The top things I have read this week 🏆

Week 10 / 2026

Hey there! Welcome to a new edition of Hybrid Hacker! 🌟

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

---

🌀 The New Pyramid of Software Engineering

12 min • by Luca Rossi

Most AI advice feels either too volatile or too tactical to me. This week I published a piece that zooms out and proposes a three-layer pyramid — Developer Experience, AI, and Product Engineering — where each layer only works if the one below is solid. It also includes a self-assessment survey to rate your team on each layer.

---

🥇 My (Hypothetical) SRECon26 Keynote

8 min • by Charity Majors

One year after delivering the SRECon25 keynote, Charity reflects on how much her stance about AI has changed, going from grudgingly acknowledging it to fully embracing it. Superb post, extremely well written (as usual).

---

🥈 The Software Development Lifecycle Is Dead

11 min • by Boris Tane

Provokative post about how AI has compressed the SDLC to the point of being unrecognizable. The new lifecycle is a tight loop: intent → build → observe → repeat. The new big skill is context engineering, and the new safety net is observability.

---

🥉 The Stranger Secret: How to Talk to Anyone

8 min • by The Guardian

We’ve lost the habit of talking to strangers. Technology and social norms have shrunk everyday exchanges to near zero, and that’s a loss — for us and for society. This is a great reminder that small, unplanned conversations are one of the most enriching things we can do, and that the skill is worth actively maintaining.

---

4) Red/Green TDD — Agentic Engineering Patterns

2 min • by Simon Willison

“Use red/green TDD” might be the highest-leverage four-word prompt you can give a coding agent. These are Simon’s words, and I confirm! Write tests first, confirm they fail, then let the agent iterate until they pass. It prevents the two biggest agent risks: code that doesn’t work, and code that’s unnecessary. Part of Simon’s excellent new Agentic Engineering Patterns guide.

---

And that’s it for today! If you find this list useful, please share it with your friends!

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca