I was listening to a non-fiction audiobook today. This is quite rare, mostly I read and listen to science fiction. It’s a book about parenting. It’s a compelling book, it’s about how Americans parent wrong and are weird more generally. Clearly this author is on to something. 

I found myself fascinated by a lot of what the author had to say noy only in application to parenting, but in application to the science fiction novel I’ve been writing in my head for the past half-decade. 

I had a sort of bone-headed realization: I can gain insight into my sci-fi novel from things other than sci-fi. After that I had another bone-headed realization: I might be pidgeon holing myself and limiting my ideas by only reading sci-fi. 

While the quality of my hypothetical science fiction novel is not a good enough reason to deny myself the immense pleasure of reading as much science fiction literature as I want, it is a valuable question, I think. 

It made me feel like an artificial intelligence being given bad training data. As we all know, AI cannot train themselves. AI trained with content from AI creates utterly surreal images and words far beyond the comprehension of even the Omnissiah. This is sub optimal for making work that is good and enjoyed by other humans.

But that brings up an interesting sidebar; people are poisoning AI. By sabotaging their artwork, either by countless shit versions of it, or by doing something to the pixels I can’t even begin to understand, it becomes not only useless to systems like Chat GPT and Midjourney, but actively detrimental. 

I think that’s awesome.
I just don’t want it happening to my brain by my own hand.