The Magnificent Seven are printing money with AI while the rest of the S&P 500 burns capital on chatbots that don't move the revenue needle.

The Summary

The Signal

Apollo's Torsten Slok just said the quiet part loud: most companies are burning money on AI with nothing to show for it. This isn't some academic pondering about future returns. This is a partner at one of the world's largest investment firms looking at actual financial data and seeing a clean split between the AI haves and have-nots.

The numbers tell a brutal story. Since 2022, the Magnificent Seven have seen revenue per employee rocket from $550 to nearly $800. The other 493 companies in the S&P 500? Still stuck at $550. Same story for profit margins and productivity. Tech companies are capturing all the gains while everyone else is just spending.

"Mainstream corporate America is starting to view AI spending not as a source of growth but as a cost center that weighs on margins."

This matters because it explains the infrastructure boom we've been watching. Nvidia, Meta, Microsoft, they're not just selling AI tools. They're the only ones who've figured out how to make AI economics work at scale. They build the infrastructure, they capture the productivity gains, they bank the margin expansion. Everyone else is:

  • Paying for enterprise AI tools that don't move revenue
  • Hiring ML engineers who can't find problems worth solving
  • Building chatbots that reduce support costs by 3% while eating 8% in implementation
  • Announcing AI initiatives to satisfy boards who read headlines

The Apollo report comes as corporate AI spending continues to climb, making the lack of returns even more conspicuous. Companies are doubling down on a bet that hasn't paid off, hoping the next quarter will be different. That's not strategy. That's momentum.

Here's what the investment world is noticing: if you're not building the picks and shovels, you're probably not striking gold. The AI gold rush is real, but most companies are buying maps to mines that don't exist yet. Meanwhile, Nvidia is selling shovels at $30,000 each and posting record quarters.

The Implication

If you're running AI strategy at a non-tech company, Apollo just handed you ammunition for the budget conversation. Stop chasing vague "AI transformation" and start asking harder questions: which specific process will see measurable productivity gains, what's the timeline to margin improvement, and why should we build this instead of buying from someone who's already solved it.

For investors, the message is clearer. The AI economy isn't diffusing across sectors, it's concentrating in infrastructure and platforms. That concentration won't last forever, but right now the smart money is following Apollo's data, not the hype decks. Watch for the first wave of companies to quietly wind down AI initiatives they announced 18 months ago. That's when the real reckoning starts.

Sources

Business Insider Tech | Bloomberg Tech | Hacker News Best