Apollo economist: AI has delivered zero financial payoff outside tech

A top economist says the AI boom has not paid off for most of corporate America. Torsten Slok, chief economist at Apollo Global Management, said that for all the heavy spending and hype, AI has produced zero financial payoff for U.S. companies outside the tech sector, TechCrunch reported.

His case rests on two simple measures. For the S&P 493, profit margins look flat, and so does revenue per worker, according to TechCrunch's account of his findings.

Key facts

• Slok's Daily Spark note said mainstream companies are beginning to treat AI token spending as a cost center that weighs on margins.
• He looked at what he calls the S&P 493, which is the S&P 500 without Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla.
• Profit margins for that group look flat, TechCrunch reported.
• Revenue per worker for the group also looks flat.
• Slok found no evidence that AI is boosting productivity, TechCrunch said.

Background

Slok made his case in Apollo's Daily Spark note. In it, he argued that companies outside tech are starting to see what they spend on AI tokens differently. Rather than a driver of growth, he wrote, many mainstream firms now view that spending as a cost center that drags on margins.

The note splits the market in two. Slok focuses on the S&P 493, a group made by removing Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla from the S&P 500. Leaving out those seven names lets him look at ordinary large companies rather than the firms that build and sell AI.

TechCrunch reported that Slok's recent note found AI has not improved the bottom line of large companies outside tech. The outlet's weekend editor, Anthony Ha, wrote the story.

TechCrunch also flagged a link between the two firms. It disclosed that Apollo, a private equity giant, once owned TechCrunch.

Slok is not alone in his doubts. TechCrunch noted that other recent studies have also found that AI has not yet produced major productivity gains.

Why Slok is skeptical

A common reply to findings like these is that non-tech firms are just slow to adopt. Slok did not buy that idea, arguing that corporate America will find it hard to turn ever-cheaper intelligence into faster revenue growth and wider margins, TechCrunch reported.

His reasoning centers on what AI lacks. He wrote that AI does not grasp accountability, trust, culture, unwritten rules, ethical trade-offs or fairness. Those factors shape reputational risk and drive revenue and margins, he argued.

In short, Slok sees business success as depending on human judgment that current AI tools do not supply. That is why, in his view, cheaper AI alone has not moved the numbers for most big companies.

The numbers behind the gap

New figures show how wide the split is. Adjusted for inflation, each Magnificent 7 employee brings in about $780,000 in revenue, versus roughly $525,000 at S&P 493 firms, per Slok's data, as Stocktwits reported on Yahoo Finance.

Margins tell a similar story. Margins at the Magnificent 7 are now close to 30%, compared with about 10% across the S&P 493, according to Slok's figures.

AI spending is also concentrated. By his count, 99.5% of model serving and inference spending, and 99% of neocloud spending, comes from just the top 10% of firms, Yahoo Finance reported. Venture backing is common at the very top: 80.4% of the biggest 1% of AI spenders have it, against 13.1% of median spenders.

Slok drew a conclusion from that. In his view, as long as agentic AI and token spending sit in a narrow corner of the economy, the wider economic impact stays very small; the real test, he said, is the S&P 493 rather than the Mag 7.

What it means for you

For investors, the note draws a line between the companies that sell AI and the ones that buy it. The seven firms Slok removed are Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. For the remaining 493, he sees flat margins and flat revenue per worker.

For workers and managers at non-tech firms, the takeaway is that AI spending has not yet shown up in the measures Slok tracked. He found no evidence of a productivity lift.

There are limits to what the note shows. Apollo's note presents its data in charts; the dollar and margin figures above come from Yahoo Finance's account of Slok's data. It is also not known whether the flat trend holds in every industry in the group Slok studied.

What happens next

The central question is whether payoffs arrive later. It is not known whether AI gains will appear as models get cheaper. Slok himself doubted that firms are simply lagging.

Spending decisions are another open question. Slok wrote that mainstream corporate America is pulling back, now seeing token spending as a cost center. The note does not say how deep any cuts are.

FAQ

Who is Torsten Slok? He is the chief economist at Apollo Global Management.

What group did Slok study? It is the S&P 500 minus Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla.

What did he find? Profit margins and revenue per worker look flat for that group, and he found no sign that AI is raising productivity.

Does he expect a turnaround? He was skeptical, saying it will be tough for companies to turn cheap intelligence into higher growth and margins.

Are firms cutting AI spending? Slok wrote that mainstream corporate America is pulling back, but the note does not quantify the cuts.