June 7th, 2026

Articles of the week

This week I picked up these articles:

Learned/Notes

Cassidy Williams on technical blogging

I think I remember watching Cassidy's streams when I first got into tech a couple years ago, could be wrong though. I don't think I've read any of her articles and I definitely don't know much about her, but when I read that she has been writing for 20 years, I was like "that's sick". She probably started really young and her main advice is simple: just write, because that's the only way you'll get better at it. It's something I've known for a while and learned from various sources but my favorite will always be the book titled "Ultralearning" by Scott H Young. In his book he talks about how to learn and get good at anything, you need to do the thing directly or as directly as possible. She also said you are the one that will care the most about what you write. Great advice from Cassidy!

What I found super interesting is how much writing did for her career and life. She landed roles because of her blogs, made friends, learned things, and even references her own posts when she runs into the same problem again — I don't write technically for now, but I want to and I hope this happens to me one day lol.

She also mentioned starting a newsletter was the most important project she's picked up. I'm tempted to do something similar once I get a better idea of what to write, I don't want to start a newsletter until I know I'm providing real value for others.

Fast is better than slow

This was a short but nice blog by Patrick Dubroy. My favorite line was: to move faster, don't waste time doing things that nobody asked for. It sounds obvious, but I think a lot of engineers (myself included sometimes) spend time thinking about the perfect solution and how to include even more than that when the ask was much simpler.

He links to a couple of articles by Jamie Brandon on speed and moving faster, and he also points to his own blog about small increments (which I read right after). He pushes back a bit on Deep Work, saying it's more of a want than a need for most engineers in tech — I still don't think I fully agree with that. I think the value of Deep Work depends a lot on the work. But the idea of breaking work down into small, shippable pieces is something I've been doing more these past couple of months and it's been working for me. I also think with AI, coming up with POCs or some sort of working demo is far easier now, which should help you get feedback far sooner than before (pre-AI times).

Getting things done in small increments

This came directly from the previous blog, same author. Patrick's argument here is that most engineers don't have the luxury of long uninterrupted blocks of time because of how tech companies are set up — meetings, teams/slack, on-calls. So instead of waiting for a perfect Deep Work window, you get comfortable making progress in smaller chunks.

His advice is simple:

  • Always know what your next step is
  • Make some progress every day
  • Block distractions for whatever small window you have

What (in my opinion) gives his advice credibility is that he built a JavaScript parser with 5k stars on GitHub following these exact principles (not easy to achieve). This isn't just productivity theory, he's actually done it. I agree with him. I've been living this approach for a while and it works!

Software after AI

In this blog Tomasz breaks down 7 components for building with AI: environment, context (such as using RAG to retrieve more information), tools (giving the model capabilities to act), a harness to keep things organized and allow teams to work on a system together safely, sandboxed environments for agents, observability, and architectural judgment (things like which model to use and when you want deterministic vs non-deterministic behavior).

To be honest, this one felt more like an overview than something that gave me new ideas, I feel like anyone who has been playing with AI for a while is aware of these. Also, I'm currently going through the book by Chip Huyen "AI Engineering" and a lot of these concepts are ones I'm learning there in more depth. I didn't really like this one.