Slok says big-tech revenue per worker is rising, and flat for the other 493

Torsten Slok, partner and chief economist at Apollo, published a note on October 24, 2026 arguing that artificial intelligence has not shown up in the financial results of large American companies outside technology. In “No Signs of AI Payoff Outside the Tech Sector”, on Apollo’s Daily Spark, he writes that revenue per worker is rising for the group he calls the Mag 7 and is flat for the other 493 companies in the S&P 500, which he calls the S&P 493. He writes the same split for profit margins. He also writes that he sees no sign of AI boosting productivity growth. The note names him as the author. What follows is his analysis. It is not a finding by this desk, and it is not investment advice.

Slok’s wording is stronger than “flat.” He writes that the technology “has delivered zero financial payoff to corporate America outside of tech,” and that with no impact on the bottom line, “mainstream corporate America is pulling back and starting to view token spending not as a source of growth but as a cost center that weighs on margins.” Those sentences are his.

What the charts are

The note is five chart pages plus that text. The revenue-per-worker chart is labeled as inflation-adjusted to the fourth quarter of 2019. Slok sources it to Bloomberg, Macrobond, and the Apollo chief economist. The margin chart carries the same three sources. The productivity chart is utilization-adjusted total factor productivity from the Federal Reserve Bank of San Francisco, which the note attributes to a 2014 series from Fernald, adjusted for cyclical variation in factor utilization, with the U.S. Bureau of Labor Statistics also listed. Two further charts are sourced to Ara Kharazian, chief economist at Ramp, and to the Ramp AI Index.

The prose does not print a percentage change for revenue per worker, margins, or productivity. This desk is not reading numbers off the chart images. The note also does not list the seven companies inside the “Mag 7” label.

On spending, Slok writes that VC-backed companies are spending more and more on tokens, while “mainstream corporate America does not need the latest model and is unwilling to pay for it.” He writes that as long as agentic AI and token spending stay in a small part of the economy, the macroeconomic effects will be very limited. He does not, in the text, give a dollar figure for that spending.

He then argues, in his own voice, that cheaper models may still be hard for those other companies to turn into higher revenue and higher margins. He writes that AI “may be a great calculator, but in business, intelligence is mostly judgment,” and that judgment depends on context. He writes that AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs, or fairness. That paragraph is his opinion about what drives margins. His closing line is: “The bottom line is that the AI story will be decided in the S&P 493, not the Mag 7, and so far, there is no evidence that revenue growth or margins are improving outside the tech sector.”

Why it matters (analysis)

Slok is comparing two aggregates over time and reading the gap as the AI story. The published text does not include a test that separates AI spending from other reasons revenue per worker or margins might differ between a handful of very large technology companies and the other 493. A firm outside his Mag 7 group could be using AI without that use showing up in the group average. A gain inside the group could come from businesses other than AI. The productivity series he cites is a whole-economy measure. His statement that it shows no AI effect is his reading of that series.

The note is useful as a clear statement of one economist’s charts on October 24. It is not a measure of any single company’s return on an AI project, and it is not a prediction this desk is adopting.

This report is not investment, tax, or financial advice. Nothing here is a recommendation to buy, sell, or hold any security.

This report is not legal, financial, or health advice. See the disclaimer.