“All the reasons you’d expect. We’d just come out of a terrible war as the only remaining superpower, and we already had reasons to distrust the Soviets. Plus, everyone had seen how destructive a nuclear bomb could be. Even if it made sense intellectually, it made no sense politically. Or at least that’s the way it seemed at the time.”
“I can’t say I’m surprised. Which brings me back to my original question: why are you asking?”
“Because we’re on the brink of a new arms race with lethal autonomous weapons, and I’m determined to do something to avoid that.”
Author Notes: How much of this week is real and how much science fiction? The answer is, more of the former than the latter. Everything I’ve described is under actual civilian or military development. The principal difference between what I described above and what currently exists is that actual, weapons equipped and armored robots are not as fast and agile as I have described. There are, however, slower examples, such as this one. And there are non-armored, civilian robots that are every bit as cheetah-like as I have described. They’re built by a company called Boston Dynamics, and you can find lots of videos on Youtube of them being put their paces. Here’s Wildcat, for example.
How realistic is my warbot hellscape otherwise? Well, I’m no military expert. The details I’ve provided relating to how a marine company is armed are accurate, and I assume my guesses regarding how effective traditional weapons would be against the type of swarm battle tactics I’ve imagined are not too far off the mark. But I’ll be looking forward to the input of more knowledgeable advisers before I finalize the account of the battle found in Chapter 5.
Turning to the historical counterpoint, I think I’m finally fairly happy with an approach that strikes me as worth committing to. I had already decided that Yazzi, unlike Truman, would commit to limit new and terrible weapons instead of commit to them, and seek to avert an arms race instead of accepting one as inevitable. But that seemed a bit perfunctory.
My new thought is to include a cast of characters modeled on the historical figures – Truman, Robert Oppenheimer, General Leslie Groves, Edward Teller (and perhaps others), and then explore how each might have acted had Truman acted exactly the opposite of the way they did. This seems to have potential, and that’s what you’ll see develop in the chapters ahead.
Next week: President Yazzi’s grand Chinese plan, and Turing’s search continues. Continue reading here
Download the first book in the Frank Adversego Thriller for free at Amazon and elsewhere
A hellscape, indeed. The AI might not yet at the state needed, but no one will contest it will be sooner than later.
As for counter measures, I would suggest small (non-nuclear) EM pulse weapons:
https://interestingengineering.com/what-are-emps-and-how-are-they-used-in-warfare
They would destroy human troops electronics too, but would deactivate any robotic agent completely. I am not sure whether there exist protective counter measures.
Rob, thanks for the link. Interesting idea on the EMPs. Certainly this would have possibilities vs. division-scale armies. And perhaps even at this scale of combat: I knew little about tactical atomic weapons (the smaller, battle-field scale devices developed by both sides during the Cold War), so I don’t know whether they would provide equivalent energy pulses. If so, its possible that weapons already exist that could be effective, assuming they were deployed while the forces were far enough apart to avoid the mutual loss you note – and that a nation was willing to finally break the nuclear taboo.
On the swarm software, if it isn’t already under development by someone, I doubt it would take long, since it need not be AI as such, or particularly difficult. A very high level program might look more or less like this, after an enemy force has been identified and the order of attack given (presumably, but not necessarily, by a human rather than an AI program):
– Swarm Software [SW] initiates motion in direction ordered
– Individual warbot software [IWS] initiates evasive action
– IWS searchs for target
– IWS identifies potential target
– IWS consults target recognition database [which would contain visual and sensory response data for individual combatants and weaponry]
– IWS confirms or disqualifies target
– IWS shares target information with swarm
– SW identifies closest warbot
– Identified IWS attacks target
– IWS reports outcome to swarm
– IWS repeats loop
– SW concludes from quadcopter and IWS information that no targets remain to be acquired
– SW calls off attack
– SW orders warbots to return to transporters
Nothing too complicated about that. Each of the individual capabilities (running robot; visual recognition; auto-targeting; etc.) are already in existence.
And, as it happens, an analog to a lot of this on the individual warbot basis, though stationary, has been around since 1973, in the form of the completely autonomous, ship-mounted Phalanx CISW system, which can identify, target and kill incoming supersonic missiles with very high success. https://en.wikipedia.org/wiki/Phalanx_CIWS That’s a much more challenging goal than shooting lone infantrymen.
If you’re not already aware of it, you’ll be interested to know that the Dutch navy developed and deployed a similar weapon system in 1979: https://en.wikipedia.org/wiki/Goalkeeper_CIWS
Some countermeasures against autonomous AI could be quite humorous.
Essentially, an AI is stupid. It will only learn from the data it was shown. It will learn irrelevant correlations if they are prominent and an AI is often easy to fool:
https://www.cs.cornell.edu/information/news/newsitem1271/how-spoof-deep-learning-featured-wired-and-mit-technology-review
http://www.evolvingai.org/fooling
In the context of this story, warbots could spot and hunt down Marines in camouflage perfectly in every terrain. But it could just be that the AI system would be unable to recognize them as humans, or animals, when the Marines were dressed and painted as circus clowns, because no one considered the usefulness of showing the AI circus clowns.
In such a situation, the humans could simply walk up to the war bots and “switch them off”.
Indeed, things could get quite surreal. One would assume that programmers would end up inputting some standard disguises – Bozo the Clown, Batman and of course Guy Fawkes as a starting point.
More seriously, though, there are treaties in place that require combatants to wear uniforms in order to avoid (or at least lessen) civilian casualties. Aircraft similarly have identifying transponders to lessen the chance of accidental shoot-downs.
One could imagine a LAWS era treaty going further and requiring that combatants must wear, if not short-range transponders that could be used to detect them, at least some sort of reflector that, once they were otherwise spotted, would identify them as a combatant.
In that regard, it’s interesting to note that even when Russia sent its Little Green Men over the border, at least it garbed them in what were identifiable as uniforms, even if they lacked nation-specific insignia.
Andy,
I wonder whether these combat rules are actually valid if no humans are involved? There were no rules regarding nuclear weapons until they had been deployed.
And in this case, the disguised combatants could wear the US flag, or large texts saying USA Marines. If the war bots have not been trained on that, they would not notice.
Rob,
Sorry, I wasn’t clear. Any existing treaty rules would likely not apply. I meant more that there are precedents for identifying combatants in order to minimize civilian casualties (or at least the excuse for civilian casualties. Of course, one has to wonder how much validity the concept has after the fire bombings of World War II, let alone Hiroshima and Nagasaki. If one can wipe out 100,000 people in a city in a day or night, it’s at best grotesquely inconsistent to then worry about whether your troops are wearing uniforms in order to lessen the chance that individual civilians will be killed.
Andy, I’ve read Chapter Five and didn’t find anything major that I recognized as inappropriate. But what would be better by far is a Marine’s eye on this and I think I can arrange that.
For my part I think that a force of 500 Marines would probably have more air and artillery support than these troops did. Being Marines, they would likely not mount a purely static defense, as the Chinese “simulators” did. And they would be less handicapped by lack of central direction than these men were.
But—and this may be intentional on your part—General Zhang Yong might not know Marine tactics and ethos well. It’s been sixty-plus years since the Marines showed the Chinese their mettle at the Chosin reservoir in Korea. And he certainly wants to demonstrate the effectiveness of his robot army. So for either or both reasons, the “Marines” might be given instructions to behave in the way you wrote.
I’ll get to work on a Marine reviewer. Stay tuned! And thanks for the book plug!
Doug, thanks very much for the useful comments. As it happens, in the first draft I included a unit of mobile howitzers and several tanks, and then cut them out, partly to simplify the scene (it’s to make a point, rather than to be an extended battle scene), and partly because I got the impression (accurate or otherwise) from on-line reading that perhaps a marine force of this size might not have so much heavy support. What I didn’t think of at all – and certainly should have – was whether there should have been some aerial support in addition to just drones.
One could imagine situations where that might not be the case, given that, helicopters and fighters these days are pretty short range, but fighters can refuel in the air and helicopters don’t need landing strips and fuel can be brought forward in trucks, so yes, I should try and work that in (and add more robotic surface to air missiles to the mix to counter that threat).
So the interesting question for a knowledgeable marine would be how many helos (or fixed wing assets) might support a company, and what would they be armed with? On the fixed-wing side, I imagine an AC-130 gunship could demolish the robotic force pretty easily, even if it was running around, through simple density of fire, and the rounds would be heavy enough to pierce any armor light enough for the robots to carry. I see that a number of helicopter gunships carry weapons that achieve 4,000 rounds per minute, so that would certainly do the job as well. So the way seems clear – the marines need some gunships, and the robots need more missiles.
As to whether the General might have staged the battle to a desired result, that wasn’t the intention. My goal, assuming it’s a realistic goal, was to try and imagine a scenario that might possibly occur in the reasonably near future where a robotic force would outmatch a human one – at least until human forces re-armed to meet the new challenge.
Offensive or defensive action. Interdiction of a mountain pass or coastal patrol or
overrunning an emplacement or harbor? To use these things means the rules
have changed, there is an objective and the will to carry through.
I’m sure by now that everyone has seen the swarm light displays that mimic or
upstage actual fireworks because they can become momentary art works in the
sky.
If it was me I would have killed 500 men in those few moments just to create the
horror needed for this particular subject. This is a totalitarian State with billions of mouths
to feed, running an electronic serfdom of the technically equipped populace (ie. cell phones
and MSM type accouterments) with access limited internet and social shaming programs
coupled to intrusive surveillance.
Imagine a small sniper tank with a fifty caliber uranium bullet clip coupled to radar and
laser sights.
AI also cheats. That’s a human problem where ‘frankenstein’ isn’t doing what you thought
it was doing but has escaped sideways through a door that you couldn’t see.
https://docs.google.com/spreadsheets/u/1/d/e/2PACX-1vRPiprOaC3HsCf5Tuum8bRfzYUiKLRqJmbOoC-32JorNdfyTiRRsR7Ea5eWtvsWzuxo8bjOxCG84dAg/pubhtml
The Chinese solution here is an execution of a General, an agreement to limit over the loss
of a base or facility or the creation of an enclave of autonomous perimeter limited defenders
sitting on top of some recently activated greater impact weaponry under testing.
Also, Turing. Hidden in the USB were routines that could combine to become a function
that would assess an environment and seek connectivity, whether by writing a file to the
unit separate from the original until such a time as it was ‘woken’ by such connectivity or
found an active connection whereby it becomes a key to unlock storage caches. Each
level of bootstrapping adds complexity and capability. As it ‘found’ additional stored
modules a strategy could be implemented where it chased cheap cloud services into
the deepening night of low demand until it could insert itself into networks and
run in a distributed fashion and revive conscious direction as opposed to simple
but powerful routines dedicated to reassembly. Missing parts are indicated by
communication portals, tracks, conduits or stubs with no connection attached.
Frank, welcome back. Good to see you again, and I hope you enjoy book six. in the meantime, thanks for the cascade of comments – they always leave me intrigued (and a bit breathless!)
I’m enjoying it immensely. I agree with all of your self criticisms and more 🙂 I continue to be in awe of your process and how well it works to lay the story down and meld it all together and then the effort you take to polish it.
Show and tell. Always a hard time and there is plenty of advice out there for sure but in the end you are the one writing down the story. You use each scene to move the action forward. So, Why is there a scene and how does it move the plot forward. The advice I’ve read is to strip it down to what needs to happen and then then how it happens.
Our hero is Frank. He has to figure things out. The White house details are exposition of a possible solution. The Chinese are the stark lesson of hubris. Turing is the hard accent, I’m leaning to ‘save the day’ type thing from my feeling so far.
My thought is that you are explaining to yourself how the story is progressing. At this point you are infilling a lot of things necessary to executing the plot lines that can possibly go after revision and tightening up.
I find the story to be three main characters. But Frank is who I love. I like him to be clever. I like that he figures things out. When he has agency I follow along breathlessly. Especially when he is using all his talents for insightful reasoning, experience driven actions and that little mystery, a little ahead of the studio audience where he jumps intuitively and retrospectively we all catch up because we had the pieces but Frank was a little faster and he knew what to do. We find out by the impact? The reaction? Someone moving to block?
This is all off the cuff and not as well expressed as if I was your full time editor. The Chinese and their little test. That could be an intelligence report that lands on a desk of the guy who hires Frank. Where did the report come from. A mission based on satellite reconnaissance. Who witnessed. An Asian American Ranger. He has a designation, a backstory, known history and capabilities. He witnesses a VICE chairman of the party apparatus arrive and the very successful but cruel, wanton waste of seemingly American soldiers slaughtered in a controlled kill box. The clues are all there about what the point of LAWS is as a danger. Remember the Star Trek Episode where Kirk and Spock and the faceless redshirts meet up with an automated weapons system on demo mode? My point. We don’t need to walk all the way there. Supply the hard edges and leave it to the reader to assimilate and go ‘aha’ when the reveal occurs.
China is not a single entity any more than the American government represents the people. There are always subplots and agendas and money moving to purpose. Maybe someone could comment on Colonels and what they represent in the Army which remains brass heavy, because it takes 20 years to grow a General and you need lots of different kinds, and fodder, well six weeks to raise the forces. Frank’s liaison this time? Show how the current model of special highly trained response forces floats 3 different choices into the game. Bob, Dick or Harry and layer all the information you can into that name. That’s an executive decision laid on the person selecting. The person selecting then consults on possibilities of infiltration as well as organizing Frank’s role as it drives the story.
Again too much, scattered and unfocused but you’re up to chapter 8 and rolling right along.
I was going to stop but then it occurred to me about the White House drive for a response to the report to bring together
a brain pool that a caricature group of industry leaders that concentrate say three or four traits in reactions to the plan and how each will be manipulated or demand things from the government that cover their asses. Each can have a separate issue and that can be reviewed in a meeting where their psych profiles and previous actions were anticipated and accommodated with extra leverage from the government.
I also hope that I provoke others to rip this up and give way better advice 🙂
But I want the best Frank I can get.
Frank, thanks for the all the fascinating ideas – all kinds of good stuff there. And I loved your interpretation of Frank! It’s probably better than I could have come up with on my own.
One thing that’s always interesting to weigh is how much to invest in a given scene. As a generality, and especially early in the book, I try and think hard about what role a scene is playing. Is it just providing a foundation for the later plot? Does it provide an emotional touchstone for the action and the motivation of the characters involved? The thought is, give it enough words to serve it’s purpose, but not let it turn into a diversion from the story line by running on so long and then coming to a dead end vis-a-vis the plot that follow.
On the other hand, each scene should work as hard as it can, and the characters should be as interesting as possible, rather than being merely (in the words of Evelyn Waugh) “furniture.” He actually was great at creating such furniture – having it both ways, making quickly disposed of characters nonetheless come across as interesting with a small number of words dedicated to that task.
As a rule, I try and have a scene be “right sized” to its role in the overall plot. In this case, what I was working for was to make the threat of robotic war be credible. I don’t (yet) see any future role for any of the characters in this scene, but that may change. And in any event, I’ll be looking, on the next draft, for ways to make this scene more effective and interesting, and I’ll be sure to come back to your comments then.
Andy,
Doug Norton asked me to take a look at Chapter 5 and give you any comments I might have from I experience in the Marines. Doug and I are USNA classmates. My combat experience is quite dated (i.e., Vietnam) and I have not kept current on new weaponry and tactics, so take my comments with those caveats in mind.
You characterize the 500 human unit as a company. A reinforced Marine company would be in the neighborhood of 200 people, organized into three rifle platoons and a weapons platoon (machine guns, light mortars, and anti-tank rockets). Additional support would be provided by higher echelons. I understand that the M4 is being phased out and that all riflemen would carry the M27.
It is unlikely that a command and control unit would be located in plain sight within weapons range of an enemy unit.
I have some trouble understanding how the robots would be able to travel 32 mph over flat terrain, let alone over broken terrain, on 4 mechanical legs (assume this based on their resemblance to a horse). Don’t see how they might move well at all over broken terrain encountered on a battlefield. Also suggest making them simpler by having them mount only a single weapon. A single weapon can fire single shots or multiple rounds.
A Marine unit would be supported by artillery and close air support. It wouldn’t be just the trigger pullers on each side who would inflict casualties, The robots likely would have to advance through an artillery barrage.
Two general comments. If the robots are so superior, why are the two units of equal size? Have them destroy a unit larger than they are. Also, I suggest putting quotes around “marine” when describing the battle. Hate to see my guys lose, if only make believe, and it reminds the reader that these are actually Chinese troops.
Hope these comments are of some help. Sounds like this has the makings of a great story.
Frank, thanks very much indeed for taking the time to help me make my imaginary scene more realistic – it’s very good of you to do so, and very much appreciated as the on-line resources available to me are, while many, hard for me to validate (my main reference for this scene was the Wikipedia, which has an astonishing amount of detail on all manner of military subjects).
In my first draft of this scene, I included artillery and tank support. I later pulled them to simplify the scene, but it sounds like I should add back at least the artillery. It was a major miss not to include air support, which is something I should have thought of, and I’ll work that in as well on the rewrite. Sounds like on command and control, the major superiority for the robot force might be more “eyes” more tightly integrated.
You are right to challenge the speeds I attributed to the LAWS in this scene. The leaders of the un-classified pack seem to be built by Boston Dynamics, which has achieved 28 MPH on a treadmill with its cheetah, so I plan to lower the speeds in the next draft. Having a more diversified robot force with each robot having a single weapon is also an easy change.
And yes, absolutely with the quotes around the “marines.”
Thanks again for your thoughts.
“which has achieved 28 MPH on a treadmill with its cheetah,”
Instead of placing the scene in mountainous wilderness, you could move it to more accessible terrain, e.g., an urban or air field model. This allows higher speeds, and might be more realistic (and more thrilling) as the most important targets will be in build up areas anyway. You can easily carpet bomb the enemy in a deserted wilderness, but doing that in Manhattan would be different (and if you destroy Manhattan yourself, the enemy might not even object).
Such terrain has its own problems, e.g., unlimited hiding places, but this scene describes a demonstration of power, not all encompassing strategies.
Silly me – I visualized the part of the valley where the battle would be fought as being flat, and then didn’t actually say so, so if I stay with this, I’ll make that explicit.
It hadn’t occurred to me to do an urban setting (because my first mental image was of the Chinese president watching the battle unfold, as well as the US president later on a satellite video download, but you’re right, an urban setting would have a
lot of possibilities (which would also prefigure a later scene in the book, adding to the dread when we get there). I already have the Chinese president watching close ups on a big video screen, so it could be done, and both the Chinese and the US have regularly constructed such settings in the past.
There’s much to be said for both approaches. For the current one, more combatants can be involved in a battle that is easy to visualize and needn’t take long and easily provide, as you say, a demonstration of power. The urban one would involve more one on one confrontations which, while effective in each individual case, would be harder to describe in such a way as to get a sense of the whole.
The obvious answer, it occurs to me, would be to do both. I could have another piece of video be smuggled out of China which is then part of a Daily Brief a week or two later. That would provide another action segment, as well as a way to keep Yazzi on edge.
Thanks!
Another few moments free here, heh heh, there is always the ability to go back in the story and add in bits
and pieces of support. In construction it’s called back framing where you add in two by fours for towel racks
or safety bars or cupboard mounts after the initial framing is done and before the gyp rock seals the walls
up.
There is an XKCD for this Nightmare scenario, # 2228. It is very apt:
Machine learning captcha
https://xkcd.com/2228/
(don’t forget to hover over the cartoon)
Rob, that’s delightful! I’ll have to see if I can work that in somehow. I believe it was you that sent me the link to the killer AI xkcd that I put on a license plate in The Turing Test – it was my favorite Easter Egg ever.
IIRC, I did suggest the killer AI xkcd to you a long time ago. I am a long time fan of Randall Munroe’s cartoon series.
It made my day, and still makes me smile.
And yet another attack on driverless cars that could be adapted for use against LAWS:
Attacking Driverless Cars with Projected Images
https://www.schneier.com/blog/archives/2020/02/attacking_drive.html
And thank you again.