🔍 Read the full analysis: Does AI Deliver Value Beyond Tech? Here’s What The Numbers Say on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Apollo economist Torsten Slok says revenue per worker and profit margins are rising for seven major technology companies but remain broadly flat across the other 493 S&P 500 companies. That supports the view that AI’s financial payoff is not yet visible in broad corporate averages, but does not prove that every non-tech company has received no benefit. Spending commitments are mounting, while research on specific tasks has found measurable gains in some workflows.
Apollo chief economist Torsten Slok says recent corporate and productivity data show no visible aggregate financial payoff from AI outside the technology sector, even as companies commit hundreds of billions of dollars to computing infrastructure. His October 24 note compares the seven largest technology companies with the other 493 S&P 500 firms; it does not establish that AI has produced no gains at any individual non-tech company.
Apollo’s analysis points to a widening gap in revenue per worker and profit margins. Its charts show inflation-adjusted revenue per worker rising for the Magnificent Seven—Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla—while remaining broadly flat for the other S&P 500 companies. The seven are near $780,000 per employee in the third quarter of 2026, compared with about $525,000 for the remaining 493, according to the figures described in the note. The largest firms’ margins have climbed to roughly 30%, while the others remain near 10%.
The note also cites San Francisco Federal Reserve productivity measures that, Apollo says, show no sign yet of AI boosting productivity growth. Separate spending data from Ramp, a corporate-card and spending platform whose customers include more than 85,000 U.S. businesses, show AI-related purchases concentrated among a small share of firms. The top 10% of firms account for 99.5% of spending on model serving and inference, 99.0% of neocloud spending and 95.6% of spending on OpenAI and Anthropic, according to the source material.
The spending commitments are substantial, but they are not the same as realized losses or AI-only costs. Meta’s third-quarter results reported $64.60 billion in revenue, up 26% year over year, and $34.91 billion in capital expenditure. Operating cash flow was $32.79 billion, leaving free cash flow at negative $2.12 billion, compared with positive $10.6 billion a year earlier. Meta forecast $135 billion to $145 billion in capital spending for 2026.
Meta’s October 22 filing listed about $825 billion in future commitments, largely leases and contracts for data centers, cloud capacity, servers and network equipment. The total includes leases not yet started, non-cancelable contracts and contingent cloud purchases; some spending is not AI-related. Separately, the source reports that SpaceX borrowed $40 billion to buy AI chips, illustrating how infrastructure expansion is increasingly tied to financing.
Infographic 01 · The “zero” in numbers
What “zero financial payoff” actually measures.
Apollo’s headline rests on five charts. Here are their numbers, and what they can and cannot prove.
Who pays for AI? Share of spending by the top 10% of firms
Ramp customer data via Apollo · non-AI categories in grey for comparison
Spending is concentrated in every category, not only AI; AI is the most extreme.
What the five charts do and do not show
shown not shown
- Shown: revenue per worker and margins rising for the 7 tech giants, flat for the other 493
- Shown: no AI lift yet in US productivity statistics (San Francisco Fed TFP, output per hour)
- Shown: AI spending sits in few, mostly venture-funded firms (Ramp, 85,000+ US businesses)
- Not shown: that AI caused the gap; rates, tariffs and wages move the same lines
- Not shown: gains inside single functions, which can cancel out in an average
- Not shown: the S&P 493 itself in the Ramp data; those are Ramp’s customers
“The AI story will be decided in the S&P 493, not the Mag 7.”Torsten Slok, Apollo chief economist, 24 Oct 2026 (opinion)
Sources: Apollo Daily Spark (24 Oct 2026) with charts by Bloomberg, Macrobond, SF Fed, BLS and Ramp; Apollo (20 Oct 2026); Ramp AI Index. Values with ≈ are read off Apollo’s charts. Assessment: Thorsten Meyer.
Powered by Thorsten Meyer AI
The Gap Between AI Spending and Returns
The figures sharpen a question for investors, employees and customers: are large AI investments producing returns beyond the companies that sell the technology? If costs rise before productivity or revenue gains appear, firms may face pressure on margins and may scale back projects that cannot show a clear business case. Meta’s negative third-quarter free cash flow and its large future commitments make the mismatch between current spending and measurable returns particularly visible, although neither figure alone shows that AI caused the cash-flow result.
The aggregate picture does not capture every use case. The source material cites research finding that customer-support workers resolved 14% more issues per hour with AI assistance, with gains of 34% among novice workers. Retail experiments reported sales changes ranging from zero to 16.3%, depending on the workflow. These results suggest that benefits can be measurable when AI is applied to particular tasks, but they do not show that such gains are widespread or sufficient to offset company-wide investment.
Products Worth Considering
Apple iPad Air (4th Generation, 10.9-inch, Wi-Fi, 256GB) – Space Gray
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
Sonos Era 100 Wireless Smart Speaker, Black
As an affiliate, we earn on qualifying purchases.
What Apollo’s Comparison Measures
Slok’s October 24 Apollo note, titled “No Signs of AI Payoff Outside the Tech Sector,” drew attention after coverage by technology and business outlets on October 25. Its headline conclusion rests on comparisons of company-wide trends and productivity indicators, rather than an accounting of AI’s effect at each company. The distinction matters: flat averages can coexist with gains at some firms and declines at others, and the data cannot isolate AI from other influences such as labor costs, tariffs or interest rates.
The source also cites a survey of nearly 9,000 executives in which 83% reported no productivity impact from AI so far; the reported average productivity effect was 0.2% a year. Those results describe respondents’ reported experience and the study’s measured period, not a definitive forecast of long-term outcomes. Ramp’s spending figures likewise describe its customer base, not the S&P 493 as a whole.
At the same time, the investment cycle is expanding. Projections cited for Amazon, Alphabet, Meta, Microsoft and Oracle put combined capital expenditure near $780 billion in 2026 and above $1 trillion annually from 2027 through 2030. These are forecasts, not booked commitments. They should not be conflated with Meta’s separately reported commitments, which cover multiple years and include some non-AI infrastructure.
““has delivered zero financial payoff to corporate America outside of tech.””
— Torsten Slok, Apollo chief economist
What the Aggregate Data Cannot Show
The evidence does not identify how much of the gap is caused by AI, or whether AI has improved performance at particular firms. Company averages can conceal differing results, while the cited productivity measures do not isolate AI from other economic forces. The source material also does not provide enough detail here to assess the methods, time periods or uncertainty ranges behind every cited study and survey.
It remains unclear whether task-level gains will spread across more industries, translate into sustained profit growth, or exceed the costs of models, computing, integration and staff training. Meta’s commitments include non-AI items, and forecasts for other large technology companies are not finalized spending totals. The long-term financial impact and the durability of the reported productivity gains are still developing.
Evidence to Watch in Coming Results
The next test will be whether companies report repeatable productivity or revenue gains alongside their infrastructure costs. Investors and business leaders will be watching future earnings releases for capital spending, cash flow, margins and specific disclosures about AI-related returns. Meta’s 2026 spending outlook and its commitments will remain a focus, but those figures alone cannot show whether the investment is paying off.
For now, Apollo’s conclusion is best read as a statement about what is visible in broad data today: AI’s financial payoff has not appeared clearly across non-tech corporate averages. More company-level evidence and longer-term productivity data will be needed to establish how widely the gains are spreading and whether they outweigh the costs.
Key Questions
What does Apollo mean by “zero financial payoff”?
It is Torsten Slok’s interpretation of trends showing rising revenue per worker and margins among seven major technology companies, but broadly flat results among the other 493 S&P 500 firms. It does not prove that every non-tech company has received no AI benefit.
Does the data show that AI has no productivity benefit?
No. The cited economy-wide productivity measures show no clear AI-driven boost so far, while some task-level research reports gains. The findings use different measures and do not establish how widespread or lasting those gains are.
What is included in Meta’s roughly $825 billion in commitments?
The total includes leases not yet started, non-cancelable contractual commitments and contingent cloud-capacity purchases, mainly tied to data centers, servers, network equipment and cloud capacity. Some commitments are not AI-related, and the total covers future years rather than a single year’s spending.
Are companies already seeing measurable AI gains?
Some cited experiments and studies report gains in specific workflows, including customer support and retail. Those results are not evidence of economy-wide returns, and outcomes vary by task and implementation.
Source: ThorstenMeyerAI.com
