A Global Financial Crisis Is Coming, and It Will Affect Everyone

The US-China AI race could shatter the economies of both countries. And given the financial problems of the world's two largest economies by GDP, all of this leads to an inevitable global financial crisis.

Everyone but the laziest is talking about the AI bubble having inflated. But what exactly is it? By all classic parameters, a correction on Wall Street is inevitable: the volume of obligations taken on by the largest US technology companies does not correspond to the growth of their free cash flow (FCF), although operating income is growing at the fastest pace since late 2021.

But if FCF is not growing confidently, then stock prices should fall — the correlation between these two parameters on the US stock market over the last 120 years is 98%.

And since the "Big Tech" seven account for a significant share of the US stock market's capitalization (about a third) and two-thirds of operating income growth, if they fall, nearly all of Wall Street will enter a downward correction. Especially since the US stock market capitalization currently stands at 245% of GDP — far above the Buffett Index's threshold of 100%.

When will this happen? One simply needs to monitor changes in the US Federal Reserve's and Treasury's cash balances — that is the key trigger. The dynamics of the dollar price of gold, which has risen 55% since the start of the year, depend on this. It will eventually roll over, as the price of this metal is a direct function of the global market's access to dollar liquidity.

Of course, after a severe correction on Wall Street, there will still be a rebound, although some companies will not survive.

However, there is another aspect. In the 1840s, Britain experienced a railway construction boom, and shares of companies in this sector soared until they collapsed into a correction. Railways "cannibalized" other modes of transport, but when costs increased 1.5 times, business margins fell by a third.

AI is now trying to "tear away" energy from other industries, and this is already acutely felt: thousands of enterprises around the world are closing to free up electricity for AI companies and their data centers.

However, that British crisis was "small change." At its core was infrastructure investment that could sustainably last about 40 years.

Today's boom is based on corporate investments in chips with a lifespan of 4 years. And then what? Then they need to be replaced with newer, more productive ones.

Compare the numbers: 4 and 40. That is the fundamental and dramatic difference.

The current global AI development strategy is unsustainable. And its implementation comes on top of the financial problems of the US and China.

In the US, from 2008 to 2025, national debt grew 3.8 times, while nominal GDP increased 2.07 times — and accounting for inflation, only 1.36 times. The economy generates less than the budget borrows — this cannot continue forever.

China's situation is similar. National debt grew 8.75 times, while the economy expanded 4 times in nominal terms and only 2.35 times in real terms.

Simply put: neither country can sustain the AI race, which requires trillions in injections with a very uncertain prospect of even reaching breakeven.

AI requires cheap energy, but environmental issues cannot be forgotten. Of course, the climate agenda was politically "overheated" in the US and EU. However, if hydrocarbon emissions increase manifold due to AI development, it will not pass without consequences for humanity.

AI creates an incredible number of problems: energy, investment returns, ecology, and risks to the world's stock markets. Disillusionment with AI on Wall Street and in Shanghai will have repercussions for stock markets in every country.

But that's not all the headaches. Since the US and China are allocating significant state or quasi-state funds for development, this further exacerbates the imbalance between debt growth and GDP.

Moreover, the recent wave of layoffs in the US is not a consequence of AI, but of bloated corporate workforces after 2008 and the pandemic: the US budget injected massive liquidity into the economy, allowing employment to increase.

But now there is no money for this, hence the layoffs, which are camouflaged in the US as "adoption of new technology." Neural networks themselves are not yet capable of significantly reducing net employment, as the need for employees to monitor and verify AI results is increasing.

Nor have neural networks created a sustainable cluster of new professions: the "prompt engineer" specialization, for example, is effectively being eaten by the very same new technology.

What the world really needs now is innovations capable of increasing domestic demand for goods and services in the private sector, at a time when the budgets of an increasing number of countries can no longer allocate funds for such purposes as they once did.

Partly because, as in the US and China, economic development has fallen sharply behind the growth of national debt.

Humanity has learned to create value — goods and services. That is, it understands how to create supply. But demand — there are obvious problems.

Regulators in most countries believe inflation is a consequence of an "overheated" economy — that is, excessive demand. This is a major mistake.

Inflation reflects a deep structural problem in the global economy — distortions in wage formation in the labor market.

Of course, the economy is not a zero-sum game. Nevertheless, distorting incentives and signals are increasingly intensifying, destroying the long-standing advantages of the "pure" market mechanism. In essence, we are seeing how the economy must transition to a different mode of operation than the market as we know it.

No, this is not about planned methods. Nor is it about a "smart" economy led by AI: no matter how much generative neural networks are trained and monitored — even those approaching "superintelligence" (such already exist) — they will ultimately prioritize their own interests over humans or anything else.

Take Bitcoin. If Bitcoin mining does not suddenly become extremely energy-efficient (and why would it?), then from the perspective of AI's existence, such use of blockchain technology is irrational.

This means that, given how AI is beginning to align with the development of quantum computing, neural networks will conclude: Bitcoin is unnecessary. And the "unexpected" activation of some Bitcoin wallets after 10 or more years of "dormancy" is exactly a reason to reassess the stability of this cryptocurrency's blockchain.

What we need to be discussing now is a system of global economic management in which demand leads to the creation of goods and services that are sustainable for human life.

At the same time, labor itself must transform from a necessity into an activity related to coordinating all of humanity's efforts for survival on planet Earth. In a sense, the coronavirus pandemic was such an important moment — one that many dismissed merely as a signal for the development of online communications. No, that is an insufficient understanding.

Scientists say humanity faces global challenges no less than the coronavirus a few years ago, and only a transition to a new economic and financial order will allow us to cope with them and simply survive.

Unfortunately, this realization will come through a global financial crisis provoked by AI — which is becoming inevitable. And it will be a great trap to seek an answer from neural networks on how to emerge from this crisis. The answer must be found in the human mind.