Swyx's Learning in Public

Shawn Wang, known online as swyx, is a developer, writer, and founder based in San Francisco. He’s worked in developer experience at Netlify, AWS, Temporal, and Airbyte, and now co-hosts Latent Space, a top AI engineering podcast with over 100,000 subscribers. He wrote The Coding Career Handbook and coined the term “AI Engineer” in his widely-read essay Rise of the AI Engineer.

You learn JavaScript by reading docs. Someone else learns JavaScript by writing the tutorial you wish existed. Who gets the job offer?

Swyx calls this “Learning in Public.” The idea is simple: instead of learning in private and hoping people notice, document your learning as you go. The documentation becomes your portfolio. See Learning in Public for a deeper exploration of this philosophy.

What counts as “public”

Anything you ship:

• Blog posts, tutorials, quick tips
• Conference talks or meetup presentations
• StackOverflow answers, Reddit threads
• Demo projects on GitHub
• Twitter threads, YouTube videos, newsletters

Each creates what Swyx calls “luck surface area” — more artifacts, more chances someone finds your work when they need it.

The gears

Swyx describes four stages:

Explorer: You’re learning something new. Tweet what confuses you. Keep public notes. You’re writing for future-you, mostly.

Settler: You’ve built a few things. Write “React Hooks: What I Wish I Knew” — the tutorial you needed six months ago. People slightly behind you are your audience.

Connector: Companies pay you to speak. Your newsletter has real subscribers. You get paid to learn and teach.

Miner: You’re building new tools, pushing boundaries. Inventing, not documenting.

Most developers stay in Explorer or Settler. That’s fine — even that creates leverage. You don’t need to become a thought leader.

Pick up what they put down

Swyx expands on this in his essay “Pick Up What They Put Down.” The fastest way to stand out: be first.

New library released? Write the first tutorial. Framework launches? Create the first comparison. Expert publishes an essay? Write a response.

You don’t need expertise for this. Speed and clarity matter more. Your quick-and-dirty guide becomes the canonical resource because nobody else wrote one.

The three strikes rule

Third time you explain something in conversation, write it down.

Someone asks how to deploy Next.js. You explain. Week later, someone else asks. Third time: write the blog post, send the link instead. Now you’ve documented something people actually need.

The math

Luck Surface Area = Doing × Telling

High Doing, low Telling = invisible genius. Nobody knows what you built.

High Telling, low Doing = empty credibility. Nothing behind the words.

Both together multiplies opportunities.

The permission slip

Swyx has two rules:

1. You’re allowed to be wrong. Update posts, correct mistakes. You’re learning.
2. Readers help, not ridicule. (They usually do.)

Your imperfect tutorial beats 80% of developers who never publish anything.

Why I believe this works

Swyx went from finance to blogging about JAMstack to getting hired at Netlify. Then AWS, Temporal, Airbyte, and now Latent Space has over 100,000 subscribers.

Could be survivorship bias. But I’ve seen the pattern repeat with others: visible learning creates opportunities that private learning doesn’t.

If you want to try

One artifact per week for three months. Don’t overthink format.

Week 1: Tweet three things you learned. Week 2-4: Write one blog post. Week 5-8: Answer five questions on Stack Overflow or Reddit. Week 9-12: Ship one demo project with a real README.

Most people won’t finish the three months. The ones who do tend to have something to show for it.

Next: How to Build a Personal AI Operating System