Ask HN: How do novelists feel about LLMs?

As a tool or an adversary?

As a tool I have found it useful a couple of times for difficult plot elements like in my last book I wanted to know if I spelt a real word on an old numeric keypad (the ones with 3 or 4 letters on the number key) what other possible word combinations could it be mistaken for. I also tried all of them proof-reading one of my free-novels for errors and inconsistencies a few weeks back and they were absolutely terrible.

As an adversary they generate pretty useless story lines but can generate so many of them, so quickly it is screwing up the algorithm especially for small indie authors who will soon have their work lost in the noise. - kind of sad really.

There's no point because there's few things as human as writing a novel. Also I think these models work great for technical generation of things like code, but it doesn't take imagination to do engineering work. Also look at the way these data centers compete with humans for resources like power and clean water, and the surveillance nightmare we're looking at with the dovetailing of authoritarian rule and total information awareness.

Do not want them anywhere near my work and other people I talk to feel the same.