Def. No bubble: Torsten Slok, chief economist at Apollo Global Management, recently made a scathing statement about the true extent of the so-called “AI revolution.” The data suggests that the macroeconomic impact of AI remains very limited, which could mean that we are in for some really interesting times in terms of financial trends and global economic outcomes.
According to Slok’s Daily Spark note, AI has not yet brought any financial benefit to American companies outside of the technology sector, despite the huge amounts of money and hype surrounding recent advances in large language model research.
Some technology-related companies are seeing noticeable returns, but the vast majority of companies in the S&P 500 are seeing little to show for their generative AI or LLM efforts.
Slok said inflation-adjusted sales per worker are only increasing for the Magnificent 7, the Big Tech-heavy portion of the S&P 500 that includes Nvidia, Tesla, Amazon, Microsoft, Google/Alphabet, Apple and Meta. Profit margins are also increasing for this group. Meanwhile, the “S&P 493,” as it calls the rest of the index, is flat in both sales per worker and profit margins.
Furthermore, the Apollo economist sees no signs that AI is boosting productivity growth. To put it simply, GenAI and LLMs have no impact on the bottom line outside of the tech industry. With nothing to show for it, mainstream S&P 500 companies are already retreating from the massive amounts AI requires, viewing token spending as a margin-sapping cost rather than a potential source of future growth.
Venture capital-backed companies are spending more and more on AI tokens, while mainstream organizations don’t need the latest models and are unwilling to pay for them. Slok said that if AI remains confined to a small portion of the overall economy, its macroeconomic impact will be very limited.
AI models could become increasingly commoditized and tokens could eventually become “cheap” enough to increase sales and increase margins. However, Slok believes this lofty goal may be harder to achieve than it sounds. He describes AI as a powerful “calculator,” but argues that businesses are based on good judgment, not numbers.
“AI does not understand accountability, trust, culture, unwritten rules, ethical compromises, or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins,” Slok said in his post.
In his view, the AI story will ultimately be decided by the S&P 493, not the financial fireworks of the Magnificent 7.
Slok’s company Apollo manages about $1 trillion in assets, so his assessment of the economy carries weight. Its outlook also echoes other cautionary reports, with Gartner recently estimating that more than $3.5 trillion in AI infrastructure investments could be written off by the end of 2030. But not everyone shares this pessimism. Jan Hatzius, an economist at Goldman Sachs, argues that AI investments will ultimately pay off in higher productivity and overall economic gains.
