IA · 25 October 2026 · 4 min read
Zero Financial Payoff for Ordinary Enterprises: Apollo’s Warning on the AI Productivity Mirage
In brief: Despite hundreds of billions of dollars invested across corporate America, artificial intelligence has delivered zero tangible financial payoff to companies outside the technology sector. A macro report published by Apollo Global Management chief economist Torsten Slok analyzes the 'S&P 493'—the S&P 500 excluding the dominant tech infrastructure players—finding flat profit margins and stagnant revenue per employee. The findings suggest that raw algorithmic computation struggles to generate enterprise value where business judgment, accountability, and trust are paramount.
by Team Mocchi's
The Chasm Between Toolmakers and Enterprise Adopters
Over the past four years, boardrooms worldwide have poured massive capital into integrating artificial intelligence into corporate workflows, betting on promises of transformative productivity. However, while chipmakers and cloud hyperscalers report record quarterly revenues by selling compute capacity, the enterprise buyers using those tools are encountering a far colder economic reality.
This structural divide was laid bare by Torsten Slok, chief economist at private equity powerhouse Apollo Global Management. In an economic briefing reported by TechCrunch, Slok analyzed the balance sheets of the "S&P 493"—the standard index stripped of the seven dominant technology giants driving the infrastructure buildout (Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla). His conclusion is stark: outside Big Tech, generative AI has so far delivered zero financial payoff.
Flat Margins and Stagnant Revenue Per Worker
Apollo’s assessment centers on two foundational metrics of corporate efficiency: operating profit margins and revenue generated per employee. If the widespread rollout of copilot systems, autonomous coding assistants, and automated customer workflows had meaningfully transformed white-collar output, these figures should have reflected measurable structural gains.
Instead, Slok’s data reveals a completely flat line. Profit margins across traditional industries have not expanded beyond historical pre-AI averages, nor has revenue per worker demonstrated any positive divergence. Rather than serving as an immediate margin multiplier, enterprise AI adoption has largely added software subscription overhead and systems integration fees without moving the needle on bottom-line profitability.
The Calculation Trap Versus Practical Business Judgment
Slok’s critique goes deeper than cyclical adoption friction. He argues that corporate leadership has fundamentally conflated computational capacity with commercial value creation. Turning increasingly cheap machine intelligence into higher margins and faster revenue growth presents an intrinsic organizational barrier.
As the Apollo economist observed, modern AI functions as a remarkable calculator, but business value relies almost entirely on judgment. Algorithmic architectures do not comprehend legal accountability, executive trust, corporate culture, unwritten relationship rules, or ethical trade-offs. Because these intangible factors govern commercial counterparty risk, brand equity, and pricing power, simply automating routine data retrieval or draft generation fails to unlock economic leverage.
Mocchi's take
Apollo's findings highlight the perils of blanket enterprise adoption: subscribing to generic SaaS copilot seats rarely translates into measurable operational leverage. For companies navigating this transition, the takeaway is clear: artificial intelligence creates genuine return on investment only when embedded within purpose-built, custom software workflows that eliminate specific architectural bottlenecks. The economic upside does not come from attempting to automate strategic human judgment, but from freeing critical talent to exercise that judgment with superior data clarity.