Apollo Global Management chief economist Torsten Slok has delivered one of the bluntest verdicts of the AI boom so far: outside the technology sector, artificial intelligence has yet to pay for itself. In a note reported by TechCrunch on October 25, 2026, Slok argued that “the technology has delivered zero financial payoff to corporate America outside of tech,” despite massive investment and widespread hype.
What happened
Slok’s test is simple. He splits the S&P 500 into two groups: the seven mega-cap technology companies — Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla — and the other 493 companies, which he calls the S&P 493. For the Magnificent 7, revenue per worker and profit margins are rising. For the S&P 493, both measures look flat, and Slok finds no evidence that AI is lifting productivity growth.
The gap is wide. Secondary coverage of the note puts inflation-adjusted revenue per employee at roughly $780,000 for the Magnificent 7, against about $525,000 for the other 493 companies, with profit margins near 30% for the tech giants and around 10% for the rest. Those figures describe the performance gap; Slok’s “zero payoff” claim is his interpretation of what that gap means for AI’s broader economic return.
Slok is skeptical that this is just a matter of non-tech companies lagging behind. He argues it will be tough for corporate America to “turn increasingly cheap intelligence into higher revenue growth and higher profit margins.” In his words: “AI may be a great calculator, but in business, intelligence is mostly judgment.” He adds that AI “does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins.”
Two cautions matter before anyone quotes the headline. First, flat aggregate figures do not prove that no individual company benefits: gains at some firms can be offset by losses elsewhere, or can take time to show up in company-wide numbers. Second, the note’s own summary closes with a question rather than a verdict for buyers: “The bottom line is that the AI story will be decided in the S&P 493, not the Mag 7, and so far, there is no evidence that revenue growth or margins are improving outside the tech sector.”
Why it matters for small business
Small businesses are not in the S&P 493, but the warning still lands. Most SMBs buy AI rather than sell it, so they sit on the cost side of Slok’s divide: subscription fees, implementation time and staff training go out now, while any return depends on work actually changing. If even large companies with big budgets cannot yet show a payoff in their margins, a small team should be suspicious of any vendor pitch that promises an automatic return.
The practical lesson is to treat AI spending like any other operating cost, not like a magic multiplier. A tool that saves an hour a week is only worth paying for if that hour turns into billable work, faster delivery, fewer errors or a better customer experience — and if someone actually measures it. Slok’s note also points to a second risk: mainstream companies are, in his words, starting to view token spending “not as a source of growth but as a cost center that weighs on margins.” For a small business, that is exactly the trap to avoid: a stack of AI subscriptions that nobody owns, nobody measures and nobody would miss if cancelled.
Pricing, availability and what to do next
This story is not a product launch, so there is no price or release date to report — and no vendor claim to take on trust. Instead, here is a short checklist to run before renewing or expanding any AI subscription:
- Name the job. Write down the single workflow the tool is supposed to improve, such as answering common customer questions, drafting quotes or reconciling invoices.
- Set a baseline. Record how long that job takes today, what it costs, and how often it goes wrong.
- Measure for 30 days. Compare the same numbers with the tool in use, including the time spent fixing AI mistakes.
- Keep what pays, cut what does not. If the measured saving does not beat the subscription cost plus staff time, cancel it and say so plainly.
- Put a human on judgment calls. Keep people in charge of refunds, hiring, pricing and anything that affects reputation — the areas Slok says AI does not understand.
If a tool passes that test, it has earned its place in your stack regardless of what any economist says about the market as a whole. If it does not, Slok’s note gives you permission to say so: paying for AI is not the same as profiting from it.
Sources
- TechCrunch, “AI delivers ‘zero financial payoff’ to most companies, Apollo economist says,” October 25, 2026: https://techcrunch.com/2026/10/25/ai-delivers-zero-financial-payoff-to-most-companies-according-to-apollo-economist/
- PPC Land, “Apollo finds zero AI payoff outside tech as S&P 493 margins stay near 10%,” October 25, 2026: https://ppc.land/apollo-finds-zero-ai-payoff-outside-tech-as-s-p-493-margins-stay-near-10/
