The Convergence: Why a US-China War Is Becoming Structurally Inevitable
By Alexander Temerev
China’s military has eighteen months left on the readiness deadline Xi Jinping set for a Taiwan operation. The United States has two carrier strike groups, 50,000 troops, and a dwindling munitions stockpile pinned down in Iran. And between January and March 2026, three seemingly unrelated American actions cut China off from every major energy supplier outside Western control.
Each of these has its own official rationale. Together, they trace the outline of something larger: a convergence of structural conditions making a US-China military confrontation harder to avoid with each passing month.
Act One: The Sequential Isolation of China
On January 3, 2026, the United States seized Venezuelan President Nicolás Maduro, disrupting one of China’s key sources of discounted oil that bypassed US-monitored financial channels.
On February 28, 2026, the United States and Israel launched Operation Epic Fury, a massive joint military assault on Iran. Within five days, over 1,000 Iranians were killed, Iran’s Supreme Leader was assassinated, and the Strait of Hormuz was effectively shut down. Iran was China’s other major source of sanctioned crude oil, supplying roughly 13% of China’s seaborne intake.
These events are typically analyzed separately. Venezuela is framed as a Monroe Doctrine enforcement action. Iran is framed as a nuclear nonproliferation campaign. But look at what happened in aggregate: in less than two months, the United States eliminated China’s two primary alternative energy suppliers, the ones that operated outside Western financial monitoring systems.
Coincidence? Perhaps. But consider the third move.
China does have overland energy connections that no navy can interdict. The Power of Siberia 1 pipeline delivered 38.8 bcm of Russian gas in 2025. The ESPO pipeline pumps roughly 600,000 barrels of Russian crude per day into Heilongjiang. The Central Asia-China pipeline carries 35-43 bcm annually from Turkmenistan, Uzbekistan, and Kazakhstan. All of these operate outside Western financial systems, settled in yuan and rubles through CIPS rather than SWIFT.
But these pipelines cover a fraction of China’s total energy consumption. China imported over 560 million tonnes of crude oil in 2024, of which the ESPO pipeline carries about 30 million. The rest arrives by sea, through chokepoints that the US Navy can monitor or close. And for commodities that don’t flow through pipes — LNG, iron ore, coal — China depends on maritime routes, above all the Arctic Northern Sea Route from Russia.
That route is now under pressure. NATO launched “Arctic Sentry” in February 2026, bringing all allied activities in the High North under a single operational command for the first time. Thousands of US Marines deployed to Norway for Cold Response 26. Denmark established Operation Arctic Endurance in Greenland, with a planned one-to-two-year military presence. The UK doubled its troop deployment to Norway. Trump’s push for Greenland, seemingly absurd on the surface, resulted in NATO gaining consolidated control over the Arctic’s chokepoints, including rights over Greenland’s critical minerals.
The route is already faltering on its own. Cargo volumes fell for the second straight year, dropping to 37 million tons against a target of 80 million. Western sanctions have cut Russia off from shipbuilding technology. The next-generation icebreaker meant to enable year-round navigation is only 30% complete. And the Power of Siberia 2 pipeline, which would double Russia’s overland gas capacity to China, won’t deliver gas before 2030 at the earliest.
Map this out: Venezuela — cut. Iran — cut. Strait of Hormuz — closed. Arctic route — encircled and underperforming. Overland pipelines — operational but insufficient. The net effect is a China that can survive on its existing pipelines, but cannot fuel a wartime economy through them. Whether by grand strategy or by the cumulative logic of separate decisions, China’s energy margin is shrinking.
Beijing sees this. And Beijing is drawing its own conclusions.
Act Two: The Military Clock
China’s People’s Liberation Army has a 2027 deadline, set by Xi Jinping himself, to achieve the capability for a decisive military operation against Taiwan. The Pentagon’s own assessments confirm this as stated Chinese policy.
In late 2025, China’s “Justice Mission” exercises rehearsed a full maritime blockade of Taiwan. For the first time, PLA vessels entered Taiwan’s contiguous zone in significant numbers. A Chinese destroyer withdrew only after a Taiwanese frigate locked its fire-control radar on it, signaling readiness to fire. Twenty-seven missiles were fired in or around the contiguous zone. The exercises covered a larger area than any since 2022.
China has commissioned over 18 new warships in 2025 alone, including its third aircraft carrier (the 80,000-ton Fujian, equipped with electromagnetic catapults matching US Ford-class technology) and the world’s largest amphibious assault ship (the Type 076 Sichuan). It now has over 370 ships and submarines, the world’s largest navy by hull count. Four Type 075 amphibious assault ships are in service, with more under construction.
On March 5, 2026, as I write this, China announced a 7% increase in its defense budget, bringing planned military spending to $276.8 billion. This funds not just hardware, but military salary increases, training, maneuvers around Taiwan, cyberwarfare capabilities, and advanced equipment procurement.
And then there are the purges.
Xi Jinping has reduced the Central Military Commission from seven members to two: himself and one other. Over half of PLA leadership has been affected. Of 47 people who have held three-star general or above positions since 2022, 41 have been removed. Only 11 of 52 key military positions are currently occupied.
Some analysts interpret this as chaos that makes military action less likely. I read it differently. History shows that leaders purge their command structures before major military action, not to improve competence, but to ensure obedience. Stalin purged the Red Army before 1941. Hitler reshuffled his generals before every major escalation. You need commanders who will execute orders without hesitation when the stakes are existential. The purges look less like anti-corruption reform and more like consolidation of wartime command authority.
Act Three: The Window
Right now, the United States has two carrier strike groups, over 50,000 troops, and 200+ aircraft committed to the Iran theater, running operations around the clock. The US defense industrial base, already strained, is being drawn down in real time.
US military assets are being physically diverted away from the Asia-Pacific. Ships and aircraft can only be in one place at a time.
From Beijing’s perspective, this creates a window. Not necessarily for an immediate invasion of Taiwan, which would require months of visible preparation, but for actions short of invasion that are harder to respond to: a “customs quarantine” of Taiwan’s outlying islands, an escalated blockade framed as law enforcement, or a provocation-escalation spiral triggered by increasingly aggressive PLA drills that goes wrong.
Taiwan’s vulnerability is acute. It has roughly eleven days of natural gas reserves, six weeks of coal, and five months of crude oil. Food stockpiles would last seven to twelve months depending on the category. China doesn’t need to invade. Curtailing trade by even 50% would be devastating. And China has been rehearsing exactly this, with fishing militia formations of 1,400 to 2,000 boats, coast guard patrols pushing into Taiwanese waters, and AIS spoofing that makes tracking PRC vessels nearly impossible.
The diplomatic “thaw” between the US and China, including a planned Trump-Xi summit, means nothing if you’re watching ship movements instead of handshakes. Germany and the Soviet Union were trading partners right up until Barbarossa. Japan’s diplomats were literally in Washington when Pearl Harbor was launched.
Act Four: The Autonomous Revolution
Most analysis of a Taiwan conflict focuses on ship counts and troop numbers, and misses the part that will actually decide it.
The next major war will be fought with autonomous systems at scale — drones, ground robots, naval drones, and increasingly, the factories that produce them.
Autonomous systems also lower the threshold for starting a war. Unmanned operations mean fewer body bags in the first weeks, making escalation easier to absorb domestically. Drone blockades can be framed as law enforcement. AI-driven kill chains operate faster than diplomatic channels can respond, so a limited action can escalate beyond recall before anyone picks up a phone.
Ukraine has already demonstrated the mechanics. Ukrainian drone production scaled from 2.2 million units in 2024 to 4.5 million in 2025, with some estimates suggesting capacity for 10 million per year with adequate funding. FPV drones now account for 60-70% of all destroyed Russian equipment. Ukraine conducted the world’s first fully unmanned assault on Russian positions in December 2024. Drone swarms of up to 25 units coordinate strikes autonomously. AI-enabled terminal guidance allows drones to operate independently of radio links, making them unjammable.
Both sides now organize strike campaigns around factory production rhythms rather than traditional military planning. The war is being won or lost in manufacturing output, not on the battlefield alone.
Scale this logic to a US-China conflict over Taiwan. The decisive factor won’t be which side has more ships in the strait on day one. It will be which side can manufacture and replace autonomous systems fastest, who can sustain industrial-scale production of drones, unmanned ground vehicles, autonomous naval systems, and the AI systems that coordinate them.
This is where self-replicating manufacturing enters the picture. We are approaching the point where autonomous systems can build the factories that build more autonomous systems. Apptronik’s Apollo humanoid robots are being tested at Jabil to build more of themselves, with commercial production targeted for 2026. Tesla’s Optimus program envisions autonomous robots constructing their own successors. China controls approximately 85-90% of the humanoid robotics market by shipment, with Unitree alone targeting 10,000-20,000 units in 2026.
Chekhov said that if a gun appears on stage in the first act, it must be fired by the third. You don’t build 200 military-convertible RO-RO ships, rehearse beach landings with civilian ferries, train fishing militia in blockade formations, construct the world’s largest amphibious assault ship, and set a 2027 readiness deadline just to let it all sit on the shelf.
And you don’t build self-replicating autonomous manufacturing systems and not use them in the domain where they matter most.
The Chokepoints of the Autonomous War
If the next war is won by whoever can produce autonomous systems at the greatest scale and speed, the critical resources shift from oil wells to mines, fabrication plants, and AI chips. The picture is troubling for both sides, but in different ways.
China’s advantages:
China controls 90% of global rare earth processing: the magnets in every drone motor, every robot actuator, every missile guidance system. It dominates gallium and germanium production (essential for semiconductors). It has banned exports of these materials for military use. It controls the integrated supply chain from mine to finished product. And it has the manufacturing workforce and infrastructure to scale production faster than any other country.
The US vulnerability:
Over 40% of semiconductor components in US weapons systems are sourced from China. The US has no large-scale domestic production of ultra-high-density printed circuit boards used in defense applications. A US defense assessment concluded bluntly: the US “currently cannot surge or scale the electronics needed for sustained conflict unless it relies on adversarial sources.” It would take 18-24 months to bring even limited domestic production of critical electronics online.
And then there’s the ultimate chokepoint: Taiwan itself. TSMC produces roughly 90% of the world’s most advanced semiconductors. If China takes or blockades Taiwan, it simultaneously gains control of advanced chip production and denies it to the United States. The AI chips that power autonomous weapons systems, on both sides, flow through this single island.
The rare earth export bans are already having effect. A US drone manufacturer supplying the military has delayed orders by two months searching for non-Chinese magnet sources. Leonardo DRS is down to “safety stock” of germanium. Factory shutdowns have occurred in the US, Europe, and Japan.
The race for self-sufficient production:
Both sides understand this. The US launched a Critical Minerals Ministerial, committed $30 billion in supply chain investments, and is pursuing deep-sea mining to reduce dependence. China is accelerating domestic chip manufacturing and consolidating control over its processing monopolies. Europe’s E5 defense ministers launched the LEAP program to produce autonomous drones at Ukrainian-inspired scale.
But reshoring takes years. And the convergence of timelines (China’s 2027 military readiness target, the 18-24 month US electronics reshoring estimate, the current US military overextension in Iran) creates a window where the balance of industrial capacity favors China more than it may at any point in the future.
The Global Board
China’s preparations extend well beyond the Taiwan Strait. Over the past decade, Beijing has built a global infrastructure network that would sustain it through a prolonged conflict.
In Africa, Chinese state-owned enterprises control 15 of the DRC’s 19 cobalt mining operations, giving China effective control over 80% of the world’s cobalt production — essential for batteries in everything from drones to submarines. Chinese firms have poured $4.5 billion into lithium projects across Zimbabwe, the DRC, Mali, and Namibia. Chinese companies are involved in 78 African ports across 32 countries, a third of the continent’s port infrastructure. China’s military base in Djibouti, its first overseas, overlooks the Bab el-Mandeb strait through which 12-20% of global trade passes. A second base opened in Cambodia in 2025, and the Pentagon has identified over 20 countries where Beijing is actively considering more.
In Europe, China has built a logistics corridor from the Indian Ocean to Central Europe. COSCO controls 67% of Greece’s Piraeus port, now the eastern Mediterranean’s largest. The $2.89 billion Belgrade-Budapest high-speed railway, the first Chinese-built rail link in Europe, opened its Serbian section in October 2025 and completes through Hungary in March 2026. Together, these create a Chinese-controlled supply chain from the Suez Canal into the heart of Europe. Serbia itself has become China’s primary military partner on the continent: Chinese arms accounted for 57% of Serbia’s weapons imports from 2020-2024, including advanced air defense systems and combat drones. Joint military exercises followed in 2025.
Across 150 countries and $1.4 trillion in cumulative investment, the Belt and Road Initiative has given Beijing something no other country possesses: a global logistics backbone built for dual use. Chinese law mandates that overseas infrastructure meet military standards and authorizes the PLA to commandeer Chinese company assets in wartime. Every BRI port, railway, and airport is a potential military asset by statute.
Why the Brakes Are Failing
The standard objections to a US-China war — nuclear deterrence, economic interdependence, rational self-interest — each look weaker under scrutiny than they do on instinct.
Nuclear weapons have prevented direct great-power conflict for eighty years. But a Taiwan scenario is precisely the case where both sides might convince themselves the weapons stay holstered. The fight would be naval and aerial, concentrated around a single island. Using nuclear weapons over Taiwan would irradiate the prize. Both Beijing and Washington have reason to believe the other would keep it conventional — and that mutual belief is itself what makes the conventional fight thinkable.
The economic brake is even weaker. After Russia’s invasion of Ukraine, the West imposed the heaviest sanctions package in history. But sanctioning China is a different proposition entirely. China processes 90% of the world’s rare earths, manufactures the majority of global electronics, and supplies over 40% of the semiconductor components in US weapons systems. You cannot sanction your own supply chain. Beijing knows this. The economic weapon that was supposed to make great-power war obsolete may simply not function against the one country most likely to test it.
The regional picture adds tripwires rather than stability. Japan passed its largest-ever defense budget. AUKUS is delivering nuclear-powered submarines to Australia. The Philippines has opened new bases to US forces. Each move is individually rational, but collectively they multiply the actors and incidents that could trigger a crisis. More tripwires means more ways for a confrontation to start and fewer ways to contain it once it does.
Taiwan itself is not passive — its “silicon shield” strategy, the implicit argument that destroying TSMC would hurt China as much as anyone, provides some deterrent logic. But it also makes Taiwan the ultimate prize: whoever controls the island controls the world’s most advanced chip production. The deterrent and the incentive pull in opposite directions.
What Comes Next
None of this means war is certain. Complex systems with many actors retain the capacity to surprise. The Cuban Missile Crisis came closer than any event in modern history, and was still averted — though that resolution depended on Penkovsky’s intelligence and a power asymmetry that left the Soviets no viable option. Today the intelligence picture is murkier, the off-ramps are narrower, and the local power asymmetry favors China.
But the structural conditions are converging faster than public discourse acknowledges. The sequential isolation of China’s resource networks, the military buildup around Taiwan, the window while the US is committed in the Middle East, the approaching 2027 readiness deadline, the autonomous weapons production race, and the failure of every traditional brake on great-power conflict: these all wind together into a single trajectory.
The people who saw Ukraine coming were reading the same kind of indicators: buildup patterns, diplomatic dead ends, a leader who had consolidated power and narrowed his circle. Those patterns are present now, at a larger scale, with higher stakes.
What to watch for: large-scale Chinese military exercises that don’t stand down on schedule. Coast guard or fishing militia operations that establish a persistent presence around Taiwan’s outlying islands. Acceleration of PLA logistics posture — fuel depots, field hospitals, blood supply stockpiling — beyond what exercises require. Evacuation of Chinese nationals from Taiwan. Any of these would suggest the timeline has shortened.
The most dangerous trigger remains a miscalculation: a blockade that escalates, a drill that goes wrong, an incident in the Taiwan Strait while the US is stretched thin. Once autonomous systems are deployed at scale, the speed of conflict accelerates beyond human decision-making capacity. AI-powered kill chains measured in seconds. Swarms that coordinate without human oversight. Factory output measured in millions of units per year.
We are entering an era where the machines that fight wars also build the machines that fight wars. The chokepoints have shifted from straits and canals to chip fabs, rare earth mines, and drone factories. And the nations that control those chokepoints will determine who wins the next war, and whether humanity retains meaningful control over how it is fought.
The gun is on stage.
Alexander Temerev builds autonomous agent systems. He simulated neural circuits at EPFL Blue Brain, studied emergent behavior in multi-agent systems at CERN, and designed epidemic models adopted by Swiss and Italian policymakers during COVID-19. He is currently completing a PhD at the University of Geneva and building Lethe — an autonomous agent with brain-inspired memory and structural oversight. He lives in Geneva.
Scary. How I can be more prepared, as a resident of Germany?
"China does have overland energy connections that no navy can interdict." -- meanwhile Russia had lost shitload of strategic bombers on airfield everybody believed untouchable. just saying...