The hype around Artificial Intelligence is undeniable, with trillions being invested in a promised productivity revolution. Yet, a top economist from investment giant Apollo is sounding an alarm: the real economic payoff has yet to materialize.
The Economist's Sobering Take
Torsten
Sløk, the chief economist at Apollo Global Management, recently made waves with a stark analysis. Despite massive investment and widespread excitement, he argues that AI has delivered 'zero financial payoff' to corporate America outside of the tech sector itself. His research highlights a growing divide. While tech giants, specifically the 'Magnificent Seven,' have seen revenue per employee and profit margins soar since 2022, the other 493 companies in the S&P 500 have remained largely flat. Sløk’s point is that the true test of AI’s economic impact isn’t the success of its creators, like Nvidia or Microsoft, but whether the companies adopting the technology see a boost to their bottom line. So far, the evidence for this broad-based improvement is missing.
Why Productivity Is the Magic Word
In economics, productivity growth is the crucial ingredient for a rising standard of living and a healthy economy. It means producing more goods and services with the same amount of input (like labor or capital). The promise of AI has always been that it will supercharge this metric, making workers more efficient and creating new avenues for growth. However, Sløk and other skeptics point to a modern 'productivity paradox'. This term, echoing a similar puzzle during the early computer era, describes the gap between the adoption of a powerful new technology and the failure of national productivity statistics to reflect its impact. While individual employees report being more productive with AI tools, this isn't translating to better overall performance for their organizations or the economy at large.
The Data Doesn't Lie (Yet)
Sløk’s argument is grounded in data showing that for most companies, AI spending is currently a cost center, not a profit driver. Many non-tech firms are reportedly pulling back, questioning if the high cost of sophisticated AI tools will ever generate the revenue to justify the expense. This is compounded by concerns that much of the current AI boom is circular, with venture-backed startups and tech companies spending heavily on each other's products. The aggregate data supports this caution. Since the AI investment boom began accelerating, overall US GDP growth has remained moderate, suggesting that the massive spending in AI is being offset by slower growth or even contraction in other sectors of the economy. Some economists even worry that the rush to build data centers is 'crowding out' other necessary investments in areas like housing and factories.
Patience or Paradox?
The counterargument, of course, is that we are simply too early in the process. Major technological revolutions, like electricity or the internet, took decades to fully reshape the economy. The infrastructure needs to be built, businesses need to learn how to redesign workflows, and workers need to adapt. Proponents believe the productivity gains will come, but they won't be instantaneous. Studies have shown that AI tools can significantly boost the performance of less-skilled workers, suggesting a potential to level the playing field over time. The World Development Report from 2026 concluded that while AI's impact is unpredictable, it holds the potential to accelerate development significantly, especially in low and middle-income countries. The question is not whether the technology is useful, but whether the current market valuation and investment frenzy have outpaced the realistic timeline for these benefits to appear.
