Every week someone asks me whether AI is going to take their job. Sometimes they say it outright. More often it shows up sideways "it'll never really replace what I do," said with the particular confidence of someone trying to convince themselves.
I get it. I work in this stuff and I feel the pull of it too. But I've started collecting something, and it's changed how I hear those conversations. I've been gathering the things people said about previous technologies. Not AI. The ones before it. The printing press, the power loom, recorded music, the first computers.
Here's the game I now play. I show someone a quote, let them assume it's a hot take about ChatGPT from last Tuesday, and then tell them when it was actually written. The faces are worth it.
Try a few yourself.
"It will replace human skill, intelligence, and soul"
Sounds like a LinkedIn post about AI art, doesn't it?
That's John Philip Sousa in 1906, and he was furious about the gramophone — recorded music. (Paraphrased from his essay "The Menace of Mechanical Music.") He genuinely believed that machines playing music would kill off amateur musicianship and hollow out the nation's soul. He even worried that mothers would stop singing lullabies to their children because a machine could do it instead.
Recorded music did not, as it turns out, destroy human musicianship. There are more people making music today than at any point in history.
"The skilled worker is being pushed out to make room for a faster, lifeless machine"
This one I use a lot, because it's the cleanest. Swap "skilled worker" for "knowledge worker" and you could publish it tomorrow.
It's Thomas Carlyle in 1829, writing about industrial machinery and the mechanised loom. (Paraphrased from "Signs of the Times.") And he didn't stop at jobs. He went further, into territory that sounds eerily like the conversations we're having now about what AI does to our minds:
It's not just physical work that's run by machines now — it's the mental and emotional too. People have become mechanical in how they think and feel, not just in how they work.
A man in 1829, worried that machines were making us think like machines. Nearly two hundred years before anyone typed a prompt.
"This invention will create forgetfulness — people will stop using their memory"
Here's my favourite, because of how far back it goes.
That's Socrates, around 370 BCE, and the terrifying new technology he's warning about is writing itself. (Paraphrased from Plato's Phaedrus — and to be fair to Socrates, he's retelling a myth to make the point, not quite saying it in his own voice.)
His argument was that if people could write things down, they'd stop remembering them. They'd rely on the marks on the page instead of their own minds. They'd seem wise without being wise — full of information, empty of understanding.
People say almost exactly this about Google. And about AI. "Nobody knows anything anymore, they just look it up." The complaint is 2,400 years old and it was originally aimed at the alphabet.
"We manufacture everything except people"
John Ruskin, 1853, on factory mechanisation and what the division of labour was doing to the men working the machines. (Paraphrased from "The Nature of Gothic.") His worry wasn't that the work would disappear. It was that the work would survive but turn the worker into a component — a cog doing one repetitive fraction of a task, all craft and judgement stripped out.
If you've heard anyone worry that AI will leave humans doing the boring bits while the machine does the interesting bits, you've heard a version of Ruskin.
"A new disease you'll hear a lot about: technological unemployment"
This is John Maynard Keynes in 1930, naming the fear so precisely that we still use his exact phrase. (Paraphrased from "Economic Possibilities for our Grandchildren.") He defined it as what happens when our ability to automate work away outpaces our ability to find new work for people to do.
Now — Keynes is a useful one but I'll be honest about him, because someone always brings it up. In the same essay he predicted that by now we'd all be working fifteen-hour weeks, drowning in leisure. We are not. So his diagnosis landed and his happy ending didn't. The productivity did arrive. It just didn't turn into free time for everyone — it got shared out rather unevenly. Which, if anything, is the more interesting lesson for the AI conversation.
"The old link between having a job and having an income is breaking"
In 1964, a group of scientists, economists and civil rights leaders — including a couple of Nobel laureates — sent an open memo to President Lyndon Johnson. (Paraphrased from "The Triple Revolution" memorandum, written about computers and automation.) They warned that the combination of the computer and the self-running machine had created a system with almost unlimited output that needed less and less human labour.
Read it cold and it's indistinguishable from a 2024 think-tank report on AI and jobs. Same argument, same alarm, same proposed fixes retraining, public works, a guaranteed income. They thought mass joblessness was imminent. They were wrong about the timing by about sixty years and counting.
I'm not saying they'll be wrong forever. That's the honest bit. But the pattern matters.
So what's the point of all this?
Not that the fear is stupid. I want to be clear about that, because the easy version of this argument is "people always panic and they're always wrong, so relax." That's lazy, and it's not what the history shows.
The weavers Carlyle was writing about? A lot of them really did lose their livelihoods. The disruption was real. Some of those communities never recovered. The fear wasn't irrational, it was just aimed at the wrong target. The machine didn't end work. It ended that work, and the pain landed hard on the people who couldn't get to the other side of the change.
So when I show people these quotes, I'm not trying to talk them out of being worried. I'm trying to do two things.
First, take the edge off the specialness. The feeling that this moment is uniquely terrifying, that we're the first humans to stand at the edge of being made obsolete — that feeling is itself a tradition. Socrates had it about handwriting. We are not as alone in this as the panic makes us feel.
Second, point at where the real question sits. It was never "will the machine replace us." Every generation asked that and the answer kept being more complicated than yes or no. The better question is the one Keynes was circling and the 1964 memo was shouting about: when the productivity gains show up, who gets them? That's not a technology question. It's a choice. It always has been.
The machine has never been the thing that decides. We are.
If you've got a favourite "wait, that's not about AI?" quote, send it my way. I'm always collecting.