AI Payoffs Are Rare Outside Tech, An Economist Says

🔍 Read the full analysis: AI Payoffs Are Rare Outside Tech, An Economist Says on ThorstenMeyerAI.com

Before you orderOffer from Amazon

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

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

Apollo chief economist Torsten Slok said an October 24 research note found no clear financial payoff from artificial intelligence outside the technology sector, based on broad measures of revenue per worker, profit margins and productivity. The assessment comes as companies commit large sums to computing infrastructure, but it does not establish that AI has delivered no benefit at individual firms or in specific tasks.

Apollo’s note, titled “No Signs of AI Payoff Outside the Tech Sector,” compared the Magnificent Seven 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 for the seven, while the corresponding measures for the remaining companies have been broadly flat. The note also points to productivity data from the San Francisco Federal Reserve that show no visible AI-related lift in productivity growth so far.

These figures describe aggregate economic and company trends; they do not isolate AI as the cause of any change. A flat average can mask companies or departments that gained and others that did not, and other forces—including labor costs, interest rates and tariffs—can affect margins and output. Slok’s “zero” framing is his interpretation of the broad data, not a direct measurement showing that every non-tech company received no return from AI.

The note also uses data from Ramp, a corporate spending platform with more than 85,000 U.S. business customers. Ramp’s figures indicate that the top 10% of firms account for nearly all recorded spending on model serving and inference, neocloud services, and products from OpenAI and Anthropic. Those figures characterize Ramp’s customer base; they are not a census of all U.S. businesses or a direct measure of the S&P 493.

The Gap Between AI Spending and Returns

The contrast matters because businesses and investors are weighing large, continuing infrastructure costs against returns that are difficult to identify across much of corporate America. Meta’s filings, cited in the source material, list about $825 billion in future commitments, including leases and contractual purchases for data centers, servers, cloud capacity and related infrastructure. That total is not all AI spending, and it represents commitments extending into future years rather than a single-year cash outlay.

For companies outside the technology industry, the Apollo findings are a warning against treating broad AI adoption as proof of higher productivity or profits. The results also do not rule out useful applications. The source material cites workplace studies reporting higher customer-support issues resolved per hour and sales gains in some retail experiments, with outcomes varying by task and workflow. Those examples are evidence of possible gains in defined settings, not proof of economy-wide returns.

The distinction affects decisions about budgets, staffing and investment. If firms cannot connect AI use to measurable improvements, software and computing costs may pressure margins before benefits appear. At the same time, aggregate data can lag behind changes inside individual operations, so the present lack of a broad signal does not settle whether returns will emerge later.

What Apollo’s Measures Capture

The October 24 note was covered by TechCrunch, PYMNTS and Bloomberg the following day, according to the source material. Apollo’s comparison focuses on financial and productivity indicators, rather than a company-by-company accounting of AI projects. Revenue per worker and margins show how the seven large technology firms compare with the rest of the index, but they cannot by themselves attribute those differences to AI.

Slok has also raised concerns about the economics of AI infrastructure. On October 20, he described the possibility that the marginal cost of computing could exceed the benefit users receive; on October 22, he referred to venture-funded token spending as “circularity.” Those are his assessments of the investment model, not findings that establish the return on every AI product or deployment.

The source material describes spending as concentrated among a small group of firms, many of them venture-backed. It also reports that some task-specific studies find gains, including a customer-support study with a reported average increase of 14% in issues resolved per hour and a larger gain for novice workers. Such results depend on the study, task and comparison used; they should not be read as a general productivity rate for all workers.

“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

The Apollo analysis does not isolate AI’s effect from other factors or report returns for every company, business unit or use case. It therefore remains unclear how many non-tech firms have obtained measurable gains, how large those gains are, and whether early benefits will appear in company-wide financial results over time. The source material also cautions that Ramp’s spending data reflect its own customers, not the full corporate sector.

The reported productivity indicators have not yet shown an AI-related lift, but that does not prove AI had no effect on specific tasks. Nor do task-level studies establish that measured improvements will persist, cover their costs or scale across different workplaces. The timing and distribution of any broader financial payoff remain uncertain.

Evidence to Watch in Company Results

The next useful evidence will come from company filings and operating data showing whether AI deployments reduce costs, increase output or generate revenue beyond the cost of software, chips and computing capacity. Investors and executives will also watch whether capital spending and contractual commitments keep rising, and whether companies can explain how those investments are expected to produce returns.

Future productivity data may offer a broader test, though they will not on their own identify AI as the cause of any shift. For now, Apollo’s note supports a limited conclusion: an economy-wide payoff outside tech is not yet visible in the measures it examined, while returns in particular companies and workflows remain possible and uneven.

Key Questions

What did Apollo economist Torsten Slok conclude?

In an October 24 note, Slok said aggregate measures show no clear financial payoff from AI outside the technology sector. The conclusion reflects broad indicators, not a review proving that every non-tech firm has received no benefit.

What data did Apollo use?

The note compared revenue per worker and profit margins for the Magnificent Seven with the other 493 S&P 500 companies, and cited productivity indicators from the San Francisco Federal Reserve. It also discussed AI spending data from Ramp, whose figures cover its own business customers.

Does the report prove AI has no value outside technology?

No. The aggregate comparison cannot separate AI’s impact from other economic factors or show results for every company and task. The source material also cites studies reporting gains in some customer-support and retail workflows.

Why are companies’ AI costs part of the story?

Building and accessing AI systems can require substantial spending on data centers, servers, chips and cloud capacity. Meta’s filings list about $825 billion in future commitments, though that total spans multiple categories and is not all AI spending.

Source: ThorstenMeyerAI.com