Microsoft participates in Open AI, which allocates 38 billion to Amazon Web Services (AWS) to ensure access to its infrastructure and equipment from Nvidia, which also invests 100 billion in Open AI. Much of the money flow of big tech companies is circular: equipment manufacturers and service vendors participate in each other’s businesses to generate a closed economic ecosystem in which investments and benefits are shared. The strategy is very old, but the investment volume is unprecedented and this financial stress was followed by announcements of thousands of layoffs and doubts about a possible bubble or the creation of an oligopoly on artificial intelligence (AI).
The formula known as seller financing or circular investing is not new. It has been used by multinationals and consists of mutual investments in which companies support each other financially and acquire shares of their distributors, buyers and even competitors to expand or strengthen markets and to present themselves to investors as a big business.
“The first thing you look at to see the value of a company is the turnover and with that (circular financing) speculators obtain an image of greater power,” explains Pedro Palos, professor of Financial Economics and Operational Management at the University of Seville and specialized in the digital sector and the knowledge society. “To advance as they are doing, they need large investments, and to achieve this, they increase the scale and volume of business,” he adds.
In this direction, two hundred scientists signed a letter to the President of the European Commission, Ursula von der Leyen, accusing companies of exaggerating the capabilities of AI (“AI language”). marketing “Instead of effectively addressing the harms (of AI) and preventing companies from exploiting workers and stealing creative work, the president is promoting the AI bubble to benefit these companies,” warns Kris Shrishak, of the Irish Council of Civil Liberties and initiator of the initiative.
The Bank of England’s Financial Policy Committee also warns of the “growing risk of a sudden correction” in the market. There are “bottlenecks” in the supply chains of energy, data and raw materials needed to meet demand, as well as a disproportionate race between the development of models and the capabilities to implement them, which can impact infrastructure investment.
Palos agrees that there is a bubble risk and not just because of circular financing or for the reasons the Bank of England warns against: “What will happen when companies improve their competitiveness and put more products on the market? Who will buy them if the deterioration of the purchasing power of families and workers continues? Obviously there is a bubble risk: excess production will be followed by massive layoffs and not just in the tech sector. Artificial intelligence can make any company much more profitable and competitive, but those that are not will disappear.”
This alleged bubble requires gigantic spending on the accelerated development and maintenance of AI infrastructure, which forces us to look for more formulas, beyond financial ones. “We’re running out of easy ways to secure more funding, so the cost reduction will continue,” Pratik Ratadiya, product manager at IA Narravance, says in X. One of these costs is the cost of labor.
Amazon has cut 14,000 jobs this year, including 1,200 in Spain. IBM, according to a spokesperson, planned in the last quarter to lay off “a low, single-digit percentage of the global workforce,” which amounts to 270,000 people and where a cut, even less than 10%, means hundreds of layoffs. Meta will lay off around 600 employees within its artificial intelligence unit. Chegg, the online education company, expects a 45% reduction in its workforce, Salesforce has already cut 4,000 customer service positions, and UPS has cut 48,000 positions. And the list goes on. In Spain, Telefónica has invited unions to negotiate an ERE for 5,000 employees.
Some companies, like Salesforce or UPS, admit that robotization and task automation have made some positions obsolete. Amazon, however, does not directly attribute layoffs at its company to artificial intelligence. Nishant Mehta, vice president of one of the divisions of this multinational, denies: “They have to do with how to achieve adequate efficiencies, how to make the team able to make the right decisions at all times and very quickly; it is not related to artificial intelligence.”
Mehta, during a meeting with the international press in Seattle, headquarters of the multinational, to which EL PAÍS was invited, refuses that the technological revolution will generate long-term unemployment: “In the end, artificial intelligence will create many more jobs.”
Sri Elaprolu, director of the AWS Generative AI Innovation Center, specifies the employment impact of innovations. “The nature of work will change,” he admits, citing changes after electrification or in transportation as an example. “But we need to understand the value of what you get and the time it frees up to do more important and valuable work,” he adds. “We need to keep improving our skills, we need to keep getting better at using artificial intelligence to tackle increasingly complex jobs. Preparation is vital and it starts in all sectors of society, from businesses to local governments. Everyone will have to do their part, collaborate and be prepared to manage all of this because the net benefit to society is great,” he defends.
As for circular investments in seemingly competing technologies (Amazon makes its own chips, such as the latest Trainium and Inferentia, but the deal with Open AI includes access to Nvidia’s GB200 and GB300 accelerators), Nishant Mehta argues that this is a mandatory practice. “To build an international supply chain, we partner with suppliers from all over the world to meet our customers’ needs,” he says.
Nor does he endorse the bubble risk warned by Palos or the Bank of England, although he understands there is speculation about that. “It is natural that, when there is an investment of this caliber (the global one expected in artificial intelligence), there is skepticism. Sometimes it is healthy and, sometimes, paranoia”, he comments. He adds: “As long as we have strong business fundamentals, whether it’s a bubble or not, we think we’ll be fine.”
The other risk of the practice of circular financing by the tech clan, led by giant companies, is that it leads to a concentration of interests that generates an oligopoly, a market in which a small number of companies dominate the offering of products or services. Palos believes this is a valid fear, but Amazon’s manager dismisses it: “We believe there’s room for everyone to grow, so we don’t think it’s going to be a one-player, two-player thing. This industry is going to be huge.”
Sri Elaprolu also dismisses concentration risk – “quite the opposite,” he replies” – and argues that they work with all players because a company’s mission is “to provide the client with the right capabilities so that they can build what is appropriate safely and profitably.”
