Analysis
Apollo Global Management's chief economist, Torsten Slok, put a number on the AI-skeptic case this week without actually giving a number: in a note titled "No Signs of AI Payoff Outside the Tech Sector," he argues AI has delivered zero financial payoff to corporate America outside of tech, full stop. TechCrunch covered the note, and the framing is worth taking seriously precisely because Apollo is not a disinterested observer -- it is one of the largest credit and equity providers funding the AI infrastructure buildout this note is implicitly questioning.
Slok's method is to strip out the seven companies, Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla, that account for most of the AI capex and most of the AI revenue story, then look at what is left: the S&P 493. There, he finds profit margins and revenue per worker essentially flat, with no visible productivity lift from a technology companies have been investing in for several years now. His own words are blunt -- AI "may be a great calculator, but in business, intelligence is mostly judgment," and it "does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness."
Room for disagreement: Slok's analysis is indirect. Flat aggregate margins and flat revenue per worker across hundreds of companies could reflect AI genuinely not paying off yet, or it could reflect AI gains being offset by unrelated cost pressures (tariffs, labor, financing costs) that would have dragged margins down even further without any AI contribution at all. The note cites no company-level AI-spend-to-return data, no survey, and no sector breakdown -- it is a macro proxy, not a controlled study, and a single quarter or two of margin improvement at a handful of S&P 493 companies could flip the conclusion without any change in the underlying trend.
What the note gets right is the question it is forcing: every AI-infrastructure financing deal, every enterprise software vendor pricing in an AI premium, and every startup pitch deck assuming AI-driven margin expansion is implicitly betting against Slok's data. If he's even partly right, the gap between the seven AI mega-caps and everyone else trying to monetize AI internally is the real story of this cycle, not the compute buildout itself.