🔍 Read the full analysis: AI Pays Off Almost Nowhere Outside Tech, An Economist Says. The Spending Keeps Exploding. What The Numbers Really Show 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 AI-related financial payoff outside the technology sector, even as spending on computing infrastructure expands. The finding describes broad trends, not proof that AI produces no gains: research and company-level examples show productivity and sales improvements in some specific workflows.
Apollo chief economist Torsten Slok says financial and productivity data show no visible aggregate payoff from artificial intelligence outside the technology sector, even as companies continue committing large sums to data centers, chips and cloud capacity. His October 24 analysis points to flat revenue per worker and profit margins among 493 S&P 500 companies outside the seven largest technology firms; it does not establish that AI has produced no gains at individual companies or in particular tasks.
Slok’s Apollo “Daily Spark” note compares the Magnificent Seven — Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla — with the other 493 S&P 500 companies. Apollo’s charts show inflation-adjusted revenue per worker rising for the seven while remaining broadly flat for the other firms. The note also shows margins near 30% for the seven and around 10% for the other 493. These comparisons describe company performance; they do not isolate AI’s effect from other economic forces.
The note cites productivity measures from the San Francisco Federal Reserve, which Apollo says show no sign yet of AI boosting productivity growth. It also uses spending data from Ramp, a corporate card and spend platform whose customers include more than 85,000 U.S. businesses. In those data, the top 10% of firms account for 99.5% of spending on model serving and inference, 99% of neocloud spending and 95.6% of spending on OpenAI and Anthropic. Ramp’s customer base is not the same as the S&P 500, and the figures do not represent all U.S. business spending.
Company disclosures show the scale of the infrastructure bill. Meta reported $34.91 billion in capital expenditures in the third quarter and forecast $135 billion to $145 billion in capital spending for 2026. Its 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’s October 22 filing listed about $355.73 billion in leases not yet started and $454.52 billion in non-cancelable contractual commitments, along with up to $14.72 billion in contingent cloud capacity purchases. The combined figure is about $825 billion in future commitments; it is not all AI spending, and includes items such as Reality Labs hardware.
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 Costs and Returns
The contrast between rising infrastructure commitments and limited aggregate evidence of returns matters to investors, executives and customers. Companies are spending on computing capacity before broad productivity gains are apparent in economy-wide indicators. That creates a financial risk if the spending does not produce enough revenue or cost savings, while also making it difficult to determine whether current investment is premature or simply not yet reflected in reported results.
Slok’s conclusion is a warning about the economics of the current AI buildout, not a verdict on every use of the technology. The source material cites studies and experiments in which defined applications have produced measurable results: customer-support workers resolved 14% more issues per hour, with a 34% increase for novice workers, while retail sales experiments ranged from no increase to 16.3%, depending on workflow. These examples suggest that returns may depend on how AI is applied, rather than appearing uniformly across entire companies.
The spending also has implications beyond the largest technology firms. Infrastructure providers and companies building AI services rely on continued access to capital, chips and energy. If returns remain uncertain while financing needs grow, investors may scrutinize debt and spending plans more closely. Conversely, an absence of gains in broad measures today does not rule out future improvements or benefits that are difficult to capture in current statistics.
Products Worth Considering
Jackery Explorer 300 Portable Power Station, 292Wh
Jackery HomePower 3600 Plus Portable Power Station, 3584Wh, 3600W
Anker SOLIX C200 DC Portable Power Station, 192Wh, 200W
As an affiliate, we earn on qualifying purchases.
What Apollo’s Comparison Measures
Apollo’s October 24 note, titled “No Signs of AI Payoff Outside the Tech Sector,” was subsequently covered by technology and financial news outlets. Its central comparison is between the seven largest technology companies and the rest of the S&P 500. Because the groups differ in business models, scale and exposure to technology, a widening performance gap cannot by itself show that AI caused the difference.
Slok has also questioned whether spending on AI services can generate sufficient returns. In an October 20 note, he described the gap between the marginal cost of computing and the benefit users receive as a core problem for the AI investment case. On October 22, he characterized some venture-backed spending on AI tokens as “circularity.” Those are Slok’s economic interpretations. The supporting charts show broad trends but do not measure the return on every AI project.
The source material also reports that 83% of nearly 9,000 executives said AI had not affected their productivity so far, while the average reported productivity effect was 0.2% a year. The survey’s precise methodology and timeframe are not detailed in the material provided, so those figures should be read as reported survey results, not as a definitive measure of AI’s economy-wide effect.
“AI has delivered zero financial payoff to corporate America outside of tech.”
— Torsten Slok, Apollo chief economist, in the October 24 Daily Spark note
What the Aggregate Data Cannot Show
Apollo’s charts do not establish that AI caused the differences in revenue per worker or margins between the largest technology firms and other S&P 500 companies. A flat average can conceal companies or teams that gained while others did not, and the comparison cannot separate AI effects from changes in labor costs, interest rates, tariffs or other business conditions.
It is also unclear how much of Meta’s listed commitments will ultimately be spent, when the capacity will become operational, and what share is directly attributable to AI. The company’s filing includes leases, cloud capacity, servers and network equipment, as well as some non-AI activity. More broadly, the available source material does not provide a common accounting method for comparing AI investment and returns across companies. Whether current spending will translate into durable profits remains unresolved.
Evidence to Watch in Coming Reports
The next evidence will come from company earnings, capital-spending disclosures and productivity data. Investors will be looking for signs that AI-related revenue, cost reductions or output gains are growing enough to offset infrastructure and operating costs. Companies’ updates on spending plans and financing will also indicate whether the buildout continues at its current pace.
For now, the central issue is not whether some AI applications can improve particular tasks; the cited examples indicate that they can. It is whether those gains can spread across businesses and appear in financial results. Apollo’s analysis gives one current reading of broad indicators, while the relationship between spending and returns remains a developing story.
Key Questions
What does Slok mean by “zero financial payoff”?
It is his description of aggregate indicators showing flat revenue per worker and margins for the 493 S&P 500 companies outside the Magnificent Seven, alongside productivity measures that Apollo says show no AI boost yet. It is not a finding that every company or AI project has had zero returns.
Does Apollo prove that AI is not improving productivity?
No. The note says the cited broad productivity measures show no visible boost so far. Those measures may not capture every company, task or early-stage benefit, and the comparison does not isolate AI from other economic factors.
What is included in Meta’s roughly $825 billion in commitments?
Meta’s filing lists about $355.73 billion in leases not yet started and $454.52 billion in non-cancelable contractual commitments, plus up to $14.72 billion in contingent cloud capacity purchases. The commitments cover infrastructure such as data centers, cloud capacity, servers and network equipment, and are not all AI-related.
Are there examples of AI producing measurable gains?
Yes. The source material cites customer-support research reporting 14% more issues resolved per hour, and 34% for novice workers. Retail experiments ranged from no sales increase to a 16.3% increase, depending on the workflow. Those results do not establish that the same gains apply across companies or sectors.
Source: ThorstenMeyerAI.com
