The Liber Man: What "Liberal Arts" Actually Means (And Why It Matters in the Age of AI)
In my last post I wrote about why I started learning the soroban — the Japanese abacus — and how it trains the brain to think in images rather than words. But as I dug deeper into the research, I kept bumping into a bigger question: what is education actually for?
Not what subjects should be taught. Not which skills are "marketable." The more fundamental question: what kind of person is education supposed to produce?
The answer, for most of recorded history, was startlingly consistent. Education was supposed to produce a liber — a free person.
The Arts of the Free
The word "liberal" in "liberal arts" has nothing to do with politics. It comes from the Latin liber, meaning free. The liberal arts — the Trivium (grammar, logic, rhetoric) and the Quadrivium (arithmetic, geometry, music, astronomy) — were literally the arts considered worthy of a free person. They weren't job training. They were freedom training.
A free person, in the classical sense, was someone who could think independently, argue clearly, detect when they were being manipulated, and govern themselves and others with judgment. The seven liberal arts were the curriculum designed to build that person. Grammar taught you to structure reality in language. Logic taught you to reason without falling for fallacies. Rhetoric taught you to move others toward truth. And the Quadrivium trained you to perceive the deep patterns of number, space, harmony, and motion that underlie the physical world.
The opposite of the liber wasn't the poor person. It was the servus — the slave. Someone whose thinking was done for them. Someone who followed instructions without understanding why. Someone easily led because they lacked the internal machinery to evaluate what they were told.
That description should sound uncomfortably familiar. We live in an era where manipulation operates at scale, and most of us don't even notice it happening.
One of the most striking things I took from Christopher Wylie's Mindfuck — his account of building Cambridge Analytica's psychographic targeting tools — was the research it was built on. In 2013, Michal Kosinski and his team at Cambridge showed that with just 10 Facebook likes, a predictive model could assess your personality more accurately than a coworker. With 70 likes, better than a friend. With 150, better than a family member. With 300 likes, better than your spouse. Cambridge Analytica took that research and weaponized it — not to sell you shoes, but to shift how you vote.
Now pair that capability with AI. In 2023, Sam Altman posted a one-line prediction: "I expect AI to be capable of superhuman persuasion well before it is superhuman at general intelligence, which may lead to some very strange outcomes." He said it casually, almost as a curiosity — superhuman persuasion framed as a feature, not a threat. But the classical tradition understood persuasion as a power. And power exercised over someone who can't recognize it is the definition of servitude.
"Once, Men Turned Their Thinking Over to Machines"
Frank Herbert saw this coming in 1965. In Dune, the Reverend Mother Gaius Helen Mohiam tells the young Paul Atreides about the Butlerian Jihad — a war, in the far future, against thinking machines. Not because the machines became evil on their own. But because they made people controllable. As she puts it: "Once, men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them."
We're not facing killer robots. We're facing something more comfortable and therefore more dangerous: the slow outsourcing of our cognitive autonomy. Every time we let an algorithm decide what's interesting, let a chatbot draft our arguments, or let a recommendation engine shape our taste, we hand over a small piece of what the classical tradition would call our selfdom. Herbert's character Leto II, the God Emperor, put it even more bluntly — humans destroyed thinking machines because they had begun to "usurp our sense of beauty, our necessary selfdom out of which we make living judgments."
Beauty. Selfdom. Living judgments. Leto's list sounds mystical, but it maps precisely onto what the Trivium trains. Selfdom is what you get when grammar, logic, and rhetoric become internal — when you can name what you see, test whether it holds, and articulate why it matters, all without waiting for someone (or something) else to do it for you. It's not mysticism. It's cognitive sovereignty.
The Question Is More Important Than the Answer
Here's the irony of our moment. We now have, for the first time in history, machines that can give us answers to almost anything. And it turns out the answers aren't the hard part. They never were.
Douglas Adams nailed this decades ago. In The Hitchhiker's Guide to the Galaxy, a civilization builds the most powerful computer ever constructed and asks it for "the Answer to the Great Question of Life, the Universe, and Everything." Not a precise question — a gesture at one. After 7.5 million years, the computer delivers: forty-two. The answer is useless, but the machine isn't broken. The input was. They'd built an extraordinary tool and given it the laziest possible prompt.
I see this every day at work. Give the same AI tool to ten people with the same task, and you get ten wildly different results. The difference isn't technical skill — everyone can type a prompt. The difference is what they bring before the prompt: the ability to frame a problem precisely, to notice when the output is subtly wrong, to push back with a better question when the first answer is shallow. The people who get extraordinary results from AI are, without exception, the ones who could have done serious thinking without it.
What surprises me most is the gap in awareness — not about how to use AI, but about what's even possible. Most people treat it as a faster search engine or a drafting assistant. Useful, sure. But the real unlock is something else entirely: using AI to build. To write code for tools you couldn't have built alone. To prototype in an afternoon what used to take a team and a quarter. To accelerate research by iterating through ideas at a pace that used to require a whole department. The gap between "help me summarize this document" and "help me build a system that solves this class of problem" — that's not a technology gap. It's a thinking gap. And it maps directly onto whether someone can define a problem clearly enough to architect a solution, or only clearly enough to ask for a summary.
This is where the Trivium becomes radically relevant again. Grammar isn't just about semicolons; it's about constructing reality clearly enough to ask precise questions — and if you can't name what you're looking for, no machine can find it for you. Logic isn't just about syllogisms; it's about navigating contradictory information without being paralyzed or deceived, recognizing when your own reasoning has holes, building an argument that doesn't collapse under scrutiny. Rhetoric isn't just about persuasion; it's about the ability to make something compelling — to find the shape of an argument that reveals why it matters, to create beauty in how ideas connect.
The Factory vs. the Forum
I wrote in my last post about how modern schooling traces back to Prussia — compulsory, standardized, designed to produce obedient citizens. It achieved universal literacy, a genuine triumph. But it did so by replacing the forum with the factory, training students to answer questions rather than ask them.
Consider what it replaced. A Roman senator's son learned grammar by dissecting speeches that had moved armies; learned logic by arguing both sides of legal disputes; learned rhetoric by standing in front of actual audiences and being judged on whether he could move them. The student wasn't a vessel to be filled. He was a future leader being forged — someone who would need to think on his feet, see through deception, and persuade free citizens to follow him. The education was hard because the stakes were real.
The stakes are real again — just different. The Roman kid needed rhetoric to hold a forum. We need it to hold our own against systems designed to bypass our judgment entirely.
What a Free Person Looks Like Now
So what does the liber look like in 2026?
Not someone who refuses to use AI. That would be as foolish as refusing to use the printing press. The free person uses every tool available — but the tool doesn't use them. You can see the difference in how they work: they don't accept the first answer, they interrogate it. They don't ask the machine what to think; they ask it to stress-test what they already think. When the output feels off, they can say why it's off — not just "this doesn't seem right" but "this conflates two things that are structurally different" or "this is optimizing for the wrong variable."
That's not a personality trait. It's grammar, logic, and rhetoric — internalized so deeply they've become instinct.
But the Trivium is only half the picture. The free person in the age of AI isn't the one with the best tools. It's the one whose thinking can't be steered by the tools — or by the people behind them. Someone who brings something no model can replicate: a sense of beauty shaped by their own life, their own experience, their own taste. That's what makes our judgment ours — and it's the one thing that can't be sourced from a dataset.
Previously: Why I want to learn how to use the Soroban (Abacus)
Next up: The incredible history of the abacus — from Sumerian counting boards to Incan quipus — and what it tells us about who taught humanity to count.