🔍 Read the full analysis: An Economist Explains The Divide Between AI Spend And Returns on ThorstenMeyerAI.com
Get the little things that make your day 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 technology companies, though that is an interpretation of broad trends, not a direct measurement of AI’s effect. Meanwhile, Meta has disclosed about $825 billion in future commitments, and examples from customer support and retail show measurable gains in some specific workflows.
Apollo chief economist Torsten Slok said in an October 24 research note that artificial intelligence has shown no clear financial payoff for corporate America outside technology, even as companies commit increasing sums to data centers, servers and computing capacity. The conclusion reflects trends in broad company-level measures, not a direct test proving AI has produced no gains at individual businesses.
Apollo’s note, titled “No Signs of AI Payoff Outside the Tech Sector,” compares the seven largest technology companies—Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla—with the other 493 companies in the S&P 500. Apollo’s charts show revenue per worker and profit margins rising among the seven, while the figures for the other companies have remained broadly flat. The charts put the seven’s revenue per worker near $780,000 in the third quarter of 2026, compared with about $525,000 for the other 493; their margins are around 30% and 10%, respectively.
The note also points to San Francisco Federal Reserve productivity measures, which Apollo says show no sign yet of AI boosting productivity growth. Slok’s “zero financial payoff” wording is a conclusion drawn from these aggregate indicators. A flat average across hundreds of companies cannot isolate AI’s effect, and it may conceal individual firms or departments with gains offset by losses elsewhere.
The investment figures underline the scale of the gap between spending and visible returns. Meta’s October 22 filing lists approximately $355.73 billion in leases not yet started and $454.52 billion in non-cancelable contractual commitments, largely for cloud capacity, servers and network equipment, along with up to $14.72 billion in contingent cloud purchases. The combined figure is about $825 billion. The commitments are not all AI-related, and the filing includes items such as Reality Labs hardware.
Meta reported third-quarter revenue of $64.60 billion, up 26% year over year, and capital expenditures of $34.91 billion. Operating cash flow was $32.79 billion, leaving free cash flow of negative $2.12 billion, compared with positive $10.6 billion a year earlier. The company narrowed its 2026 capital-spending outlook to $135 billion to $145 billion. Those numbers describe Meta’s finances; they do not establish that AI spending caused the change in free cash flow.
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 Cost of Scaling AI
The divide matters because AI infrastructure requires large, continuing investment, while many companies have yet to show a clear return in their overall productivity or margins. If the gains remain concentrated among technology firms or a small set of users, other businesses may face pressure to spend to keep pace without knowing whether the investment will improve their results.
Evidence from particular tasks is more positive than the broad company-level picture. The source material cites research finding customer-support agents resolved 14% more issues per hour with AI assistance, with a 34% increase for novice workers. Retail experiments showed sales changes ranging from no increase to 16.3%, depending on the workflow. These results suggest that benefits can be measurable when AI is applied to a defined process, but they do not establish the effect across entire companies.
The financial risk extends beyond whether AI tools work. Major technology companies are funding data centers and compute capacity at a time when the payoff is difficult to quantify. The source reports that 83% of nearly 9,000 executives said AI had not affected their productivity so far, while estimated average productivity growth was 0.2% a year. The survey figure reflects reported experience, not a forecast of what AI will deliver later.
Products Worth Considering
Blueair Blue Pure 211i Max Air Purifier
Garmin Forerunner 165 GPS Running Smartwatch, 43 mm, Black
Withings Body Scan Smart Scale with Segmental Body Composition and ECG, Black
As an affiliate, we earn on qualifying purchases.
What Apollo’s Measures Show
Apollo’s comparison is based on broad economic and corporate measures, including inflation-adjusted revenue per worker, profit margins and productivity data. It also cites spending data from Ramp, a corporate-card and spending platform with more than 85,000 U.S. business customers. In that dataset, the top 10% of firms account for 99.5% of model-serving and inference spending, 99.0% of neocloud spending and 95.6% of spending on OpenAI and Anthropic. These figures describe Ramp customers, not all U.S. companies.
The source material also says 80.4% of the top 1% of AI spenders in the Ramp data are venture-backed, compared with 13.1% of median spenders. Slok has described some venture-funded token spending as “circularity” and argued that the cost of compute can exceed the benefit users receive. Those are his assessments of the economics, rather than a settled finding about every AI business or deployment.
Meanwhile, the source cites forecasts that Amazon, Alphabet, Meta, Microsoft and Oracle may spend nearly $780 billion in 2026 and more than $1 trillion a year from 2027 through 2030. These are projections, not reported spending or binding commitments. Meta’s approximately $825 billion figure is a separate total of future contractual and lease commitments, with components extending through 2036.
“Artificial intelligence has delivered zero financial payoff to corporate America outside of tech.”
— Torsten Slok, Apollo chief economist, in the October 24 Daily Spark note
Where the Payoff Is Still Unclear
Apollo’s comparison does not establish that AI caused the performance gap between the seven technology companies and the rest of the S&P 500. The data combine many influences, including labor costs, interest rates and other business conditions. Nor does a flat average for 493 companies show whether some have achieved gains that are being offset by weaker results elsewhere.
The scale and composition of future spending are also uncertain. Meta’s commitments include non-AI items, and forecasts for spending by large technology companies may change. It is not clear from the cited figures how quickly the investments will generate returns, how much of the spending will be passed on to customers, or whether the reported gains in selected tasks will persist when AI is used more widely.
The source material says SpaceX borrowed $40 billion to buy AI chips and reports that credit protection tied to Meta and CoreWeave became more expensive. Those details point to financing pressures, but the supplied information does not establish the full terms of the borrowing or whether higher insurance costs signal a lasting change in lenders’ views.
Returns and Financing to Watch
The next test will be whether companies can turn improvements in particular tasks into sustained gains in revenue, margins or output per worker. Investors and executives will also be watching upcoming financial results and productivity data for signs that the benefits have spread beyond a narrow group of technology businesses.
For Meta, future reports can show how capital spending, cash flow and contractual commitments change as the company builds out infrastructure. More detail on how much capacity is used for AI and what revenue or cost savings it supports would help clarify the economics. For other companies, evidence from controlled deployments and comparable performance measures may show whether gains extend beyond customer support and selected retail workflows.
Financing conditions are another developing indicator. The source material cites a forecast from Elon Musk that more than 200 neocloud providers could shrink to a few, but it does not provide evidence that such a consolidation has occurred. Whether computing demand can support the number of providers and the scale of investment remains an open question.
Key Questions
Did Apollo prove that AI produces no financial returns outside technology?
No. Apollo’s note interprets broad measures such as revenue per worker, margins and productivity. Those figures do not directly isolate AI’s effect or rule out gains at individual firms.
What does Meta’s $825 billion figure represent?
It combines about $355.73 billion in leases not yet started, $454.52 billion in non-cancelable contractual commitments and up to $14.72 billion in contingent cloud purchases. The commitments are mostly related to data centers, cloud capacity, servers and network equipment, but are not all AI spending.
Are there examples of measurable gains from AI?
Yes. The source material cites a customer-support study showing 14% more issues resolved per hour, rising to 34% for novice workers, and retail experiments with sales effects ranging from zero to 16.3%. Results varied by task and workflow.
What remains uncertain about the AI investment boom?
The timing and scale of returns remain unclear. The available figures do not show how much broad company performance is caused by AI, how much of Meta’s commitments is specifically for AI, or whether gains in selected tasks will hold at larger scale.
Source: ThorstenMeyerAI.com
