Key Takeaways

  • Doomsayers warned that AI would rapidly eliminate offshore labor, but sectors like Philippine call centers have remained remarkably steady.
  • Joe Weisenthal's framing of “AI as normal technology” explains how major technical shifts integrate gradually rather than detonating labor markets overnight.
  • Alarmists like Tristan Harris warned of sudden economic wipeouts, but call centers account for only about 3.5% of the total Philippine job market.
  • The internet altered every desk job on Earth without creating a visible spike in macro productivity data, and AI is following the exact same curve.

The Ghost of the White-Collar Wipeout

Two years ago, tech commentators treated the white-collar labor collapse as an established fact. The narrative was simple: generative models would instantly write code, draft copy, and answer support tickets, leaving millions of knowledge workers and offshore teams without a paycheck.

That catastrophe never arrived. As John Coogan pointed out on TBPN, the conversation drifted quietly away from mass layoffs toward broader debates about safety and existential risk because the initial panic failed to materialize: “We sort of moved past the job apocalypse SAS apocalypse narrative which was predicted from somewhat of the same community into actual apocalypse.”

Look at the offshore markets that pundits claimed were directly in the blast radius. Coogan cited warnings from prominent alarmists: “I remember seeing, I think it was Tristan Harris on Modern Wisdom, he is now sounding the alarm bells about existential risk, but he was saying something like the Philippines would see 90% of their economy go away.”

That prediction fell apart on basic arithmetic. As Coogan noted, “Call centers are actually only like three and a half percent of the Philippines job market.” Even within that segment, demand did not vanish. Companies did not fire their offshore operations overnight. They gave them software to handle higher ticket volume, faster turnaround times, and messy edge cases.

The Internet Precedent: Normal Technology

To understand why mass unemployment never happened, Coogan pointed to financial journalist Joe Weisenthal's concept of "AI as normal technology."

“Was in a conversation yesterday with some folks during which it became necessary to distinguish the people who see AI as normal technology from the AI as normal technology people,” Coogan said. The core of the thesis is simple: when real technology enters the economy, it diffuses through existing systems instead of vaporizing them.

Think about the rollout of the internet in the late 1990s and early 2000s. It completely restructured communication, commerce, and media. Yet economists spent decades arguing over the productivity paradox. As Coogan asked, “Where is the economic impact of the internet? Why can't you see a kink in the graph of really any economic data when the internet takes off? It's not like productivity went way up.”

When a tool makes a task five times faster, companies rarely eliminate 80% of their staff. Instead, the baseline volume of work expands. An engineer with an AI coding assistant writes more tests, refactors older codebases, and builds features that were previously left on the backlog. The total demand for output absorbs the efficiency gain.

What to Do With This

Stop freezing headcount decisions in anticipation of autonomous software running your operations next quarter. Audit your customer support and engineering workflows this week: if you are holding off on hiring support reps or junior engineers because you expect an AI agent to replace them entirely by year-end, pull that backlog item forward. Hire the talent, equip them with modern tooling, and plan for higher operational output rather than a phantom zero-employee team.