The most valuable thing a young person can do right now is write. It has a low ROI on the timescale of a week. It is the highest ROI I know on the timescale of a life.

I've started asking friends a question. Name a journalist you admire. Name a beat you read every week. Name one writer whose judgment, on anything, you trust enough to update your priors. They can name Twitter accounts. They can name Instagram accounts. The very online ones can name a Substack or two! Almost none of them can name a journalist, a historian, a biographer. I'm not catching them out — I genuinely want the recommendation. The recommendation isn't there.

The other thing I notice is that fewer and fewer of them are choosing to become journalists. This seems to be a trend globally: fewer young people are choosing to become writers. Writing has acquired the romance of the starving artist — the thing you do if you don't hold OpenAI or Anthropic stock, the worst way to put in enormous amounts of effort for no return. So the people who would have been good at it go into something where the paychecks clear, and the field thins, and the field thinning is itself the argument for not entering it.

This is an extremely educated, extremely well read generation. This is a generation that consumes information constantly, in every spare minute, on every available surface. Inputs are not the problem. The thing that's missing is the more important act of synthesis — the one where you take what you've read and have to defend a position about it to someone who won't let you get away with the easy version. The output muscle — where someone asks why do you believe this and where does the argument line up and what would you have to see to change your mind, and you find out which parts you actually believe and why – has atrophied.

I do this for a living. It is the most useful thing that has ever happened to my thinking. And I notice, when I push my friends two layers down on something they're sure about (usually a simple why, and then why again), the position usually crumbles. Not because they're not knowledgeable, but because no one has ever asked them. The inputs are infinite and the practice of defending a belief is nonexistent.

I don't want to make the whole argument here about why. I'll write about that somewhere else. Some of it is financial. Some of it is the AI story, the why bother of doing slowly what a model does instantly. Some of it is what a tweet has done to the concept of a position. The hot take, the ragebait, the quippy one-liner that came to you in the shower — fired off, applauded for ten minutes, called a belief and called a day.

It isn't. Value creation and value capture are difficult to balance in writing. A piece worth reading is the residue of something laborious — a writer pushing on the seam of what's known (often for weeks and weeks at a time), an editor pushing back, a fact-checker confirming, the position getting held to the chest and pulled at until it survives. Many are like small, condensed PhD dissertations. The reason the writing is good is because the process is expensive. If you don't pay for it financially, you pay for it in another currency — the writer who could have been good stops, the editor lays off, the work shrinks to whatever can be produced for free. You pay some way or another. You usually don’t notice it immediately.

The second-order effect is the one I think about more. When the people who would have been writers don't become writers, the people around them lose the friend who pushes back. The friend who calls you out saying that's not quite right! and stays at the table until you get to the version that is. Writing is a public craft and a private one. Most of what it teaches is private. Sit with yourself. Think critically with what is going on in the world. Notice when you don't actually know. You don't have to publish to write. You should probably write if you want to think better.