The wage hierarchy of automation is inverting, and the trades may be winning. - Psychology Today

I want to tell you about two career predictions from the year I was born.

In the late 1980s, a young person looking at the labor market was told the prestigious path was a four-year college degree leading to white-collar work — law, finance, consulting, technology. The dwindling path was the trades. Plumbing, electrical, HVAC, mechanical repair, and construction. Those were industries to escape. Knowledge work was the future.

Forty years later, almost every assumption baked into that advice is being inverted. If you have a teenager in your house wondering what to do with their life, this is information worth having.

Moravec's Paradox

In 1988, the roboticist Hans Moravec noticed something strange. Computers could already beat humans at chess and prove mathematical theorems, but they couldn't reliably pick up a coffee cup or walk across a cluttered room. What's hard for humans is easy for computers, and what's easy for humans is hard for computers.

That observation, now called Moravec's paradox, has become the single most important fact about which jobs AI takes over and which it doesn't.

Document review, contract drafting, basic medical diagnosis, financial analysis, code completion — the cognitive work that requires years of education and pays six figures — turns out to be relatively tractable for AI. Folding laundry in a stranger's house, repairing a leaking pipe behind a finished wall, calming a frightened toddler, diagnosing a strange noise in an aging Camry — work that pays $25 an hour if you're lucky — turns out to be extraordinarily hard.

Knowledge work happens in a controlled environment with structured inputs and a vast training corpus. The physical world is unstructured, variable, and full of one-off problems. Every job a plumber walks into is a new job.

What the Next Decade Looks Like

A drone slides through a small access panel and sends video back to a tablet. An electrician in the truck identifies the short and powers down the right circuit before walking in. The dangerous part of the job is no longer for humans.

A roofer in Phoenix in August wears an exoskeleton that takes 40 pounds of strain off his lower back. He'll work another 15 years before his body gives out, instead of seven.

A long-haul trucker still has a human in the cab — partly because cargo insurers aren't yet willing to underwrite fully autonomous freight, partly because the last mile, from highway to loading dock, involves a thousand judgment calls a self-driving system can't yet make. But the human spends more of the drive supervising and less of it white-knuckling through traffic.

This is what most of the next 20 years of work in the physical world looks like. Not replacement. Augmentation that makes the work safer, the worker more productive, and the wage more defensible.

Why the Story Isn't Being Told

Journalists, policy researchers, and tech commentators are knowledge workers themselves. The displacement story is the story of their own peer group. It feels urgent because it is — for them.

"Robots will take the trades" makes for better headlines than "robots will help plumbers diagnose blockages faster." Disruption sells; quiet improvement doesn't.

And demography is forcing the issue. The median construction worker is 42, and Deloitte projects 41 percent of the current workforce will retire by 2031. The country needs more skilled tradespeople than it can produce. Automation in this sector isn't replacing scarce workers; it's helping the workers we have stay productive longer.

Who's Been Quietly Winning

A journeyman electrician now earns a median wage of $62,350 with zero student debt and four to five years of paid on-the-job training. A four-year college graduate enters the workforce with a median starting salary of $59,384 and an average of $39,000 in federal loans. The gap will widen.

What We Owe the Conversation

For parents wondering what to tell their teenagers: the trades are real work, dignified work, and increasingly well-paid work. The young person who becomes a master electrician at 28 will likely have a more stable career than the young person who graduates with a marketing degree the same year. That sentence wouldn't have been true in 2005. It's almost certainly true in 2026.

For policymakers, apprenticeship programs, community colleges, journeyman pathways, and skilled-trades credentialing deserve the same cultural status and policy attention we've poured into four-year degrees for 40 years.

For the rest of us, the work is internal. Many of us were raised in a status hierarchy that treated college-educated work as more important than work done with one's hands. That hierarchy was always wrong, and it is now also empirically backwards. The plumber who comes to your house on Saturday is doing harder work than most knowledge workers, in conditions most would refuse, with skills that took longer to develop than most master's degrees require.

The least we can do is notice.

References