What Can We Learn from the Luddite Revolution?
In 1811, a group of skilled textile workers in Nottingham, England began smashing knitting frames. They weren't stupid. They weren't anti-progress as a philosophy. They were highly trained craftsmen who looked at automated looms and saw a very specific future: one where the expertise they'd spent years developing was worth less. They organized under the banner of a probably-fictional "King Ludd," and for a brief window, they were the most feared labor movement in England. Two centuries later, the looms are large language models, the craftsmen are software engineers, and the frames are being smashed on Reddit.
I've Seen This Movie Before
I've had a front-row seat to every major technology wave of the last four decades. In the 1980s, I watched personal computers go from hobbyist curiosities to business necessities while plenty of smart people insisted they were toys. In the 1990s, cellphones and the internet arrived to a chorus of "why would I need that?" The dot-com boom of the 2000s was supposedly all hype — until it became the foundation of the modern economy. The cloud revolution of the 2010s was "just someone else's computer" right up until it was running everything.
Every single wave had the same arc: emergence, resistance, grudging acceptance, and then a collective pretending that everyone was on board from the start.
I'm all in on AI. Not because I'm a blind optimist, but because I've lived through enough of these cycles to recognize the pattern. The technology that gets this much resistance from people who should know better is usually the technology that changes everything.
The Unlikely Resistance
Here's what genuinely surprises me: the most vocal resistance to AI isn't coming from the people you'd expect. It's not coming from workers in non-technical fields worried about displacement. It's coming from software engineers — the very people who build technology for a living.
Spend ten minutes in any major programming subreddit and you'll find the sentiment served hot and daily. AI-generated code is garbage. It introduces bugs. It makes junior developers lazy. Real engineers don't need it. The arguments are delivered with the confidence of someone who has never been wrong about a technology trend, posted by people whose entire careers exist because previous generations of skeptics were wrong about technology trends.
Reddit, to its credit, is at least self-aware about this. The platform practically invented the concept of the "hive mind" to describe its own behavior. And yet, watching the consensus against AI harden in real-time is like watching someone narrate their own fall down a staircase without grabbing the railing. The awareness doesn't produce a course correction — it just makes the fall more articulate.
One well-upvoted post declaring that AI coding tools are a fad gets reinforced by hundreds of agreeing comments. Dissenting voices — engineers who are quietly using AI to double their output — get downvoted into the basement. The consensus isn't built by independent evaluation. It's built by social proof, which is an ironic engine for a community that prides itself on empirical thinking.
It Was Never About the Looms
The original Luddites had a legitimate grievance. The textile industry did displace skilled workers. Wages did fall. The transition was painful and real people suffered. History has been too quick to turn "Luddite" into a punchline when the underlying anxiety was completely rational.
But here's what the Luddites got wrong: they assumed the future would have fewer jobs of the same kind, when it actually produced entirely different kinds of jobs that hadn't been imagined yet. The industrial revolution didn't just eliminate roles — it created categories of work that a 1811 textile worker couldn't have conceived of.
The same dynamic is playing out now. The engineers who refuse to touch AI tools aren't wrong that the landscape is shifting. They're wrong about what the shift means. AI isn't eliminating the need for engineering skill — it's redefining what that skill looks like. The engineer who can architect systems, orchestrate AI agents, validate outputs, and apply judgment at the points where it matters most isn't less valuable. They're dramatically more valuable.
But you can't become that engineer if you won't open the tool.
Identity as Obstacle
There's a deeper layer here that the Luddite parallel exposes perfectly. For many experienced engineers, technical skill isn't just a job qualification — it's an identity. "I'm someone who can write elegant code from scratch" is a statement about who they are, not just what they do.
AI code generation doesn't just threaten a workflow. It threatens a self-concept. And when identity is on the line, people don't evaluate evidence objectively. They entrench. The Luddites didn't smash looms after a careful cost-benefit analysis. They smashed them because the looms told them something about themselves they didn't want to hear.
When a subreddit collectively declares that "real engineers don't use AI," that's not a technical assessment. That's identity protection wearing a technical costume.
The Pattern Always Resolves the Same Way
Every technology resistance movement I've witnessed in forty years has ended identically. Not with the skeptics being proven right. Not with the technology failing. But with the skeptics quietly adopting the thing they swore they'd never use, and then acting as though they were early adopters all along.
The engineers who dismissed PCs eventually got desktops. The ones who mocked the cloud eventually deployed to AWS. And the ones who are currently posting five-paragraph arguments about why AI is overhyped will, within a few years, have AI tools integrated so deeply into their workflow that they'll forget there was ever a debate.
The only question is whether they'll be leading that transition or catching up to it.
The Luddites, for all their passion, didn't stop the looms. They just missed the chance to be the ones operating them.