🔍 Read the full analysis: The Cost Of AI Is Rising Faster Than Its Broad Economic Returns on ThorstenMeyerAI.com
Get the latest gadgets 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 chief economist Torsten Slok says aggregate data show no clear financial payoff from AI outside the technology sector, while spending on infrastructure and computing capacity continues to climb. The finding is an interpretation of broad trends, not proof that AI has produced no gains: studies report measurable improvements in some tasks, while many firms say productivity has not yet changed.
Apollo chief economist Torsten Slok says recent company and productivity data show no clear aggregate financial payoff from AI outside technology, even as companies commit growing sums to data centers, servers and computing capacity. His October 24 note contrasts gains among seven major technology companies with broadly flat revenue per worker and profit margins across the other 493 S&P 500 firms; the comparison does not establish that AI itself caused either trend.
Apollo’s charts put the seven companies it calls the Magnificent 7—Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla—near $780,000 in inflation-adjusted revenue per employee in the third quarter of 2026. The other 493 S&P 500 companies were around $525,000, in a range Apollo says they have occupied since 2021. Apollo’s figures also show margins around 30% for the seven, compared with roughly 10% for the rest. Those figures describe the companies’ performance; they do not isolate the effect of AI.
The note also points to San Francisco Federal Reserve measures of productivity, which Apollo says show no sign yet of AI lifting productivity growth. Separately, data from Ramp, a corporate spending platform with more than 85,000 U.S. business customers, suggest that AI-related spending is concentrated: the top 10% of firms account for nearly all spending in several measured categories, including 99.5% of model-serving and inference spending. Ramp’s customer data are not a census of U.S. companies.
The costs are visible in corporate filings. Meta reported $34.91 billion in capital expenditures in the third quarter and narrowed its 2026 spending forecast to $135 billion to $145 billion. Its cash flow from operations was $32.79 billion, leaving free cash flow at negative $2.12 billion, compared with positive $10.6 billion a year earlier. Meta’s October 22 filing listed about $355.73 billion in leases not yet started and $454.52 billion in non-cancelable contractual commitments, mostly for cloud capacity, servers and network equipment. The source calculates roughly $825 billion when those items and contingent purchases are combined; the total is not all AI spending.
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
Costs Rise Before Gains Show Up
The gap matters because firms are committing capital and signing long-term infrastructure contracts before the returns are clear across the wider economy. Infrastructure spending is measurable, but Apollo’s comparison finds no corresponding broad rise in revenue per worker or margins outside the biggest technology companies. If those investments do not lead to sufficient savings or new revenue, they could weigh on company finances and eventually affect spending on other priorities.
That does not mean AI has no economic value. Results vary by task and organization, and aggregate company averages can conceal gains in some teams alongside costs or losses elsewhere. The source cites studies reporting 14% more customer-support issues resolved per hour, rising to 34% for novice workers, and retail sales results ranging from no increase to 16.3%, depending on the workflow. These are task-specific findings, not evidence that the whole economy is seeing comparable gains.
For workers and customers, the distinction is practical: adoption may change particular jobs or services before it lifts company-wide productivity statistics. For investors and executives, the open question is whether returns will catch up with the costs of computing, energy, facilities and financing.
Products Worth Considering
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
Apple iPad Air 11-inch (M2), 512GB, Wi-Fi, Space Gray
Bose SoundLink Flex Bluetooth Speaker (2nd Gen), Portable Waterproof Speaker, Black
As an affiliate, we earn on qualifying purchases.
What Apollo’s Numbers Measure
Apollo’s October 24 “Daily Spark” note, titled “No Signs of AI Payoff Outside the Tech Sector,” was discussed by news outlets on October 25. Slok’s “zero financial payoff” wording is a conclusion drawn from aggregate indicators, not a direct measurement showing that AI has generated no revenue or savings at any company. Flat averages across hundreds of firms cannot reveal whether gains in some businesses are offset by declines in others, and they do not separate AI from other influences such as labor costs, tariffs or interest rates.
The evidence is mixed at the level of adoption, too. The source reports that 83% of nearly 9,000 executives said AI had not affected their productivity so far, while the reported average productivity effect was 0.2% a year. These findings describe surveyed executives and the study’s measurement period; they should not be treated as a definitive reading of every firm or as a forecast. Apollo also cites Ramp’s concentration figures, but those reflect Ramp customers’ spending, not spending by all companies.
On the cost side, Meta’s commitments cover multiple years and include data centers, energy infrastructure, network capacity and non-cancelable purchases. They are not equivalent to cash already spent. Meta’s quarterly results also show revenue of $64.60 billion, up 26% year over year, alongside the negative free cash flow figure. The juxtaposition illustrates why a single spending or cash-flow measure cannot, on its own, establish whether the company’s AI investments are profitable.
“AI “has delivered zero financial payoff to corporate America outside of tech.””
— Torsten Slok, Apollo chief economist, in the October 24 Daily Spark note
Returns Are Hard to Isolate
It remains unclear how much of the spending will produce lasting returns, and when. The Apollo charts track company-level outcomes, not the portion caused by AI. They cannot show whether reported changes stem from AI, other business decisions or broader economic conditions. Nor does the absence of a visible change in aggregate productivity prove that no individual company or worker has benefited.
Meta’s roughly $825 billion figure combines obligations with different terms and purposes, and includes items that are not all AI-related. The source also reports that SpaceX borrowed $40 billion to buy AI chips and that insurance costs on debt at Meta and CoreWeave increased, but the material provided does not give enough detail to establish how those developments affect the overall financing outlook. Future spending estimates for major technology companies are projections, not amounts already committed or spent.
Watch Spending and Productivity Data
The next evidence will come from company disclosures on capital spending, cash flow and use of AI services, alongside future productivity and survey data. Investors will also watch whether large infrastructure commitments translate into revenue, cost savings or improved output per worker. Those outcomes are not yet settled; spending totals and executive reports of limited impact describe the current picture, not the eventual return.
For now, Apollo’s note is best read as a warning about the gap between investment and visible broad-based gains, rather than a verdict on every AI deployment. More detailed reporting on individual workflows and longer-term company results will be needed to determine where returns are emerging and whether they can justify the expanding costs.
Key Questions
Did Apollo prove that AI produces no financial returns?
No. “Zero financial payoff” is Slok’s interpretation of aggregate trends. The figures do not isolate AI’s effect at individual firms, and gains in some businesses may not be visible in broad averages.
What does Meta’s roughly $825 billion figure include?
It combines about $355.73 billion in leases not yet started, $454.52 billion in contractual commitments and up to $14.72 billion in contingent cloud purchases. The source says the total is not all AI-related and does not mean the entire amount has already been spent.
Are there signs that AI can improve work?
Yes, in some measured tasks. The source cites customer-support research finding 14% more issues resolved per hour, with a larger result for novice workers, and retail experiments with sales results ranging from zero to 16.3%. Those results depend on the workflow and do not establish economy-wide gains.
Why might productivity gains not appear in broad data yet?
AI adoption and process changes may take time to affect company-wide output, and gains in particular teams can be obscured by costs or weaker results elsewhere. The available figures also cannot separate AI’s effect from other business and economic factors.
Source: ThorstenMeyerAI.com
