Learn to code, lead with AI. Build real software on https://teapawt.net/t.co/4gMaLHG1H3. Courses for the AI era, in your browser. Projects: closed beta for Premium 👇
Joined June 2026
- Tweets229
- Following7
- Followers427
- Likes175
Expressions vs statements, the one rule behind ternaries, arrow bodies, hoisting and JSX:
start.dev/blog/javascript-ex…
Add braces to an arrow function and it can quietly return undefined 🤫
(n) => n * 2 hands back the value. (n) => { n * 2 } is a block, and a block gives nothing back unless you write return.
The same rule explains a lot of JavaScript syntax 👇
With Prisma course being live, Start.dev now covers each layer of a AI product development with TypeScript track -
- The database: courses on SQL and Prisma
- The language: Typescript course
- The server: Building APIs and routes with Node course
- The interface: Build beautiful UI components with React course
- The code that lasts: Course on Design Principles of Software
- AI features: Adding AI features to your full stack app using Vercel AI SDK
Learn Prisma is live on Start.dev 🎉
Work with a real PostgreSQL database from TypeScript using Prisma 8. You'll write typed queries and relations, use transactions, and run migrations that change a table without losing data.
Taught by @traversymedia . Course link below 👇
You'll need to be comfortable with SQL and TypeScript, but you don't need any Prisma experience.
start.dev/course/learn-prism…
Learn Prisma is live on Start.dev 🎉
Work with a real PostgreSQL database from TypeScript using Prisma 8. You'll write typed queries and relations, use transactions, and run migrations that change a table without losing data.
Taught by @traversymedia . Course link below 👇
Why width: 300px comes out 350px wide, how box-sizing fixes it, margin collapse, and checking it all in DevTools:
start.dev/blog/css-basics?ut…
Every element on a page is a box with four layers 📦
- Content, then padding, then the border, then margin.
- Padding is inside the border, margin is outside it.
- The background fills content and padding, never the margin.
Diagrams, box-sizing and the gotchas 👇
Pretty sure if you just loan your thinking to an AI, that will erode your skills soon.
Use AI as an enabler and accelerator.
If you are just staring at Codex/Claude Code's outputs while prompting the AI agents with stuff like "fix this" or "I don't like this" (with zero understanding of what code it has generated) - let me tell you, you're NGMI.
If you ask us, we believe it is lot more relevant in 2026 than it was ever before.
The people who can read and fix what AI has broke on prod are the ones who are still shipping next year in 2027.
Secret to Pewdiepie's success is that he just enjoyed building and has actually shipped stuff that he would personally use.
That's it. That's the whole strategy!
Lesson - just go learn some concepts and ship a project by weekend. Use AI to accelerate your learning and shipping.
What not to do - "staring at Claude Code/Codex's output without an iota of idea of what the agents wrote"
"Play with the tools" - may be yes. We don't deny it either.
But "it won't matter" only holds when the AI is right. Someone still has to know when it is wrong as to "why it is wrong".
This is f*cking insane. Boris Cherny, the man who built Claude Code, was asked if people should still learn to code. His answer:
"In a year or two, it's not going to matter"
In 73 seconds, he explains what to do instead: play with the AI tools, learn them, don't be scared of them, stay on the frontier
On his team, everyone codes now. The PM, the designer, even the finance guy
He hasn't edited a single line of code by hand since last November. He ships 10 to 30 pull requests a day
Writing Code → Prompting → Agents → Harness → Loops → Everyone on the Team Ships
Claude Code is a harness: the tools, memory, rules and loops around the model. His team changes it with every new model and cut 80% of its system prompt
Save this before it disappears from your feed
Every doc, guide and repo below is free:
Claude Code best practices
code.claude.com/docs/en/best…
Effective harnesses for long-running agents
anthropic.com/research/effec…
Building effective agents
anthropic.com/research/build…
Claude Code docs
code.claude.com/docs/en/over…
Memory: how Claude remembers your project
code.claude.com/docs/en/memo…
Custom subagents
code.claude.com/docs/en/sub-…
Hooks: rules the agent can't skip
code.claude.com/docs/en/hook…
Skills
code.claude.com/docs/en/skil…
Run Claude Code on a schedule
code.claude.com/docs/en/rout…
Free course: Claude Code in Action
anthropic.skilljar.com/claud…
Claude Code on GitHub
github.com/anthropics/claude…
Claude Code inside your PRs and issues
github.com/anthropics/claude…
Build your own agent on the same harness
github.com/anthropics/claude…
The best Claude Code resources in one list
github.com/hesreallyhim/awes…
Plugins, agents and skills for Claude Code
github.com/wshobson/agents
He compares it to the printing press. Writing went from a few scribes to everyone. Building software is next
95% will keep asking whether to learn to code, the 5% who build a harness will ship 30 pull requests a day
Bookmark it and read the first link tonight
Then use the guide below to build your harness before everyone else does ↓
Yes!
AI writes code super fast, but you will still need to read it, debug it and know when it is wrong!
As we often say, learn to code first and then you can lead super fast with AI.
Yes.
The chat writes the code, fundamentals tell you when it's wrong. Learn them while building something real.
Monday: pick one small thing you actually want, build it with AI, then make sure you can explain every line.
Repeat.
You will get hired. Or you can use @startdotdev.
When you have more people to play with, it is always a good outcome.
The ones who cared about their craft are always helping new comers, and we hope such good people would continue to do it!
Managing your agents is "easy" when you know what and why of the job you are managing.
This is something, beginners still have to learn in a structured way just like good olden days but also by deeply leveraging AI.
Hey unc, we are back and as always we love your takes!
We agree the power tools line is the whole alpha for beginners. We strongly suggest novices, beginners to learn what good looks like first, then let the agent go fast.
As we often say - learn to code first, you can then lead fast with AI.
Disengaging from syntax is not disengaging from engineering.
I’m working on a significant project at the moment. It’s in a pre-production state so far. More than a prototype, less than ready for real users.
I’ve been making multiple passes over the structure for each of the following concerns: Architecture, coverage, race & concurrency, security, API, etc. Each of these requires significant interrogation and review of verification procedures; but not review of the code.
At base are the /ub-mode (GitHub.com/unclebob) rules that keep the agents disciplined as far as coding structure and tests are concerned. But above that it’s me making sure that all the bases are covered well enough to keep my client safe from stupid sloppy mistakes.
I am the engineer. I am responsible. The agents are just my tools. They are power tools, to be sure. But power tools in the hands of novices result in missing fingers. So I am very careful.