Being more explicit with the limited AI usage in my blog posts' content

I don't want to read lengthy blog posts that have been spat out by a Large Language Model (LLM), because the author hasn't felt they want to write the words to share their point.

I'm not going to rehash why that's (generally) a bad thing, as a lot of brilliant writers like Bryan Cantrill have written eloquently on the topic.

What is important is that I don't want my readers to feel they're needing to work out what of my content is AI-assisted, and so I want to make it super explicit assisted - and when it is, in what ways - so the reader can decide if it's worth their time to read.

I've previously noted that I track my LLM usage quite carefully, so I already have metadata on my blog posts to note where I am using AI model(s) to support the post, but until today, it wasn't clear what usage was for what purpose.

I've updated my metadata as best as I can (especially for some of the older posts where I can't necessarily remember exactly what bits of content may have been edited) and will continue to do this with posts going forwards.

You can see an example of this more in-depth metadata i.e. with the AI usage notice on my Logs as UX post.

I've also created a /ai page to note this, and provide a running list (and count) of posts that have had any AI involvement in them. We'll see how that list increases over time.