I don’t use AI on this blog (or any of my writing) to the best of my knowledge. I know other people use AI for various reasons and this is not me picking a fight with them; this is me talking about my process. I acknowledge that I probably don’t know every single AI-using aspect of the tools that I use. However, when I detect them, I avoid them. I like to write, I like to type, and I can’t imagine wanting that process to be different. This is up to and including:
- My content management system. I use WordPress’s Classic editor and just type the HTML into the box.
- Spell checker or “help me write this” tools, though I do use Firefox’s “wiggly underline” to help make sure I’ve got things spelled right
- Speaking of Firefox, I’ve taken advantage of their “block AI enhancements” option. You can too.
- My laptop computer has Apple Intelligence turned off, as does my iPhone. My desktop computer is too old to even have it.
- I get most of my images from Flickr (or they’re my own) and try to avoid ones which have been obviously AI-generated or enhanced.
I’ve used AI/LLM tools before. I used to try DeepSeek occasionally for some of my work at the Flickr Foundation (think “what is this image” sort of things), found it wanting, and did not use it again. We had an AI clerkbot which we were encouraged to use to keep track of our daily/weekly tasks and I used it for a bit, found it wanting, and did not use it again.
The one bit of AI I have in my life now is that the code which makes up my reading list has been partially built by someone who was using Claude. My host upgraded their version of PHP, my booklist broke. I reached out to the man who had helped me fix it before. He asked my permission to have Claude help him update and document the code and I said okay.
If any of this changes, I’ll probably say something about it. I do have a lot of ethical and practical concerns with AI/LLMs companies as well as the tools they build. For now, I don’t use them because they solve no problems for me.
Outside handwriting recognition and optical character recognition, both of which used neural networks long before the GPT chatbots, I’ve been stumped trying to find a use for the bots. I’m supposed to teach students about the tech and outside that very limited use case related to OCR of handwriting that suffers from dysgraphia I’ve got nothing.
I think some AI tools can be useful when used to do computer things–for instance, “Computer, take all the things in this program that are tagged A and put them into this other program and tag them A” is the sort of job computers were designed to do but that I do not have the programming chops to accomplish on my own. That means I usually spend many hours moving things that probably a computer could do more quickly. If I were a programmer, I might find using an AI tool to do that as unnecessary and offensive as I find using an AI tool to write things. Or maybe not.
I am generally an AI resister, which I say proudly until I remember that I’ve used AI transcription tools. So. We are all complicated.