While the sharp falls of a week ago have abated somewhat, uncertainty over the impact of artificial intelligence (AI) on share market valuations and the financial system more broadly continues.
One of the features of last week’s turbulence was that while there was an uptick, there was no significant movement to “buy the dip,” which has been a significant feature of the stock market in the recent period as it has powered its way to record highs.
As the Financial Times noted, after the release of new AI tools directed at a series of software-based business models, ranging from wealth management, real estate, trucking to advertising, the market is “wracked with uncertainty about what comes next.”
Remarks by Robert Schramm-Fuchs of the global asset management firm Janus Henderson made clear why.
The world was changing “very, very quickly,” he said. “The AI models today are substantially more powerful than the ones from six or 12 months ago. What seems protected as a business model today might not be [in the future]. It makes it even harder to buy the dip.”
Fears about the long-term profitability of the massive investments made by the so-called “hyperscalers,” including Alphabet (the owner of Google), Microsoft, Meta (the owner of Facebook) and Amazon, are reflected in the shifts in their share price.
In the wake of the release of ChatGPT by OpenAI in November 2022, the share prices of the big tech stocks surged. Meta rose by nearly 450 percent from the end of 2022 to the start of this year, and Alphabet increased by 250 percent.
It has been a different story since then. From the end of last month, Microsoft and Meta have dropped by around 16 percent and Amazon stock has been falling. Alphabet is down by 11 percent. In total, nearly $1.5 trillion has been wiped off the value of this group of companies, leading to the tech-heavy NASDAQ index falling into negative territory for the year.
Even the chipmaker Nvidia has been impacted. After two consecutive years of triple-digit gains in its share price and a 40 percent jump in 2025, it is down this year.
The market sentiment was summed up in the remarks of a market strategist cited by Bloomberg.
“Investors were comfortable saying, ‘so long as it happens in the future, I’m comfortable with Microsoft, or Amazon or Alphabet spending the money.’ Now they want to know more immediately when the payback will come—and we don’t have a clear picture.”
The spending is not only continuing but accelerating. It has been estimated that it will reach around $700 billion this year, an increase of 65 percent over last year.
In the initial investment phase, the hyperscalers were able to use their large cash reserves. They have now been largely exhausted and there are fears that the rising level of spending will not be sustainable.
In a note to clients at the start of this week, Ulrike Hoffmann-Burchardi, chief of American investment for UBS Wealth Management, drew attention to the change in the financial landscape.
“This level of capex [capital expenditure] will consume 100 percent of hyperscalers’ cash flow from operations compared with a 10-year average of 40 percent. That spending is now increasingly being funded by external debt or equity financing,” she wrote.
Amazon, which has said it expects to spend $200 billion this year, is looking at a cash flow deficit as much as $28 billion this year, according to Bank of America analysts.
While the market is becoming increasingly concerned that AI may fail to deliver sufficient profit in relation to the massive investment spending, what could be termed AI “success” is also causing major turbulence.
The fall in the value of software-based companies began with the release by Anthropic of an AI tool to speed up work in legal and financial firms among others. This was followed by the release from a little-known start-up company, Altruist, of an AI tool which impacted major wealth management firms.
Shares in the investment firm Charles Schwab dropped by 11 percent and those of the commercial real estate firm CBRE fell by 16 percent.
Then came the release of a white paper by a tiny former karaoke company (now a freight company), which said its AI tool could increase freight volumes by up to 400 percent without the same increase in the labor force.
As the FT reported: “The note ignited fears that the new technology would destroy the market value of some of the industry leaders, sending logistics companies CH Robinson and Landstar both down by about 15 percent in a single day.”
More broadly, the development of new AI tools—and there will be more to come—are impacting on firms which supply software for a range of business management procedures.
Companies that sell software as a service have been hit. The shares of Salesforce, one of the leading companies in this area, are down by 30 percent so far this year.
And the problem in this sector could have major financial implications for private equity firms that have invested heavily in software firms. According to an estimate by Bain & Co, between 2020 and the first half of last year, around one fifth of all buyouts in North America were tech deals. Business development companies, financed by private equity, are heavily invested in this area. But they now face the risk that the firms they have invested in could be made obsolete by the development of new AI tools.
These twin developments—on the one hand, fears that AI investments will not produce the revenue streams necessary, and on the other, that where AI does bring about major productivity gains this will produce major economic and financial disruption—has left some commentators and analysts scratching their heads.
As a report on Bloomberg put it: “The turmoil unleashed by AI reflects fears that are increasingly at odds. One is that it’s poised to disrupt entire segments of the economy so dramatically that investors are dumping the stocks of any company seen at the slightest risk of being displaced by the technology. The other is a deep skepticism that the hundreds of billions of dollars spent in AI will deliver big payoffs anytime soon.”
Another Bloomberg article cited the comments of an investment officer at the major financial firm Nomura, who said there was “a contradiction” in what investors were worried about and “these two things can’t be true at the same time.”
In fact, they are because they are different expressions of the same underlying process. The fundamental driving force of capitalism is not the development of productivity, per se, and the increased production of material wealth which AI makes possible, but the accumulation of profit at a rate sufficient to ensure the expansion of capital.
The fear is that revenues from the investments by the hyperscalers will not generate a sufficient rate of return and the capital invested, running into trillions, increasingly financed by debt, will be devalued.
In the case of the disruption caused by AI tools, the fear is that the capital already invested in vast areas of the economy, also financed by debt, will likewise be devalued.
And both cases contain the potential for a major financial crisis, with devastating economic and social consequences, because of the massive debt in funding assets that could have their value slashed virtually overnight. The case for the socialist reorganization of the economy, which would involve not least the bringing of AI development into public ownership under democratic control, and the conscious application of its benefits, could hardly be clearer.