In 2011, Marc Andreessen wrote the essay that defined a generation of innovation: “Software is eating the world.” Fourteen years later, one of his partners at a16z, Alex Rampell, has updated the line for the AI era — “Software is eating labor.”

That single word swap changes everything.

For over a decade, software meant tools that helped people work faster — CRMs, ERPs, workflow platforms. But now, software isn’t just helping with the work. It’s starting to do the work.

The Real Market Isn’t SaaS — It’s Labor

Rampell’s comparison is blunt but brilliant. The global SaaS market pulls in about $300 billion a year, with a total market cap of roughly $2.2 trillion. But the U.S. labor market alone represents $13 trillion in annual spending.

That’s the real opportunity AI has unlocked — not in selling software subscriptions, but in capturing the economic value of human labor, one repetitive process at a time.

From Software as a Tool to Software as a Worker

This shift reframes how every business — from banks to manufacturers — needs to think about productivity. In the old model, digital transformation meant licensing new tools. In the new model, it means deploying digital labor.

AI systems can now perform the kind of structured work once reserved for entry-level analysts, admin staff, or operators: generating reports, managing orders, screening leads, or tracking compliance data. Each agent replaces not a person, but a task flow. And that’s the quiet revolution underway.

For Leaders: A New Kind of Workforce Planning

AI doesn’t just reduce costs. It reshapes organizations. Leaders now need to think in terms of hybrid workforces — part human, part digital. That means building processes where:

  • AI agents handle routine cycles with audit trails and transparency.
  • Humans make judgment calls and own relationships.
  • Both sides share context, data, and feedback loops.

The companies that master this balance will move faster than their peers — not because they have “AI,” but because they have restructured how work gets done.

What This Means for South Africa and Emerging Markets

Labor-heavy industries like manufacturing, retail, and professional services stand to gain the most. In many cases, productivity is held back not by effort, but by system inefficiency — the mess between data silos, outdated software, and manual reporting.

AI changes that. By connecting directly to the source — finance, CRM, HR, or production data — intelligent systems can surface real-time insights that are cleaner than any spreadsheet. It’s not just automation. It’s clarity.

For South African businesses, this is an opening. Our markets are flexible, our teams creative, and our cost bases lean enough to experiment. We can test and scale faster than most corporates burdened by legacy systems.

The Takeaway

AI’s real disruption isn’t in the tools we use — it’s in who (or what) is doing the work. As software starts hiring itself, leaders face a new kind of strategic question: not which AI to buy, but how to design work for a mixed human-and-machine team.

That’s the trillion-dollar shift — and the next decade’s biggest management challenge.

This piece draws on Alex Rampell’s 2025 a16z LP Summit talk, “Software is Eating Labor.”