Dirty Work: The Return of Blue Collar

ai is automating white-collar junior roles, but there are lots of blue-collar jobs out there, and many are resistance to the first wave of automation. status has to change.

I can’t stop thinking about the crisis of youth unemployment. The first rung of the white-collar career ladder is vanishing. It's not just an economic problem; it's a social time bomb. A generation without a stake in society becomes increasingly disconnected from its institutions and norms. It’s an urgent problem that needs urgent action. The good news? We already have the jobs. We just don’t respect them yet.

Lawrence xx

Note, Claude tell me many references in the essay will “confuse readers and distract from the core message.” lol.

Sarah's been working at the care home for five years. This morning she's with Mrs. Chen, who's 82 and deep into dementia. Mrs. Chen can't really talk anymore, but she's gripping Sarah's hand tight and looking uncomfortable. Sarah's gotten good at reading these signs. She adjusts the pillow again, gets a cool flannel for her forehead. Mrs. Chen finally relaxes.

This is what Sarah does all day. She's learned to understand what people need when they can't tell her. She makes £24,000 a year. Without her, the care home falls apart. She’s a contractor on a zero-hour contract, obvs.

Her brother James? He was looking at a £150,000 entry role as a graduate at a Magic Circle law firm. Was. They paused the graduate scheme last month while they better understand staffing requirements “going forward”. Truth is, they’ve gone all in on GPT-4 for contract review. And the Partners can prompt it 7-days a week for revisions. Another Partner says Claude Opus is even better, he tried it yesterday. The junior associate still makes fewer mistakes. But at 99% higher cost. The numbers add up.

Here's the thing: we've built our entire status hierarchy backwards. The more necessary you are, the less respect you get. If your job involves fixing things, building things, or caring for people, the work that actually keeps society running, we treat you like you've failed. If you shuffle papers and send emails, we treat you like you've made it. At least, we used to.

I: “Skilled” Work

For fifty years, we've told young people a consistent story. Pursue knowledge work over manual labour. Choose clean over messy. We directed our education systems around this premise. Universities produced lawyers, consultants, analysts. We called it "skilled" work. It's still called "skilled" work on the immigration forms.

The work we labelled "skilled" turns out to be automatable. Document review. Financial modelling. Marketing copy. Market research. Software programming. Tasks existing entirely in AI's native environment: digital, pattern-based, systematic. Meanwhile, the work we dismissed as "unskilled" represents sophisticated human capabilities. Intuition. Gut-feel. Emotional intelligence. Reading when a dementia patient is becoming distressed. Fixing a leaky roof on an 1890s Victorian semi. Finding the words that comfort someone facing their death. I mean many M&A lawyers do something similar I guess...

I keep hammering home this number: 78% of employers plan to reduce or eliminate graduate hiring because AI can do the work. Entry-level positions in law, finance, and consulting are rapidly disappearing. Things are going to get worse.

And yet. And yet, the UK has 31,000 unfilled nursing positions. 131,000 vacant care roles. 140,000 construction jobs seeking workers. Run the numbers for your own country, o3 will help, and you will find hundreds of thousands of unfilled jobs. But generally they are low-prestige and low-paid.

So we are on a sticky wicket. We've created a generation trained for contracting job markets while sectors that desperately need meat bodies go understaffed. You grew up wanting to be a lawyer. (I’m not sure why. Maybe Suits?) But the jobs are in plumbing. Choose your adventure.

II: Why We Pay Backwards

The psychiatrist earns three times what the psychiatric nurse earns. But when someone’s in crisis, who provides the hour-by-hour care? The person prescribing from an office, or the person who knows how to calm without condescension, how to approach without triggering fear? We clap for the carers, but forget them when budgets are set.

This isn’t just about nurses. It’s about the broader logic of our economy. Trace the payscale across any set of essential jobs, nurses, scaffolders, sanitation workers, bricklayers, care home staff, insulation installers: the closer the work is to physical reality, the lower the status and pay. The people who build our homes, keep our water clean, protect us from cold, or stay by our side in moments of crisis earn less and have lower social standing than those who can manipulate abstractions from a distance. You can get rich moving numbers. But not moving people, or bricks, or beds.

That’s not an accident. It reflects centuries of cultural bias. Since the Enlightenment, we’ve elevated mind over body, abstraction over sweat, rationality over touch. (Thanks Descartes) We built an economy that rewards distance from the material and emotional realities of life. And the further from death, dirt, or dependency your job is, the more you’re likely to be paid. The highest earners are the ones most insulated from the physical world. Ivory Towers, etc.

And then there’s gender. There always is. Many of the jobs that get dismissed as “natural” or “unskilled”, caregiving, cleaning, early childhood work, are disproportionately done by women. As if moving someone without causing pain is instinct, not expertise. As if managing fear or confusion is personality, not practiced competence.

And scale. A programmer writes code once, and it serves millions. A construction worker lays each brick by hand. An insulator installs each panel one wall at a time. So we equate value with reach, forgetting that some things matter precisely because they can’t be mass-produced. Comfort. Shelter. Safety. These are one-to-one human acts. They don’t scale. That’s why they’re precious. Do things that don’t scale?

III. Inversion

Professional prestige is a story we tell ourselves until the story stops making sense. Take the master mason. In the 1200s, they were social equals of nobility. Their knowledge, how to build vaults and cathedrals, was esoteric and irreplaceable. Then came the Industrial Revolution. Their secrets got codified, published in architectural treatises. Once mystery became manual, the architect became the gentleman and the builder, labour. Knowledge was no longer embodied; it was abstracted, and that abstraction got the prestige.

Then there is the Black Death. When half of Europe died, the labourers who survived suddenly had bargaining power. Wages did not rise, they exploded. Laws were passed to cap pay, but they were unenforceable. When work is essential and workers are scarce, the hierarchy flips.

AI is now inverting the pyramid again, but in reverse. Instead of too few workers for the fields, we have too many graduates for the desks. Each year, 2.4 million Americans, 500,000 Britons, and 3.3 million Europeans graduate with degrees aimed at a knowledge economy that is rapidly automating. GPT-4 can draft contracts and conduct basic research. Claude can summarise depositions, analyse legal risk, and write case strategy. Mid-level skills, once safe, are being eaten by cheap inference.

Here is the difference from past disruptions: AI is not spawning new pointless jobs to replace the old ones. When factories collapsed, we invented middle management. When middle management ballooned, we created consultants to manage the managers. Each automation wave invented a new layer of abstraction to soak up labour. This one will not. Or at least not much. One person can now orchestrate 1,000 agents. And soon one agent will orchestrate 1,000 more.

Meanwhile, it is the "unskilled" jobs that are hardest to fill. The care sector cannot hire fast enough. Construction is bleeding workers. Tradespeople are booked out for weeks. I cannot find a plasterer for love or money, and I have genuinely considered founding a plastering firm just to meet my own demand. But where do I get the staff?

This is not just a supply-demand mismatch. It is a status problem. Parents who sacrificed to send their kids to university cannot admit that the degrees might be functionally obsolete. Universities, whose entire business model is credentialling white-collar futures, cannot pivot overnight. Our cultural machinery, TV, media, school guidance counsellors, still glamorises consultants and lawyers, not plumbers and geriatric care specialists.

We are in the middle of another inversion. And like the others, it will not be led by thought leaders or policy memos. It will happen in the margins first. When a roofer can out-earn a project manager. When your cousin the HVAC technician buys a house before your friend the analyst. When the job no one wanted becomes the job no one can get. The last time this happened, it took a plague. This time, it is taking an algorithm. Or maybe, when we look back we will realise it took both.

IV: Status Inversion

If this analysis is correct, let me do some nearcasting. It’s January 2027.

Sarah's getting recruitment calls, but so is everyone who works with their hands. The electrician who rewired James's flat? He's booked solid for three months. Just raised his rates 40%. Again. The plumber who fixed the boiler at the care home? He's hiring apprentices, can't keep up with demand. Turning down jobs.

My neighbour’s son just chose trade school over university. His parents were mortified. Until they saw the starting salaries. Until they realized he'll be earning while his friends are accumulating debt. Until they understood he's learning skills that can't be uploaded to a server.

The dating apps tell the story. "Management consultant" used to be gold. Don Cheadle in House of Lies. Now it's a yellow flag. Meanwhile, "electrician" or "HVAC technician" signals stability, real skills, actual knowledge. Construction sites are changing too. The old image of the wolf-whistling builder? Gone. These are skilled professionals making serious money. Solar panel installers driving BYDs. Heat pump engineers with waiting lists like surgeons. Insulators—yeah, insulators—becoming the climate heroes, making Britain's leaky homes liveable.

Even the language is shifting. They're not "manual labourers" anymore. They're "critical infrastructure specialists." Not "tradesmen" but "technical specialists." The rebrand is happening naturally because the reality has changed. When you can't get someone to fix your roof for six weeks, you stop calling them "just a roofer." This is how status actually shifts. Not through think pieces or policy papers. Through who's buying houses. Through who feels secure enough to start a family. And probably through TikTok honestly.

V: Build, Baby, Build + Blue Collar

What was “white-collar” work, really? Most of it was Office Space. Moving digital bits, sitting in meetings about meetings, typing into boxes no one reads. Entire economies were built on roles that produced nothing you could touch, fix, or take pride in. Graeber called them bullshit jobs. He was right. Looking back, 1990 to 2025 may come to seem like the Great Age of Abstract Toil.

Now contrast that with the trades. Plumbers, electricians, roofers. They show up, solve a real problem, and leave the world better than they found it. They don’t optimise metrics, they fix broken boilers. They keep the lights on. These jobs offer something white-collar roles rarely did: visible impact. You can point to the thing you built or fixed.

Young people today have a choice my generation didn’t. Skip the bullshit. Go where the demand is. Build something that lasts. Learn skills that AI can’t imitate, let alone replace. Become indispensable. Earn a living. Earn respect. Make a tangible difference. Choose life?

If we really mean “Build, Baby, Build”, then we need to get honest about who’s doing the building. Not software engineers, but scaffolders. Not founders, but electricians, insulators, civil and structural engineers. This isn’t nostalgia. It’s foresight.

---

So, the takeaways:

1. ai replacing high-status routine white-collar tasks but they were mainly bullshit

2. lots of low-status blue-collar jobs which can be more meaningful

3. blue collar jobs become relatively higher status in time, we should encourage young people into these jobs now

4. … health, happiness and profit?

---