No Signs of an AI Payoff Outside Tech, Apollo’s Charts Show
Apollo’s chief economist says there is still no sign of an AI payoff outside tech: gains remain confined to a handful of giant technology firms, while the other 493 S&P 500 companies show flat revenue per worker, flat margins, and no productivity lift.
PetaTech24 Technology Desk | October 25, 2026
There is still no sign of an AI payoff outside tech, according to an Apollo Daily Spark note from October 24, 2026, by chief economist Torsten Slok. The note argues that massive investment and hype have not produced measurable financial gains beyond the technology sector.
Slok’s test is simple. He splits the S&P 500 into the Magnificent 7, conventionally counted as Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, and Tesla, and the remaining 493 companies, the S&P 493. If AI were lifting the broad economy, the second group should show it in revenue per worker, margins, or productivity.
One transparency note matters. Apollo’s public page exposes only the note’s summary, not the full chart data, so readings here come from secondary coverage of its five charts. Those readings are approximations where they rely on plotted lines. The core conclusion is Apollo’s wording: “Despite massive investments and widespread hype around artificial intelligence, the technology has delivered zero financial payoff to corporate America outside of tech.”
Key Takeaways
- No AI payoff outside tech signal yet: revenue per worker rising for the Magnificent 7, flat for the S&P 493.
- Profit margins approach 30% for the Magnificent 7, near 10% for the S&P 493.
- Productivity growth shows no AI-driven step up, on the San Francisco Fed’s series.
- AI spending is extremely concentrated: the top 10% of firms account for 99.5% of model serving and inference (Ramp).
- Among the top 1% of AI spenders, 80.4% are venture capital backed, vs 13.1% of median spenders.
- Slok says mainstream firms now treat token spending as a cost center.
What the AI Payoff Outside Tech Charts Actually Show
The first chart tracks inflation-adjusted revenue per employee from early 2019, priced at 2019 fourth-quarter prices. The Magnificent 7 started near $695,000 per employee, dipped to about $565,000 around the turn of 2022 and 2023, then climbed to roughly $780,000. The S&P 493 line stayed in a narrower band, from about $495,000 to roughly $527,000.
The second chart, on profit margins, reaches back to 2015. The Magnificent 7 margin rose from around 11% to about 22% in early 2022, fell to roughly 14% in 2023, then climbed to close to 30%. The S&P 493 margin stood near 8.5% in 2015, dipped to about 6.5% in the pandemic, peaked near 11.7% in early 2022, and has hovered between 9% and 10% since 2023.
The third chart widens the lens. It plots Bureau of Labor Statistics output per hour and total factor productivity (TFP) from January 1985. The TFP series is the San Francisco Fed’s utilisation-adjusted measure. Output per hour spiked to around 7% during the 2020 pandemic, then dipped below zero. Recently, output per hour growth runs at roughly 2% to 2.5%, while TFP growth has slipped to around zero.
TFP is the measure economists watch for a general-purpose technology, because it captures output growth that cannot be explained by adding more workers or machines. A flat TFP line, set against record AI capital spending, is the core of Slok’s AI payoff outside tech case.
Why the AI Payoff Outside Tech Has Not Arrived Yet
The fourth and fifth charts explain why the payoff looks so narrow, using spending data from Ramp, the corporate card and expense management company behind the Ramp AI Index. Model serving and inference, the bill for running AI models once trained, sits almost entirely with the top 10% of firms at 99.5%, steeper than conventional software spending.
The fifth chart adds the funding twist. Among the top 1% of AI spenders, 80.4% are VC-backed and 19.6% are not. Among median spenders, the flip holds: 13.1% VC-backed, 86.9% not. Slok links that to the circular financing debate, in which investors fund AI startups that then spend much of that money on cloud and model access from tied companies.
That is where behaviour enters. “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,” Slok wrote. Token spending means fees paid to AI providers; the note does not quantify the pullback. A KPMG survey of 2,145 leaders found 49% had paused or narrowed AI rollouts after costs outweighed value; only 7% reported established ROI.
“VC-backed companies are spending more and more on tokens, but mainstream corporate America does not need the latest model and is unwilling to pay for it,” Slok wrote. “As long as agentic AI and token spending are confined to a small part of the economy, the macroeconomic effects of AI will be very limited.” Agentic AI acts in multi-step ways, so narrow use keeps the AI payoff outside tech out of reach.
The Bigger Picture
Earlier Apollo notes traced the same gap. On June 30, 2026, Slok argued that healthcare, banking, and energy face longer productivity timelines than software, and that AI valuations depend on S&P 493 margins eventually rising. On August 9, 2026, he argued the capital spending boom showed up in AI sellers’ margins, not buyers’.
There are fair counterarguments. The charts show correlation in time, not cause. Rising Magnificent 7 margins could reflect AI-driven sales, but also cost cutting, layoffs, pricing power, and cyclical factors.
What to Watch Next
The verdict on an AI payoff outside tech will come from earnings reports. Three signals deserve attention first. The first is S&P 493 revenue per worker: if that line starts rising, Slok’s thesis weakens. The second is S&P 493 margins: a sustained move above the recent 9% to 10% range would be first real evidence of a payoff. The third is total factor productivity: a sustained step up would show AI lifting the economy.
Corporate spending behaviour is the second thing to watch. If mainstream companies resume token spending growth and pay for premium models, that suggests they found uses worth the price. If they keep treating token bills as a cost to control, the AI payoff outside tech will stay limited, exactly as Slok predicts.
Quick Answers
Spending behaviour is the second thing to watch. If mainstream companies resume token spending growth and pay for premium models, they found uses worth the price. If they keep treating token bills as a cost to control, the AI payoff outside tech stays limited, as Slok predicts.
Who is making money from AI so far? The Magnificent 7, whose revenue per worker has risen to roughly $780,000 and whose margins approach 30%, plus the infrastructure sellers behind them. Buyers in the rest of the index have not yet seen an AI payoff outside tech gain.
Does flat aggregate data mean AI helps no ordinary company? No. Gains at individual firms or sectors could be offset by losses elsewhere, or could take longer to appear in index-wide figures.
Verdict
Slok’s bottom line is a test, not a prophecy: “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.” It does not prove AI will never pay off for mainstream companies. It proves the payoff has not arrived in the numbers that matter most.
For businesses, the practical takeaway is restraint with a purpose. Do not buy the most expensive model by default, do not treat token spending as growth spending until it shows up in revenue or margins, and measure AI projects like any other investment: higher revenue per worker, wider margins, or genuinely higher output.
Sources / Attribution
- Apollo Global Management, Torsten Slok, “No Signs of AI Payoff Outside the Tech Sector,” Daily Spark, October 24, 2026. Public page exposes summary only.
- PPC Land, “Apollo finds zero AI payoff outside tech as S&P 493 margins stay near 10%”, October 2026, chart readings, Ramp table, and KPMG survey summary.
- Apollo Daily Spark source note.
- Bureau of Labor Statistics output per hour and San Francisco Fed total factor productivity series, as cited in Apollo’s note; Ramp AI Index data, via Apollo.
