April 12th, 2026
Articles of the week
This week I picked up these articles:
- The #1 mistake you're making with your mentor (and how to avoid it)
- The Git (log) commands I run before reading any code
- The unwritten laws about Software Engineering
- Eight years of wanting, three months of building with AI
Learned
The #1 mistake you're making with your mentor (and how to avoid it)
This was one of my favorite articles I've read about these topic, not only was it short and to the point, but it had a good example to higlight the importance of what it talks about.
- Getting a mentor is important, but it's more important to get value from it, which most people don't do correctly. You don't want to waste your time or your mentor's time, so be proactive about what you'll discuss with them.
The git commands I run before reading any code
- I never thought to use git log this way, it's literally a master piece. I think everyone should read this blog, it's super short.
The unwritten laws about Software Engineer
I honestly didn't really like this article, I rated it like a 5/10 I think? It's just a list of things that you learn by doing prod deployments and I didn't think it was worth reading. There was only one thing I liked about this one:
- You need to test your data backups, not only seeing that they exist, but like actually making sure they bring the data back. It does say most people never do this... and I'm afraid they're right.
Eight years of wanting, three months of building with AI
I relate to this article hard. I'm someone who overthinks and is super indecisive, AI definitely helps when you're this way.
- They use AI as an accelerator (no surprise there), but also mentions how AI can make you addicted to "one more prompt" which sounds funny but while reading it, I realized I've done this. Thinking "let me just prompt it a different way to see if that works", and just wasting time when you could get it done yourself.
- It's just even more important now to know Software Architecture since AI can produce a lot of code, but it still will produce shitty code unless you prevent that, by specifying a good architecture.