Aghh, I've lost two days of my holiday because I've had to change roles at work. This year I get to look forward to more hours, more busy work, less pay and a multitude of hoops to jump through to prove I can do the job I had been doing for the past three years because I didn't take a straight forward path to teaching. I have a day and a half to not think about all that, tomorrow evening will be scooping together everything I need to go back to work and prepare for my course, so today I'm going elsewhere with my thoughts.
Last weekend I went on a writers' retreat with some friends my writing group. Well, writers' retreat is a loose label. We went to a summer cottage in the Turku archipelago, in the west of Finland. This was our fourth year of having a Helsinki Writers Group retreat since Covid, though we missed the last two years, I think, due to low numbers. It's hard to organise 6 or more people to do anything together, but totally worth it when we manage.
We did a few writing things, had writing related talks and did writing prompts. I did a lot of journaling, but mostly we just hung out and did Finnish summer cottage things: picked berries and mushrooms, walked in the woods, sauna'd and jumped in the sea, drank a bit, played some games and ate a lot of food. All in all, a lovely relaxing weekend, except for the long drive.
One of the presentations from a lovely writer that works in IT was on AI witch hunts for writers: how to recognise AI writing (most of which they have already found ways around) and avoid our writing being mistaken for AI. How is this even a problem?!
But I know it is, it's a debate that has popped up around the edges of this blog. I have started posting this blog on Substack as well as Blogger and Substack has recently introduced an AI checker for posts, via Pangram. I don't think posters can opt out of this as it's supposedly to protect the reader from believing AI slosh is human-generated. (Maybe you can, I'm still trying to read the instructions.) I'm not sure how often it happens with writing, but I do see a lot of AI-generated images everywhere. It's sad that on a site for writers and creatives, posters are so hungry for likes and subscribers that they would use AI at all, but I guess they do.
I've been trying to ignore it, but it's making me reconsider whether I want to post this also on Substack because I fear this checker will be feeding my work directly into Pangram's LLM AI. I don't want to help train AI to write more like a person. (Checking this before posting - Substack says it doesn't use the detector to train AI and that it offers a block from external training models. Why do I not feel reassured?)
When I voiced this concern, my friend pointed out that my writing was probably being gobbled up in other ways. They are finding countless ways at consuming as much raw data as they can, possibly even scanning copies of last available hardcopies of books and then shredding them. Maybe this is just scaremongering, but with AI nothing feels implausible, just a warning.
This is where we are at, in my lifetime we've gone from the excitement of Pong and writing simple 001 code commands to AI writing our stories and poems and faking photographs of sunsets to make someone money through little digital hearts. We have to prove that we are writing our thoughts ourselves, that we were the one behind the camera or the pen.
The easy answer is don't post online and again I'm left to wonder why I do. Why do I like seeing the finished post, why do I check my stats to see if anyone is reading? I couldn't tell you how many subscribers I have (hello out there, if there are any), I don't charge to read my posts, but I'm not immune to the dopamine rush from being published, from being read.
So for that moment of chemical pride, do I have to just accept that AI is reading this and eating it up in its greedy digital maw to regurgitate my meandering human style at some later date, that bots are most of my visits rather than real eyes? In 2026, I guess I do.
How long will I hesitate before I hit Publish?
. . .
Not even an hour and some of that time was spent looking for Blogger's AI policy (doesn't seem to have one) and reading Substack's AI detector instructions and info. Sad, but here I am.
No comments:
Post a Comment