The Revenge of the English Major

The people building AI keep recommending the education we just spent a decade walking away from.

Most mornings I start with coffee and a small discovery practice: my AI and I sweep what crossed the wires overnight, hunting for whatever stretches my thinking instead of confirming it. Not long ago it turned up an older Sam Harris clip I haven't been able to shake. Expect, he said, "something like the revenge of the humanities." The world is going to need "well-educated generalists with good taste. People who have read good books, gone to good museums, had good arguments, can make good arguments." He's now more bullish on a degree in philosophy, even English literature, than on computer science.

He has company. The president of the company that built the AI I work with every day majored in English literature. In February she sat down with ABC News and said it straight out: "I actually think studying the humanities is going to be more important than ever." Daniela Amodei runs Anthropic, which filed for an IPO this month at a valuation just shy of a trillion dollars. Asked what Anthropic hires for, she listed great communicators, people skills, kindness, curiosity.

Jensen Huang, whose chips power the whole buildout, gets the what-should-my-kid-study question everywhere he goes. His answer this spring: it won't matter. All the things that used to matter are still going to matter, and he made a point of including storytelling. The one thing he urges students to ask is how AI can elevate their learning, their craft, their purpose.

Now put those voices next to a calendar. The New Yorker ran its long obituary, "The End of the English Major," in early 2023, three months after ChatGPT launched. Three years ago. Enrollment in English and history had been sliding for a decade by then, and every dinner-table conversation about college still ended the same place: be practical. A professional program, a STEM degree, the business school. "Learn to code," the decade's confident catch-all, had already curdled into an ironic meme. In the three years since, the people closest to the technology have reversed the verdict entirely. And the hiring follows the talk. Anthropic put philosophers on the team that wrote its model's constitution. Sam Altman says OpenAI consulted hundreds of moral philosophers on how ChatGPT ought to behave. In 2013, one philosophy job posting in a hundred mentioned AI; last year, one in six did. The people building these systems live and work a little way into the future. They are seeing what the rest of us will see in time.

The easy read here is career advice. The pendulum swings, the guidance counselors update their slides, philosophy becomes the new computer science, and four years from now we run the same anxious cycle on some other major. I've had an unusual seat for this one: professor in the classroom, builder the rest of the day, businesses for most of my life and AI systems now. I've watched pendulums cross higher education before, and this doesn't move like one. Work with the technology daily and you can feel the pace and the immensity of it in your hands. So I want to offer a different read, because the career framing misses what changed.

What changed is that the product schools have actually been selling, certified intelligence, is now just a subscription, twenty dollars a month.

For a few centuries now, polished cognitive output has been the scarce resource. Fluent prose, solid argument, smart analysis, a professional presentation: these were expensive to produce, so we built schools, created credentials, and did most of our hiring around the people who could produce them. AI has now made them close to free - and it improves by the day. When intelligence costs nothing, a demonstration of intelligence stops telling you anything about the person who handed it to you.

So the question quietly shifts from what you can produce to who is doing the intending. Taste. Judgment. A self that knows what it's about. AI has none of this to give you. I've come to think of AI as a mirror with no one behind the glass: no stakes, no history, nobody who has to live in the room after the decision. What the mirror reflects is decided entirely by who looks in. A person with depth of their own finds that depth amplified. A person without it finds the average of everything looking back, and takes the reflection for a face.

Which brings me back to the English majors.

The strength of the humanities has never been about information, and treating them as information is what nearly killed them. If a literature degree is a database of dates, themes, and respectable opinions, then a search bar beats it, and students drew the obvious conclusion by the hundreds of thousands. But nobody ever read Middlemarch for the data. A philosophy seminar is really a set of reps: hold a question that refuses to be resolved, in a room full of people who disagree with you, and learn to stay in it without grabbing the first exit. A novel is hours inside a mind that isn't yours. A second language rewires what you're able to notice. Travel breaks the assumption that your normals are the world's. None of it was ever about moving knowledge into a person. It was about forming a person.

There's an old word for what that forming produces. Wisdom. We say it constantly, drop it into commencement speeches, and screenshot the quote cards drifting through our feeds, and almost never stop to look at what it really is. It's never something a person finally knows. It's something a person becomes, and you can recognize it only from how they live. Nobody has found a way to install it, download it, or prompt it, and I don't expect anyone to, because it grows the way fitness grows: from reps, under load, over years. The humanities, run properly, have always been the gym.

I watch this play out in my own classroom. Since leaving tenure, I periodically teach business law at a small college in the Colorado mountains, and my exams are open-book and open-AI, on purpose. When AI can write everyone a competent answer, competence stops being the difference between an A and a C.

I've designed the back half of the course to leave the territory of settled law for the open country. A factory in a small town is failing: close it and three hundred families lose their paychecks, automate and half do, move it overseas and the town hollows out. Every option is defensible, and no statute answers it. AI will hand you a balanced memo in seconds. It cannot stand in the town and choose. So I grade for the part no model can fake, whether the student stood somewhere. The strongest exams are never the most polished ones. They have fingerprints on them.

Now the part the headlines skip, because it's harder to sell. This formation effort used to just be a required part of living. The work came bundled with going through any given day: you wrote the essay yourself because nothing else would write it, you sat with the hard book because there was no other door into the ideas, you worked out what you believed because nobody would do it for you. The struggle was never just the hard part. It was the training.

We have run this experiment before. The car made walking optional, and within two generations we were paying monthly to walk indoors on a machine. GPS made wayfinding optional, and the map we each used to carry of our own town quietly dissolved. Technology improves life by removing difficulty, and sometimes the difficulty was doing something for us that we only discover once it's gone. AI is that same pattern, aimed for the first time at the interior. You can now route around your own formation journey entirely, and the output will still look fine. The cost arrives later, in a currency that's hard to name until you're short of it: a voice that's fluent and somehow not yours, opinions you can't remember forming.

But the same tool cuts the other way. Let AI carry what's genuinely thoughtless, and those thoughtful hours can come back to you. The hours are the whole game. Spend them on what forms you, the reading, the arguing, the scary, slow building of an actual self, and everything you then do with AI gets deeper. Same amplifier either way. The difference is what you bring to it.

This is why the people building AI keep saying the quiet thing out loud. From where they sit, the new scarcity is already so apparent. Intelligence is abundant now; they sell it by the token. Direction is not. Courage is not. Discernment is not. They have built the amplifier. They cannot build the person it amplifies. A formed human being, someone with taste, judgment, and a story that is theirs, is becoming the scarcest input in the entire system. Amodei put it gently: "The things that make us human will become much more important instead of much less important."

So take the comeback of humanities personally, whatever you majored in. Pick up the novel you've been circling for years. Take the trip that has no purpose. Have the argument at dinner, in person, and stay in it past the comfortable exit. Those hours are the serious work now.

And when you do hand something to AI, watch what it smooths away. The textures, the odd angles, the particular shape of how you see: that isn't roughness to buff out. That's you. Learning to see it, in a novel, in a stranger, in yourself, is what the humanities have been teaching all along. AI will still be there when you get back from the trip, and it will have more of you to amplify. That has become my working definition of getting your life back: when the amplifier arrives, somebody's home.

So which is it for you: depth amplified, or the average looking back?
WILDiQ is eight minutes to find where you actually stand. No cost.