AI usage
This page outlines how I use Large Language Models (LLMs) on this site. I believe in transparency around AI usage, especially as it becomes more prevalent in content creation.
Whenever I use LLMs on this site, I manually review suggestions before I accept them and verify that outputs are accurate and functional, and done the way I would do them myself.
Code and engineering
It might be hard for me to quantify how much of my code has been written using LLMs. Primarily, I am using LLMs here as playground to help me better understand how I can think about and use them at work. I try to write to as much as possible about the process and how I’m thinking about it along the way.
I still write all of my own CSS and make all decisions about interaction design, particularly where accessibility is concerned. My experience with agentic coding here has not been great.
As a front-end developer, I have found that using an LLM is helpful to me in working through issues that are outside of my wheelhouse. For example, handling issues related to image optimization and uploading to the CDN in a way that optimizes developer experience.
Content writing
All em dashes are my own. I do not use LLMs to write any content on this site. The only time I use it in writing is correct grammar mistakes when I am done writing a post. I include instructions not change content, voice, or tone; only fix grammar.
Image creation
None unless it is contextually relevant.
Alt text
I will often use LLMs to write alt text for images, particularly when it comes to screen shots, graphics, or images for photo essays. I always review and edit the alt text to ensure it is accurate and descriptive.