ANALYSIS: The boom in AI spending and its impact on the US economy

The U.S. economy faces a complex outlook, with recent revisions by the Department of Labor showing a weaker-than-expected labor market, reflected in a drastic reduction in new hiring in May and June.

In this context, tech giants - Google, Meta, Amazon and Microsoft - are emerging as a crucial pillar for the economy, with unprecedented spending on artificial intelligence (AI) that could redefine the country's economic growth.

These companies plan to invest more than $350 billion this year to build and equip AI data centers, notes a report from the Wall Street Journal. A figure that exceeds the cost of NASA's Apollo program, adjusted to current dollars, which amounted to approximately 180 Billion.

A significant economic boost

Economist Jens Nordvig, founder of Exante Data, estimates that AI investments could boost U.S. economic growth by 0.7% by 2025, which would represent half of the Federal Reserve's (Fed) 1.4% growth projection for the year.

This capital flow is funding the construction of data centers, generating infrastructure jobs, and increasing demand for chips, servers and networking equipment.

Along those lines, regions such as Texas and Northern Virginia could particularly benefit from construction jobs, although data centers require fewer staff once operational.

Tech giants are leading this transformation. Google has raised its budget from $75 billion to $85 billion, Amazon expects to exceed $100 billion, Meta projects up to $72 billion, and Microsoft plans to invest $30 billion in the current quarter alone, exceeding its annual forecast of $80 billion.

This level of investment exceeds the cumulative spending by these companies between 2010 and 2022, a period of global expansion that included office construction, cloud services, and transoceanic cables.

Promises of jobs and scientific breakthroughs

AI proponents argue that this massive spending will not only boost the economy but also generate jobs and scientific breakthroughs. OpenAI, for example, asserts that its new data centers will create "hundreds of thousands of jobs" in the U.S.

For his part, Mark Zuckerberg, CEO of Meta, has stepped up hiring for his "superintelligence" division, with salaries reaching hundreds of millions of dollars. In a July 30 manifesto, Zuckerberg predicted that AI will usher in a “new era” of productivity and economic development.

Nvidia, the leading supplier of AI chips, reported revenue of $44 billion last quarter, and its stock market value has soared, benefiting employees and executives with stock options.

Wall Street has rewarded these bets: Microsoft reached a $4 billion valuation, and Meta rose by 11% after announcing higher-than-expected AI spending.

An AI-dependent economy

However, this growing reliance on big tech raises concerns, Callie Bost, a strategist at Ritholtz Wealth Management, warns the Wall Street Journal that "the AI complex seems to be carrying the economy on its back now,” and notes that a healthy economy requires participation from a variety of sectors, something that is not currently happening.

If the AI boom slows, the impact could be severe, affecting not only big tech but also the 401(k) retirement plans and investment portfolios of millions of Americans.

Along those lines, Bost stresses that "recessions can snap your focus back to reality quickly, and what gets hit the worst in recession-fueled market sell-offs is high-valuation growth stocks.” 

For his part, Paul Kedrosky, an investor and research affiliate at MIT, questions the sustainability of these investments, arguing that long-term return assumptions are not guaranteed.

Moreover, the viability of some projects is uncertain. For example, OpenAI and SoftBank announced plans to build $500 billion worth of data centers over the next four years, but funding is not yet secured.

But ambitious tech projects, such as Foxconn's plant in Wisconsin during President Donald Trump's first term, reduced its billion-dollar investment, a fact that fuels skepticism.

A race against China

In China, President Xi Jinping has criticized overinvestment in AI data centers, according to reports from the Financial Times, while seeking to compete with the U.S. in this technology race.

However, the scale of U.S. spending is unique: AI investments account for between 1.2% and 2% of GDP, according to Paul Kedrosky. Also, AI data centers have short life spans and high maintenance costs, raising questions about their long-term profitability.

A bubble?

Skeptics warn of a possible bubble. Noah Smith, an economist, warns that past tech bubbles, such as the telecom bubble, anticipated real innovations but often prematurely.

For his part, Kedrosky adds that AI, while useful in certain scenarios, has yet to prove revolutionary, and massive investments could be driven more by fear of being left behind than by a robust market. In addition, training costs for AI models, such as GPT-4, exceed $100 million, and future versions could cost more than $1 billion, intensifying financial doubts.