Even though this blog is my own writing, it’s not published, and I’m not beholden to any publisher rules, I don’t really enjoy having AI write for me. I find the writing, no matter how advanced the models get, to be too plain, too perfect, too banal.
I hate having things ‘land’, or ‘delving’ into ideas.
I find the fence-sitting, hedged, nuanced nature of AI writing frustrating to read.
I don’t copy and paste any AI text into my blog, because I don’t think it’s good writing for what I’m trying to express – which is my own thoughts, reflections, and ideas.
What I do use AI for in this blog is:
- Checking for spelling and grammar errors, because it’s better than Word.
- Discussing ideas and bouncing around concepts. AI never gets bored with me debating sociology and philosophy, and it sends me down rabbitholes I didn’t know about before.
- GIving me ideas or ‘jumping off points’ for experiments with the glasses and what I could do to write about them.
For reference, I tend to use Claude Opus 4.8. I used Fable briefly, and sometimes I’ll bounce between ChatGPT, Gemini, and others, depending on what I’m doing.
For images, I will sometimes use ChatGPT Images 2.0. if no reasonable image is available copyright-free. I am aware of the environmental and social critiques of these, and I will address these in another post.
Leave a Reply