AI Doesn't Reduce Work, It Intensifies It
– hbr.org
A Berkeley Haas study of 200 employees found that AI makes workers take on more, not less. The productivity gains are real, but exhausting.
Read more [hbr.org]
Posts tagged with ai
– hbr.org
A Berkeley Haas study of 200 employees found that AI makes workers take on more, not less. The productivity gains are real, but exhausting.
Read more [hbr.org]
I've been using Claude Code as my daily driver for coding tasks. Over time, I've built up a pretty specific configuration that makes the whole experience better. I keep everything in my dotfiles repo under config/claude/, so it's easy to sync across machines. In this post I'll walk through my setup.…
Join 9,500+ smart developers
Get my monthly newsletter with what I learn from running Spatie, building Oh Dear, and maintaining 300+ open source packages. Practical takes on Laravel, PHP, and AI that you can actually use.
"Freek publishes a super resourceful and practical newsletter. A must for anyone in the Laravel space"
No spam. Unsubscribe anytime. You can also follow me on X.
We recently released the Flare CLI, a command-line tool to manage your errors and performance data. It also ships with an agent skill that lets AI coding agents use Flare on your behalf.
The CLI has dozens of commands and hundreds of options, yet we only wrote four commands by hand. Our laravel-openapi-cli package made this possible: point it at an OpenAPI spec, and it generates fully typed artisan commands for every endpoint automatically.
Here's how we put it all together.

The Flare CLI lets you manage errors and performance monitoring from the terminal. It was built with almost no hand-written code, generated from our OpenAPI spec. Having a CLI is useful on its own, but where it gets really interesting is when you let an AI coding agent use it.
The Flare CLI ships with an agent skill that teaches AI agents like Claude Code, Cursor, and Codex how to interact with Flare on your behalf. Let me show you how it works.
At Flare, we track errors and monitor performance for your applications. Until now, that meant opening the Flare dashboard in your browser whenever you wanted to check on things.
We just released the Flare CLI, a command-line tool that lets you manage your errors, projects, and performance monitoring data directly from the terminal. It also ships with an agent skill that lets AI coding agents like Claude Code and Cursor use Flare on your behalf. And the fun part: the entire CLI was built with almost no hand-written code, generated from our OpenAPI spec.

Let me walk you through how to install and use the CLI.
Matt Pocock shares his ideal Claude Code status line setup, showing repo name, git branch info, and context window usage percentage. The post includes step-by-step instructions using bash scripts and the ccstatusline package.
Read more [www.aihero.dev]
A comprehensive guide to making your Laravel app work well with AI agents. Covers llms.txt, markdown responses, structured data, and coding guidelines.
Read more [hafiz.dev]
Geocodio shares practical advice on getting engineering teams to adopt AI tools like Claude Code, addressing the growing gap between eager adopters and skeptical engineers.
Read more [www.geocod.io]
Hafiz Riaz walks through building a document analyzer agent with the Laravel AI SDK. The tutorial covers structured output, streaming responses, and tool use, all in about 30 minutes.
Read more [hafiz.dev]
We just published a new package called Laravel Markdown Response that lets your Laravel app serve markdown versions of your HTML pages. Your existing controllers and views stay exactly the same.
Let me walk you through what the package can do.
My Oh Dear co-founder Mattias Geniar was a guest on the WP Legends podcast. They talked about how we built Oh Dear over the past eight years, common misconceptions about monitoring, and how AI is changing developer workflows.
After posting a screenshot, I often get questions about which editor, font or tools I'm using. Instead of replying to those questions individually I've decided to just write down the settings and apps that I'm using.
Pete Koomen argues that most AI apps are "horseless carriages," bolting AI onto old paradigms instead of letting users shape the prompts that drive them. He makes the case that the best AI apps should be agent builders, not just agents.
Read more [koomen.dev]
– x.com - submitted by Aaron Francis
Aaron Francis shares how he used Claude to create an entire launch video for his new app, without using any traditional motion graphics software. No Remotion, no MCPs — just clever prompting and iteration.
Read more [x.com]
Amit Merchant walks through practical use cases for the Laravel AI SDK: mining user data with Eloquent models as context, building a code review bot that comments on PRs, and creating adaptive quiz systems for EdTech apps. Nice examples of agents, structured output, and conversation memory beyond the typical chatbot demo.
Read more [www.amitmerchant.com]
Martin Fowler on the overuse of bold in technical writing — and how LLMs have picked up and spread this practice. The more you emphasize, the less power it has.
Read more [martinfowler.com]
Ray 3.0 is here! Completely rebuilt for better performance (60% less memory), a fresh new look, message archiving, and MCP support so AI agents can interact with Ray directly.
Read more [myray.app]
Christoph reflects on how personal AI assistants have changed his daily workflow — from calendar updates to searching the web less. An interesting take on where this technology is heading.
Read more [christoph-rumpel.com]
At Laracon India, I launched a major update of Ray. For that talk, I needed a little demo project to showcase Ray. I built a simple website about a then-fictional mobile app to play a Scrabble-like word game called WordStockt.
But then I got curious: how far could I push AI-assisted development? Could I actually just create the whole game? After about 10 days, WordStockt is a fully functional word game that's 98% vibe-coded. It's available for iOS and Android. In this post, I'd like to tell you more about it.

If you're using AI tools like Claude Code to help write code, you've probably noticed they don't automatically know your team's coding conventions. The AI might write perfectly valid PHP, but it won't follow your specific style guide unless you tell it to. That's the problem Laravel Boost solves. It…