A 9-minute video for Premium members. I tested Claude Opus 4.6 and GPT-5.4 Codex on an older Laravel project called Bookstack. I tried with 5 different prompts: did LLMs deliver? And which one did a better job?
A usual weekly newsletter covering what is happening in AI coding community, including a few of my videos and tutorials. Enjoy!
“What will we do as devs in five years? Will our jobs survive against vibe-coding wave and agentic coding?” - these are just a few questions I heard over last two days, at a conference in Amsterdam called Laracon EU for Laravel/PHP developers. I have FOUR thoughts and takeaways from Laracon audience.
Another week full of AI coding experiments. It's never really simple, with all those models, IDEs, tools, workflows and recommendations keep changing. So I'll keep experimenting, join me by watching the videos below!
A 13-minute video for Premium members. Version 2 of my Laravel Evals for AI coding models, with 6 new projects and challenges. One clear and surprising winner.
I decided to write a prompt so AI agent would research the market for useful Laravel/PHP packages suitable for specific project, based on its existing codebase.
I spent ~3 days on a methodology to test LLMs with Laravel projects, on different use-cases, with evaluations. In other news, OpenClaw was acqui-hired, while many people release their Claude Code tricks and prompts.
A 17-minute video for Premium members. My biggest AI models comparison ever, so far. I've prepared the Evals (tests) for many different use-cases that I will use now and in the future "showdowns".
I've found a great article by G-J van Rooyen where he suggests two prompts for the case when you inherit an old codebase and need to quickly understand the flaws and the backstory of main code decisions.
Claude Code can give you a hint on what you can do better next time.
You'll also get TWO free tutorials:
"My Favorite 10 Tips & Tricks" on Claude Code and Codex CLI!
Sent every Wednesday. No spam, ever. Unsubscribe anytime.