At the recent CWI summer school in Berlin there was considerable discussion about the world economic crisis and the increasing instability in financial markets, the impact of tariffs, the Iran War, and the speculation mania surrounding the AI boom. The general consensus at the school was that while the timing could not be predicted, the accumulating contradictions including the AI bubble meant an economic crash was highly likely, leading to social and political upheaval and intensifying class conflict.
Marxism is a science of perspective. The last 40 years of capitalist development and crisis has proven the validity of Marx and Lenin’s analysis of the laws of capitalist development, and breakdown. One task of the revolutionary party is to warn the working class that the underlying trends in the economy eventually must lead to a deep crisis and likely collapse into crisis, and only the timing cannot be determined with mathematical precision. This article is an expanded version of a contribution to the discussion at the Berlin summer school.
The rise of finance capital to dominance under Volcker
In the 1980’s the US ruling class launched a counterrevolution, essentially, against all the economic and social gains made by the working class over the previous 5 decades since the New Deal, which has continued to this day.
Paul Volcker’s appointment as Chair of the Federal Reserve in 1979 signalled a major shift in the economic policies of the United States. The paramount concern of government policy became protecting and advancing the interests of capital rather than safeguarding high levels of employment and broad income growth generally.
Volcker forced up interest rates to 20% to drive out less profitable industry, especially in manufacturing and to secure the ascendency of finance capital through the policy mechanism of “monetarism”. The ‘ultimate goal’ of that policy was to eliminate inflation, which was then reaching over 14%, at the expense of high employment levels and living standards.
Industries of the American Midwest were devastated, unemployment surged above 10%. Volcker’s policies had the desired effect of profoundly weakening labour and strengthening the power of capital. As a side-effect of capital’s counterrevolution in the US, surging interest rates caused the under-developed economies to be plunged into crisis, skyrocketing debt and impoverishment ensued. IMF “restructuring” tightened imperialism’s economic stranglehold over the neocolonial world. What became known as “neo-liberalism” was born, and its first leading political representatives were Ronald Reagan in the US and Margaret Thatcher in Britain. Reagan’s election slogan was “Let’s make America great again”.
Neoliberalism’s onslaught on labour began with Reagan’s attack on the air traffic controllers’ union (PATCO) in 1981, when he fired 11,000 striking air traffic controllers. In Britain it began in earnest with Thatcher’s well prepared attack on the mineworkers union (NUM) in 1984.
1980’s Financialization in the US
The Reagan administration unleashed an aggressive wave of deregulation across major industries, airlines, telecommunications, media, banking and finance and others in a programme of radically reshaping the entire US economy to promote corporate consolidation and profit driven restructuring. The policies were specifically geared to promote monopolisation and financialization. Taxes for the wealthy were slashed. In the first eight years from 1981 to 1989 the top marginal tax rate was slashed from 70% to 28%. At the same time social services, most of which had been created in Roosevelt’s New Deal, were gutted in the name of “fiscal discipline”.
The most important deregulation in terms of the future trajectory of the economy was in the field of banking and finance. A swathe of regulations going back to the depression era were abolished and new laws enacted facilitating the unfettered activity and massive growth of financial institutions from 1982 to the turn of the century – and beyond. The ascendency of finance capital and the processes of monopolisation were promoted by Republican and Democratic administrations alike. Among the most significant Acts passed was on 12 November 1999 when President Bill Clinton signed the Gramm-Leach-Bliley Act which abolished the 1933 Glass-Steagall Act thereby removing the legal barrier to a single financial institution acting as a combination of a commercial bank, an investment bank and an insurance company. The removal of this major piece of New Deal legislation designed to curtail monopoly and financial recklessness, allowed the creation of massive financial conglomerates – and fundamentally changed the character of the financial markets by permitting deposit taking banks to engage in high-risk Wall St. trading.
The organic decline US industry faced through increasing competition from Japan and Germany through the 1970’s and the effects of the Volcker shock and financial deregulation through the 1980’s and 1990’s, led US capital into a process of escalating the “financialization” of the economy where corporate profit was increasingly found in non-productive financial activity including “leveraged buyouts”, mergers and acquisitions, general speculation and financial parasitism. The stock market, banks and investment houses became the central loci of the American economy rather than the factories, mines and mills and productive economy generally.
Manufacturing as a share of real US GDP fell from 25%-28% in 1960 to 11% in 2010. The percentage of the workforce employed in manufacturing has fallen from about 30% in 1960 to 9% in 2010.
The transformation in the character of the US economy as a result of deindustrialization and the neoliberal counterrevolution is reflected in financial sector profits. They surged from approximately 13% of total domestic corporate profits in 1985 to over 40% in 2002. These levels were lost following the 2008 financial crisis which eliminated a large portion of financial sector earnings, but they have since risen consistently and between 2016 and 2026 have averaged 25% of total corporate profits. Central to financialisaton was the development of new and complex financial products such as derivatives, swaps and margin lending, the value of these investments exploded into the tens of trillions, and all essentially linked to some form of parasitic financial activity.
1990’s
The triumphalism of capitalism following the collapse of the Soviet Union was hollow and short lived. The deepening problems of a lack of productive investment, the deindustrialisation and the inevitable instability of unfettered finance capital quickly led to crises. In 1997 the Asian financial crisis erupted, reverberating throughout the world, followed by the economic meltdown in Russia in 1998. Within the US a massive speculative bubble developed in the late 1990’s. In 1998 a major hedge fund Long Term Capital Management collapsed as a result of Russia’s default on its debt – threatening a global crisis requiring a Federal Reserve rescue. The cause of LTCM’s collapse was emblematic of the US economy’s new financial “architecture”; the hedge fund had huge borrowings controlling over $100 billion in assets and more than $1 trillion in derivatives with only $4.8 billion in equity.
The speculative mania which fuelled the dot com bubble was unprecedented since the speculation leading to the crash of 1929. Driven by the prospect of large profits, billions poured into companies associated with the development of the internet including companies involved in the infrastructure and thousands of companies (“startups”) with .com in their name – invariably ignoring traditional valuation metrics. When funding dried up and the surge collapsed the NASDAQ dropped 78% wiping 5 trillion off market valuations. Hundreds of companies went bust as well as major telecom and fibre-optic network providers like WorldCom and 360networks. While most failed a small number such as Amazon and eBay survived and became tech giants. A recession followed over the next 8 months, and unemployment continued to rise thereafter. The 1990’s crises and the 2000 dot com crash were precursors to the devastating slump of 2007 – 2008.
“Primitive Slumpism”?
The late Peter Taaffe, former General Secretary of the Socialist Party, in his lengthy work “From Militant to the Socialist Party” (Socialist Publications 2017) made a prescient analysis of the economic crises of the 1990’s and the dot com bubble. Based on that analysis, which had taken place contemporaneously with the crises, the Socialist Party and the Committee for a Workers’ International maintained that while capitalism had stabilised following the post dot com recession the deepening contradictions embedded in US and world capitalism meant that future crisis, breakdown and slump were inevitable. Even some self-proclaimed Marxists criticized this foresight as dogmatic “primitive slumpism”, overstating and exaggerating the prospect of economic collapse. In answer to those who insisted that the world economy was set to soar once more into a bright future of stability and prosperity Peter Taaffe responded;
“This is entirely false. The period ahead was likely to be a continuation of insecurity and uncertainty … Our conclusion was that lodged in the situation today is the possibility of another devastating slump like 1929.” Was such a prognosis “primitive slumpism”? asked Peter;
“On the contrary, the more farsighted capitalists were making the same points and moreover, this disaster came to fruition in the economic meltdown of 2007-2008. They avoided this temporarily once more through the massive bail out of the banks and the financial system. This meant that a highly unstable situation persisted which clearly in the future could produce another 1929 economic scenario and one that capitalism would not be able to so easily avoid. We also sketched out the likely economic scenarios to follow; A period of stagnation, of deflation, of an extended period of economic depression with only small anaemic growth in production and a growth in the social malignancy associated with this, of poverty, rising unemployment, …. We drew all the necessary political and social conclusions from this, which would be manifested in increased conflict between the classes.”
The AI boom and the new crisis
Ten years on from this analysis, and twenty years since the sub-prime collapse based on the housing bubble and financial parasitism on a whole new scale, the general global economic crisis and the AI bubble confronts the world once again with the prospect of a 1929 style crash, slump and social upheaval. The massive bailout undertaken in the 2008 global financial crisis and the government expenditure required to stabilise the economy during the pandemic, coupled with massive military expenditures, led to an explosion of US government debt to astronomical levels. In 2006 total US national debt was $8.5 trillion, 20 years on it has just passed the $40 trillion mark, representing nearly 126% of annual GDP.
The interest on the debt runs at $24 billion per week. Over 20% of the federal government’s total tax revenue annually of $5 trillion is directed towards interest payments to bondholders, amounting to in excess of $1 trillion per year. A further $1 trillion, it should be noted, goes to the military budget. Given the structural limitations and strains on the US economy, including higher interest on US debt demanded by the bond markets, the inflationary pressure of Trump’s tariff war, and Iran war (and more recently the crisis of the Japanese Yen) many bourgeois financial analysts refer to the US government’s financial position as a classic “debt death-spiral”.
The mountains of debt which have piled up in both the private and public sectors, the deepening problems of laggard growth in the productive economy, intense competition from China, and rising interest rates all guarantee that the next collapse, in which the AI bubble will almost certainly be central, will be catastrophic.
When the US stock market crashed in 1929 the US government debt was 16% of GDP, in today’s dollars an amount of 330 billion. As indicated earlier it is now 122% and rising rapidly – that is almost 8 times its indebtedness in 1929. It is worth considering the capacity of the US government to mount another major bail out – and equally worth considering is the consequences of attempting such a bail out – given its almost Argentine-style fiscal position. The 1929 stock market crash led to a loss of around 89% of total market value by 1932 – having lost around 35% in the first few weeks. A fall of that magnitude today would wipe 40 to 65 trillion dollars off current stock market valuations. In the 1930’s monopoly capitalism suffered a potentially existential crisis. With the giant reserves it had at its disposal US capitalism was able to stabilise through the New Deal. In Germany the political crisis intensified by the depression led to the rise of Hitler.
In his final work “Marxism in Our Time” written in 1939, Trotsky remarked on the Marxist theory of capitalist crisis and collapse in the following brilliant passage which perfectly describes the history of US capitalism in the last 40 years;
“The life of monopolistic capitalism in our time is a chain of crises. Each crisis is a catastrophe. The need for salvation from these partial catastrophes by means of tariff walls, inflation, increase of government spending and debts lays the ground for additional deeper and more widespread crises. The struggle for markets, for raw materials, for colonies makes military catastrophes unavoidable. All in all they prepare revolutionary catastrophes.”
Trotsky further observed on the social and political effects of this reality;
“The uncontrollable deterioration in the living conditions of the workers makes it less and less possible for the bourgeoisie to grant the masses the right of participation in political life – even within the limited framework of bourgeois parliamentarism.”
It was not only Trotsky and the Marxists who recognized the inexorable organic drive of monopoly capital to destroy both living standards and democracy. More astute bourgeois contemporaries also recognized the reality. Roosevelt’s Attorney General Homer S. Cummings, who led the defence of the New Deal anti-trust legislation said in 1937;
“Monopoly will find ways to destroy most of our reform and, in the end, lower the standards of our common life.”
Justice Louis Brandeis on the US Supreme Court, seeking to uphold Roosevelt’s New Deal legislation against the Republication right and big business declared;
“We can have great concentration of wealth in this country, or we can have democracy, but we cannot have both.”
Some fundamental features of the AI boom
Like all historical booms in capitalist history relating to technological development and breakthroughs, the phenomenon is generated out of the proprietary context of the technology – that is it is privately owned – and the broadly held assumption that there are large profits to be made through its sale in the marketplace. Historically this has created booms through the vast application of capital for development and infrastructure and an accompanying frenzy of speculation. This occurred with the railway mania in the 19th century, the automobile and radio boom of the 1920’s and the dot com bubble in 1995-2001, which was of course based on the development of the internet.
As with the two significant previous boom and busts this century centred on the United States, the dot com crash and the sub-prime crash of 2007-2008, the underlying root cause of the phenomenon is, as has been discussed, the decades-long erosion of productive investment through the hollowing out of industry, along with the rise of monopolistic finance capital and its attendant speculative and parasitic character. There is presently in the US economy virtually no economic growth outside of AI related activity, which in 2025 accounted for around 35% of US GDP. In 2025 capital expenditure by the 4 big tech “hyperscalers” Amazon, Microsoft, Google and Meta, was in the vicinity of $410 billion and is projected to be around $750 billion in 2026. These amounts are predominantly deployed in the building of data centres and the purchase of memory chips for them.
The top 10 tech companies account for 40% of the market valuation of the S&P 500 (which includes about 80% of the entire market).
The big 4 “hyperscalers” – Microsoft, Google, Amazon and Meta – alone account for almost 20% of the S&P 500 reflecting the incredible monopoly concentration in the sector.
The struggle to dominate the AI market
One of the reasons, perhaps the central reason behind the enormous speculative bubble in AI, is that there is, although not openly stated, a huge battle taking place between the tech giants as to who will ultimately dominate and control the commercial market for AI. Once again, as Lenin outlined in ‘Imperialism: The Highest Stage of Capitalism’, the essence of monopoly is to destroy competition. It is frequently observed that small investors are often motivated to buy stocks because of what is described as ‘FOMO’ – fear of missing out. But in actuality it is not restricted to them. Giant corporations act on the same impulse. None of the big 4 has historically been an AI business as such, in the sense that it develops and invests in AI for the purpose of selling AI into the market as a commodity the way the startup labs – and in particular the two leading frontier model labs Open AI and Anthropic do. The tech giants fund the massive infrastructure – the data centres – which the AI startups require to operate their models and provide that computing power for a return – either in cash, or commitments or equity, a form of ‘circular financing’. The CEOs of these companies have said that the AI buildout is required for the development of their own businesses as they “shift to a different platform” but many highly respected analysts have rejected these explanations as inadequate, failing to recognise the drive for AI monopoly. Only Meta has acknowledged that it is building its own AI business as a competitor.
By the end of 2026 the tech giants will have outlaid in the order of 2 trillion dollars. These funds go to building computing power required to train and run the AI models. The breakdown of projected AI related spending for 2026 by the big 4 is as follows:
- Amazon $200-220 billion
- Microsoft $190 billion
- Google $175 to 185 billion
- Meta $115 to 145 billion
Three of the big 4 have acquired massive stakes in open AI and Anthropic (both still private startup companies dependent entirely upon external financial support) through equity acquisition, financial guarantees, infrastructure contracts and other arrangements. Below is a table showing the position of Microsoft, Amazon and Google in Open AI and Anthropic.
| Microsoft | · 27% equity stake in Open AI and $13 billion in commitments and provision of supercomputing infrastructure. |
| Amazon | · Amazon is the principal strategic backer of Anthropic with around $40-50 billion in capital and infrastructure commitments.
· 5% equity ownership · Provides use of Amazon Web Services computing power. · 5% equity ownership in Open AI. |
| · 14% equity stake in Anthropic backed by up to 4$0 billion in capital commitments. |
Meta has taken a different strategic approach. Rather than seeking to control Open AI and Anthropic, it is pouring billions into its own proprietary “open weight” AI models such as Llama. Meta aims to undercut Open AI and Anthropic closed models with a cheaper open source commodity.
SpaceX led by Elon Musk is also in the race to dominate the AI market. In the release of its latest AI model, Grok 4.6, it claims technical superiority and lower cost over the models of Open AI and Anthropic. The frenzied speculation in the recent SpaceX IPO is symptomatic of the broader AI boom phenomenon. When listed SpaceX had a market valuation of $2 trillion, more than the annual GDP of Australia, the twelfth largest economy in the world.
In addition to the investments of the 3 tech giants tens of billions have been invested by venture capitalists, institutional investors, and private credit firms into Open AI and Anthropic. In all it is estimated that around $3 trillion has been invested in the boom since 2022. The Economist says that the AI buildout is the largest investment surge in history. Many tens of billions have been invested in Open AI and Anthropic apart from the tech giants but the exact figures are not known because as private companies their disclosure is not required. All of this capital investment is entirely speculative because there is no evidence to date that there is a viable market for AI which can generate a commercial rate of return on investment.
An uncertain market and the quest for monopoly guarantees a collapse
A fundamental uncertainty regarding the AI boom is the question of the actual potential market for AI, and the fact that this is not genuinely quantifiable gives rise to its completely speculative character. The size of the market determines the potential rate of return on the trillions being invested, but the fact is nobody knows. Promoters, such as Goldman Sachs and Morgan Stanley, float estimates in the tens of trillions, but there is no real evidence to support it, and they have a direct interest in “the story”. All the financing and speculation – including on the stock market – remains a bet on the likelihood that AI can become profitable.
The huge sums being invested in the companies involved in AI (and there are many more than just the well-known ones) is driven by investors “pricing” each company as if it is going to be the AI winner – dominating the AI market while the others lose out – because of what Professor of Business at New York University, Aswath Damodaran, calls the “Big Market Delusion” which accompanies new breakthroughs in technology in the history of capitalism. Clusters of investors are essentially gambling on who they think will end up being the future AI giant. At the same time the boards of the monopoly corporations are planning their strategies of takeover, control and domination in the sector. However, it is clear that all the investors cannot be right. Obviously, the ultimate market size is not what the companies, combined, expect the market size to be – and accordingly what earnings they anticipate. Under capitalism the market for any commodity has definite limits. To put the matter another way, the AI market is currently being “priced” (including market valuations, capital investment and other financial inputs) for a future that is not actually possible. As one serious analyst recently observed, in order for the AI sector to make a profitable return on the levels of investment being made it would need to earn revenues greater than the revenue of the entire global technology sector today. However an additional major factor in respect of the uncertainty of the profitability of the potential AI market is the ferocious competition from Chinese open source AI model.
Almost certainly the two major frontier labs will not survive because they are saddled with debts and obligations which they will never generate sufficient returns to sustain. Their rush to IPO is seen (although not disclosed) as the only salvation. The tech giants have their own plans for them. In any event, the inescapable structural truths are going to bring a terrible reckoning. As one analyst recently put it, the numbers are catching up with the story. Ultimately however, it is the fundamental state of the real economy which drives everything. Booms and busts, and their frequency and depth, are rooted in the real conditions of economy, and the US has been in a terminal state of decline, decay and death agony for decades. Monopoly capitalism is likely to face a potentially existential crisis as it did in the 1930’s with all its extreme social and political consequences.
Imperialism, the rule of monopolies and finance capital, is once again leading mankind into a catastrophe. The working class is faced with the urgent need to build new mass working class parties with a revolutionary socialist programme and leadership to end this outdated, rotten and violent social system. It must bring the rule of reason and planning into the sphere of economic and human relations, on a national and international basis, so as to secure the future equality, progress and happiness of our species.