Apollo: S&P 493 sees no AI payoff as giants' revenue per worker hits almost $800,000

Apollo's analysis, published on Saturday, sets the Magnificent Seven's gains against the rest of the S&P 500 on three measures: revenue per worker, profit margins and productivity growth. The gap is starkest in revenue per employee, which at the tech firms climbed from an average of $550,000 in 2022 to almost $800,000 this year, while the S&P 493 has lingered at about $550,000 since 2019. Profit margins tell a similar story, soaring at the tech firms since the third quarter of 2023 and staying roughly flat elsewhere.

Slok's reading is that this is not simply a lag. He notes some venture capital-backed startups are going against the grain and increasing spending on AI tokens, but more established firms are pulling back, treating token spending as a cost center rather than a source of growth. His conclusion is that mainstream corporate America does not need the latest model and is unwilling to pay for it, which caps AI's macroeconomic reach while agentic AI and token spending remain confined to a small part of the economy.

The research also challenges predictions that AI will usher in an "age of abundance." Slok argues that business intelligence is mostly judgment, and that AI still has to overcome significant challenges before it can be trusted — including its inability to understand business culture, accountability, ethical trade-offs, fairness and trust, which all factor into reputational risk and drive revenue and profit growth. Even as models become commoditized, he writes, turning increasingly cheap intelligence into higher revenue growth and higher profit margins may be harder than it sounds.