Had a project a few weeks back. A client needed a proof-of-concept for a new feature set, which meant spinning up a bunch of custom Gutenberg blocks, and quickly. I handed it to a junior dev on my team, a sharp guy, but I watched him get bogged down in the details: the block.json boilerplate, the registration functions, all the scaffolding.
My first instinct, baked in over 14+ years, was to tell him to muscle through it. That is how we learned, right? But he was burning hours on problems that were already solved. My old-school thinking was the bottleneck. We were not building character, we were wasting the client’s money. It was time to take a serious look at the AI tools for web development that everyone keeps talking about.
The AI tools I actually kept using
You can find dozens of listicles out there. The idea for this post actually came from a decent one on the official WordPress.com blog. But most of them are just a laundry list. You do not need eight tools. You need one or two that fit straight into your workflow without adding friction. For me, it came down to a couple of code assistants.
The first was GitHub Copilot, right inside VS Code. Instead of looking up the exact syntax for register_block_type, the junior dev could type a comment describing what he needed, and Copilot generated the whole PHP function for him in seconds.
My first thought was, “the code is probably garbage.” And you do have to watch it; you cannot accept everything it suggests. But for boilerplate and standard functions, it is about 90% of the way there. It gets the structure right and you guide it from there. We asked it for a basic block registration function, and this is what it produced:
<?php
/**
* Plugin Name: BBioon Custom Blocks
* Description: A plugin to register custom Gutenberg blocks.
* Version: 1.0
* Author: Ahmad Wael
*/
if ( ! defined( 'ABSPATH' ) ) {
exit; // Exit if accessed directly.
}
/**
* Register the custom block.
*/
function bbioon_register_custom_cta_block() {
register_block_type( __DIR__ . '/build/cta-block' );
}
add_action( 'init', 'bbioon_register_custom_cta_block' );Is it perfect? No. But it saved the dev 15 minutes of digging through docs and typing, and across a dozen blocks that adds up fast. The other tool we tried was Cursor, which is basically an IDE built around AI. It is better at holding the context of the whole project, not just the file you are in, which helps a lot when you are refactoring or debugging across several files.
So what’s the point?
These AI tools are not going to replace experienced developers. They will not architect a complex WooCommerce checkout flow or debug a race condition. What they are very good at is taking the boring, repetitive parts off your plate. They let a junior dev punch above their weight and free a senior dev to focus on the hard problems that clients actually pay us to solve.
This gets complicated quickly. If you are tired of untangling someone else’s mess and just want your site to work, get in touch with my team. We have probably seen it before.
The point is to stop treating them as a replacement for skill and start using them to amplify it. Once you do, it changes how you approach the work.
