Fairy tales, then, are not responsible for producing in children fear, or any of the shapes of fear; fairy tales do not give the child the idea of the evil or the ugly; that is in the child already, because it is in the world already. Fairy tales do not give the child his first idea of bogey. What fairy tales give the child is his first clear idea of the possible defeat of bogey. The baby has known the dragon intimately ever since he had an imagination. What the fairy tale provides for him is a St. George to kill the dragon. Exactly what the fairy tale does is this: it accustoms him for a series of clear pictures to the idea that these limitless terrors had a limit, that these shapeless enemies have enemies in the knights of God, that there is something in the universe more mystical than darkness, and stronger than strong fear.
As a teen I was once wandering around a church thrift sale and ran across *hundreds* of Harlequin Romance novels. It would seem that LLM's threaten this market more than others. That and the jobs of anyone writing Scooby Do episodes.
Romance has a very well-defined language of what exactly is expected from certain types of stories. From the readers I've talked to, that's a _feature_, not a bug. It's what they want, the same way that you want comfort food to taste the same way your mom used to make it.
Harlequin Romance books are probably at the far end of the formulaic spectrum. Who doesn't like some romance? Adds a lot of interest to almost any story. Though I would imagine any publisher of "cheap" formulaic novels, AI must be pretty exciting. Feed in some character descriptions and a few other details and turn the crank. In fact, why not have an automated user-driven vending machine for them? Sounds practical to me.
Though I have also joked about it for awhile that they do strike me as a rendition of "from each according to their ability, to each according to their needs." Hm. Funny the overlaps I see of the biggest objectors of them....
You know what's fascinating about the gap between Organic and Clanker?
It's shrinking exponentially. Almost daily.
Fair warning to all the author friends. Up your game, or you're already extinct. Working on an article about this very thing, Sam. Always appreciate your thoughts on it.
The juice ain't worth the squeeze for me when it comes to LLM's, mainly due to the fact that if I have to watch it's output as closely as I am required (to make sure my writing still says what I meant to say) I might as well just edit it myself. This may change if I write more or longer works, but not yet or maybe even ever. I still don't know if I'm capable of being a writer that others will respond to, but I'm holding out hope. 🙂
Yes. A subject for another post: my experience observing others is that if you know what you are doing in a problem domain, an LLM can be a tremendous help. Every job has scut work, and LLMs are great at helping get that out of the way so you can do the FUN parts. They’re especially good at doing repetitive, detailed work that requires a larger context set - checking data (manuscripts) for contradictions and inconsistencies, for example.
~~Every job has scut work, and LLMs are great at helping get that out of the way so you can do the FUN parts. They’re especially good at doing repetitive, detailed work that requires a larger context set - checking data (manuscripts) for contradictions and inconsistencies, for example.~~
Yes, very true. BUT, scut work has a purpose to it too, and if we're not careful, LLMs taking that will mess things up too.
There is something to be said for not using automation to do a job until you understand how (and why) to do it yourself. @Holly MathNerd made that point recently in one of her articles.
You're an idiot (not really, but I felt compelled to obey for some reason).
LLMs are here to stay. I use my pet LLM as an inexpensive editor and it does a pretty good job of dismantling my prose. It's not as creatively destructive as a biologically biased editor, but sometimes it can be a little snarky. Or perhaps it's just reflecting its operator...
Heh. :) I am an idiot, in a large number of problem domains. My only saving grace is that I can generally recognize that and not make a complete fool of myself.
LLMs are useless to me because there is no there there to trust. I haven't even explored their possible uses because I don't want to develop a reliance on their output. Schneier wrote an excellent book on this topic, "Liars & Outliers" that helped me formalize my thinking on this topic, and with a trust score of zero, there's nothing with any worth to me that I could ask them for.
*clears throat*
The actual G.K. Chesterton quote:
Fairy tales, then, are not responsible for producing in children fear, or any of the shapes of fear; fairy tales do not give the child the idea of the evil or the ugly; that is in the child already, because it is in the world already. Fairy tales do not give the child his first idea of bogey. What fairy tales give the child is his first clear idea of the possible defeat of bogey. The baby has known the dragon intimately ever since he had an imagination. What the fairy tale provides for him is a St. George to kill the dragon. Exactly what the fairy tale does is this: it accustoms him for a series of clear pictures to the idea that these limitless terrors had a limit, that these shapeless enemies have enemies in the knights of God, that there is something in the universe more mystical than darkness, and stronger than strong fear.
Thank you!
LLM's are definitely Artificial but so far I have yet to see them behave as intelligent
I have absolutely noticed the hypocrisy you mentioned: "AI is ok for me but not for you... " And that is why it is the future.
As a teen I was once wandering around a church thrift sale and ran across *hundreds* of Harlequin Romance novels. It would seem that LLM's threaten this market more than others. That and the jobs of anyone writing Scooby Do episodes.
Romance has a very well-defined language of what exactly is expected from certain types of stories. From the readers I've talked to, that's a _feature_, not a bug. It's what they want, the same way that you want comfort food to taste the same way your mom used to make it.
I do look forward to Scooby Doo on Epstein Island. "I sense a mystery here, gang!"
Harlequin Romance books are probably at the far end of the formulaic spectrum. Who doesn't like some romance? Adds a lot of interest to almost any story. Though I would imagine any publisher of "cheap" formulaic novels, AI must be pretty exciting. Feed in some character descriptions and a few other details and turn the crank. In fact, why not have an automated user-driven vending machine for them? Sounds practical to me.
A bit of what I wrote on LLMs myself.
https://natewinchester.wordpress.com/2025/09/08/dont-fear-the-bot/
Though I have also joked about it for awhile that they do strike me as a rendition of "from each according to their ability, to each according to their needs." Hm. Funny the overlaps I see of the biggest objectors of them....
That's some interesting insight. I'm going to have to give it some though, especially your point about exceptions.
You know what's fascinating about the gap between Organic and Clanker?
It's shrinking exponentially. Almost daily.
Fair warning to all the author friends. Up your game, or you're already extinct. Working on an article about this very thing, Sam. Always appreciate your thoughts on it.
The juice ain't worth the squeeze for me when it comes to LLM's, mainly due to the fact that if I have to watch it's output as closely as I am required (to make sure my writing still says what I meant to say) I might as well just edit it myself. This may change if I write more or longer works, but not yet or maybe even ever. I still don't know if I'm capable of being a writer that others will respond to, but I'm holding out hope. 🙂
Yes. A subject for another post: my experience observing others is that if you know what you are doing in a problem domain, an LLM can be a tremendous help. Every job has scut work, and LLMs are great at helping get that out of the way so you can do the FUN parts. They’re especially good at doing repetitive, detailed work that requires a larger context set - checking data (manuscripts) for contradictions and inconsistencies, for example.
~~Every job has scut work, and LLMs are great at helping get that out of the way so you can do the FUN parts. They’re especially good at doing repetitive, detailed work that requires a larger context set - checking data (manuscripts) for contradictions and inconsistencies, for example.~~
Yes, very true. BUT, scut work has a purpose to it too, and if we're not careful, LLMs taking that will mess things up too.
https://natewinchester.substack.com/p/mentorship-forgotten
There is something to be said for not using automation to do a job until you understand how (and why) to do it yourself. @Holly MathNerd made that point recently in one of her articles.
To make an analogy: All tools can be hammers, but there is such a thing as the right hammer for the job. Choosing the right "hammer" is the hard part.
Yep. Whatever the unintended consequences of AI are, it's not going to be what everyone is banging on about in the centres of approved discourse.
It'll be hell, the world being what it is, so we should seize the opportunity while we've got it to make something heavenly of the tool.
You're an idiot (not really, but I felt compelled to obey for some reason).
LLMs are here to stay. I use my pet LLM as an inexpensive editor and it does a pretty good job of dismantling my prose. It's not as creatively destructive as a biologically biased editor, but sometimes it can be a little snarky. Or perhaps it's just reflecting its operator...
Heh. :) I am an idiot, in a large number of problem domains. My only saving grace is that I can generally recognize that and not make a complete fool of myself.
LLMs are useless to me because there is no there there to trust. I haven't even explored their possible uses because I don't want to develop a reliance on their output. Schneier wrote an excellent book on this topic, "Liars & Outliers" that helped me formalize my thinking on this topic, and with a trust score of zero, there's nothing with any worth to me that I could ask them for.