Marx explained the AI bubble as far back as 150 years ago

When OpenAI CEO Sam Altman said in San Francisco this year that a bubble had formed around artificial intelligence, the US tech market reacted instantly.

With up to 95 per cent of AI pilot projects failing, investors took Altman's words as a signal of overheating. Although he was speaking mainly about private startups, many took it as a diagnosis for the entire industry.

Billionaire Peter Thiel sold off Nvidia shares, investor Michael Barry (known for The Big Short story) bet large sums against companies like Palantir and Nvidia.

Writing in The Conversation, researcher Elliot Goodell Ugalde argues: it's not just about the vulnerability of individual companies. What is happening to the AI market was well described by Karl Marx almost 150 years ago - through the concept of surplus capital, which no longer finds profitable application in real production.

How Marx described crises

The future of AI technology itself is not in question - just as the Internet survived the dot-com crash, AI will remain. What is in question is something else: where capital will flow when AI companies' stocks stop making their current speculative profits.

Marx explained crises by the overaccumulation of capital:
when there is too much accumulated money, but it is no longer possible to invest it in production in a way that generates new surplus value.

If additional investment does not create new "real" wealth, capital begins to flow into speculation - into assets whose growth is disconnected from the real output of goods and services.

The growth of AI masks the weakness of the real economy

Years of ultra-low interest rates and "cheap money" during the pandemic inflated corporate balance sheets. Much of this liquidity has gone into the technology sector and has been concentrated in the so-called "magnificent seven" - Amazon, Alphabet, Meta, Apple, Microsoft, Nvidia and Tesla.

Without these giants, the author emphasises, the overall market dynamics would be negative. That is, we see not so much a "flourishing of technology" as a concentration of capital in a narrow group of overvalued assets.

In Marxian terms, this looks like "money thrown into circulation without a material basis in production": value rises not because more goods are created, but because the price of paper rises.

The "real economy" suffers from this:

• less investment in production and infrastructure,

• stagnation and cost-of-living crises grow,

• but for GDP, the picture is not so dramatic - the bubble in technology "pulls up" the formal indicators.

AI as a temporary "patch" for excess capital

Economic geographer and Marxist theorist David Harvey developed Marx's ideas into the concept of the "spatio-temporal fix " - a temporary patch for the system:

• capital pulls the problem into the future (long-term projects, promises of future profits),

• or takes it to new geographical areas (new markets, resources, labour).

The overabundance of money, capacity and labour, which cannot be used without losses, is diverted to:

• long-term projects where profits are promised "someday later" (in fact - fictitious capital),

• new territories - data centres, chip factories, mining minerals for electronics.

The AI boom, the author emphasises, works as both a temporal and spatial patch:

• promises future super-profits that may never materialise;

• requires huge investments in hardware and infrastructure, which temporarily "soaks up" excess capital.

But, as Altman himself admits, the limits are already visible. And US President Donald Trump's protectionist measures, which complicate global trade, are only increasing the pressure on these channels.

The price of speculation: who pays for the bubble

Marx warned: the overproduction of capital also means the overproduction of means of production and the goods of life, which cannot be sold at a profit at the current purchasing power of the population.

When profitability falls, the system is "levelled" by:

• job destruction,

• losses for households whose savings are tied up in equity markets.

History provides examples:

• the dot-com crash swept away many small investors and consolidated the power of the surviving giants;

• the 2008 crisis deprived millions of people of their homes, while key financial institutions were rescued by the state.

Today, major asset managers are already preparing for new shocks: Vanguard, for example, is shifting noticeably toward bonds and "defensive" instruments.

When speculation crowds out "normal" growth

Marxist Rosa Luxemburg asked herself at the beginning of the 20th century where the ever-increasing demand necessary for the expanded reproduction of capital would come from.

The answer, echoing Marx and Harvey:
when normal productive avenues narrow, capital goes either outward (into new geographies) or into speculation. The US is increasingly betting on the latter.

Corporate spending on AI infrastructure now contributes more to GDP growth than consumer spending by households. This is a historically unusual situation: economic growth is increasingly tied to speculative investment rather than to the expansion of real output.

Thus the overall rate of return is falling, and when the flow of speculative money reverses, recession will inevitably follow.

Tariffs, restrictions and capital lock-in

The classic ways of "relieving pressure" - moving production to new countries, expanding markets - work worse.

• Tariffs,

• export restrictions on chips,

• mirror trade measures from other countries

- all narrow the global space for expansion.

Capital, unable to find a way out, increasingly:

• shifting debt forwards,

• inflating asset values,

• relies on financial instruments that temporarily mask losses but make the system more fragile.

Fed Chairman Jerome Powell, by announcing his willingness to cut rates, effectively opens the way to a new cycle of cheap credit. This allows:

• "to flood the problems with money,

• create another round of speculation,

• but it also builds up household and corporate debt.

Marx described this as the logic of interest-bearing capital: financial instruments create claims to future profits over and above what can actually be produced.

What happens if the AI bubble bursts

If the bubble around AI collapses at a time when:

• states have little ability to divert investment overseas,

• the economy is already on fragile credit,

the consequences could be severe.

Capital won't disappear - it will flow:

• into debt markets,

• credit instruments backed by the Fed through low interest rates.

This will not cancel the crisis, but will only shift its burden down the social ladder: to workers, borrowers, pension funds.

In this sense, the author emphasises, the bubble around AI is not a unique anomaly, but another mechanism for how the system is trying to soak up excess capital.

If protectionism continues to block spatial outputs, and "solutions over time" rely on increasingly risky credit and financial leverage, we will be moving through a cycle:

asset inflation → collapse → government intervention → new bubble.

Artificial intelligence itself will survive and grow. But the speculative bubble around it is just a symptom of a deeper structural problem. And, as is almost always the case in such stories, the main price for its bursting, according to Marx, will fall on the shoulders of the working class.