The Reset Is Coming. But This Time It’s Different

Three of the most powerful people on the planet stood up last Friday and said the same thing. At the Munich Security Conference, German Chancellor Friedrich Merz told the room that the post-1945 world order “no longer exists.” Macron said Europe needs to become “a geopolitical power.” Essentially admitting the old security architecture is finished. And U.S. Secretary of State Marco Rubio, representing the country that built the bloody system, said “the old world is gone.” Not hedged. Not diplomatic. Gone. When the people who designed the plumbing tell you the pipes have burst, you don’t stand there debating water pressure. You start looking for higher ground.

Dalio Goes Viral (And LinkedIn Loses Its Mind)

Ray Dalio’s piece on the back of Munich has been doing the rounds. And predictably, half of LinkedIn has reposted it with a “🔥 THIS” and zero additional insight. (You knew there’d be one of those paragraphs. Sorry.)

Right. Before anyone mistakes this for another breathless “the sky is falling” post from a bloke with a newsletter, let me be clear about what Dalio actually said. Because most of the takes I’ve seen have butchered it. His argument is built around something he calls the “Big Cycle.” Stage 6, specifically. The disorder phase. Rising powers challenge incumbents. Debt piles up. Internal divisions widen. Economic wars escalate. Eventually the whole thing resets and a new order crawls out of the wreckage. In his model, the last big reset was 1945. The historical parallels he draws between the 1930s and today are genuinely unsettling. Protectionism. Populism. Currency wars. The slow escalation from trade conflict to actual shooting. If you haven’t read Chapter 6 of Principles for Dealing with the Changing World Order, it’s worth your evening. Not because Dalio has all the answers. But because the pattern recognition is bloody hard to argue with.

Here’s the thing, though. This idea isn’t new. Not remotely. Thucydides wrote about the same dynamics 2,400 years ago. Paul Kennedy mapped it across five centuries in The Rise and Fall of the Great Powers. Dalio’s financial lens is genuinely useful. He’s good at the money bit, as you’d expect from a man who runs the world’s largest hedge fund. But the broader thesis of cyclical rise and fall? That’s been around since before anyone had a LinkedIn profile to repost it on.

Where I agree with him: we are closer to a reset than most people want to admit. Where I part company: what this reset actually looks like. Because Dalio’s model has a blind spot. A big one.

The Blind Spot (Or: What Happens When the Cycle Itself Changes Speed)

Dalio’s framework accounts for debt, military capability, trade routes, demographics. All of which still matter. But it treats technology as a feature of the cycle rather than a force that changes the cycle itself. That’s like modelling traffic patterns and ignoring the fact that half the cars just became self-driving.

Previous resets played out over decades. Sometimes generations. This one has three forces converging simultaneously that no previous empire has faced. And the fact that they’re hitting at the same time is what makes this fundamentally different from anything Dalio describes.

Force One: AI (The Thing That Changes the Speed of Everything Else)

Right. Let’s not mess about. AI is not “another tech wave.” It’s not the internet circa 1997. It’s not mobile. It’s not even cloud. It’s a compression variable. It accelerates the speed at which markets react, labour displaces, capital moves, and political narratives shift.

In previous cycles, the world had decades to absorb disruption. Institutions moved at geological timescales. Labour markets adjusted over generations. Political coalitions took years to fracture and reform. AI collapses that timeline. And not theoretically. Right now.

Microsoft laid off 15,000 workers in 2025. Its CEO, Satya Nadella, openly told investors that AI now writes 20-30% of the company’s code. Meta’s Zuckerberg said AI could function as a “mid-level engineer.” Anthropic’s Dario Amodei, the man building one of the most powerful AI systems on the planet, predicted it could wipe out half of all entry-level white-collar jobs within five years. Ford, Amazon, Salesforce, JP Morgan. All saying some version of the same thing in slightly more polished language.

In 2024, over 150,000 tech workers were laid off across more than 525 companies. In 2025, the pace accelerated. An 80% increase in the first half compared to the same period prior, according to Challenger, Gray & Christmas, who’ve been tracking this stuff since 1989. And here’s the number that should properly keep you up at night. The Burning Glass Institute found that entry-level job postings in AI-exposed fields (software development, data analysis, consulting) dropped from 43% to 28% between 2018 and 2024. Not because companies were hiring fewer people overall. Because they were skipping new graduates entirely and hiring experienced workers who needed no training.

The white-collar immunity myth? Dead. Buried. Flowers on the grave. It’s not just factory work and call centres anymore. It’s analysts, designers, recruiters, project managers, junior developers. The middle layers of the modern economy. Shopify now requires managers to prove AI can’t do a job before they’re allowed to post a vacancy for it. Let that sink in for a second.

Force Two: Peak Debt (Or: Why Nobody Has Any Room to Manoeuvre)

We are entering the disorder phase carrying more financial baggage than any previous generation of empires. And somehow, nobody in a position of power seems particularly bothered by this. Which is, frankly, bonkers.

The numbers are stark. Global debt, public and private, now sits above 235% of world GDP according to the IMF’s September 2025 Global Debt Monitor. Sovereign bond issuance in OECD countries hit a record $17 trillion in 2025. Nearly double the 2007 level. Government interest payments across the OECD reached 3.3% of GDP in 2024. Just to be clear, that is more than most countries spend on defence. Governments are now spending more to service their debt than to protect their citizens. And it got there fast. Interest costs went from their lowest point in 20 years to their highest in just three years, between 2021 and 2024. That’s not a slow bleed. That’s a structural shift happening at a speed that central banks are still pretending they’re in control of.

The 1945 reset, the one Dalio keeps referencing, began with young populations, high growth potential, industrial expansion, and relatively clean balance sheets. Today? Ageing societies. Slower growth. Political paralysis. High debt. Rising interest costs. And precisely zero political appetite in any major economy to cut spending or raise taxes. Not exactly the starting conditions you’d choose if you were designing a resilient system.

High leverage doesn’t just reduce policy flexibility. It makes shocks contagious. When debt loads are this high, a crisis in one sector doesn’t stay politely contained. It cascades. Fast.

Force Three: Demographics (The Slowest Variable and the Most Brutal)

Demographics are the thing nobody wants to talk about at conferences because the charts are depressing and there’s no quick fix to announce. China’s population has fallen for three consecutive years. It peaked at 1.4 billion in 2022. The UN projects it’ll drop to 1.26 billion by 2050, with 40% of the population over 60. By 2100? 633 million. The Washington Post called it population decline “on a scale and at a speed the world has never seen.”

The U.S. Census Bureau projections are even more dramatic. Axios reported a projected fall to 662 million by end of century. The steepest decline of any nation in recorded demographic history. And China isn’t alone. Europe is ageing. Japan has been stagnating for decades. Russia, South Korea, Italy, Spain. All facing what demographers are calling structural population implosion. The U.S. remains comparatively stronger, but it’s deeply polarised and running its own fiscal deficits that would make your accountant weep.

Oxford Economics estimates China’s shrinking workforce alone could shave 0.5% off annual GDP growth over the next decade. Fewer workers. Lower productivity. Lower consumption. Mounting pressure on pension and healthcare systems that are already underfunded. And China is ageing twice as fast as Japan did. Its over-65 share went from 7% to 15% in just ten years.

Now here’s where it gets properly interesting. And where Dalio’s model really misses something. Ageing societies don’t go to war. Not like they used to. You don’t send your only child into a conflict. You don’t sustain a decades-long military campaign with a workforce that’s shrinking by millions annually. The demographic conditions that made total war thinkable in the 1940s (young populations, high birth rates, nations with millions of men to spare) simply don’t exist in any of the major powers anymore.

So demographics simultaneously increase economic fragility and reduce the appetite for total war. That’s a paradox Dalio doesn’t spend nearly enough time on. And it’s arguably the single biggest reason this cycle won’t replay 1945.

The Taiwan Question (The Bit Everyone’s Thinking About)

You can’t take Dalio’s thesis seriously without addressing Taiwan. He identifies it as the most explosive flashpoint in global geopolitics. He’s right. The Pentagon’s 2025 China Military Power Report, released quietly on December 23rd (because nothing says “happy Christmas” like a 200-page document on the PLA’s invasion capability), stated flatly that China expects to be able to fight and win a war over Taiwan by the end of 2027.

The PLA has been running large-scale blockade exercises, live-fire drills and simulated island encirclement with increasing frequency. The U.S. responded with its largest-ever arms sale to Taiwan. $11.1 billion in December 2025. Eighty-two HIMARS, 420 ATACMS missiles, drones, Javelins, the works. The Pentagon also noted that China’s commercial AI sector has narrowed the gap with U.S. models, and that Beijing is integrating AI and cyber capabilities into military systems. Not eventually. Now.

So does AI make a Taiwan conflict more or less likely? Honestly? Both. AI increases the capability gap in surveillance, autonomous weapons, and decision-making speed. That could embolden whichever side thinks they’re ahead. But it also raises the stakes of miscalculation astronomically. An AI-assisted military response that fires in seconds rather than hours changes the calculus of deterrence entirely.

And then there’s the semiconductor dimension. Taiwan produces the majority of the world’s advanced chips. One analysis estimated a conflict would set global microprocessor production back to early 2000s levels for up to 15 years. My read: the Taiwan risk is real, but the shape of the confrontation is more likely to be economic coercion, sustained blockade, and grey-zone pressure than full-scale invasion. The demographics don’t support a 1940s-style campaign. Neither do the economics. China can’t afford to level the island that makes most of the world’s chips, and it can’t afford a prolonged occupation with a population that’s contracting every year. That doesn’t mean it’s safe. It means the danger looks different from what Dalio implies.

Two Futures (And Why I Think We Get Both)

When you stack AI, debt and demographics on top of each other, two very different scenarios emerge.

The Slow Bleed. AI improves forecasting. Governments detect risks earlier. Demographics temper aggression. Instead of collapse, you get stagnation. Instead of war, managed competition. Instead of reset, a prolonged plateau. Decades of lower growth, tighter borders, more surveillance, brittle politics. Not Hollywood. Just grey and grinding.

The Compressed Reset. AI accelerates labour displacement faster than institutions can adapt. Debt limits fiscal manoeuvrability. Political systems fracture under economic strain. Capital weaponises. Supply chains reconfigure. Instead of 20 years of gradual decline, you get 3-7 years of sharp correction. I lean toward the compressed version. Here’s why I anchor at 3-7 years specifically. Because I’ve had enough LinkedIn prophets throwing timelines around with nothing behind them.

Enterprise AI adoption curves are running at roughly 18-24 months from pilot to full deployment. That’s dramatically faster than previous technology waves, which typically ran 5-10 years. Corporate restructuring cycles that used to play out over years are now measured in quarters. And the sovereign debt refinancing wall (the point at which governments must refinance pandemic-era debt at today’s much higher interest rates) peaks between 2026 and 2028 across most OECD economies. That convergence of technological disruption, corporate restructuring, and fiscal stress hitting the same window is what creates the compression.

The pain will be real. But the same forces that make the disruption sharp will also make the recovery faster. This isn’t a 20-year depression. It’s a restructuring cycle. And that distinction matters enormously.

So What Do You Actually Do? (The Bit Without the Bollocks)

Right. Before this turns into one of those “10 steps to thrive in the new world order” posts that LinkedIn is absolutely drowning in, I’ll keep this short.

For individuals: AI literacy is no longer a nice-to-have. It’s baseline professional competence, like knowing how to use email was in 2005. Stack skills across technical and human dimensions. Build income diversity. Invest in the things AI still can’t do. Strategic thinking. Systems-level judgement. Leading humans through uncertainty. Competing against AI is a losing strategy. The only play that works is being the person who knows how to use it better than the person sat next to you.

For businesses: adopt AI early. Scenario-plan for multiple outcomes, not just the comfortable one. Strengthen balance sheets now, while capital’s still available. Shorten decision cycles. The winners over the next five years won’t be the biggest companies. They’ll be the ones that adapted fastest. That’s it. That’s the whole thing.

The Bigger Picture (No, Really)

Dalio is right about cycles. But this may be the first reset unfolding under the simultaneous pressure of artificial intelligence, record global leverage, and demographic stagnation. No previous empire faced all three at once. None of them had to contend with disruption that moves at machine speed. Previous empires declined slowly. This one might restructure quickly. That doesn’t mean collapse. It means transition. And transitions, if you’re paying attention, are where the real opportunities sit.

The next five years will test resilience like nothing in recent memory. Your move.

If you want real insights, zero hype, and none of that inspirational LinkedIn nonsense, The Sunday Blueprint lands in your inbox every week.

P.S. If you’re reading this and thinking “but Dalio’s a billionaire and you’re a bloke with a Substack,” fair point. But I’d remind you that Lehman Brothers had a lot of billionaires too. Pattern recognition isn’t a function of net worth. It’s a function of paying attention.

Sources I used for this blog (Correct as of 16th Feb 2026):

• Ray Dalio’s original post (Feb 14, 2026) on X/Twitter via BusinessToday coverage (Feb 16, 2026)
• Ray Dalio, Principles for Dealing with the Changing World Order, Chapter 6: The Big Cycle of External Order and Disorder (book and LinkedIn series)
• Munich Security Report 2026: “Under Destruction” (Executive Summary)
• CNN: Rubio and European leaders agree the old world order “no longer exists” (Feb 13, 2026)
• Euronews: Merz warns Munich Security Conference freedom “is no longer a given” (Feb 13, 2026)
• TIME: Germany Issues Warning to United States Amid “Deep Rift” (Feb 13, 2026)
• PBS: European leaders warn of urgent need for self-defence (Feb 13, 2026)
• France 24: Macron tells Munich conference Europe must become geopolitical power (Feb 13, 2026)
• IMF Global Debt Monitor 2025: Global debt remains above 235% of world GDP (Sep 2025)
• OECD Global Debt Report 2025: Record $17 trillion sovereign bond issuance (2025)
• Pentagon China Military Power Report 2025 (via USNI News) (Dec 23, 2025)
• DefenseScoop: Pentagon report notes China’s progress on LLMs (Dec 26, 2025)
• AEI: China & Taiwan Update, December 23, 2025. $11.1bn U.S. arms sale (Dec 23, 2025)
• AEI: China & Taiwan Update, January 2, 2026. Justice Mission 2025 exercises (Jan 2, 2026)
• Washington Post: China’s shrinking population could crimp global ambitions (Oct 2025)
• Axios: China on track for massive population decline (Dec 2025)
• Pew Research Center: Key facts about China’s declining population (updated Apr 2025)
• Fortune: Tech layoffs 2025. How Microsoft, Google, and Meta are plotting for the AI era (Jul 2025)
• CNBC: AI-washing and the massive layoffs hitting the economy (Nov 2025)
• HBR: Companies are laying off workers because of AI’s potential, not its performance (Jan 2026)
• SF Standard: Use AI or be fired. The tech industry push to shrink workforces (Jul 2025)
• InformationWeek: Major tech layoffs tracker (2024-2025) (updated Jan 2026)
• Burning Glass Institute / HR Executive: The AI layoff trap (Dec 2025)
• Visual Capitalist: Visualizing the State of World Debt in 2025 (Nov 2025)