Defective control: naval warfare in the Pacific in 2036
8 Sep 2023|

ASPI’s annual conference takes place on 14–15 September at the National Convention Centre in Canberra. The theme is ‘Disruption and deterrence’.

Effective deterrence, built upon a strong Australian Defence Force that works closely with allies and partners, will be a vital part of ensuring a sustainable strategic balance in the Indo-Pacific. The simultaneous disruptive power of rapid changes in technology carries risks and opportunities for Australia and its partners as they invest in superior capabilities and look to integrate them into a seamless force.

The conference will bring together Australian government ministers, senior defence officials, leading industry figures and international experts to tackle these challenges and explore key trends and areas of innovation.

In keeping with the conference theme, The Strategist is publishing the second of two pieces by Jeffrey Becker, leading futurist for the US military. ‘Defective control’ is a fictional account of a naval clash in the Pacific in the 2030s. It aims to illustrate the potential of reusable rockets, directed energy and space capabilities—in the hands of adaptive and innovative people—to win a fight on earth. Hyperlinks are embedded throughout the story to illustrate how some of the ideas are already being put to use.

Aboard USS Miguel Keith—4 March 2036, 0900 hours

‘It’s not often you get to be around for the birth of an entirely new kind of warship,’ Captain Snyder thought as he looked down from the bridge of the USNS—well, now USS—Miguel Keith. The ship was ungainly, a thin deck perched atop steel pillars. Below was an open but very busy hanger. Above, the narrow flight deck was littered with a forest of blunt, stubby missiles standing tail-first on squat landing legs. Looking like an unfinished construction site, the flight deck was flanked by cobbled-together cranes and gantries welded to each side of the ship, giving the Keith something of the appearance of a gigantic fishing trawler.

USNS Miguel Keith, before conversion to a rocket carrier (designation CVR-1). Source: US Navy/Wikimedia Commons.

From the bridge, Snyder watched the peculiar, alien dance playing out on the Keith’s deck. Sailors, many in powered exoskeletons, were pushing and shoving a dozen AMR-1 Anvils around the deck, preparing for flight operations. ‘AMR’ designated the US Navy’s new ‘attack missile—rocket’ weapons.

AMR flight ops felt completely unlike the aggressive, low-slung—and decidedly horizontal—operations one would expect of sharp-nosed jet fighters on a nuclear-powered aircraft carrier. Instead, Snyder watched a study in the vertical. Below the bridge, he saw a complex dance of what appeared to be mobile Greek columns that were, in fact, 50-foot-tall rocket boosters.

Half a dozen of the Anvil rockets were lined up, ready to be secured by one of the many jury-rigged cranes and gantries. The gantry tipped an Anvil horizontal, dropping it to the lower hanger deck to be armed and fuelled.

‘She’s ugly, all right. But lethal doesn’t have to be pretty.’

The busy ship spread out before him was the centrepiece of a new kind of fleet. The Miguel Keith and the other expeditionary mobile base ships were originally designed as forward lily pads to support air and amphibious operations over the shore. In this war, however, they were pressed into service in an entirely new role.

They might just give the United States an edge as the war with the Chinese Communist Party moved into a new and more dangerous phase.

‘We’re going to need to think very differently if we’re going to get back in this fight,’ Snyder thought.

Twin sharp supersonic cracks interrupted his thoughts. Snyder looked up. Two Anvil boosters plummeted tail-first from the edge of space. Snyder watched them as they fell, comet-like, fiery streaks marking their path back down towards earth. At 300 feet, four spider-like legs unfolded downwards as a bright purple needle of fire spit out from the centre engine for the retro burn. Shock diamonds in the exhaust underscored the massive power of the Anvil boosters as they settled gently and almost simultaneously onto the deck of CVR-2 Montford Point two miles to port.

In the distance, rolling seas gave way to a bright blue sky. Snyder took in the view beyond the missile-studded deck of the Keith and the ongoing recovery operations aboard the Montford Point. From the bridge, the whole of Joint Task Force 216 was in view. The motley fleet was cruising deep in the Pacific Ocean north of Wake Island. Spread out over just a few miles, three new CVRs were joined by LHA-7 Tripoli, a large-deck amphibious assault ship that was also pressed into service as an Anvil-carrying warship. The four attack ships were huddled around CVN-80 Enterprise in a formation far more compact than the navy’s former aircraft carrier fleets.

Snyder looked across the bridge wing as the Enterprise pulled alongside the Keith just a few feet away. The Enterprise was hastily brought to San Diego after the Ronald Reagan and Gerald Ford were nearly sunk.

‘We’re probably going to change “CVN”, too,’ he thought. The Enterprise wasn’t really an aircraft carrier anymore, its role far different—but no less important—from when it launched fighter aircraft from its deck. Now, its nuclear power plant was pressed into service manufacturing the massive quantities of liquid oxygen needed to fuel each Anvil attack rocket.

‘Sir, ready to start UNREP,’ the XO called out.

‘Go ahead,’ Snyder replied.

With this command, minivan-sized quadcopters lifted off the deck of the Enterprise carrying with them deep-cold cryogenic refuelling lines. The quadcopters shuttled back and forth between external valves on the side of each ship, automatically and precisely connecting the liquid oxygen and methane lines and allowing each to flow from the Enterprise to the Keith.

As the replenishment proceeded, Snyder looked hard at the repurposed aircraft carrier. The Enterprise was really more than just a refueller and held another surprise for the enemy. Catapults, arresting gear and combat aircraft were removed, replaced by a dozen large, white domes. Topside, the ship looked like an astronomer’s telescope farm. Some of the domes were open, revealing large optical mirrors and devices inside. Instead of gathering light from faraway stars, Snyder knew that these ‘telescopes’ concentrated megawatts of power into targets. Each was a long-range laser ready to intercept and destroy incoming attacks from air and space with speed-of-light, nuclear-powered precision.

In the centre of the deck, a much larger dome stood above the laser arrays, nearly the full height of the Enterprise’s island. It was faceted and a metallic dark grey colour. Inside, Snyder knew, lay a massive, multidirectional grid of phased-array slabs, able to attack and destroy electronic systems over dozens of miles. This electronic warfare system was also nuclear powered and was—he hoped—strong enough to fry the electronic guts out of any missile, aircraft or ship unwise enough to approach.

Instead of spreading the fleet out, Admiral Lane kept the CVRs in tight formation with the Enterprise, which provided high-density laser and electronic fire to fend off missile-borne aerospace attacks. Precisely hitting a moving target at long range depends on precision electronics and optical systems, which are fragile. Nor can the complex and fragile control surfaces on the aircraft and missiles themselves be armoured enough to defeat high-power laser attack. The Enterprise would exploit that and put up a protective bubble for the CVRs while closing on the Chinese coast.

Snyder believed that if the People’s Liberation Army could locate the fleet, they would attack with cruise and air-launched ballistic missiles, shot from the new and stealthy long-range H-20 bombers. Getting targeting right at this range would be tough for the PLA Air Force. China’s long-range anti-ship ballistic missile force was sidelined by the orbital blockade that had stopped most space-based surveillance data from getting through. Even now, the Enterprise was attacking overhead Chinese satellites with laser and electronic fire, contributing to the US Space Force’s orbital blockade.

Snyder’s ‘fishfinder’, the onboard personal comms devices every sailor carries, squawked at his belt. A display floated at eye level. It was the admiral. Snyder gestured at the holo, answering the call.

Admiral Lane’s quiet voice came over the line. ‘Captain, could you meet me down in the print shop? I’d like to talk to you privately. I’d appreciate your take on something.’

‘Yes, ma’am. Ten minutes?’ The admiral agreed, gesturing to her own fishfinder and cutting off the connection.

Snyder wondered what the impromptu meeting was all about, and more, how the admiral managed to project calm determination in a world that was about to explode.

Operation Posui Lian (Shattered Chain)—1050 hours

Admiral Zhao’s fleet charged east, a mirror image opposite Admiral Lane and her own hidden fleet. The massive vessels under his command were heading away from their assembly area southeast of Shanghai. Zhao’s flagship, the aircraft carrier Guangdong, along with its protective screen of cruisers, destroyers and submarines, had until recently been supporting army operations on Taiwan. However, the Central Military Commission had called the force back to Ningbo to rearm and refuel, adding more combat power in the process—much more—for an ‘operation of even greater consequence’.

Admiral Wei, the fleet political officer and Zhao’s equal and counterpart in all decisions, stepped to the centre of the Guangdong’s command information centre. Zhao knew that the PLAN was not really a ‘Chinese’ navy, but rather an armed wing of the CCP. As such, every ship was led by both a commander and a political officer, as were the fleet itself, ensuring that key warfighting decisions were in line with the CMC’s party directives.

This was natural. What was less natural was Wei’s taking over the situation brief from the intelligence officer.

‘Must be important,’ Zhao thought. ‘Wei must be delivering something straight from Beijing.’

Wei turned to the holo display projecting up from the floor in the middle of the room. The Guangdong’s combat information centre was arranged in a circle around the central display. The whole of the PLAN’s East Fleet commanded from here.

‘As you know,’ Wei began, ‘Taiwanese defences are crumbling. The war on the island is moving to a new phase as the army moves on Taipei. Splittist forces are still fighting in and around the city, but our army and rocket forces are breaking the remaining resistance. The chairman anticipates the imminent success of these operations due to an extremely favourable and improving correlation of forces on this front.’

Wei gestured to the 3D display. Units from both sides floated by as the Warbot battle command system rapidly calculated Taiwanese battle positions and anticipated future movements and engagements, including the probability of success. Warbot did this in near real time, anticipating branches and sequels, and making recommendations to the human commanders.

‘As you can see from the data, this fleet’s combat power is no longer needed on Taiwan. Instead, I am pleased to inform you that we have a new mission. The objective of this mission is only within reach because of the swift and decisive successes of this fleet.’

Wei paused as the holo imagery moved outward into the central Pacific. ‘The Guangdong, together with the rocket force, delivered decisive blows against the US Navy. Our space assets have caught several glimpses of the fleet. We know the Reagan and the Ford were each struck by waves of our long-range missiles near Guam and are either sunk or on fire. Several destroyers have also been sunk. What remains is limping back to Honolulu. Social network analysis run through Warbot predicts that they have sustained over 4,000 casualties in the initial assault.’

Another pause. ‘Our attacks were far more successful than even Warbot anticipated. We re-ran our models with these updates. Warbot gave us some very surprising results.’

Wei scanned the room, filled with commanders, captains, admirals and political officers charged with operating the great fleet.

‘Comrades,’ he said, ‘with the bulk of US combat power pushed out of reach of the mainland, the party has assembled this armada to break the oceanic chain that surrounds us once and for all and instead use it to hold Japan and the Philippines in our own grip. With the Ryukyu Arc and Northern Luzon in hand, we set in motion the final stage of our great rejuvenation of the Chinese nation, communist party leading.’ A cheer went up from the assembled crowd.

Zhao listened, examining the map as Wei spoke, eyes following the long string of islands that loomed just 150 miles to the west. The founder of the modern PLAN, Admiral Liu Huaqing, spoke of these islands stretching from Japan to Singapore as a steel chain that shackled China to Asia, blocking it from its rightful place on the global stage.

‘Time to think bigger,’ Zhao thought.

Wei continued. ‘The CMC fully embraces Warbot’s conclusion that the bulk of US combat power in the Western Pacific has been damaged beyond repair. I’ll admit to being surprised. I suppose our situation is much like that of the US military after its stunning success in the first Gulf War. That success was, ironically, the impetus for this very force, a force fine-tuned to dominate information- and AI-based warfare. We will build on this success, and now execute Operation Shattered Chain.

Operation Shattered Chain. Source: Google Maps, annotated by the author.

Shattered Chain will consist of two independent task forces, Luzon and Ryukyu, aimed at seizing the northern and southern anchors of the first island chain. TF Luzon’s objective is to seize a strip of Luzon and its offshore islands to drive off the US Army’s Multidomain Task Force. This force is currently impeding our access to the South China and Philippine Seas. Our own element, TF Ryukyu, is designated the main effort. Our objective is the seizure of entire of the Ryukyu island chain, with our initial target the seizure of the US base on Okinawa itself.’

Zhao looked on as the projected operation unfolded on the display. The Warbot AI’s confidence indicators glowed green above the unit and engagement icons as they moved, denoting a very low risk for each manoeuvre. To Zhao’s eyes, Okinawa and the rest the island chain turned a very satisfying shade of red as they fell under PLAN control.

Wei seemed very pleased by the chain of events plotted out by Warbot. All PLAN admirals, Zhao included, placed great faith in scientific precision and the use of calculations to reveal the true nature of warfare. Warbot enabled commanders to predict the course of events if they just looked hard enough, accelerating warfare faster than the enemy could adapt. Massive amounts of data, huge processing centres and powerful, networked battle AIs like Warbot brought all these calculations into reach at inhuman speed. Central Military Commission planners embedded intelligent systems throughout the PLA’s command-and-control structures, certain that Warbot would spit out the mathematically correct answer.

Events over the past months appeared to bear out Wei’s overweening confidence. The PLA’s command systems, pulled together and amped up with AI, had been delivering blows from China to Guam, from space to the electromagnetic spectrum, faster, harder and with more coordination than even he, an experienced military commander, had ever expected.

Still, Zhao was uneasy. Had China really unlocked the secrets of war that had eluded the great commanders? Even great commanders were surprised sometimes. Was Warbot that much better?

‘Secretary,’ he addressed Wei. ‘I understand the new mission. May I ask a question? I have a concern I’d like to share with this group.’

‘Go ahead, Admiral,’ said Wei, looking a bit nonplussed by the question after the delivery of his speech.

‘Comrade,’ Zhao said, ‘I’ve seen something that bothers me, something not well represented in the Warbot assessment. Our sensors are very spotty over the Central Pacific, but at least one piece of data is very strange. Radar and infrared have detected several bright and fast-moving targets at very high speed. Probably at least one rocket body. However—and this is what I don’t fully understand—instead of coming at us directly, it tipped upward at the edge of the atmosphere and started falling back to earth, moving east, rather than in this direction. This doesn’t follow any known flight profile for US long-range weapons.’

‘Admiral,’ Wei said, cutting him off. ‘It sounds like either a failed submarine attack or perhaps a feint.’

‘Interesting,’ Zhao thought, the tiny fissure of doubt widening. He knew that the IR and radar returns were only part of the story. Sensor data sent from the Strategic Support Force told him that just before the tip-up, the missiles ejected at least 16 small vehicles each. Those also fell far short of the fleet, just east of Okinawa. Not dangerous, but it was still a concern and left him unsettled. The launch location, the flight profile of the twin rockets, and the fast-moving vehicles falling short—none of it was right.

‘Yes, that is probably the case,’ Zhao said agreeably, not wanting Wei to lose face in front of his command. However, he resolved to take a closer look, running these observations through Warbot to see where they might lead.

Contact—1120 hours

Aboard the Miguel Keith, Captain Snyder met Admiral Lane below decks. They walked together through the printshop, examining stacks of gleaming 10-foot-long shards of 3D-printed metal. Each was an independently targetable munition and they would be loaded 12 at a time atop an Anvil booster. The printers allowed intricate hypersonic scramjet-powered engines to be grown, inside out, aboard the ship. As if they were on some steampunk coal-fired warship, exoskeleton-clad sailors shovelled titanium alloy powder into hoppers feeding the machines. On the other side, sailors in forklift-sized powersuits lifted the shards as they were spit out of the printer, stacking them with the others, steaming as they cooled.

The two walked side by side, silently examining the newly made weapons. It was hot in the print shop—the titanium had to be heated to 3,000 degrees Fahrenheit. Hopefully the whole system would one day be automated. For now, sailors in the heavy, but weirdly graceful, powersuits moved the shards between the printers, an assembly area where they would be fitted with an electromagnetic-pulse-hardened electronic brain and radar and optical sensors, and then mated to a fuelled-up Anvil to be hurled at the enemy.

But the robotic systems would have to wait for now. Snyder and Lane paused, examining one of the newly printed shards. Someone had painted ‘With love, from Honolulu’ on it. ‘Nice touch,’ Lane remarked. Of course, the art wouldn’t last long, vaporising the moment the shard hit the atmosphere after being ejected from an Anvil’s aeroshell.

‘Captain,’ Lane said, breaking the silence between them. ‘What do you think Zhao’s fleet is going to do next?’

‘Admiral, the PLAN is now—by an uncomfortable margin—the largest naval force on the planet. If I were them, I would cross the East and South China Seas as fast as I could and seize and fortify the first island chain. They’ve got to be feeling pretty pleased with themselves. For them, another easy roll of the iron dice puts everything from Kyushu to Luzon in their grasp. Hell, I might not stop there. From their perspective, there’s nothing between them and Guam. Maybe even Midway and Hawaii.’

‘I agree. Where do you think they’ll go first?’ Lane asked.

Snyder said, ‘I think the PLAN is assembling a northern fleet here.’ He gestured to a large-scale map of the Western Pacific projected from his pocket holo display. A satellite view swept inward, zooming in from space and focusing on a massive assembly of over a hundred ships. The high-resolution imagery of the ships hung in the air in front of them.

‘When was the last time you had a good look and updated this?’ asked the admiral.

‘About six hours ago,’ Snyder responded. ‘Too long ago to get a good firing solution.’

Lane turned to Snyder, gesturing to push a data packet to Snyder’s fishfinder. ‘Take a look at this. We got a very good look from those two scout-configured Anvils that just came back.’

Snyder was surprised at the diverging trajectories of the two Anvils. Each dropped off a swarm of high-endurance sniffer drones. The first was expected: deploying stealthy drones to stalk the northern fleet. Snyder had wondered why the admiral had risked the discovery of JTF 216 by launching the Anvils. But now he understood.

‘You sent the second Anvil south.’

Lane nodded. ‘Since two reconnaissance shots would be just as detectable as one, I had the Montford Point look to the south as well. The Multidomain Task Force has been hitting a lot of Chinese hardware coming south and east. It’s got to be giving them headaches. If it were me, I’d spare a force to clean them up.’

Snyder’s eyes widened. ‘Did you find it?’

‘Yes. Targeting quality. I want you to get your AI working on an attack-management plan immediately. But first, I want your take on how fast we can turn around an Anvil alpha strike force. Can we get a second strike in the air quickly?’

‘Admiral, the short answer is, two hours. However, I don’t think we’ll need a follow-up strike. I think we can put enough Anvils with enough shards on the northern fleet to put most of them on the bottom, or at least make them think twice about trying for Okinawa. I’d recommend getting our licks in, recovering the Anvils and retreating beyond H-20 range to preserve the fleet to support the task force when we can really reset. This second set of Anvil data should really help them out.’

Lane looked across the hanger as the sailors continued stacking the shards. She was thinking. Her eyes stayed on the stacks and narrowed. ‘I don’t intend to hit the northern force twice. I intend to kill both fleets—northern and southern. Today.’

Raid—1250 hours

The general quarters alarm broke the fevered preparation throughout JTF 216. The Miguel Keith had already detached from the Enterprise, but it was still close. Overhead, a Chinese Dark Sword drone was screaming by the fleet. The high-speed reconnaissance drone was detected by the Enterprise, which quickly smashed the control computer with a high-powered microwave pulse, sending it crashing into the sea. However, the electronic attack wasn’t fast enough to stop a data burst transmission to three H-20 stealth bombers on combat patrol north of Task Force Okinawa.

The H-20 flight command Warbot received the message, analysing the data in milliseconds and beginning on-board mission attack planning. Within seconds, Warbot had identified the contact as the Enterprise and assessed the five-ship formation as a critical strategic strike priority. The lack of air-defence destroyers in the formation was highly irregular, but this fact pushed the assessed probability of a kill for a combined attack on the Enterprise at close to 100%.

Before the human flight lead understood what was happening, the three aircraft turned sharply northeast with pinpoint precision. The bombers straightened, pitched slightly upward, and the bomb bay doors opened and simultaneously dropped six air-launched ballistic missiles in rapid succession. The missiles fell away several hundred feet before the rocket motors ignited. The aircrew watched as the missiles sped away, arcing high above the bomber formation at Mach 5, only now made aware of the nature of the target after being alerted by Warbot to return home.

The Enterprise’s high-altitude aerostat, cruising alongside the fleet at 100,000 feet, caught the bright infrared signatures of the missiles and passed the data to the Enterprise. An automatic alarm sounded in all the ships of JTF 216: ‘Crew alert. Six vipers inbound. Laser operations commencing momentarily. Don eye protection immediately.’ Snyder reached below his command console on the bridge and put heavy goggles over his eyes, as did everyone else in the fleet.

Aboard the Enterprise all 12 laser domes opened and snapped to the southwest. Invisible beams immediately crossed the 20-mile gulf, striking the inbound missiles almost instantaneously. Occasionally a beam would cross through cloud and sea spray, splashing high-intensity flashes of light around the fleet and blanketing the sky like a silent lightning storm. The beams found the missile bodies, heating up and warping the metal just enough to cause the weapons to tumble and break apart, pieces careening and cartwheeling into the sea.

Lane, connected by line-of-sight lasers to Snyder and the other ship commanders, came up on the battle network. ‘JTF 216, the enemy is aware of our positions. You have your targets. Fight these ships and destroy the Chinese fleet,’ she said.

Alpha strike—1332 hours

Lane delivered the order from the bridge of the Miguel Keith and flight deck operations began in earnest. Snyder was busy monitoring the complex movements of the Anvils to get them off the deck as close together as possible. Within seconds, eight rockets roared off the decks of the CVRs, two apiece, climbing skyward. Robotic grabbers moved two more rockets into position on each ship. Wave after wave of missiles leapt upward from the deck, the sequence repeating until each ship had had put 18 boosters into the sky.

The admiral looked up as her four ships threw 72 Anvils configured to deliver a shower of 864 half-ton, independently targetable shards at Zhao’s force. At Mach 8, the shards would cover the 2,500 nautical miles in under 20 minutes. Streaking upwards in minutes to the edge of the atmosphere, the Anvils spit out the shards, which immediately snapped away in many different directions. As the Anvils rose, the onboard AIs coordinated trajectories for each grouping and commanded the shards to converge on TF Okinawa in two waves less than two minutes apart.

As the first wave of shards descended, they began scanning Zhao’s fleet, each shard picking out the most lucrative targets and coordinating with the other weapons to focus on the air-defence ships, ensuring the second wave would get through. The emphasis on air-defence ships would turn out to be unnecessary. Zhao and Wei had placed Warbot in full command of the automated defence systems, but the highly manoeuvrable shards moved too fast and unpredictably to be intercepted.

The result of the first wave was catastrophic. Multiple shards punched all the way through the ships from top to bottom, setting fires and detonating ordnance. Just before impact, the first wave passed data to the second, allowing them to pick undamaged or high-value ships and to maximise damage to the remaining ships.

Back at JTF 216, Snyder heard a sharp succession of supersonic cracks as the Anvils returned from the strike. As they again landed tail-first, squat, crablike robots scuttled across the deck, attaching to the base of each returning rocket. Lifting the Anvil slightly, the bot moved the missile out of the way to the side gantries, which tipped the rocket horizontally.

With surprising speed, the gantry dropped the rocket to a cradle on the lower deck. The cradle slid into the ship, where weapons techs waited in their exoskeletons, sliding a new bundle of shards into the Anvil’s forward conical fairing. Locking the warheads down, the tech swept an exoskeleton-clad arm forward, moving the rocket to the opposite elevator, while the cradle slide outward to receive the next Anvil to rearm and refuel.

Anvils, reloaded and refuelled, began leaping off the decks of all four ships once again. This time, they headed for the smaller task force aimed at Luzon.

The aftermath—11 November 2043

Retired admiral Zhao took the interview via holo from a reporter in now-independent Hong Kong.

‘Admiral, thanks for joining us. This is the first time we’ve had someone at your level talk about the PLAN experience during the Great Pacific War. Can you reflect a bit for our audience on the disastrous Battle of the East China Sea?’

‘Thank you for the opportunity. It’s very gratifying to tell this story. As you well know, that engagement turned the tide of the war and ultimately sent the Chinese Communist Party to its grave.’

Zhao paused.

‘The missile assault that sunk not one, but two fleets is well known. What is perhaps less well known, but was really the most important factor in our defeat, was the great stock we placed … well, our utter fixation, actually, on our ability to hammer the nuclear-powered aircraft carrier, its escorts and all its means of support far out to sea. We knew our rain of missiles would send the US Air Force fleeing Japan eastward. We worked through scenario after scenario with Warbot, predicting victory after victory. We had built a symphony of destruction we were sure would give us a firm grip on the first island chain and beyond.’

‘But ultimately, Warbot didn’t work?’ the reporter asked.

Zhao paused again and sighed.

‘The party, in true scientific-socialist fashion, calculated everything the Americans would do. Everything except the new missile ships. Those were a surprise.’

‘But why did missing just a few ships used in this new way cause your plans to unravel?’

‘For all of our centralised, automated and AI-based force planning, we couldn’t control an adaptive, motivated America that could respond and strike out in an entirely new way.’

As the interview continued, Zhao recounted the long war that dragged on over the next four years, with over 100,000 dead and wounded on each side and more than a few nuclear near-misses along the way.

‘By that time, we should have known that bad information ultimately would lead to disorganisation and chaos. But even after the twin calamities off of Okinawa and Luzon, we couldn’t adapt. Our confidence that we understood modern battle was broken shell that would never recover. The worst part for me was that we clung to our mistaken beliefs, standing by, waiting for Warbot to give us the right answers, replacing our own judgement and experience with those of the machines.’