AI policy

This is my personal policy around use of AI (as at 2026), specifically large language models. Like many people in tech, I have “thoughts”. I’m definitely not a hardline anti-AI absolutist, but I have serious concerns about the impact of slop on both the reader and “creator”.

Where I won’t use AI

  1. No content generation! The stuff you read on this website will never be generated by AI. Everything will have been crafted by a human (me), and reflects my thought processes.
  2. No vibe coding! I create the technology that runs this site for fun, and to keep my hand in with some lightweight coding and web technologies. I don’t want to hand off that responsibility to a model trained on other peoples code. Let me tinker in peace!
  3. No planet-burning search! When I search the internet, I will use a search engine, not a making-things-up machine. I have just switched to using Ecosia as my search engine, and have disabled the AI overviews.
  4. Making art of any kind! I occasionally write music, and sometimes create designs for things. For the same reason I won’t vibe code stuff, I won’t lean on GenAI to create artworks. If I need an image for a blog post or presentation, I’ll either take photo, deploy my own limited drawing skills or find something via a Creative Commons search.

Where I might use AI

Pretty much my only use of LLM chats is to help structure my thoughts. I’ll conduct a chat about an idea that I’ve had and use the back-and-forth as a way of refining my ideas. This will only ever be used by me in the form of rough notes.

Why I find this useful? I recognise I’m a better editor than I am originator when it comes to writing. Faced with a blank page I can (very!) easily be distracted by something else. The short sessions of braindumping and kicking ideas around with an apparetly intelligent parrot helps me get ideas out of my head and into the world.