In an AI-driven development world, speed is easy to gain but control is easy to lose. This session shows how to stay in charge of your code and your craft.

Prerequisite

Basic experience with modern development workflows and an interest in using AI tools such as Claude or Cursor. No advanced AI knowledge required.

Target Audience

Mid-level and senior PHP/Drupal developers who are already using or experimenting with AI tools like Claude or Cursor in their daily workflow. Also relevant for team leads and architects who want to keep strong technical ownership while benefiting from AI productivity gains. Anyone who has felt “the code works but I don’t fully own it anymore” will get the most value.

Outline

AI tools like Claude and ChatGPT are changing how we build software at high speed. We ship more with less effort, follow our intuition, and let AI fill in the gaps. It feels powerful.
But there is a downside we rarely discuss. We are no longer just handing over repetitive tasks — we are also delegating understanding. Code appears before we fully grasp it. Architectures emerge that we didn’t consciously design. Everything works, yet we risk losing our grip on what we are building.
This talk explores how to reclaim your autonomy as a developer in an AI-accelerated world, without rejecting the tools that boost productivity.
Vibe coding lets us move fast and trust our gut, but when speed becomes the default we lose real ownership. Can you explain the code you shipped? Do you understand the trade-offs? Would you feel confident maintaining this system in six months? And if it breaks, can you take responsibility?
These questions matter especially in Drupal, where long-term maintainability, collaboration and transparency are essential.
This is not an anti-AI talk. It is about developing with intention.
We will look at how AI-assisted workflows shift responsibility away from the developer, why “it works” is no longer enough, and the risk of losing architectural awareness. Most importantly, we cover practical strategies to stay accountable: how to critically evaluate AI-generated code and integrate AI without outsourcing your own judgment.

Learning Objectives

Understand the hidden trade-offs of relying heavily on AI-assisted coding
Recognize when you are losing architectural awareness and ownership
Learn practical techniques to validate, challenge, and deeply understand AI-generated code
Build habits that keep you accountable and in control of your codebase
Use AI tools more critically and effectively without giving up your autonomy

Experience level
Intermediate