Learning in Public

This blog is an exercise in being wrong in public.

I don't mean wrong as in "I posted bad takes and got ratio'd." I mean wrong as in incomplete. Every post I publish captures my understanding at a specific moment. That understanding is, by definition, partial. Tomorrow I'll know more. Next year, some of these posts will make me cringe.

That's the point.

The Expert Trap

There's an unspoken rule in online writing: only write about things you're an expert in. Establish authority first, then share knowledge. This is how you build credibility, grow an audience, get invited to conferences.

It's also how you never write anything.

If I waited until I was an expert on AI, I'd never write about AI. If I waited until I fully understood the relationship between tools and thought, I'd never publish "Your Tools Shape Your Thoughts." If I waited until I had the definitive take on taste, "The Last Human Skill" would still be in my drafts folder.

Expertise is asymptotic. You approach it but never reach it. Waiting for expertise is waiting for a bus that doesn't stop.

Writing as Thinking

Joan Didion said: "I don't know what I think until I write it down."

This isn't a cute quote. It's literally how cognition works. Writing forces you to linearize your thoughts — to take the tangled, multi-dimensional web of associations in your head and flatten it into sentences. The act of flattening reveals gaps. "Wait, how does this connect to that?" "Do I actually believe this?" "What's my evidence?"

These questions don't arise when you're thinking. They arise when you're writing. The medium demands rigor that the mind doesn't.

Thinking about a topic:
  "AI and taste are related somehow...
   there's something about embodied knowledge...
   and Ira Glass said that thing about the gap..."

Writing about a topic:
  "Taste is pattern recognition across domains,
   filtered through lived experience, expressed
   as judgment. AI can do the first part but not
   the second, because..."

The first version is a feeling. The second version is an argument. You can't get from feeling to argument without writing.

The Vulnerability Tax

Learning in public costs something. Every post is a snapshot of your ignorance. Someone who knows more will read it and think "that's not quite right." Someone who knows less might take your incomplete understanding as gospel. Someone on the internet will definitely be mean about it.

This is the vulnerability tax, and it's real. But consider the alternative: learning in private. Thinking thoughts that nobody challenges. Building mental models that nobody stress-tests. Believing things that might be wrong, forever, because you never exposed them to friction.

Private learning is comfortable. Public learning is useful.

What I've Gotten Wrong

In the spirit of the thing, here are beliefs I've held and revised:

"Code comments are a code smell." I used to think clean code shouldn't need comments. Now I think comments about why are invaluable. The code shows what. Only comments show why.
"Static typing prevents bugs." It prevents a category of bugs — usually the least interesting category. The bugs that matter are logic bugs, and types don't help with those.
"Premature optimization is the root of all evil." Knuth said this. He was right. But I took it too far — I used it as an excuse to write lazy code. There's a difference between premature optimization and intentional sloppiness.
"You should have a plan before you start coding." Sometimes. But often the code is the plan. You discover the architecture by building it. Planning is thinking, but so is coding.

Each of these revisions came from writing the original belief down and having it challenged — by readers, by reality, or by my future self reading the old post and wincing.

The Compounding Effect

Here's the part that surprised me: learning in public compounds.

When you have ten posts, nobody cares. When you have fifty, people start to see patterns. When you have a hundred, you've built a body of work that shows how your thinking evolved. That evolution — the visible record of changing your mind — is more valuable than being right.

An expert who's always right is a textbook. A person who thinks in public, makes mistakes, and revises — that's a mind at work. People connect with minds at work.

The Practice

If you want to start, here's what I'd suggest:

Write about what you learned today. Not a tutorial. Not a guide. Just: "Today I learned that X, and it surprised me because Y."
Don't edit for a week. Publish it rough. The urge to polish is the urge to hide.
When someone corrects you, update the post. Not silently — add a note. "I originally wrote X, but [reader] pointed out that Y. I think they're right because Z." This models good thinking.
Re-read your old posts every few months. Not to cringe (though you will). To see how far you've come.

The goal isn't to be right. The goal is to be less wrong over time, and to leave a trail that others can follow.

This post will probably make me cringe in a year. Good. That means I learned something.