Whenever I think about a project made with love and respect for the craft, Web Weekly is one of the first things that pops into my mind. Thanks Stefan for doing it 🙌

DHH: “We’ve seen peak programmer” YouTube video by The Pragmatic Engineer

100% agree! The future belongs more to excellent product managers than devs. A good PM is not the one who stays 10h/day in meetings, but the one who has a hyper-deep knowledge of an industry & knows where to place bets. And this comes from a hardcore dev such as DHH:
youtube.com/shorts/7FmYJ...

this === true

Chasing “Build your AI startup in 5 minutes” YT thumbnails === maximizing your exposure to a "Red Wedding" event. If you don’t understand the the non-spectacular fundamentals you are like a turkey waiting for the next GPT launch to be your Thanksgiving.

Morning listen: www.youtube.com/watch?v=Jj0M...

Large Language Models explained briefly YouTube video by 3Blue1Brown

Great, succinct video explaining what's inside the magical black box of LLMs: www.youtube.com/watch?v=LPZh...

Don’t rely on AI for everything!

Traditional “boring” code it’s much more reliable, faster and cheaper.

Use AI/LLMs only for tasks that truly need them, like Natural Language Processing, image recognition and similar tasks.

TypeScript BEATS Python when building AI Agents (Mastra's YC Journey) YouTube video by CodeRabbit

Great morning listen: podcast with Abhi Aiyer from Mastra AI, on how hard & important is to fiind real market fit, how they used printed books to increase their reach (yeah, people still like to read), MCP courses & more! Python trains, Javascript ships 💪
youtu.be/6AbmRwKoF1U

podcast Archives - Js Craft Tips about React, CSS and Javascript

🎙️ I’ve come to realize that being able to formulate and ask the right questions is a crucial skill. In the age of AI, that skill matters more than ever. Running a podcast really pushes you to learn how to choose the right questions and express them clearly: www.js-craft.io/blog/categor...

Hyper cool :D

Halfway through the "Scaling Fastbook: software Engineering Through the Hockeystick" book by @swizec.com . Great read, full of timeless advice that will still be relevant a long time after the AI hysteria will have cooled down.

hey look, it still totally makes sense to make features that are not "AI something ... "! Great work Chris & team!

Wow, indeed, really cool!

Brian this is 100% true! It's not just about open source but in general who will produce valuable content on the internet if there is no incentive left? Why publish anything if it will just be used by the some model?

ha! super cool! 😎

🎙️ Swizec Teller — Tech Lead at Plasmidsaurus & Author of Scaling Fast YouTube video by Js-Craft Academy

My interview with @swizec.com in now online. We covered:
- his new book Scaling Fast: Software Engineering Through the Hockeystick
- starting again from scratch
- EU vs USA tech scene
- why good writing matters, especially in the age of AI & more

Watch it here: 👉 youtu.be/zN0SjFvtr8w

Lindy Effect = the future life expectancy of non-perishable items such as ideas, technologies, books is proportional to their age. The longer something has survived, the longer it is likely to continue surviving.
🏛️ Learn timeless stuff, things don't change as fast as we think!

I’ve just seen this post. To be honest, I would be extremely happy if I were able to truly study and assimilate only 1 or 2 of these timeless & deep books.

Oh boy, definitely, I need to check this out!

I've updated my popular persona-based AI-assisted programming post with the models I'm currently using for each role.

https://humanwhocodes.co

Knowledge Work Is Dying—Here’s What Comes Next While AI devours information-based roles, OpenAI, Alphabet, and Apple are investing in wisdom work—and you can, too

This article has one of the best ideas I’ve read recently. Moral of the story: don't compete with AI on marginal improvements & instead, do things it cannot do. Knowledge work is dying, welcome to the age of wisdom work!

every.to/thesis/knowl...

More than your regular Frontend newsletter Web Weekly covers what the web platform gives you – no package.json required.

Happy Monday party people!

You know the drill. 🫣 Later tonight a new Web Weekly will hit all the inboxes!

If you think that knowing and understanding how the web works remains important in the wild times of AI "stuff" (I do at least), you should check it out!

webweekly.email

LLMs with their NLP features are becoming the UI! You won’t need to wrestle with Excel menus or formulas anymore. You can just say something like, “What’s the average of the top half of the salary column?” and via some MCP style tools you should get the result.

🎙 Chris Coyier — Making CodePen & CSS Tricks, Lessons from 700 Podcasts, Business as a set of bets YouTube video by Js-Craft Academy

I'm over the moon excited to share my interview with @chriscoyier.net ! We talked about building CodePen & CSS Tricks, why having fun is crucial for a long sustainable career, AI, and much more! Thank you, Chris!

youtu.be/-bLimodCd4M

Using LangChain and LangGraph with local models YouTube video by Js-Craft Academy

Think of LLMs the way you think about DBs. For some cases, a simple like SQLite is ideal; for others, you need a behemoth like Oracle. Context determines the right fit. Bigger isn’t automatically better & smaller, local models will play a bigger role soon.
youtu.be/pKK07kObVnw

VS Code: Inline Suggestions Delay YouTube video by Neutron Dev

This tip in VS Code / Cursor made my life better. Setting a 5-second delay between when you stop typing and when the code completion appears gives you time to think and makes AI-assisted coding much more enjoyable. Thanks @stefanjudis.com for the ideea!
www.youtube.com/watch?v=KZZz...

Next week, I’ll record an episode for the JS Craft podcast featuring Swizec Teller, author of the Scaling Fast book. To get some context about him, I recommend checking out this really good talk between Bas Steinss and Swizec:

👉 youtu.be/hJ4Yytccl4Q

✍️Pen and paper are becoming a lost luxury! When preparing podcast questions, I’ve found it more enjoyable to take YouTube transcripts, run them through an LLM for formatting, then study & highlight them in print. Less screen time, better focus.

GitHub - paoloricciuti/mcp-add: Universal cli to add an MCP server to a variety of clients Universal cli to add an MCP server to a variety of clients - paoloricciuti/mcp-add

I got tired of figuring out how to add an MCP server over and over to various clients...so I've built mcp-add.

A CLI that knows the right questions to ask and where to put the info in most clients.

You can run it in interactive mode with `npx mcp-add@latest` or by passing args to do it in one go!

Pydantic, Jason Liu & MCP Meetup - April 8, 2025 YouTube video by Pydantic

RAG is the most economically impactful application of gen AI. Cursor & Claude Code are, at their core, RAG systems. For example one of Cursor improvements comes from using embeddings to predict where the cursor should move on next. Details in this talk by Jason Liu:
youtu.be/GqufkaLJV44