They say AI will crush knowledge work but trades will boom. "Just become a plumber." They forgot the part where the people who pay plumbers are the ones losing their jobs.
A new consensus is forming among tech CEOs, investors, and business pundits: AI will obliterate white-collar knowledge work but physical labor is safe. The pitch is simple — get your kids into trades, not college. Plumbers will outearn lawyers. Electricians will thrive. Blue-collar is the new gold.
These are not fringe voices. They are the most powerful people in technology and business, speaking from stages, podcasts, and Fortune covers.
They are also wrong. Not about AI's destructive power — that part they have right. They are wrong about what happens next.
Let's hear it in their own words.
Every one of these predictions treats AI displacement as though it happens in a vacuum. As though you can collapse half the knowledge economy and everything else stays the same.
It doesn't.
Here's what actually happens when AI replaces half of white-collar work:
Lawyers, accountants, paralegals, analysts, marketers, copywriters, designers, junior developers, customer service reps — the jobs that pay $60K–$200K and form the backbone of the middle class.
These were the people buying houses, renovating kitchens, calling electricians, hiring plumbers, eating at restaurants, and paying for everything that funds the trades economy. Their income is gone. Their spending follows.
Nobody is calling a plumber when they can't make rent. Nobody is renovating a kitchen they're about to lose. The "labor shortage" in trades evaporates because the customers evaporated first.
Millions of laid-off knowledge workers, told to "learn a trade," do exactly that. Vocational enrollment surges (already up 16% in 2023). The trades labor shortage becomes a trades labor glut.
More supply of tradespeople + less demand for their services = wages plummet. The very thing they promised would be the safe harbor becomes another flooded lifeboat. Plumbers don't outearn lawyers — nobody earns anything.
The US economy is 70% consumer spending. When the people who do the spending lose their income, the entire downstream economy contracts. This isn't theory — it's what happened in 2008, except AI displacement is structural and permanent, not cyclical.
Khosla says "$15 trillion of U.S. GDP is labor that will mostly go away." That's not a knowledge-work problem. That's a civilization-level demand collapse. And he frames it as "deflationary abundance" — as though making everything cheap matters when nobody has income to buy it.
Consider what happens to the construction and trades ecosystem specifically:
| If This Happens... | Then This Follows |
|---|---|
| 50% of white-collar jobs eliminated | Mortgage defaults spike; housing demand crashes |
| Housing market contracts | New construction halts; renovation spending dries up |
| Construction freezes | Electricians, plumbers, HVAC techs have no projects |
| Displaced workers retrain into trades | Supply of tradespeople surges into a market with no demand |
| Oversupply + no demand | Trade wages collapse to minimum viable levels |
The trades don't exist in isolation. They exist because employed people with disposable income need physical things built, maintained, and repaired. Remove the income, and you remove the trades.
Ask yourself: who profits from telling you trades are safe?
If the public believes "just learn a trade" is a solution, there's no urgency for regulation, automation taxes, or displacement programs. The AI companies get to automate without consequence.
If AI displacement has a clean "solution" (trades!), then the $7T+ invested in AI doesn't face regulatory headwinds. Khosla's "abundance" narrative protects his portfolio.
If displaced workers can "just become plumbers," then mass layoffs aren't morally fraught. It's just the market rebalancing. The CEO gets to cut headcount and sleep at night.
If the trades narrative holds, governments don't need to implement automation taxes, transition income, or structural reform. They can point at vocational programs and call it done.
The "trades will boom" narrative is convenient for every powerful entity that wants to automate without accountability. It's not economic analysis. It's a sedative.
This isn't the first time the powerful have promised that displacement would sort itself out.
"New jobs will replace the old ones." They did — eventually. After 40 years of poverty, child labor, and social upheaval. The transition cost was paid entirely by workers, not factory owners.
"Just move to California." Millions did. Wages collapsed from oversupply. Hoovervilles formed. The "just go where the jobs are" narrative produced misery at both ends.
"The service economy will replace manufacturing." It did — with jobs paying half as much, no pensions, no unions, and no stability. The Rust Belt never recovered.
"Learn to code." The mantra for every displaced factory worker. Twenty years later, AI is coming for the coders. The cycle accelerates.
"Just become a plumber." The latest version of the same lie: that structural economic upheaval has a simple individual solution. It doesn't. It never has.
Every technological revolution produces the same narrative: "the displaced workers will find new work." And every time, the transition is paid for in decades of human suffering while the owners of the new technology accumulate the gains.
Here's the honest timeline if the current path continues unchecked:
Entry-level knowledge jobs contract. Companies post record profits from AI efficiency. Layoffs are called "restructuring." Media runs "become a plumber" stories. Vocational enrollment surges.
Consumer spending drops measurably. Retail, restaurants, and services contract. Housing market softens. Trades start feeling it — fewer renovation calls, fewer new builds. But the narrative hasn't caught up yet.
Newly trained tradespeople graduate into a market with shrinking demand. Competition for remaining work drives prices down. The plumber-earns-more-than-lawyer narrative dies in the data. Structural unemployment hits 15%+.
Governments scramble for emergency measures. The people who built the AI fight regulation. Social cohesion fractures. Eventually — after enormous human cost — a human-first economic model is forced into existence. Not because the powerful chose it, but because the alternative is civilizational collapse.
Through FairMind's Value Dynamics Model, the "AI Delusion" is a textbook case of Dimension Collapse — where optimizing one value dimension (Functional efficiency) destroys the other three:
| Value Dimension | What the Delusion Ignores |
|---|---|
| Sentimental (a) | Human identity is tied to purpose and work. Mass displacement doesn't just reduce income — it destroys meaning. "Just become a plumber" dismisses the psychological devastation of forced career upheaval. |
| Intrinsic (b) | Physical and mental health deteriorate with unemployment. Depression, substance abuse, family breakdown, and suicide rates all spike in displaced populations. The trades "solution" ignores the 5–10 year human cost of transition. |
| Functional (c) | The entire argument is about Functional value — who can do what task. But Functional value requires demand. A plumber's skill has zero Functional value if no one can pay for it. |
| Compressed (d) | Decades of accumulated human expertise — in law, medicine, finance, engineering — are being compressed into AI parameters and sold back as a service. The knowledge workers' lifetime of skill development is liquidated, not preserved. |
The AI Delusion exists because it's easier to say "become a plumber" than to do what's actually required:
These are structural solutions. They require policy courage. They threaten AI company margins. That's why you're hearing "become a plumber" instead.
The richest and most powerful people in technology are telling you that AI will destroy knowledge work and that the solution is physical labor. They are modeling half the equation and ignoring the other half.
When the people who pay for services lose their income, the services die too. This is not complicated economics. It is arithmetic.
You cannot collapse the demand side of an economy and expect the supply side to thrive. An economy of plumbers with no customers is not an economy. It is a fantasy told by people who will never need a job.
The return to a human-first economy is inevitable. Not because the powerful will choose it — they won't — but because the math will force it. The only variable is how much damage is done first.
It is going to get really bad before it gets better. Not because it has to. Because the people with the power to prevent it are too busy selling the delusion.