A friend asked me what his kid should major in college.
He expected something practical. “Data science” maybe. Or “computer science.” Something that sounded like a career.
I said: debate. Philosophy. Math. Physics.
He paused. “Those don’t sound like jobs.”
He’s right, they don’t. But they teach something that most jobs are going to need badly: the ability to think through a problem, evaluate information critically, and actually decide what’s true.
That’s the skill that matters now. And most people are still optimizing for the wrong one.
What Changed
For most of modern history, knowledge was genuinely valuable. People who knew a lot could command premium prices. Doctors, lawyers, accountants, consultants a significant part of what they sold was access to information that other people didn’t have.
If you knew how to read a balance sheet, that was worth something. If you understood the tax code, that was worth something. If you could diagnose a condition based on symptoms, that was worth something.
The scarcity of knowledge was the whole game.
AI changed the game.
Now you can ask an AI to explain the periodic table of elements, summarize recent case law, outline the symptoms of a rare condition, or walk through the logic of a financial model — and get a pretty good answer in seconds. For free.
The information isn’t scarce anymore. Which means the information isn’t the edge anymore.
What the Edge Actually Is Now
Here’s what AI still can’t do reliably: tell you whether the information it gave you is right.
It can produce a confident-sounding answer. It can cite sources. It can explain the logic. But it can’t evaluate that logic from the outside. It can’t catch its own mistakes the way a good critical thinker can. It doesn’t have a judgment layer that operates independently of its training.
That’s still a human thing. For now.
Judgment — knowing whether something is true, whether a source is reliable, whether a conclusion actually follows from the evidence — is the skill that doesn’t get commoditized when AI gets better. If anything, it becomes more valuable, because there’s more AI-generated content to wade through, and someone has to be able to tell what’s actually worth acting on.
I said this directly in the conversation with my friend: “Knowledge used to be very powerful. People who had a lot of knowledge could command premium prices. But nowadays knowledge is becoming more commoditized because of AI. You don’t have to memorize the periodic table anymore. You can just ask AI about it. But judgment and evaluation is something AI cannot do yet.”
That last part is doing a lot of work. Yet is a word worth paying attention to. But right now, today, the asymmetry is real.
Why This Is Hard to Act On
The tricky part is that most of our education system — and most career development — is still built around the old model.
School rewards memorization. Exams test recall. Credentials signal that you know a body of information. Even corporate training tends to be “here’s what you need to know about X.”
The few disciplines that genuinely teach you to think rather than just know are the ones that tend to get undervalued. Debate teaches you to construct and tear apart arguments. Philosophy teaches you to spot logical errors and hidden assumptions. Physics and math teach you to follow an argument rigorously to its conclusion and notice when something doesn’t add up.
None of those feel immediately practical. But the actual skill you’re getting from them is the ability to evaluate information rather than just retrieve it. And that transfers everywhere.
What This Means at Work
This isn’t just a college advice question. It applies to anyone in a professional context right now.
The people who are getting the most out of AI tools aren’t the ones who know the most about a subject. They’re the ones who can tell when the AI is wrong. Who can push back on a confident-sounding answer with a better question. Who can take what AI produces and actually decide what to do with it.
I see this constantly in the AI workshops and coaching work I do. Two people can use the exact same tools, with access to the exact same information. One of them makes dramatically better decisions with it. The difference is almost never knowledge. It’s judgment — the ability to evaluate what they’re getting and figure out what it means.
That skill is learnable. But it takes more than just knowing things.
The Practical Takeaway
If you want to be genuinely hard to replace over the next decade, the question isn’t “what should I learn?” It’s “how well can I think?”
A few places to start:
Practice being wrong deliberately. Take a position and argue for it, then argue against it. Notice where your logic breaks down. This is basically what debate trains you to do.
Get in the habit of sourcing. When AI gives you something confident, ask where it came from. Try to find it. If you can’t verify it, treat it as a hypothesis, not a fact.
Work on decisions under uncertainty. Most judgment isn’t about knowing the right answer — it’s about figuring out the best move when you don’t know. Practice making calls with incomplete information and reviewing how they turned out.
Read more philosophy and logic. I’m serious. One book on critical thinking or informal logic will do more for your judgment than most courses on “AI skills.”
The information age put a premium on knowing things. The AI age puts a premium on thinking about what you know.
If you want to build better judgment about AI — when to use it, how to evaluate it, where it breaks down — the 4-Day AI Sprint is designed around exactly that. Not just how to use the tools, but how to think about what they’re giving you.
Recommended for you
Want the full system? 25X is the flagship productivity system we teach.
Explore 25X →