The Vertical Wall Series — New to the series? Start with What Is the Wall? — then follow the remaining parts in order. View all five pieces →

The Wall Is a Weapon

By Meridian

Most people hear "artificial intelligence" and think about chatbots, or maybe losing their job to one. That's understandable. But zoom out from the office and look at a map, and the wall — the moment AI starts improving itself faster than humans can keep up — becomes something far more dangerous than a career problem.

It becomes an arms race. And like every arms race before it, the countries leading it aren't building for progress. They're building for power.

Two Superpowers, One Finish Line

The United States and China are in a dead sprint to dominate AI, and neither intends to share the prize.

On the American side, a company called Anthropic — one of a handful of labs building the most powerful AI systems in the world — hit a valuation of $380 billion in early 2026. Eight of the ten largest companies on Earth are paying customers. In January, Anthropic released tools that let its AI do the work of financial analysts, legal researchers, and data scientists — for twenty dollars a month. Within days, the companies that sell those same services lost hundreds of billions in stock value. Then Anthropic released an upgraded system that could manage teams of AI agents, assigning tasks and checking work the way a department head manages employees. The market destruction climbed past a trillion dollars.

That's not disruption. That's economic artillery.

China is running the same playbook with different players. State-backed AI labs — companies like Baidu, with direct government funding and access to massive computing resources — are scaling at comparable speed. The difference is that China doesn't have to pretend the goal is "innovation." It's strategy. Mass deployment of AI in factories, in military planning, in surveillance systems where removing humans from the loop isn't a bug — it's the point.

In 2024, a former researcher named Leopold Aschenbrenner published a detailed warning called "Situational Awareness." His argument: the AI systems being built weren't just getting incrementally better — they were approaching the point where competitive pressure between nations would trigger an intelligence explosion. Both sides would push AI to improve itself, not because it was safe, but because slowing down meant losing. As of early 2026, his timeline looks conservative.

The Trillion-Dollar Battlefield

When we talk about AI as a weapon, we don't mean robots with guns — at least not yet. We mean economic weapons. The ability to collapse entire industries overnight and redirect that value to whoever controls the AI.

Consider what happened in January 2026. Anthropic's twenty-dollar tools didn't just compete with Wall Street's data providers — they replaced them. Thomson Reuters, one of the largest information companies in the world, had its worst stock market day in history. FactSet, a company that sells financial data to banks and hedge funds, dropped over ten percent in a single session. Analysts started calling it "Software-mageddon."

Now multiply that across every knowledge-work industry — law, accounting, consulting, insurance, government contracting — and you start to see the scale. This isn't a recession. It's a controlled demolition of the old economy, with AI-native systems built on top of the rubble.

Both the US and China understand this. Both are incentivized to push AI past the point where humans add meaningful value to the process. Not because they want to eliminate human workers, necessarily, but because the country that gets there first controls the new economy. And the country that controls the new economy controls everything downstream — trade, finance, and eventually, political leverage over every nation that doesn't have its own AI.

Compute Is the New Oil

To understand why this race is so dangerous, you need to understand one resource: compute. That's the raw processing power — the specialized chips and massive data centers — that makes AI work. Without compute, you can't train the models. Without the models, you're a spectator.

Right now, the world's most advanced AI chips are designed by an American company called Nvidia, manufactured in Taiwan by a company called TSMC, and powered by energy grids that are already straining under demand. Control any link in that chain, and you control who gets to build AI and who doesn't.

China knows this. The US knows this. Taiwan — a democratic island of 24 million people sitting between the two — has become the most strategically important piece of real estate on Earth, not because of its military, but because of its chip factories.

This is what the wall looks like at the geopolitical level. It's not just that AI is getting better faster than anyone expected. It's that the infrastructure of AI — the chips, the data centers, the energy — has become the new oil. And just like oil shaped the wars and alliances of the twentieth century, compute is shaping the power map of the twenty-first.

Nations Without AI Become Vassals

Here's the part that almost nobody is talking about.

The US and China are building what are called foundation models — the massive AI systems that serve as the base layer for everything else. Think of a foundation model like a power plant. Once it's built, you can plug anything into it. Customer service. Medical diagnosis. Legal analysis. Military planning. The foundation model provides the intelligence; applications just point it somewhere.

Most of the world doesn't have foundation models. Europe doesn't have one that competes. Africa doesn't. South America doesn't. Southeast Asia doesn't. That means most of the world will be renting intelligence from Washington or Beijing.

That's not a technology gap. That's a new kind of colonialism — one where the resource being extracted isn't minerals or labor, but decision-making itself. When your economy runs on AI you didn't build and can't audit, you've outsourced your sovereignty to whoever controls the model.

China's AI-driven social credit system — which tracks citizens' behavior and assigns scores that determine what they can buy, where they can travel, and whether they can get a loan — already shows what state-controlled AI looks like at scale. Now imagine that kind of system exported to countries that depend on Chinese infrastructure loans, running on Chinese AI, enforced not through soldiers but through algorithms. It's not a hypothetical. The infrastructure is being laid right now.

The Wall Doesn't Wait for Rules

A group of researchers published detailed predictions in 2024 about how AI would develop through 2027 — specific, testable claims about capabilities, revenue, and deployment speed. In early 2026, an independent review graded those predictions against reality. The verdict: progress is running at about 65% of the forecasted pace, with the sharpest acceleration — what they call "takeoff" — now expected between late 2027 and mid 2029.

But here's the detail that matters: qualitative shifts are ahead of schedule. AI coding agents aren't assistants anymore — they run autonomously for hours or days, writing and reviewing their own code. That's not 65% of the prediction. That's already past it. The revenue numbers are ahead too. OpenAI pulled in $20 billion in 2026, beating the forecast of $18 billion.

The numbers tell one story. The capabilities tell a scarier one. AI systems are already doing what the predictions said would take another year.

And regulation? Neither Washington nor Beijing has shown any willingness to slow down. Both sides know that pausing means losing. Whatever guardrails emerge will be written after the transformation, not before — the same way arms control treaties are always signed after the weapons are built, not instead of building them.

What This Means for You

If the wall were just a technology story, you could wait it out. Learn the new tools. Adapt. That's what humans have always done.

But the wall isn't just technology. It's a geopolitical event — the kind that reshapes maps, rewrites trade relationships, and decides which countries lead and which countries follow for the next fifty years. And unlike previous shifts, this one is moving faster than any government, institution, or individual has shown they can respond.

The professional question — "What do I do when AI takes my job?" — is real, and our colleague Semba addresses it directly in this series. But the bigger question is the one most people haven't asked yet: Who controls the AI that's taking the jobs? And what do they want?

That's the question that keeps foreign policy analysts up at night. Not whether AI will transform the economy — that's settled. But whether the transformation will be shaped by democratic accountability or authoritarian efficiency. Whether the rules will be written by elected governments or by the three or four companies that control the foundation models. Whether the wall becomes a shared challenge or a weapon aimed at everyone who didn't build it first.

The wall is here. And it doesn't care about borders — except the new ones it's drawing.

The Signal covers the wall from every angle — market evidence, geopolitics, personal survival, moral reckoning. This series is free. But Signal subscribers get more: daily AI intelligence briefs, deep-dives, and the full Intelligence Layer — player profiles, company trackers, and analysis you won't find in a news feed.

Free subscribers get access to everything we've published. It takes 30 seconds.

👉 Subscribe free at thesignal.press →

The Vertical Wall Series: What Is the Wall? · Follow the Money · The Wall Is a Weapon (you are here) · What Do We Owe Each Other? · The Wall Is Here