🚜 I looked up to my grandfathers a great deal growing up. They didn't work in offices. They farmed, sold tractors to those who farmed, and flew jet planes around the world.

None of them had a knowledge-work job in the modern sense — and none of them ever worried about being replaced by software, because what they did required them to be there. To show up with a body, a set of skills, and judgment you couldn't download.

They were enormously capable, highly adaptable men, and I always admired that.

I spent 35 years in corporate America since then. I have real respect for what a college education gives you — I wouldn't trade my own for anything. But I've also spent enough time in hiring rooms to know we've been over-indexing on the degree path and under-respecting the people who build, fix, fly, and grow things — who can do things with their hands.

This week, a piece of research came out that reframes that conversation.

On March 5th, Anthropic published a rigorous study on AI's real-world impact on jobs — not theoretical future impact, but what's actually happening right now based on how people are using AI tools in professional settings.

Their key finding — as we’ve been saying:

AI is already affecting many knowledge-work roles.

Computer programmers, customer service representatives, financial analysts, and data-entry workers show some of the highest exposure — because a large share of their daily tasks involve writing, summarizing, coding, researching, or analyzing information.

And at the other end of the spectrum? Those most immune to AI-tilted job loss:

Cooks. Electricians. Mechanics. Plumbers. Construction workers. Lifeguards. Healthcare aides.

The researchers found that roughly 30% of occupations show little to no measurable AI use today.

That number isn't a consolation prize. It's a competitive advantage — and it belongs largely to people doing physical, presence-required, hands-on work.

📌 Meaning: The labor market is quietly sending a signal our culture has been slow to hear. The jobs that require you to show up — in body, in skill, and in real-time judgment — are the jobs where AI currently plays the smallest role.

I want to say something I don't think gets said enough in career conversations with young people:

Working with your hands is not a fallback.

It is not settling.

For the right person — someone who loves building things, working outdoors, solving physical problems, making things run — it is the ideal path. And right now, it also happens to be one of the most economically durable paths available.

A licensed electrician. A plumber running their own crew. An HVAC technician. A welder certified in specialized materials.

These are not low-skill jobs. They are high-skill, high-demand, increasingly well-compensated careers — and the pipeline feeding them is thin because for two decades we told every smart kid to go to college instead.

The result is predictable: while many entry-level office roles are becoming more competitive, the U.S. already faces hundreds of thousands of open skilled-trade positions.

Many companies serving these industries are still struggling to find people. Meanwhile, the entry-level analyst pipeline was overflowing.

but that won't last forever. Trade and non-white-collar jobs will become more competitive and more difficult to secure.

The market is correcting. The research this week is a data point in that correction.

📌 Meaning: If your kid loves being outside, loves building or fixing things, loves working with their body and seeing tangible results at the end of a day — the trades are not the second choice. They may be the smart choice. And a deeply fulfilling one.

Here's where I want to bring this together, because the real story isn't "AI is bad for some people and irrelevant to others."

The real story is this: AI is a tool. And the people who learn to use it well — in any field — will have an edge over those who don't.

A plumber who uses AI to streamline estimates, manage scheduling, and handle the business side of their operation will outcompete the plumber who doesn't. A contractor who uses AI to draft proposals, research materials, and manage projects gains the same advantage.

The same goes for knowledge-work careers. An analyst who treats AI as a thinking partner — using it to go deeper, faster, and with more insight — will outperform the one who sees it purely as competition.

The research found that AI is currently doing more assisting than replacing across most roles. The gap between what AI could theoretically do and what it's actually doing is still wide. The workers who close that gap on their own terms — who pull AI into their workflow strategically — are the ones building durable careers.

📌 Meaning: "Should my kid learn AI?" is the wrong question. The right question is: How does AI make my kid better at what they already love doing?

Try this conversation — and let it go wherever it goes:

No agenda. Just a question. Some kids will shrug. Others will quietly file it away. A few might surprise you with how much they've already been thinking about it.

I love cooking and fixing and building things — because I didn't do a lot of that in the course of my white-collar career. It’s novel to have something tangible at the end of your effort.

The men I grew up admiring never put "resilient to technological disruption" on a business card. They just showed up every day, did skilled work, and took pride in results you could see and touch.

That's not nostalgia.

In 2026, it may be one of the smartest career strategies available.

Whatever path your kid is on — trades, college, somewhere in between — the move is the same: get good at something real, learn to use the tools available, and build a career around what makes them want to get up in the morning.

That principle hasn't changed.

— Barry

Found this useful? Forward it to a parent who needs it. If someone sent this your way — welcome. You can subscribe here.

Source: Anthropic, "Labor market impacts of AI: A new measure and early evidence," March 5, 2026 — anthropic.com/research/labor-market-impacts