AI in art, in writing, is becoming a real polarizing hot debate, with accusations of gatekeeping, accusations of slop-making, accusations of people being without talent, now able to just generate what others call slop. The name-calling and the insults are endless but there’s a lot of real sentiment behind it, on both sides, from what I can see.
On the one side you have folks who have worked to learn a manual skill of one kind or another, whether it is drawing, sculpting, graphic design, or many other arts. They’ve invested a sizable chunk of their lives. It may be their living. It may even be quite a core part of who they are. They’re able to express themselves through the medium of their skill that has taken time and dedication to build.
This is pretty understandable friction because their livelihoods, their identity, and many other aspects of who they are are under threat from somebody who can sit and blurt out whatever’s on their mind to a suitably sophisticated model and get something that looks like it’s probably competent. The typical response is going to be abrupt, at the least, and on social media that’s the mildest description. You get a lot of almost puritanical gatekeeping.
On the other side you have lots of folks: very casual folks, people running small businesses, and more, who’ve never really had the time nor the inclination to develop these skills. They still don’t have the time but do have a need for the skills and quite often don’t necessarily have the money to spare to buy a professional. That’s not everybody. That is some.
Many others are casual. Many others do have the money but would just rather cheap out, and others feel that the art world is just gatekeeping. Sadly if you read the social media posts, it certainly has that tone to it and gatekeeping is never attractive.
My argument isn’t for or against either side here. People fight about definitions all the time. Technology moves on all the time. Change happens all the time. Tools evolve and grow. Change is not always positive for all concerned. In fact I’d argue it’s quite often quite a neutral force, causing as much harm as good.
You’re not going to find a stubborn single side in this post, rather my rambling thoughts on the subject. I do have a position for my writing and I do have a position for my consumption of fiction. I don’t like it in my writing. I’ve tried it. I got an AI to improve a paragraph of one of my stories. It was competent, it was well written, and it was a cancer.
I happen to like writing, whether I use dictation software or use a keyboard. I like writing. I like trying to construct the sentences, trying to get some sense of sound, syntax, rhythm, and variation but I’m far from a professional. I’m somebody who’s read a few books on rhetoric.
Nevertheless if I write fiction, it’s my own rambling stumbling rhetoric that’s used, not the polished mid-average tones of an LLM trained on one entire internet. As for consumption of fiction, having been caught by quite a few AI-written books, my conclusion has come through experience: I really don’t like it. It has that mid, well-polished overuse of contrasts, of three-part patterns and others, and a weird use of metaphors, yet it does not keep quite a coherent character or story together.
I have no doubt the mechanical competence will get better over time but I still would rather it came from a mind. No AI currently can be described as having a mind.
On these pages there are still a few hangovers. Where Claude wrote an article, one or two of the maths ones, one or two of the technical ones, they will either be removed or rewritten over the coming weeks. All of the fiction of any kind will always be handwritten or dictated manually and polished a little by me.
As I like rhetoric I would expect the sum of my text to appear to have patterns found in AI because AI learned from us. I refuse to rewrite or remove patterns just because AI likes it to. In the end the reader decides.